UEU-co logo

AccessSurgery – Print

  Print  |  Close Window

Note: Large images and tables on this page may necessitate printing in landscape mode.

Skandalakis’ Surgical Anatomy > Chapter 18. Large Intestine and Anorectum >


The large intestine is formed by the following anatomic entities:


Ileocecal valve



Ascending colon

Hepatic flexure

Transverse colon

Splenic flexure

Descending colon

Sigmoid colon


In this book, the appendix has been presented in the preceding chapter.

Following our presentations on the above list of topics is a special section written by Professor Ahmed Shafik. Dr. Shafik, who is chairman of the Department of Surgery and Experimental Research at the medical school of Cairo University, Egypt, is such an original thinker in regard to the large intestine and anorectum that we invited him to share his unique perspective with our readers. We are grateful to him for interpreting his research and offering his philosophy to us, and we are honored to provide his excellent work on the large intestine and anorectum to our readers. We encourage careful thought about Dr. Shafik’s innovative ideas.

NOTE: Some writers use the word “colon” to mean the collectivity of four anatomic entities: ascending, transverse, descending, and sigmoid colons. Occasionally we will employ the word “colon” or “colonic” in that manner to mean “large intestine” (e.g. “colonic” wall for “large intestinal” wall).


The anatomic and surgical history of the large intestine and anorectum is shown in Table 18-1.

Table 18-1. Anatomic and Surgical History of the Large Intestine and Anorectum

Hippocrates ca. 400 B.C. Treated hemorrhoids with white-hot iron or by burning them off; treated fistulas by use of seton
Aristotle (384-322 B.C.)   Used the word “colon”; noted that congenital malformations occurred more often in boys than girls
Soranos of Ephesus (A.D. 98-138)   Performed digital rupture of membranes of the anal canal of newborns
Galen (121-201 A.D.)   Named the rectum “apefthismenon”
Paul of Aegina (625-690 A.D.)   Collected lost works of Heliodorus, Leonidas, and Antyllus, and presented first description of a surgical technique for anal atresia
Albertus Magnus (1193-1280)   Works indicate that anal atresia was well known during Middle Ages
John Arderne 1367 Wrote paper that contained the basis of anorectal surgery
Antonius Benivenius (1513-1572)   Studied recto-anal agenesis in autopsy; reported recto-anal agenesis with vaginal fistula
Tobias Cneulinus ca. 1580 Unsuccessful perineal operation for recto-anal agenesis because the rectum could not be identified
Fabricius Hildanus 1593 Successfully opened anal agenesis by incision, introduction of rectal speculum, dilatation of the opening, and application of lead carbonate and meninge-dye
J.S. von Grafenberg 1609 Reported the first case of anal agenesis with urethral fistula; reported isolated rectal atresia without anal atresia
G.T. Dürr 1668 Described anus copertus with perineal fistula; incised the anal membrane and the obstructed anus, curing the child
Hendrik van Roonhuysen 1676 Ruptured anal stenosis with a knife and successfully maintained the opening with salves and instruments
Physician to King Louis XIV 1686 Conceived of and performed fistulotomy on the king
Frederik Ruysch (1638-1731) ca. 1700 Reported spontaneous rupture of anal atresia in a 5-day-old child (who died soon thereafter)
Saviard 1702 In a child with no trace of an anus, he inserted a lancet, and entered the blind rectal pouch. Meconium was released and the child survived.
Littré 1710 Successfully treated imperforate anus by opening colon in left lower quadrant of abdomen
Heister 1718 Operated – using a trocar – on two children whose rectum ended at the level of upper sacrum; both died
Barbout 1739 to 1775 Reported two cases of recto-anal agenesis with recto-cloacal fistula
Percivall Pott 1765 British doctor whose writings pointed up the advances Britons had made in colorectal diseases and surgery
Pillore 1776 Performed cecostomy for cancer of the lower bowel
Petit 1781 Incising for recto-anal agenesis, a tumor, but no rectum, was found. After incision of the tumor, meconium discharged. Patient died.
Benjamin Bell 1787 Successfully treated 2 cases of recto-anal agenesis; newly formed orifices tended to shrink from scar tissue
Duret 1793 Successfully treated imperforate anus by opening left lower quadrant of colon
Latta 1795 Successfully treated recto-anal agenesis, but dilatation therapy was required for nine full months
Callisen 1798 Suggested extraperitoneal colostomy at the lumbar area for imperforate anus
Meckel 1817 Studied embryology of the normal colon
Lisfranc (1790-1847)   Operated on colonic tumors
Frederick Salmon 1835 Salmon was founder and chief surgeon of St. Mark’s Hospital in London, which was the pinnacle of knowledge and treatment of colon and rectal diseases
William Allingham   At this hospital William Allingham wrote the first textbook devoted entirely to anorectal disease; it describes hemorrhoid treatment by excision and ligature
Amussat 1835 Described surgery for imperforate anus.
1839 Adopted Callisen’s procedure.
G.M. Bushe 1837 Wrote first American proctology book that later acquired international acclaim
Nélaton 1839 Exposed, fixed, and incised distended loop of bowel proximal to obstruction
Miller 1857 First successful operation for recto-anal agenesis with bladder fistula
J.H. Bigelow 1858 Wrote that at the present state of the art of surgery, children with rectal or anal atresia should die without operation
W.H. Bodenhamer 1860 First to produce a clear classification of rectal and anal malformations
Teale (1801-1868)   Favored exploratory surgery for bowel obstruction
Mason 1873 Reported 80 cases of colostomy for obstruction with 32.5% mortality
Wilks 1875 Described ulcerative colitis
Hutchinson 1878 “. . .exploratory operations for the relief of abdominal obstruction, the cause of which cannot be diagnosed, are not warrantable.” Stated that by the time a surgeon is called, “the stage at which abdominal taxis is most hopeful has passed.”
J.W. Matthews 1878 American physician began teaching in America the principals of anorectal surgery he had learned at St. Mark’s Hospital, London
Edmund Andrews & E.W. Andrews 1878 Co-authored one of the first textbooks on anorectal surgery
Billroth 1879 Performed sigmoid resection and exteriorization of the proximal bowel as permanent colostomy
Kraske & Kocher ca. 1880 Perfected the sacral approach to rectal tumors
Parker 1883 Urged extension of protracted palliative management of intestinal obstruction
Reybard 1884 Reported survival after resection and anastomosis for cancer of the colon
Greves 1885 Advocated operative intervention for intestinal obstruction
Bryant 1885 Reverted to the intraperitoneal maneuvers of Pillore and Duret
Tait (1845-1899)   Remarked that accurate diagnosis could only be made by exploration, “which is better performed before than after death”
H.O. Thomas 1885 Published monograph against surgical intervention for intestinal obstruction, “The Collegian of 1666 and the Collegians of 1885.” Stated that cases of acute intestinal obstruction “belonged to the department of medicine, the surgeon was a mere assistant. . . An operation may be required in a few hours or it may not be required for weeks.”
Mikulicz 1886 Reverting from his formerly more aggressive surgical approach, decided that laparotomy often offered no hopeful prospect for relief of acute intestinal obstruction
Hirschsprung 1886 Described autopsies of two infants who died from congenital megacolon (Hirschsprung’s disease)
Fitz & Senn 1888 Advised 48-72 hours observation before patient diagnosed with intestinal obstruction was turned over to the surgeon
Retterer 1890 Studied urorectal septum and cloaca
Bloch & Paul 1892 to 1895 Development of “obstructive resection,” in which portion of bowel with tumor is brought outside abdominal cavity; 1-2 days later, it is divided by cautery, forming a “loop” or “double-barreled” colostomy
Mall 1898 Studied the development and position of the human intestine.
1899 Observed the extraembryonic growth of the intestines and their return to the abdomen.
J.W. Matthews 1899 Became one of founders of American Proctologic Society. Later called “the Father of Proctology.”
Mikulicz 1905 Perfected and popularized obstructive resection
W.E. Miles 1908 British surgeon who described combined abdominoperineal resection for rectal cancer
Chilaiditi 1910 Described hepatic flexure between liver and diaphragm
Morton 1912 Offered a description of congenital absence of the colon using segmental arterial ligation in dogs
Johnson 1913 Studied development of colonic mucosa
Dott 1923 Developed a classification of abnormalities of intestinal rotation based on embryologic observations
Henri Hartmann 1923 Described treatment for obstructing carcinoma of the distal colon
Miles 1926 Developed combined abdominoperineal resection
Rankin 1930 Improved the double-barreled colostomy
Cuthbert Dukes 1932 Originated classification for carcinoma of the rectum
Ladd & Gross 1934 Created a new classification of anorectal anomalies that was the standard for many years
Kirschner 1934 German doctor demonstrated a combined synchronous approach to the abdominoperineal resection
Gilchrist 1938, Published studies of retrograde lymphatic spread in regard to gastrointestinal carcinoma
Swenson 1950 Designed operation for anal sphincter-preserving removal of aganglionic segments of colon
Stephens 1953 Recognized that the puborectalis muscle is the most critical to continence
Scott 1959 Investigated autonomic supply of the rectum and anorectum
Duthie & Gairns 1960 Investigated sensory innervation of the lower anal canal
Skandalakis et al. 1962 Collective review of cases of smooth muscle tumors of the colon, appendix, and rectum as reported in the world literature
Painter & Burkitt 1971 Studied relationship of low-residue diet to diverticulosis
Stephens & Smith 1984 Classified anorectal anomalies as high, intermediate, low cloacal, and rare
Peña 1990 Recommended posterior sagittal anorectoplasty

History table compiled by David A. McClusky III and John E. Skandalakis.


Estrada RL. Anomalies of Intestinal Rotation and Fixation. Springfield IL: Charles C. Thomas, 1958.

Kevorkian J. The Story of Dissection. New York: Philosophical Library, 1959.

Muldoon JP. History of colorectal surgery. In: Mazier WP, Levien DH, Luchtefeld MA, Senagore AJ (eds). Surgery of the Colon, Rectum, and Anus. Philadelphia: WB Saunders, 1995.

Schärli AF. Malformations of the anus and rectum and their treatment in medical history. Prog Pediatr Surg 11:141-172, 1978.

Skandalakis JE, Gray SW, Shepard D, Bourne GH. Smooth Muscle Tumors of the Alimentary Canal: Leiomyomas and Leiomyosarcomas, a Review of 2525 Cases. Springfield IL: Charles C. Thomas, 1962.

Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994.

Wangensteen OH, Wangensteen SD. The Rise of Surgery. Minneapolis: University of Minnesota Press, 1978.

Warren R. Surgery. Philadelphia: WB Saunders, 1963.


Normal Development

During herniation of the intestines into the umbilical cord (Fig. 18-1), a slight local enlargement of the portion posterior to the superior mesenteric artery marks the site of the future cecum. Growth and differentiation of this postarterial limb lags behind that of the proximal prearterial limb when the intestines return to the abdomen.

Fig. 18-1.

Diagram of the growth of the embryonic midgut into the umbilical cord. Fifth week of gestation. The distal limb shows a swelling at the site of the cecum. The three great arteries supplying the stomach and intestines are remnants of the earlier vitelline arteries. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Goblet cells and epithelial cells with a striate border may be found in the colon by the eleventh week. During the third month, villi and glands appear. The villi reach their maximum development in the fourth month and gradually shorten and disappear with the enlargement of the colon in the seventh and eighth months.

Langman and Rowland3 reported that the estimated total number of lymphoid follicles in the large intestine is between 12,761 and 18,432. The average follicular density is 18.4 per cm2 in the cecum, 15.0 per cm2 in the colon, and 25.4 per cm2 in the rectum. Previously reported numbers were very low; perhaps these new numbers will remind the physician about the diagnosis of lymphoid hyperplasia.

The circular layer of muscularis externa appears caudally in the ninth week and spreads cranially. Ganglion cells of the myenteric plexus of Auerbach reach the colon in the seventh week, and innervation appears to be complete by the twelfth week.4 The first longitudinal muscle fibers are present at the anal canal in the tenth week. Above the sigmoid colon, the longitudinal fibers extend cranially only along the mesenteric border of the colon, reaching the cecum in the eleventh week. By the fourth month, the entire colon is covered, but growth of the muscle coat does not keep up with increasing colon diameter. By the fourth month, the longitudinal muscle coat becomes separated into three bands, the taeniae coli. Meconium gradually fills the colon and the lower ileum until birth.


The colon is produced by both the midgut and the hindgut. The midgut is responsible for the genesis of the cecum, the ascending colon, and the proximal ⅔ of the transverse colon. The hindgut is responsible for the remainder of the colon, the rectum, and the proximal part of the anus. To be more specific, the distal ⅓ of the transverse colon, the descending colon, the sigmoid colon, the rectum and the proximal part of the anal canal develop from the hindgut.

The distal part of the surgical anal canal is not related embryologically to the hindgut. It most likely originates from the anal pit, which is of ectodermal origin. As O’Rahilly and Müller5 wrote, “the anal canal is probably derived from the cloaca.” To be more anatomically correct but still speculative about the embryology of the anorectum, we present the following information by Rowe et al.6 about the cloacal region.

In the fifth week, the embryonic cloaca is an endodermal sac receiving the hindgut dorsally and the allantoic stalk ventrally. The cloaca (Fig. 18-2A, B) is separated from the outside by a thin cloacal membrane (proctodeum), which occupies the embryo’s ventral surface between the tail and the body stalk. During the sixth week, a septum of mesoderm divides the cloaca into a ventral urogenital sinus and a dorsal rectum (Fig. 18-2C) . This mesodermic septum fuses with the cloacal membrane in the seventh week to form the perineal body. The cloacal membrane is divided into a larger, ventral urogenital membrane and a smaller, dorsal anal membrane. Externally, the anal membrane becomes slightly depressed, forming the anal dimple.

By the eighth week, the anal membrane ruptures, leaving no trace of itself (Fig. 18-2D). The pectinate line in the adult is often considered to be at the level of the anal membrane, but little evidence exists to either support or contradict this view. Whatever the exact line of demarcation, the rectum and the upper anal canal are endodermal and are supplied by the inferior mesenteric artery, while the lower anal canal is ectodermal and is supplied by branches of the internal iliac artery.

On either side of the anal membrane, the somatic mesoderm forms a pair of anal tubercles. These tubercles fuse dorsally into a horseshoe-shaped structure. By the tenth week, the ventral tips of the horseshoe fuse with the perineal body. Striated muscle in this horseshoe-shaped structure will later become the superficial portion of the external anal sphincter. The anal sphincter will form at the normal location even if the rectum should end blindly or should open at another site.

Fig. 18-2.

Diagram of stages in development of the anus and rectum from the fifth to tenth weeks of gestation. A, Closing plate (proctodeum separates the cloaca from the outside). Urorectal septum (arrow) grows downward to divide the cloaca. B, Cloaca almost separated into dorsal rectum and ventral urogenital sinus. Tailgut is vanishing. C, Fusion of urorectal septum with closing plate to form the perineal body. D, Closing plates rupture. E, Division into rectum and urogenital sinus by the perineal body is complete. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

The superior mesenteric artery (SMA) and the inferior mesenteric artery (IMA) provide the blood supply of the entire colon. In the surgical anal canal, the branches of the internal pudendal artery participate.

Congenital Anomalies of the Colon

Stenoses and Atresias

Stenoses and atresias occur in the same pattern as those of the small intestine (see “Stenoses and Atresias” in the small intestine chapter). They are less common in the large intestine, with an incidence ranging from 4.6 percent7 to 11.7 percent8 of all intestinal atresias. More type I (diaphragmatic) atresias occur in the ascending and sigmoid colons, and more type III (complete segmental) atresias occur in the transverse colon.9 Treatment is the same as that for atresias of the small intestine.

Dalla Vecchia et al.10 reported 277 cases of intestinal atresia and stenosis, detailing the treatment and results. The obstruction was duodenal in 138 (50%) [79 (57%) female, 59 (43%) male], jejunoileal in 128 (46%) [61 (48%) female, 67 (52%) male], and colonic in 21 (8%) [8 (38%) female, 13 (62%) male]. Patients with colon atresia were managed with initial ostomy and delayed anastomosis in 18 of 21 patients (86%) and resection with primary anastomosis in 3 (14%).

We quote from Lambrecht and Kluth11 on hereditary multiple atresias of the gastrointestinal tract:

Hereditary multiple atresias have several unique features: (1) the abdominal x-ray shows signs of gastric or duodenal atresia combined with typical large rounded or oval homogeneous calcifications in the abdominal cavity, (2) intraoperatively widespread atresias (exclusively type I and II) extending mostly from the stomach to rectum are found, (3) cystic dilatation of the bile ducts can be present in cases with both complete pyloric and duodenal or proximal jejunal atresia, (4) the pathogenesis is still speculative; a combined immunodeficiency should be excluded, and (5) a fatal outcome is the rule.

Congenital Aganglionic Megacolon (Hirschsprung’s Disease)

Aganglionic megacolon is the result of an absence of ganglion cells in a distal segment of colon. Neurenteric ganglion cells normally originate in the neural crest, enter the cranial end of the esophagus, and then follow vagus nerve fibers caudally until the entire gut is innervated. Why the migrating cells sometimes stop short of the rectum is unknown.

As seen in Fig. 18-3, only 4%12 of the aganglionic segments of the colon are found proximal to the splenic flexure. It is interesting to note that the neural crest cells forming these intestinal ganglia follow the vagus with a low failure rate to the end of that nerve’s distribution, i.e., approximately up to the splenic flexure. The remainder of the colon and rectum receive their parasympathetic innervation via the second, third, and fourth sacral nerves which have a more or less diffuse pathway to the colon themselves. Perhaps they do not serve as effectively as a “guide” or “transporter” of the neural crest cells for their migration into the hindgut derivative. Postganglionic fibers from proximal normal ganglia, as well as preganglionic parasympathetic fibers, are usually present. The aganglionic segment usually extends into the sigmoid colon, but the whole of the large intestine and even part of the small intestine can be affected (Fig. 18-3).13

Fig. 18-3.

Relative frequency of aganglionic segments by length of segment affected. Most cases involve the rectum and sigmoid colon only. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

The greatly dilated proximal segment is normal; the narrowed distal segment is without ganglia (Fig. 18-4). The line of resection must be within the area in which ganglion cells are present. Because aganglionosis (Fig. 18-5) is not the only cause of megacolon, a biopsy is necessary to demonstrate the absence of ganglion cells in the narrowed segment and their presence in the dilated segment.

Fig. 18-4.

Gross appearance and biopsy findings in aganglionic megacolon. On gross examination the normal colon is thought to end at A, but biopsy findings reveal that the aganglionic segment extends to B. Resection must be through the dilated proximal portion in which ganglion cells are present. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Fig. 18-5.

Possible causes of megacolon which must be considered in diagnosis. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

Remember to differentiate between congenital megacolon and massive fecal impaction. Because massive fecal impaction may lead to megarectum causing abdominal compartment syndrome and colorectal obstruction, perforation, or necrosis, Lohlun et al.14 recommend prompt manual disimpaction or appropriate operative treatment.

Wulkan and Georgeson15 recommend primary laparoscopic endorectal pull-through surgery as a safe and effective procedure for Hirschsprung’s disease in infants and children.

Colonic Malposition

There are several types of malposition. In Chilaiditi syndrome, the hepatic flexure is situated between the liver and the diaphragm due to anomalies of the hepatic ligaments. We quote from Balthazar:16

Positional anomalies of the colon may be explained by an arrest in the normal development of the distal midgut. Aberrations involving the incipient stages of rotation lead to severe malpositions, while those involving the latter stages to milder forms. . . There is a high incidence of associated failure of fixation resulting in mobile colons that can be demonstrated radiographically. In addition, the great majority of colonic malrotations demonstrate rotational abnormalities involving the proximal intestinal tract. Their clinical implication is related to the presence of other incidental congenital anomalies or to complications derived from faulty mesenteric fixations such as peritoneal bands, adhesions, kinking, or intestinal volvulus.

DePrima et al.17 reported reversed intestinal rotation due to abnormal rotation and fixation.

Retropsoas positioned bowel (colonic positioning posterior or posterolateral to the psoas muscle at a level below the lower kidney pole) may occur in the ascending or descending colon. Prassopoulos et al.18 advised that this condition be considered when performing percutaneous diskectomy or other interventional procedures in the posterior retroperitoneum.

Congenital Short Colon

Congenital short colon is total or partial replacement of the colon by a pouch, as well as associated anorectal malformation and colourinary fistula. A case of congenital short colon with imperforate anus was reported by Herman et al.19 An excellent paper about congenital short colon by Wakhlu et al.20 advised that the initial procedure be a window colostomy, followed by pouch excision/ coloplasty and pull-through by a combined abdominal and posterior sagittal approach when the baby is 6 months old.

Congenital Anomalies of the Anorectum

Multiple classifications of anorectal anomalies exist; none is perfect. We present a highly diagrammatic review of these anomalies with the hope that the student of embryology will be able to visualize these enigmatic malformations topographicoanatomically (Table 18-2 and Figs. 18-6, 18-7, 18-8, 18-9, 18-10, and 18-11).

Table 18-2. Anatomic Classification of Anorectal Malformations

Female Male
High High
  Anorectal agenesis   Anorectal agenesis
    With rectovaginal fistula     With rectoprostatic urethral fistulaa 
    Without fistula     Without fistula
  Rectal atresia   Rectal atresia
Intermediate Intermediate
  Rectovestibular fistula   Rectobulbar urethral fistula
  Rectovaginal fistula   Anal agenesis without fistula
  Anal agenesis without fistula  
Low Low
  Anovestibular fistulaa 
  Anocutaneous fistulaa 
  Anocutaneous fistulaa,b 
  Anal stenosisa,c 
  Anal stenosisc 
Cloacal malformationsd 
Rare malformations Rare malformations

aRelatively common lesion.

bIncludes fistulae occurring at the posterior junction of the labia minora often called “fourchette fistulae” or “vulvar fistulae.”

cPreviously called “covered anus.”

dPreviously called “rectocloacal fistulae.” Entry of the rectal fistula into the cloaca may be high or intermediate, depending on the length of the cloacal canal.

Source: Skandalakis JE, Gray SW (eds). Embryology for Surgeons, 2nd Ed. Baltimore: Williams & Wilkins, 1994; with permission.

Fig. 18-6.

Classification of anorectal malformations. (Modified from Raffensperger JG (ed). Swenson’s Pediatric Surgery (5th ed). Norwalk CT: Appleton & Lange, 1990; with permission. Prepared by Kascot Media, Inc., for the Department of Surgery, Children’s Memorial Hospital, Chicago IL.)

Fig. 18-7.

Types of anal (“low”) defects. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

Fig. 18-8.

Types of anorectal (“high”) defects. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

Fig. 18-9.

High rectal atresia. These forms do not arise from abnormal partition of the cloaca; they are related to other intestinal atresias. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

Fig. 18-10.

Bizarre combinations of developmental arrests, atresias, and fistulas. All are rare. A, Two fistulae. B, Persistent cloaca. C, D, and E, Bizarre types of anorectal atresia with fistulae. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

Fig. 18-11.

Embryogenesis of anal and anorectal defects. A, Anal stenosis resulting from disproportionately small anal portion of the cloacal membrane. B, Membranous atresia from persistent anal closing plate. C, Covered anus. The perineal body has not fused with the persistent cloacal plate, and a perineal fistula is present. D and E, Anorectal agenesis with and without arrested descent of the urorectal septum. F, Anal agenesis with failure of midline fusion of the folds forming the urorectal septum, leaving two fistulous openings. G, Anal agenesis with rectovaginal fistula. (Modified from Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994; with permission.)

While most anorectal malformations are diagnosed at birth, a significant number of mild lesions may not be recognized until later. Kim et al.21 recommend that older infants and children with cardiac, genitourinary, or VACTERL (vertebral, anal, cardiac, tracheosophageal, esophageal, renal, limb) anomalies who present with constipation be evaluated for low anorectal malformations.

Connaughton et al.22 presented a case of rectal duplication cyst with a large perineal hernia which presented as recurrent perineal abscesses.

Imperforate Anus and Related Anomalies

Immediate relief of the acute colonic obstruction is provided by colostomy, after which further repair can be planned. Remember that the first operation has the greatest chance for a successful outcome. The best procedure may be less than perfectly successful if it follows an earlier, inadequate attempt at repair.

The procedures following colostomy vary with the specific anatomy of the defect to be treated. They are outside the field of general surgery.

Another anomaly is the posteriorly situated retroperitoneal colon (see Descending Colon, Surgical Considerations).

Susuki et al.23 presented two cases of nonrotation in adults. The authors of this chapter are in agreement with Susuki et al. that the anomaly is more common in infants, but we have seen several cases in adults in the lab of the Emory Medical School as well as on barium enemas with x-rays at Piedmont Hospital.

Idiopathic Anal Incontinence

Peveretos et al.24 reported 3 cases of idiopathic anal incontinence. They supported the theory that the etiology is secondary to degeneration of the nerves supplying the pelvic floor muscles resulting in partial or total disappearance of the double right angle between the anal canal and the rectum which is essential for anal continence.

We quote from Skandalakis and Gray13 on cloacal exstrophy, the most severe of the ventral wall defects, as well as the most rare:

Superficially, the anomaly resembles exstrophy of the bladder, but the defect is larger…The lateral portion of the exposed mucosa represents the posterior wall of the bladder, but the central portion is intestinal epithelium. Superiorly, just below the umbilicus, the ileum opens to the surface and usually is prolapsed. On the exposed intestinal mucosa, one or occasionally two vermiform appendices open. At the inferior end of the mucosal surface, a segment of colon, which usually ends blindly, opens. The ureters open low on the lateral bladder mucosa, and in males there may be two penes or hemipenes. Females may have müllerian duct orifices on the bladder mucosa, or two vaginae may end blindly. Exomphalos may extend the defect cranially, and the pubic bones are separated, as in exstrophy of the bladder. Spina bifida, myelomeningocele, and a single umbilical artery are common…

Exstrophy of the cloaca arises from the failure of secondary mesoderm from the primitive streak to cover the infraumbilical wall. It differs from exstrophy of the bladder in that midline rupture has occurred earlier (about the fifth week), before the fusion of the genital tubercles (hence the double penis) and before the descent of the urorectal septum, which separates the cloaca into bladder and rectum. As the individual cloaca is exposed, its central portion is the posterior wall of the gut, while its lateral portions receive the ureters and differentiate into bladder mucosa.

Retrorectal Lesions

According to Gordon,25 retrorectal lesions may be congenital, inflammatory, neurogenic, osseous, or miscellaneous. More than half of all presacral lesions are congenital, and approximately two-thirds of these are of developmental origin. Table 18-3 describes the formation of these benign or malignant developmental cysts from an embryologic and histologic point of view.

Table 18-3. Germ Layer Origin of Developmental Cysts

  Epidermoid Dermoid Enterogenous Teratomatous
Tissue of origin  Ectoderm Ectoderm Endoderm All three layers
Histologic characteristics  Stratified squamous Stratified squamous with skin appendages (sweat glands, sebaceous glands, hair follicles) Columnar or cuboidal lining; may have secretory function Varying degrees of differentiation between cysts and cell layers of single cyst
General state Benign* Benign* Benign Benign or malignant

*Malignant variants are rare.

Source: Goldberg SM, Gordon PH, Nivatvongs S. Essentials of Anorectal Surgery. Philadelphia: JB Lippincott, 1980; with permission.

Surgical Anatomy of the Colon

Topographic Anatomy and Relations

The large intestine extends from the terminal ileum to the anus. To be more embryologically and anatomically correct, it extends to the pectinate (dentate) line, in other words, to the proximal 2 cm of the anal canal. The classic divisions of the colon are the cecum, the colon proper, the rectum, and the anal canal. The first 6 cm of the large intestine just below the ileocecal valve, the ascending colon, and the hepatic flexure form a surgical unit, the right colon (right colectomy). The distal transverse colon, splenic flexure, and descending and sigmoid colons constitute the left colon (left colectomy).

Length and Diameter of the Large Intestine

Textbooks of anatomy offer no agreement about the length of the segments of the large intestine. We have used the lengths and diameters from several studies to present averages for Figure 18-12. Estimates of the length of the large bowel average about 1.3-1.8 m. According to Gray’s Anatomy (37th ed.),26 the length from the end of the distal ileum to the anus is about 1.5 m. Goligher27 estimated the length of the colon to be 4 ½ ft (1.25 m).

Fig. 18-12.

Average lengths and diameters of the segments of the large intestine (based on an average of estimated lengths presented by studies of many researchers.)

Saunders et al.28 reported intraoperative measurements of colonic anatomy in 118 patients, reporting a mean total colonic length of 114. 1 cm (range 68-159 cm). A free sigmoid loop was not present in 20 patients (17%) because of adhesions. Ten patients (8%) had a descending mesocolon of 10 cm or more, and 11 patients (9%) had an ascending mesocolon of 10 cm or more. Twenty-four patients (20%) had mobile splenic flexures. The mid-transverse colon reached the symphysis pubis in 34 patients (29%).

The caliber of the large bowel is greater close to the cecum; it gradually gets smaller toward the rectum, then dilates again at the rectal ampulla just above the surgical anal canal. A sigmoid colon loop is occasionally as wide as a loop of terminal ileum.

Sadahiro et al.,29 in a study of Japanese patients, reported that the transverse colon was the largest in length and surface area of the 6 segments of large intestine. They also reported that the length of the entire colon tended to become longer with age. Is this, perhaps, one of the etiologic factors for the phenomenon of geriatric constipation? Sadahiro and colleagues also found the diameters of the descending colon, sigmoid colon, and rectum to be larger in males. However, the length of the entire intestine was shorter, and the surface area was smaller, in males than in females.

A discussion of the colonic wall is presented in this chapter in the section on histology.

Vascular Supply

The large intestine is supplied by the superior and inferior mesenteric arteries and branches of the internal iliac (hypogastric) artery; it is drained by the superior and inferior mesenteric veins and tributaries to the internal iliac vein.

Arterial Supply

Superior Mesenteric Artery

The cecum and the ascending colon receive blood from two arterial branches of the superior mesenteric artery: the ileocolic and right colic arteries (Figs. 18-13, 18-14). These arteries form arcades from which vasa recta pass to the medial colonic wall. As the vasa recta reach the surface of the colon, they divide into short and long branches. The short branches serve the medial or mesenteric side of the colon; the long branches serve the lateral and antimesenteric side. The long branches send twigs into the epiploic appendages (Fig. 18-15).

Fig. 18-13.

Schema of the arterial blood supply to the large intestine. There are many variations of this basic pattern. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Fig. 18-14.

The collaterals of the superior mesenteric artery. (Modified from Van Damme JP, Bonte J. Vascular Anatomy in Abdominal Surgery. New York: Thieme Verlag, 1990; with permission.)

Fig. 18-15.

Diagram of the transverse colon showing long and short branches of the vasa recta. On the left is a normal epiploic appendage; on the right, a diverticulum extending into an epiploic appendage. Inset: Effect of too much traction on an epiploic appendage resulting in injury to one of the long branches of vasa recta followed by antimesenteric ischemia. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The transverse colon is similarly supplied by the middle colic artery from the superior mesenteric artery. Steward and Rankin30 found the splenic flexure supplied by the middle colic artery in 37 percent of their specimens. In the remainder, the flexure and the left portion of the transverse colon were supplied by the left colic artery, a branch of the inferior mesenteric artery.

Inferior Mesenteric Artery

The inferior mesenteric artery (Fig. 18-16) arises from the aorta opposite the lower portion of the 3rd lumbar vertebra, at or near the lower margin of the transverse segment of the duodenum. The origin tends to become lower with age.31 The length of the artery prior to its first branch varies from 1.5 cm to 9.0 cm.32

Fig. 18-16.

The inferior mesenteric artery and its branches; origin distance in millimeters. If the inferior mesenteric artery is divided at “a”, above the last full anastomosis, collateral circulation toward the rectum is still possible. Division at “b” would interrupt the collateral circulation. (Modified from Van Damme JP, Bonte J. Vascular Anatomy in Abdominal Surgery. New York: Thieme Verlag, 1990; with permission.)

The branches of the inferior mesenteric artery (Figs. 18-13 and 18-17C, D, E) are the left colic artery, with its ascending and descending branches for the descending colon, 1 to 9 sigmoid arteries for the sigmoid colon, and the superior rectal (hemorrhoidal) artery for the rectum. An accessory middle colic artery is present in about 38 percent of subjects. The left colic artery may reach the splenic flexure (86 percent of cases) or may join the marginal artery short of the flexure (14 percent of cases).

Fig. 18-17.

Variations of the arteries to the right and left colon. A, Usual pattern to the right colon. B, The marginal artery is incomplete at “X.” C, Arteries to the left colon. There may be fewer sigmoid arteries than shown here. D and E, Two patterns of the sigmoid arteries. The “critical point” of Sudeck is marked with an “X.” RCA, Right colic artery. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

In the anatomy of the colon vasculature, the ascending branch of the left colic artery is the primary supplying vessel (96.91%).33 According to Furst et al.,34 an intact left colic artery, including its collaterals at the splenic flexure, will supply sufficient blood to the proximal ascending colon after central ligation of the middle and right colic artery. By including the ascending colon in their alternative colon interposition procedure, they were able to obtain a sufficient graft length without mobilization of the left flexure.

Is the splenic flexure vulnerable to ischemic injury due to a compromised blood flow, as Griffiths35 stated? Van Damme and Bonte,36 without supporting or denying the dogma of Griffiths, stated that the splenic flexure has 3 types of arcades: paracolic, anastomotic in the mesentery, and possibly a small intermesenteric arcade or a left accessory colic vessel close to the duodenojejunal flexure.

Dworkin and Allen-Mersh37 found that the “significant blood flow reduction after ligation of the inferior and distal mesenteric arteries supports the hypothesis that anastomotic leakage after restorative rectal excision may result from ischemia associated with inadequate blood flow in the marginal artery-dependent sigmoid colon. Improvement in inadequate intraoperative colonic perfusion from increased collateral circulation is unlikely to develop in the marginal-artery dependent colon during the first five postoperative days.”

Adachi et al.38 stated that poor bowel function after low anterior resection is associated with high ligation of the inferior mesenteric artery and injury to the pelvic autonomic nerve, and urged less aggressive surgery. Poor bowel function after sigmoid colectomy was correlated with length of the resected colon.

Remember the long and short vasa recta (Fig. 18-18). The long vasa recta branches bifurcate and anastomose at the antimesenteric border of the bowel after encircling it. The short ones, branches of the marginal artery, are responsible for the mesocolic two-thirds of the colonic circumference.

Fig. 18-18.

The terminal arterial supply to the colon and its relation to the taenia coli and appendices epiploicae.

Marginal Artery (of Drummond)

The marginal artery (Fig. 18-17A, B, C) is composed of a series of anastomosing arcades between branches of the ileocolic, right colic, middle colic, left colic, and sigmoidal arteries. These form a single, looping vessel. The marginal artery courses roughly parallel with the mesenteric border of the large intestine, from 1 cm to 8 cm from the intestinal wall. It may or may not terminate at the superior rectal artery (Fig. 18-17C).

The blood supply to the colon is adequate, but without much margin of safety. Anastomosis between the right colic and ileocolic arteries is absent in 5 percent of subjects (Fig. 18-17B).30

Griffiths35 described a point of circulation weakness at the splenic flexure. In 200 subjects, Michels and colleagues32 found good anastomosis of arteries in this area in 61 percent, poor anastomosis in 32 percent, and no anastomosis in 7 percent.

Haigh and Temple39 presented a case of a dilated marginal artery (Fig. 18-19B) providing collateral blood supply from the middle colic artery to the ileocolonic artery to compensate for the occlusion of the superior mesenteric artery secondary to chronic volvulus of the small intestine. The authors advise 1) complete knowledge of mesenteric collateral circulation; 2) relocation and repositioning of the small bowel, if necessary, prior to closing the abdomen; and 3) delineating the dilated marginal artery before performing a resection.

Fig. 18-19.

A, Normal mesenteric vasculature and collateral blood supply. B, Dilated marginal artery supplies collateral blood flow to small bowel in case of superior mesenteric artery occlusion.

Critical Point of Sudeck

The critical point of Sudeck is no longer considered to be as “critical” as was once thought. Its location is unimportant in present-day abdominal and abdominoperineal resections. We present it here only for historical reasons.

Sudeck40 described a point on the superior rectal artery at which ligation of the artery would not devascularize a long rectosigmoid stump. This point is just above the origin of the last sigmoid artery; its position varies with the number of such arteries (Fig. 18-17D, E). In about 50 percent of individuals, the marginal artery continues downward to join the superior rectal artery. Sudeck’s point would be just proximal to that junction.

The significance of Sudeck’s point depends on the surgical procedure to be performed. In 1907, Sudeck was interested in perineal excision of the rectum, the procedure of choice at the beginning of the century. Ligation of the superior rectal artery was necessary for mobilization of the rectum.

In some cases of left colectomy in which a long sigmoid stump is left, it may be difficult to bring the transverse colon down to the stump because of a short transverse mesocolon. This problem could be solved by ligation of the superior rectal artery above Sudeck’s point.

The concept of Sudeck’s critical point fails to recognize two other sources of blood to the rectum. One is the intramural network of arteries in the submucosal layer of the wall. The other is from the middle and inferior rectal arteries, especially the latter. Goligher41 stated: “Experience with sphincter-saving resections for carcinomas of the upper rectum and lower sigmoid shows that after division of the inferior mesenteric/superior hemorrhoidal trunk, the middle and inferior hemorrhoidal [rectal] arteries are capable of nourishing a distal rectal stump up to a point at least 8 to 10 cm above the peritoneal reflection.”

According to Michels et al.,32 in addition to the three pairs of rectal (hemorrhoidal) arteries, other sources of collateral blood supply to the rectum and sigmoid might include:


branches of the inferior vesical artery

arteries supplying the levator ani muscle

the middle sacral artery

the posterior retroperitoneal arterial plexus uniting the parietal and visceral circulation

The inferior rectal artery is responsible for the arterial blood supply of the distal 2 cm of the anal canal.

We strongly advise the reader to consult the excellent works of Van Damme1 and Bertelli et al.42

Venous Drainage

Veins of the colon (Fig. 18-20) follow the arteries. On the right (the cecum, ascending colon, and right transverse colon), the veins join to form the superior mesenteric vein. Veins of the hepatic flexure and the right portion of the transverse colon enter the gastroepiploic vein or the anterior superior pancreaticoduodenal vein. Voiglio et al.43 studied and defined the gastrocolic vein and reported two cases of avulsion secondary to abdominal trauma. It is present in 70% of patients, it is short (less than 25 mm), and its calibre ranges from 3 mm to 10 mm. It is located at the anterior surface of the head of the pancreas and beneath the root of the transverse mesocolon as a confluence between the right gastroepiploic and right upper colic veins. The above authors emphasize the surgical importance of this vein during surgery of the pancreas, in portal hypertension, and in abdominal trauma. Drainage from the left portion of the transverse colon enters the superior mesenteric vein. The superior rectal vein drains the descending and sigmoid colons; it passes upward to form the inferior mesenteric vein.

Fig. 18-20.

Venous drainage of colon and rectum. Dark blue, systemic venous drainage. Light blue, portal venous drainage. (Modified from Gordon PH, Nivatvongs S (eds). Principles and Practice of Surgery for the Colon, Rectum, and Anus. St. Louis, MO: Quality Medical Publishing, 1992; with permission.)

The rectum is drained by the superior rectal veins (Fig. 18-21), which enter the inferior mesenteric vein. This drainage is to the portal system. The middle and inferior rectal veins enter the internal iliac vein and thus drain into the systemic circulation.

Fig. 18-21.

Detailed diagram of the venous drainage of the rectum and anus. The superior rectal vein drains to the portal system, and the middle and inferior rectal veins drain to the systemic veins. The venous plexus between the veins forms a potential portacaval shunt. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)



The inferior rectal vein is mainly responsible for the venous return of the distal 2 cm of the anal canal.

Anastomoses occur between the superior rectal vein (portal) and the middle and inferior rectal veins (systemic). These constitute a potential portosystemic shunt.

Because lower intestinal venous malformations may cause significant chronic and acute gastric hemorrhage, Fishman et al.44 recommend colectomy with mucosectomy and endorectal pull-through before the development of large transfusion requirements.

Lymphatic Drainage

Lymph nodes of the large intestine have been divided into four groups (Fig. 18-22): epicolic (under the serosa of the wall of the intestine); paracolic (on the marginal artery); intermediate (along the large arteries [superior and inferior mesenteric arteries]); and principal (at the root of the superior and inferior mesenteric arteries). This last group includes mesenteric root nodes (which also receive lymph from the small intestine), aortic nodes, and left lumbar nodes. The number of nodes of the large intestine is shown in Table 18-4.

Table 18-4. Number of Lymph Nodes in the Mesentery of the Large Intestine by Regions

Nodes Average number
Ileocolic 29
Right colic 11.1
Midcolic 22.4
Left colic 25.2
Sigmoid and rectal 32.8

Source: Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.

Fig. 18-22.

The lymphatics of the large intestine follow the arteries and drain to the principal nodes at the root of the mesentery. The path is by way of epicolic, paracolic, and mesocolic lymph nodes. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Although lymph flow follows the arteries, there are cross-connections at the level of the arcades that are parallel with the intestine. In addition, communication between the lymphatics of the transverse colon and those of the stomach, and between lymphatics of the ascending and descending colon and the body wall have been described in the medical literature.

Wide resection of the colon should include the entire segment supplied by a major artery. This will also remove most, but not all, the lymphatic drainage of the segment (Figs. 18-23, 18-24).

Fig. 18-23.

Resection of the large intestine should include the entire area served by a major artery as well as the lesion itself. Most of the lymphatic drainage will be included. Areas of resection (stippled) for lesions in various segments of the large bowel are shown in A-F. An arrow indicates the site of the lesion. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Fig. 18-24.

Lymphatic drainage of the sigmoid colon, rectum, and anus. Above the pectinate line, drainage is to inferior mesenteric nodes. Below the line, drainage is to the inguinal nodes. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Yada et al.45 analyzed the vascular anatomy and lymph node metastases in carcinoma of the colon:

Because the ileocecal artery always arises from the superior mesenteric artery and lymph node metastases of cecum cancer were limited to nodes along the ileocolic artery, cecum cancer can be cured by ileocecal resection. The right colic artery has various origins, and ascending colon cancer shows various patterns of lymph node metastases. Therefore a right hemicolectomy should be performed for ascending colon cancer. The middle colic artery forks into right and left branches, and each branch has different branching variations. If the right colic and middle colic arteries have a common trunk, a right hemicolectomy should be performed for transverse colon cancer on the right side. If the left branch of the middle colic artery has an independent replaced origin, lymph node dissection should be modified according to the variant origin. If the left colic artery and the first sigmoidal artery have a common trunk, the lymph nodes along the common trunk should be removed for sigmoid colon cancer and for descending colon cancer. Of the patients with sigmoid colon cancer, 6.3% also had lymph node metastases along the superior rectal artery. Given that the lymph nodes along the superior rectal artery are skeletonized, sigmoid colon cancer can be also cured by partial sigmoidectomy.



The lymphatic vessels of the distal 2 cm of the anal canal drain to the inguinal nodes.


Intrinsic Innervation

From the esophagus to the anus, the digestive tract is supplied with two intramural nerve networks. The myenteric plexus (Auerbach) (Fig. 18-25) controls motility, and the submucosal plexus (Meissner) controls secretion. Vagal and sympathetic fibers synapse with these intramural ganglion cells. Innervation of the gastrointestinal system is now considered to have its own intrinsic set of nerves known as the intramural plexus or the intestinal enteric nervous system.46 The number of neurons this system comprises is estimated to be 100 million, equal to that of the whole neuronal population of the spinal cord. Postganglionic sympathetic and parasympathetic neurons synapse on neurons of both myenteric and submucosal plexuses. Postganglionic sympathetic fibers also synapse directly on the epithelial cells.

Fig. 18-25.

Neural control of the gut wall, showing 1) the myenteric and submucosal plexuses; 2) extrinsic control of these plexuses by the sympathetic and parasympathetic nervous systems; and 3) sensory fibers passing from the luminal epithelium and gut wall to the enteric plexuses and from there to the prevertebral ganglia, spinal cord, and brain stem. (Modified from Guyton AC, Hall JE. Textbook of Medical Physiology, 9th ed. Philadelphia: WB Saunders, 1996; with permission.)

The neurons of the two plexuses, generally regarded as postganglionic parasympathetic fibers, form a reticulum of neurons with widespread synapses; and within these plexuses there are also interneurons. Some sensory neurons with receptors in the epithelium are now known to collateralize into both plexuses and then continue on so that some follow parasympathetic fibers back to the brain or sacral segments of the spinal cord and sympathetic fibers back to the T1-L2,3 spinal cord segments. However, in addition to those sensory fibers, others have been identified which synapse in the collateral (prevertebral) (sympathetic) ganglia.

Sympathetic Innervation

The sympathetic supply to the right colon originates from the lower six thoracic segments of the spinal cord. Preganglionic fibers pass through the sympathetic chain ganglia, then pass as thoracic splanchnic nerves to synapse in the celiac, aortic, and superior mesenteric plexuses. From the plexuses, postganglionic fibers pass with the arteries in the mesentery to the small intestine and the right colon.

On the left, preganglionic fibers arise from the first two (or three) lumbar segments of the cord, then travel as lumbar splanchnic nerves to the aortic plexus and the inferior mesenteric plexus. From ganglia in this diffuse plexus, postganglionic fibers follow branches of the inferior mesenteric artery to the left colon and the upper rectum.

Parasympathetic Innervation

Vagal fibers from the posterior trunk pass as the celiac division to and through the celiac ganglion without synapse. From the ganglion, preganglionic fibers pass on the superior mesenteric artery to the small intestine and the right colon, where they synapse with ganglion cells of the intramural plexuses.

The left colon receives parasympathetic fibers from pelvic splanchnic nerves, which arise from the 2nd, 3rd, and 4th sacral nerves. These fibers follow the course of the presacral nerve to reach the inferior mesenteric plexus. From this plexus, the preganglionic fibers follow the branches of the inferior mesenteric artery to the left colon and the upper rectum.



The proximal 2 cm of the anal canal is innervated by visceral autonomic fibers.

The distal 2 cm of the anal canal is innervated by somatic rectal nerves.

Detailed Anatomy of Colonic Segments

Cecum and Ileocecal Valve

The cecum lies in the right iliac fossa; in about 60 percent of living, erect individuals, it lies partly in the true pelvis. Anson and McVay47 distinguished six types and several subtypes of peritoneal reflections of the cecum. In their series, almost the entire posterior surface of the cecum was attached to the posterior abdominal wall in 19.6 percent of cases. At the other extreme, the cecum was wholly unattached in 24 percent of cases. Among the latter group are cases of true “mobile cecum,” in which the cecum and the lower part of the ascending colon are unattached. Pavlov and Pétrov48 stated that the cecum was mobile more often in females than in males by a factor of 20 percent.

Because the cecum is located at the right iliac fossa, it is related to the right iliacus muscle posteriorly. In most cases, it is covered 90 to 100 percent by peritoneum. The cecum usually is not fixed by the posterior peritoneum which covers the right iliac fossa. The relations and variations of the cecum at the iliac fossa are shown in Figures 18-26 and 18-27.

Fig. 18-26.

Cross sections at the level of the cecum and the iliac fossa showing variations in the degree of fusion of the cecum to the peritoneum. A, The cecum is unattached at this level and quite free to move about. B, The cecum is held by a narrow mesenteric fold, permitting moderate mobility. C, Cecum with a retrocecal appendix is bound down to the iliac peritoneum over an extensive area. (Modified from McVay CB. Anson & McVay Surgical Anatomy, 6th ed. Philadelphia: WB Saunders, 1984; with permission.)

Fig. 18-27.

Attachment of the cecum, ascending colon, and ileum to the dorsal body wall, shown by type and percentage of occurrence (areas of fixation shown in black), based on 300 laboratory specimens. Types I and I-a, Almost complete dorsal fixation of the cecum. Type I-b, Examples of retrocecal recess (spaces occurring as offsets of the cecal fossa) arranged in order of descending fixation. Types II and II-a, Fixation is chiefly medial (in II-a, with continuing attachment of the terminal ileum). Types III, III-a, and IV, Cecum free of dorsal attachment, in varying degree. In several cases, the proximal segment of the ascending colon also found without posterior parietal attachment. Type V, Nonfixation of the cecum, with continuing mobility of the greater part of the ascending colon. Type VI, Complete absence of dorsal attachment. (Modified from McVay CB. Anson & McVay Surgical Anatomy, 6th ed. Philadelphia: WB Saunders, 1984; with permission.)

A fold of peritoneum from the mesentery of the terminal ileum may cross the ileum to attach to the lower colon and the cecum. This is the superior ileocecal fold (Fig. 18-28); the anterior cecal artery lies within it. This fold, the mesentery, and the ileum may form a superior ileocecal fossa. Inferior to the terminal ileum, an inferior ileocecal fold may lie anterior to the mesentery of the appendix. Between them is the inferior ileocecal fossa (paracecal herniation). Both the superior and inferior ileocecal folds are inconstant, and the associated fossae can be shallow or absent. Some types of cecal attachment to the body wall may form a retrocecal fossa. In 78 cadavers dissected by Skandalakis,49 12 had a fixed terminal ileum and 1 had a common ileocecal mesentery.

Fig. 18-28.

Superior and inferior ileocecal folds forming fossae. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Four types of human cecum were postulated by Treves50 in a developmental model (he referred to them as the “first” through “fourth” types).


First (Infantile): During the neonatal period, the cecum grows to a conical formation. The appendix hangs from its apex. The 3 taeniae are present, beginning at the appendiceal base. This form persists in approximately 2 percent of the population.

Second (Childhood): The cecum becomes quadrate by forming saccules medial and lateral to the anterior taeniae. The appendix hangs from the area between the saccules and not from the apex. This form persists in about 3 percent of the population.

Third (Adult): A new apex is formed by rapid growth of the right lateral saccule. The old apex is pushed to the left lateral area toward the ileocecal valve. The appendix hangs between the anterior and posterolateral taeniae. This form occurs in roughly 90 percent of the population.

Fourth (Geriatric): There is atrophy of the left saccule and enlargement of the right saccule. The apex and the appendix are now close to the ileocecal valve. The fourth form, actually a progression of the third form, is found in approximately 4 percent of the population.

Pavlov and Pétrov,48 after studying the ceca of 126 subjects, classified them as follows: infantile type (which they named “infundibular”), 13%; adult type (which they called “ampullary”), 78%; “intermediate” type, 9%.

The cecum and ascending colon are related posteriorly to the following anatomic entities, which should be protected during mobilization of the right colon:


psoas major muscle

nerves (lateral femoral cutaneous, femoral, genitofemoral)

gonadal arteries and veins


Sappey51 attributes the discovery of the ileocecal valve to C. Varolius, based on the following lines which were published by Varolius in 1573:

Where the ileum joins the colon, there is a certain membrane which protrudes into the cavity of the latter. Of this membrane, which is the very end of the ileum extending to this junction, I who am the inventor, name it operculum of the ileum.

DiDio and Anderson52 used the term ileal pylorus, which is anatomically correct, rather than the term ileocecal valve. They stated that the ileum opens into the transitional zone between the cecum and the ascending colon, thereby asserting that the name “ileocecal valve” does not have any anatomic meaning. However, for the sake of convenience, we will continue to call this anatomic entity the “ileocecal valve.”

The cadaveric valve is formed by two parallel transverse folds (labia, lips, or flaps) which project within the colon and form a transverse slit of 1-1.5 cm. At each end, the folds are united and form the frenulum. For all practical purposes, the protrusion of the small bowel into the colon is formed by the circular muscle of the terminal ileum which is covered by the mucosa.

In living subjects, the ileocecal valve has a papillary shape with a conoid ileal projection or intrusion into the cecum.

Kumar and Phillips53 stated that the superior and the inferior ileocecal ligaments are responsible for the competence of the ileocecal valve.

The purpose of the ileocecal valve is not definitely known. Bogers and Van Marck54 stated that “the ileocecal junction remains a controversial region of the gut.” They wrote in a review in 1993, “Based on the available data from the literature, evidence is accumulating for a sphincteric function.”

The definition of the ileocecal valve differs with various specialties. Anatomists define it as a bilabial type with a horizontal slit, formed by the upper and lower lips which are united medially and laterally to form the frenula. Endoscopists consider it a papillary type, which is like a uterine cervix, a round or conical projection with a starlike orifice, but without the frenula. To radiologists,55 the ileocecal valve appears as a round or oval protuberance, which usually arises from the posteromedial wall of the cecum. It is composed of superior and inferior lips which, at the corners of fusion, taper to form a part of the cecal wall.

Identification of the ileocecal region can be aided by a CT scan. In a study by Silverman et al.,56 the region was identified using CT scan in 18 of 25 patients (72 percent) without pathology.

Arterial Supply, Venous Drainage

The ileocolic artery (Fig. 18-29), which arises from the right side of the superior mesenteric artery, is the chief blood supply of the cecum. This artery divides into two branches before approaching the cecal wall. The colic branch anastomoses with the right colic artery. The ileal branch anastomoses with the terminal intestinal branch of the superior mesenteric artery. Close to its bifurcation, the ileocolic artery gives off two more branches: the anterior and the posterior cecal arteries.

Fig. 18-29.

The branches of the ileocolic artery. The ileocolic artery can be traced as a stem until it ends in two cecal branches. The other branches are direct collaterals of the main stem. (Modified from Van Damme JP, Bonte J. Vascular Anatomy in Abdominal Surgery. New York: Thieme Verlag, 1990; with permission.)

The ileocolic vein is a tributary of the superior mesenteric vein.


The lymphatic vessels of the cecum (Figs. 18-30, 18-31) drain to lymph nodes along the anterior network of the ileocolic arteries. There are two groups of lymph nodes. The ileocolic lymph nodes are found along the ileocolic artery. The cecal lymph nodes are located in the vicinity of the anterior and posterior cecal arteries.

Fig. 18-30.

The lymphatics of the ileocecal region, ventral view. (From Schaeffer JP (ed). Morris’ Human Anatomy (11th ed). New York: Blakiston, 1953; with permission.)

Fig. 18-31.

The lymphatics of the ileocecal region, dorsal view. (From Schaeffer JP (ed). Morris’ Human Anatomy (11th ed). New York: Blakiston, 1953; with permission.)


The sympathetic innervation of the cecum arises from the celiac and superior mesenteric ganglia. The parasympathetic innervation of the cecum comes from the vagus nerve.

Surgical Considerations


Volvulus of the cecum is an extremely rare phenomenon. The term “cecal volvulus” is frequently misused. The correct terminology is “volvulus of the right colon.” Cecal volvulus in a 2 month-old boy was reported by Khope and Rao.57 Laparoscopic cecopexy for intermittent cecal volvulus was reported by Shoop and Sackier.58 Frank et al.59 reported the first description of the CT diagnosis of cecal volvulus, emphasizing the significance of the “whirl sign.” Moore et al.60 reported a case with synchronous cecal and sigmoid volvulus, an extremely unusual phenomenon. Theuer and Cheadle61 reported a similar case.

The wall of the cecum is thin in comparison with the wall of the other colonic segments. The best and most secure part of the wall for cecotomy, cecorrhaphy, anastomosis with other viscera, and cecopexy is the taeniae (especially the anterior one, which is most approachable).

Ileocecal intussusception can be idiopathic or secondary to benign or malignant tumors of the terminal ileum and of the ileocecal area generally. Idiopathic intussusception is a disease of the neonatal period; it may be secondary to the hypertrophy of the patches of Peyer. VanderKolk et al.62 studied cecal-colic adult intussusception and recommended surgical reduction; colectomy was recommended only if the bowel is gangrenous. However, the authors of this chapter believe that cecopexy is a wise choice when the bowel wall is healthy.

Cecal diverticulitis is rare. It is always solitary, in contrast to that of the left colon, where the diverticula are multiple. Efforts must be made to save the ileocecal valve if possible when operating in the ileocecal area. Two such cases with acute diverticulitis among 7 cases of benign lesions of the right colon were reported by Lear et al.63 The following material is taken from that paper.

Many benign conditions occur in the right colon, and the differential diagnosis of these lesions is frequently difficult. This report concerns seven patients seen at Piedmont Hospital in the past six years, whose ultimate diagnoses illustrate several of the many benign conditions which may occur in the right colon. Patients with acute appendicitis have not been included in this report, nor have those with neoplasms which are frequently malignant such as carcinoid and polyps.

Astler et al.64 divided benign lesions of the right colon into four groups as follows:


1. Benign neoplasms (benign variants of carcinoids, hemangiomas, lymphangiomas, lipomas, fibromas, adenomas, etc.)

2. Inflammatory and parasitic conditions (regional enteritis, mucocele of appendix, ameboma, tuberculoma, actinomycosis, etc.)

3. Anatomic and congenital abnormalities (diverticula, hypertrophic mucosal folds, inverted appendiceal stump, etc.)

4. Physiologic defects (hypertrophied ileocecal valve, fecalith, prolapsed ileal mucosa, etc.)

The grouping is not exclusive, and a lesion may fall into more than one class. A benign neoplasm such as a lipoma may involve the ileocecal valve, and this may become ulcerated and produce a nonspecific inflammatory mass. Table 18-5 shows pre- and postoperative diagnoses.

Table 18-5. Benign Cecal Lesions

Patient No. Age Sex Preoperative Diagnosis Postoperative Diagnosis
1 51 F Probable carcinoma of cecum Lipoma of valve
2 49 F Appendiceal abscess Inflammation of cecum
3 38 F Probable carcinoma of cecum Non-specific inflammation
4 50 M Possible carcinoma of cecum Mucosal fold
5 47 F Possible carcinoma Diverticulum
6 46 F Volvulus Diverticulum
7 59 F Probable carcinoma of cecum Constriction ring

Source: Astler VB, Miller EB, Snyder RS, McIntyre CH, Lillie RH. Benign surgical lesions of the cecum. Arch Surg 1963;86:435; with permission.


The appendix, colon, anorectum, and surgical anal canal follow the pathologic destiny of other parts of the alimentary tract, developing cancers of epithelial and nonepithelial origin. Two studies by Hatch et al.65,66 of stromal tumors of the appendix, colon, anorectum, and anal canal (leiomyomas and leiomyosarcomas) are recommended to the interested student.

Isolated cecal infarction should be included in the differential diagnosis of acute right lower quadrant pain.67

Nelson68 stated that surgeons routinely divide the colorectum into anatomic subsites when treating colorectal cancer. He finds that the best classification is as follows:

1. Proximal colon: cecum, transverse colon, descending colon

2. Distal colon: sigmoid, rectosigmoid, rectum (above anal canal)

3. Anal canal

Justification for these three divisions is embryologic (midgut-hindgut).


Poon and Chu69 reported the following:

Most inflammatory cecal masses are caused by benign pathology, and ileocecal resection is the procedure of choice. We do not recommend routine right hemicolectomy, as it requires conversion to a midline incision and is associated with a longer operation time, higher morbidity rate, and longer hospital stay. However, careful intraoperative assessment and, in particular, examination of the resected specimens is essential to exclude an underlying malignancy that would require right hemicolectomy.


The competence of the ileocecal valve depends on the presence and action of the sphincteric mechanism in this area, which is the result of the formation of the sphincter of the valve (the valve itself is a thickening of the circular muscle). In most cases, a retrograde state exists.

With total colonic obstruction, if the ileocecal valve is competent, the colon proximal to the obstruction is dilated. This type of valve –which does not permit the colonic contents to enter the ileum, but permits ileal contents to enter the colon– will produce a closed, looplike formation with increased tension. With an incompetent ileocecal valve, the closed-loop phenomenon does not exist: the small bowel receives the colonic contents and becomes distended, and nausea and vomiting follow.

Mann et al. presented a case of right thigh abscess secondary to perforated adenocarcinoma of the cecum.70 The pathway of the cecal contents most likely followed the anatomy of the iliopsoas fascia which may be continuous with the fascia lata of the thigh.71

Ascending Colon

Normally the ascending limb of the right colon is fused to the posterior body wall and covered anteriorly by the peritoneum (Fig. 18-32). There are variations of incomplete fusion, ranging from a deep lateral paracolic groove to the persistence of an entire ascending mesocolon. A mesocolon long enough to permit volvulus occurs in approximately 11 percent of cases.72 In cadavers, the ascending colon may be mobile in approximately 37 percent of cases.48 A mobile cecum, together with a mobile right colon, may be present; this is the justification for volvulus of the right colon.

Fig. 18-32.

Degrees of attachment of the right colon to the abdominal wall. A, Normal retroperitoneal location of the colon. B, Paracolic gutter. C, Mobile colon with mesentery. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The older literature50 reported a much higher incidence of right mesocolon from autopsy specimens. Symington73 observed that laxity of the parietal peritoneum in the unembalmed body will permit the colon to be pulled anteriorly to produce a pseudomesentery. This cannot be done in the embalmed cadaver. To what extent this colonic mobility can occur in the living patient is not known.

Where the mesocolon is present, the cecum and proximal ascending colon are unusually mobile. It is this condition that is termed mobile cecum (Fig. 18-33); it can result in volvulus of the cecum and the right colon. In a study of 87 cadavers,49 the cecum alone was mobile in 55. In 6 of these, enough of the right colon was mobile so that volvulus could have occurred.

Fig. 18-33.

Mobile cecum, distal ileum, and proximal right colon. This configuration is subject to volvulus. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Two conditions must be present for right colon volvulus to occur:74 (1) an abnormally mobile segment of colon, and (2) a fixed point around which the mobile segment can twist. The first condition is often present in many individuals (Table 18-6). The second can be provided by the normal attachments of the colon or by postoperative adhesions.

Table 18-6. Percent Frequency of Persistent Right and Left Mesocolon

Source Right Left
Adults (Treves, 1885) 26 36
Infants (Smith, 1911) 30.75 37.1
Adults (Harvey, 1918) 13
Adults (Hendrick, 1964) 11.2
Adults (Skandalakis, 1949) 36.9

Source: Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.

Rogers and Harford75 reported on five patients with mobile cecum syndrome. In all these patients, the cecum and ascending colon were not attached to the lateral peritoneum for 15-18 cm. According to these authors, cecopexy is the procedure of choice to reduce the colonic mobility.

In a study by Wright and Max76 of 12 patients with cecal volvulus, 2 who were treated by simple detorsion without fixation did not experience recurrence. Nevertheless, the authors of this chapter strongly advise fixation, with or without cecostomy.

Ballantyne et al.77 reported 137 cases of colonic volvulus between 1960 and 1980 at the Mayo Clinic: 52 percent were cecal; 3 percent were transverse; 2 percent were at the splenic flexure; 43 percent were sigmoid. Mortalities with cecal volvulus involving the cecum, transverse colon, and splenic flexure were 17 percent; with sigmoid volvulus, 11 percent.

Rabinovici et al.78 reviewed 568 cases of cecal volvulus. The mean age of the patients was 53 years. The female/ male ratio was 1.4:1. There was gangrenous cecum in 20 percent of the cases. These authors emphasized that cecostomy had more complications in comparison with other techniques. They advised that with a necrotic cecum, resection should be done; with a viable cecum, detorsion and cecopexy is the procedure of choice. They also advised that cecostomy should be abandoned. Other authors disagree, however. For cecal volvulus with a viable colon, Jones and Fazio79 recommended detorsion, cecopexy, and tube cecostomy as a combined procedure.

Benacci and Wolff80 of the Mayo Clinic stated the indications for catheter cecostomy as follows:


Colonic pseudo-obstruction

Distal colonic obstruction

Cecal perforation

Cecal volvulus

Preanastomotic decompression

Miscellaneous usage

These authors conclude that good selection of patients, good cecostomy technique, and good postoperative care will provide good results.

Marinella81 reported colonic pseudo-obstruction complicated by cecal perforation in a patient with Parkinson’s disease.

Decreased mobility of the colon can result from abnormal connective tissue bands that pass across the ascending colon beneath the peritoneum. If the band is broad, covering most of the colon, it is designated as Jackson’s membrane or veil (Fig. 18-34). It may or may not be vascularized. The origin of this membrane is not known with certainty, but probably results from improper fixation of the bowel in embryologic development. Indeed, the gradation from normal to abnormal is so subtle that its incidence cannot be stated. Such bands and membranes can constrict the colon, reduce its mobility, and provide a basis for volvulus.

Fig. 18-34.

“Jackson’s veil” may contain many small blood vessels from the second lumbar or renal arteries. The extent of the “veil” is variable. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Arterial Supply, Venous Drainage

The right colic artery participates in the blood supply of the ascending colon. It is usually said to be a branch of the superior mesenteric artery, and anastomoses with the colic branch of the ileocolic artery. Variations of the right colic artery are frequent, however.

According to Van Damme,1 the vascularization of the ascending colon is through a paracolic arcade supplied by the ileocolic and middle colic arteries, with the potential reinforcement of two right colic arteries. Van Damme stated that the ileocolic artery is the most consistent branch leaving the right side of the superior mesenteric artery. The ileocolic artery is also an important landmark for the interpretation of arteriograms.

Bertelli et al.42 found the number of right colic arteries to be variable. In most cases, there are two or three.

The right colic vein, a tributary of the superior mesenteric vein, provides the venous return for the ascending colon.


The lymphatic drainage is to the right colic nodes, which are located along the network of the right colic artery and the marginal artery of Drummond. According to Woodburne and Burkel,82 the total number of lymph nodes of the ascending colon and cecum averages 75.


The sympathetic innervation of the ascending colon comes from the celiac and superior mesenteric ganglia. The parasympathetic innervation of the ascending colon arises from the vagus nerve. Cell bodies of sensory fibers from this part of the bowel are located in dorsal root ganglia of the eighth and ninth spinal nerves; therefore, pain is referred to the paraumbilical region of the abdominal wall.

Surgical Considerations


The ascending colon can be mobilized by incision of the right lateral peritoneal reflection.

Dangers from incision are: bleeding from the gonadal vessels, injury to the right ureter, and injury to the duodenum.

Hepatic Flexure, Transverse Colon, Splenic Flexure, and Transverse Mesocolon

The hepatic flexure is located under the 9th and 10th costal cartilages in the vicinity of the midaxillary line between the anterior surface of the lower half of the right kidney and the inferior surface of the right hepatic lobe. The gallbladder is located anteriorly, and the duodenum is located posteriorly.

Occasionally, there is a peritoneal fold between the hepatic flexure and the gallbladder (cystocolic ligament). Also, occasionally, there is another peritoneal fold which starts from the hepatogastric or hepatoduodenal ligament and ends at the right part of the hepatic flexure (hepatocolic ligament). A similar, but rare, peritoneal fold starts from the right lobe of the liver and extends over the entire hepatic flexure. For all practical purposes, this is a wide hepatocolic ligament.

The transverse colon begins where the colon turns sharply to the left (the hepatic flexure), just beneath the inferior surface of the right lobe of the liver. It ends at a sharp upward and then downward bend (the splenic flexure) related to the posterolateral surface of the spleen. The tail of the pancreas is above. The anterior surface of the left kidney lies medially.

The transverse colon, unlike the ascending and descending colon, has a mesentery which has fused secondarily with the posterior wall of the omental bursa (Fig. 18-35). At the beginning of the mesentery, there may be additional bands of peritoneum, the hepatocolic and cystocolic ligaments. These are adhesion bands, not persistent remnants of the ventral mesentery. At the splenic flexure, the colon is supported by the phrenocolic ligament, a part of the left side of the transverse mesocolon.

Fig. 18-35.

Diagrammatic sagittal section showing the developmental relation of the transverse colon and its mesentery to the omentum. The two are fused at “X” to form the transverse mesocolon containing the middle colic artery. SMA, Superior mesenteric artery. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Between the hepatic and splenic flexures, the transverse colon hangs in a U- or V-shaped curve. It may lie above the umbilicus, but is often lower, even extending into the true pelvis. The transverse colon varies with individuals and with body position.

The transverse mesocolon is formed by a double peritoneal fold which extends upward and attaches to the anterior pancreatic border, suspending the transverse colon from the pancreas. It ranges in length from 3 to 12 cm. The transverse mesocolon contains the middle colic artery and vein, and lymph nodes as well as nerves. Occasionally, the superior fold of the transverse mesocolon is fixed by adhesions to the posterior wall of the stomach. Gastric ulcers and benign or malignant tumors may be firmly fixed to the mesocolon and despite all efforts to preserve integrity, the middle colic artery may be injured when the stomach wall is separated from mesocolon.

The transverse mesocolon and transverse colon provide the barrier between the supracolic and infracolic compartments of the peritoneal cavity; they are responsible for supracolic or infracolic collections of fluid.

The splenic flexure has an acute angle. It is located higher than the hepatic flexure, at the level of the 8th interspace in the midaxillary line. This high position is due not only to the small left hepatic lobe, but also to the multiple splenic ligaments and other ligaments in its vicinity (see the chapter on the spleen). The splenic flexure is related posteriorly to the left kidney and anteriorly to the left costal arch and occasionally to the stomach.

Charnsangavej and colleagues reviewed CT studies of the mesocolon in health83 and disease,84 and reported the following:

The mesocolon of the cecum and ascending colon can be identified by following the ileocolic vessels at the root of the mesentery and the marginal vessels along the mesocolic side of the colon, whereas the IMV [inferior mesenteric vein] and the marginal vessels of the descending colon serve as landmarks for the descending mesocolon. These vessels . . .can be readily identified on CT scans. The transverse mesocolon and the sigmoid mesocolon are less constant because they are more mobile, but the vessels in them can also be traced on CT scans. The middle colic vessels that originate from the marginal vessels of the transverse colon and run toward the pancreas to drain into the SMV (superior mesenteric vein] serve as landmarks for the transverse mesocolon; the marginal vessels of the descending colon and the sigmoid colon and the superior hemorrhoidal vessels form the IMV and serve as landmarks for the sigmoid mesocolon.

The anatomy of the colon and mesocolon is well defined and can be recognized by following the vascular anatomy and by understanding the relationship between the colon and mesocolon and its attachment to retroperitoneal structures, particularly to the pancreas. Common pathologic conditions in the mesocolon result from the spread of disease either from the colon or the pancreas. The pathway for the spread of disease may be seen along the mesocolon via the lymphatic vessels in the mesocolon and involvement of mesocolic vessels.

Arterial Supply, Venous Drainage

In the “typical” arrangement, the right colic artery (Fig. 18-36) bifurcates into ascending and descending branches. The ascending branch anastomoses with the right branch of the middle colic artery. The descending branch anastomoses with the ileocolic artery. The left colic artery (Fig. 18-36) also bifurcates; its ascending branch anastomoses with the left branch of the middle colic artery, and its descending branch anastomoses with the first branch of the sigmoid artery.

Fig. 18-36.

Nomenclature of the colic arteries. Only colic vessels arising directly from the superior mesenteric artery deserve the name of arteria colica. The right colic artery is an exceptional vessel (13%). The middle colic is not a single vessel but a complex system of five different vessels that behave as arteries or as branches: (1) middle colic artery dividing into a branch for the right angle and one for the transverse colon; (2) artery for the right angle of the colon; (3) artery for the transverse colon; (4) accessory artery for the transverse colon; (5) accessory left colic artery. (Modified from Van Damme JP, Bonte J. Vascular Anatomy in Abdominal Surgery. New York: Thieme Verlag, 1990; with permission.)

The main blood supply of the transverse colon is the middle colic artery (Fig. 18-36), but the ascending branch of the left colic artery contributes circulation for the distal part of the transverse colon. Anatomically, the entrance of the middle colic artery into the antimesenteric border of the transverse colon is very close to the pancreatic neck. If adhesions of benign or malignant origin are present in this area, this is another point of danger.

The middle colic artery bifurcates from 3 cm to 11 cm from the colonic wall, and may be absent in 5-8 percent of individuals. In most cases, the middle colic artery originates from the superior mesenteric artery. It can arise as a stem from which the inferior pancreaticoduodenal artery, right hepatic artery, jejunal artery, or other branches take origin. It may be absent.85

Van Damme1 stated that the middle colic artery is not a single artery, but a complex system of different vessels. He described the “composed system of the middle colic artery” as follows:

The middle colic artery is not a single vessel. Five different vessels can be discerned (Fig. 18-36) behaving as arteries or as branches: (1) the middle colic artery (46% of specimens), dividing into a branch for the right angle and one for the transverse colon; (2) the artery for the right angle of the colon (32%); (3) the artery for the transverse colon (12%); (4) the accessory artery for the transverse colon (3%); and (5) the accessory left colic artery (7%).

Bertelli et al.42 noted a middle colic artery in more than 95% of their subjects.

The venous return is formed by right and left networks. The right network drains into the right gastroepiploic vein or the superior mesenteric vein. The left network drains into the inferior mesenteric vein.


The lymphatics of the transverse colon can be subdivided into two networks: the right serves the proximal two-thirds, and the left serves the distal one-third.

In the right network, the lymphatic vessels follow the right and middle colic arteries to the lymph nodes of the superior mesenteric artery, and then to the aortic nodes, and finally reach the intestinal lymphatic trunk.

In the left network, some of the lymphatic vessels follow the middle colic artery, while most accompany the left colic artery to the lymph nodes of the inferior mesenteric artery and lumbar lymph nodes.



Lymphatic drainage of the proximal transverse colon including the hepatic flexure usually drains into the middle colic or right colic system; rarely it drains into the ileocolic system.

The middle transverse colon is served by the middle colic system.

Lymphatic drainage of the distal transverse colon including the splenic flexure drains to the middle colic and left colic systems.

Retrograde systems may be present.


The innervation of the transverse colon is by the autonomic system. Postganglionic sympathetic fibers arise from the superior mesenteric ganglia. Preganglionic parasympathetic fibers from the vagi supply two-thirds or somewhat more of the transverse colon. The remainder arise by way of the pelvic splanchnic parasympathetic outflow from the intermediolateral cell column at levels S2, S3, and S4. Postganglionic cell bodies are present within the terminal ganglia of the wall of the colon.

Surgical Considerations


Mobilization of the hepatic flexure can be accomplished by: (a) incision of the upper right lateral peritoneal reflection (as is done for mobilization of the ascending colon), (b) placement of the colon in a medial position.

Mobilization of the right colon (cecum, ascending colon, hepatic flexure) can be combined with duodenal mobilization by kocherization (see the duodenum chapter).

The retroperitoneal anatomic entities with which the hepatic flexure is related are the right kidney and the ureter, and the descending part of the duodenum and the gonadal vessels.

The right gonadal vein is more vulnerable than the right gonadal artery. This is because the entrance of the vein to the inferior vena cava is higher, and the origin of the artery is lower. Both vessels are very close together just above the right iliac fossa. They travel downward together.

The anterior relations with the gallbladder can be seen very well by the orthodox or laparoscopic approach.

Separation of the posterior gastric wall from the transverse mesocolon requires great care.

It is also essential to separate the transverse mesocolon very carefully from the pancreas at the neck of the pancreas.

The left mesocolon is less vascular than the right. These avascular areas should be used in surgery for posterior gastrojejunostomy, Roux-en-Y, and other procedures. Remember that the “arc of Riolan” may be present in the more cranial part of the mesocolon. This is an artery of respectable size interconnecting the inferior mesenteric or left colic arteries with the superior mesenteric or middle colic arteries. Hollinshead86 states that Griffiths reported the incidence of the arc of Riolan to be 10%.

The posterior area of the splenic flexure is usually not covered by peritoneum and is fixed to the posterior abdominal wall. Occasionally, it is covered completely by peritoneum and its mesocolon is fixed to the distal body and tail of the pancreas.

The splenocolic ligament, a remnant of the left end of the transverse mesocolon, is a peritoneal bridge between the splenic flexure and the lower spleen. The left gastroepipolic artery and some aberrant inferior polar vessels may be in the vicinity. Incise the ligament between clamps and ligate. Should excessive traction be applied to this ligament, tearing of the splenic pole can result, with profuse bleeding.

The phrenocolic ligament, between the splenic flexure and the diaphragm at the level of the 10th rib at the midaxillary line, forms a hammock in which the spleen rests. The ligament should be incised for mobilization of the splenic flexure.

Mesenteric cysts of the small or large bowel are uncommon, but may be benign or malignant; thus the surgeon should be alert to them. Horiuchi et al.87 state that retroperitoneoscopic resection of such cysts has a lower risk of traumatizing the bowel than laparoscopic intra-abdominal excision, with a further advantage of not compressing other intra-abdominal organs. Forty percent of all mesenteric cysts occur in the colon. A complete discussion of mesenteric cysts can be found in the chapter on the small intestine.

Descending Colon

The descending colon is related to the following anatomic entities: the quadratus lumborum muscle, left adrenal gland, left kidney and left ureter, left gonadal vessels, and the iliohypogastric and ilioinguinal nerves.

Like the ascending colon, the left or descending colon is covered anteriorly and on its medial and lateral sides by peritoneum. It normally has no mesentery. During the last decade, several authors, such as Dixon et al.88 have investigated the superficial or deep retroperitoneal topographical anatomy of the descending colon with the use of CT scans and MRIs. Hadar and Gadoth,89 Sherman et al.,90 Hopper et al.,91 Helms et al.,92 and Prassopoulos et al.93 investigated the retroperitoneal relations between the descending colon and the left kidney. Because the goal of these articles is to avoid anatomic complications and injury to the left kidney during percutaneous renal and vertebral procedures when the descending colon is too deep, they should have presented more details of the anatomy of the descending colon.

LeRoy et al.94 stated that a descending colon that is located more posteriorly than expected may be injured if too posterior an approach is used. According to Dixon et al.,88 in rare cases, the descending colon (which is situated within the retroperitoneal fat) passes between the psoas major and the quadratus lumborum muscles in a posterolateral location, which makes it possible for the bowel to be injured during posterior percutaneous procedures. LeRoy et al.,94 Helms et al.92 and Bonaldi et al.95 emphasized the possibility of puncturing the descending colon.

According to Dixon et al.,88 there are several explanations for these variations in topography and location of the descending colon. These include: obesity, variations of the architecture and topography of the psoas major muscle, and the more posterior location of the colon in women. The authors of this chapter emphasize that these variations are not of congenital origin and we concur with Dixon et al. that they constitute anatomic variations.

When the left or descending colon has a mesentery, it is rarely long enough to permit a volvulus to occur. The surgical unit of the left colon consists of the splenic flexure and the descending and sigmoid colons.

Arterial Supply, Venous Drainage

The descending colon is supplied by the left colic artery. The inferior mesenteric vein provides venous drainage. This then passes into the portal vein. The route may be via the splenic vein, the superior mesenteric vein, or venous flow may pass directly into the origin of the portal vein at the junction of the two previous veins. These pathways occur with nearly equal frequency.

According to Grant and Basmajian,96 “a large branch . . .not rarely connects the stem of the superior mesenteric artery with the left colic artery on the posterior abdominal wall.” This branch is the arc of Riolan.

Van Damme1 reported the following observations on the arterial supply of the left colon:

The left colic artery. . . has a long S-like course and ends at the splenic flexure. Its behavior is influenced by the presence of an accessory left colic artery or branch (14% of specimens) from the middle colic artery group, yet they arise from a completely different area. If this accessory left colic vessel is present, the normal left colic artery is absent (12%), atrophic, or displaced. In the absence of the left colic artery, both the colosigmoid artery and the paracolic arcade are usually reinforced by an accessory vessel. . . The left colic artery gives rise to the colosigmoid (38%) and sigmoid (4%) vessels and seldom to an intermesenteric arcade. We reintroduced the name “colosigmoid artery”97 because this vessel plays a key role in the supply of the descending and sigmoid colons. It arises close to the angle between the left colic and the inferior mesenteric artery from one of these constituents and points to the colosigmoid transition area. Sigmoid branches arise from the colosigmoid, the left colic, the inferior mesenteric, or the superior rectal arteries. The last sigmoid artery is a troublesome heritage from the discussion about the critical point of Sudeck.40 It deserves no special attention.

Bertelli et al.42 agree with several authors that the left colic artery is the only true colic artery.


The lymphatic drainage is to nodes of the superior mesenteric artery.


The autonomic system is responsible for the innervation of the descending colon. Preganglionic and postganglionic sympathetic fibers are carried by lumbar splanchnic branches of the sympathetic chain, the preganglionic fibers synapsing in inferior mesenteric ganglia. Preganglionic parasympathetic fibers are carried by the pelvic splanchnic nerves from the ventral primary rami of S2, S3, and S4; these fibers synapse in intramural (terminal) parasympathetic ganglia in the rich intermuscular and submucosal plexuses.

Surgical Considerations


Mobilization of the descending colon is accomplished by incising the peritoneal reflection at the left gutter along the “white line of Toldt.” During mobilization, the most vulnerable anatomic entities are the left ureter and the left gonadal vessels.

Remember that the inferior mesenteric vein may also lie, often well-concealed, behind the mesentery of the descending colon, if a mesentery is present.

Sigmoid Colon

At the level of the iliac crest, the descending colon becomes the sigmoid colon and acquires a mesentery. The sigmoid colon is described as having two portions: (1) the iliac portion, which is fixed and located at the left iliac fossa; and (2) the pelvic portion, which is mobile. This entity is called “sigmoid” because of its “S” shape in many people. For all practical purposes, the sigmoid colon begins at the iliac crest and ends at the 3rd sacral vertebra. The iliac part is the downward continuation of the descending colon. It does not have a mesentery, and it ends at the pelvic brim, where the pelvic colon and its mesentery (pelvic mesocolon) start.

The mobile, omega-shaped () pelvic colon begins at the medial border of the psoas major muscle. It has a mesentery (the pelvic mesocolon) that is fixed to the posterior pelvic wall, its fixation being like the capital Greek letter lambda (). The pelvic colon terminates at the rectosigmoid junction, which is located at the area of the 3rd sacral vertebra. At this point, its mesentery ceases. The middle of the base of the lambda is located at the point where the left ureter crosses the pelvic brim –at the intersigmoid mesenteric recess– just lateral and posterior to the fossa in which the left ovary rests. The left leg of the lambda is attached to the pelvic brim. The right leg travels medially and downward to the 3rd sacral vertebra. The superior rectal vessels are within the mesentery of the sigmoid colon.

Shafik98 suggested that the rectosigmoid junction acts as a functional sphincter: opening reflexively upon sigmoid contraction (rectosigmoid inhibitory reflex) and allowing feces to pass to the rectum, and closing upon rectal contraction (rectosigmoid excitatory reflex) and preventing stool reflux into the sigmoid.

Arterial Supply, Venous Drainage

The sigmoid branches of the inferior mesenteric artery supply the sigmoid colon. There can be a left colic-sigmoid branch of the left colic artery that supplies the proximal sigmoid. The inferior mesenteric vein drains the area.


The course of the lymphatics of the descending colon is as follows: the lymphatic vessels drain to nodes along the left colic artery, then to inferior mesenteric artery nodes, then to left lumbar nodes or left aortic nodes.


Autonomic nerves, as described for the descending colon, innervate the sigmoid colon. Pain fibers for the sigmoid colon and the descending colon pass upward via lumbar splanchnic nerves and the sympathetic chain to the upper 1 or 2 lumbar segments of the spinal cord (the site to which pain may be referred from the descending colon). Their path is via white communicating rami, connecting the sympathetic chains to the spinal nerves at those levels.

Surgical Considerations


During mobilization of the ileal and pelvic parts of the sigmoid colon, it is necessary to protect the left ureter and the left gonadal vessels (just as with the descending colon).

The inferior mesenteric vein can be ligated with impunity. If possible, ligate it close to its termination or to its entrance to the splenic or superior mesenteric vein.

Transection of the colon 3-4 cm below the tumor is adequate for rectosigmoid tumors because the lymphatic flow in most cases is not retrograde. Most, if not all, involved lymph nodes are 2 cm or less below the tumor.

The promontory of the sacrum is perhaps a landmark for the termination of the sigmoid colon. Remember that in rare cases there are tongues of short mesentery for both the proximal rectum and the terminal sigmoid colon.

The inferior mesenteric artery also can be ligated with impunity if normal anastomoses are present between the right colic artery (if present), middle colic artery, and the marginal artery interconnecting them with the left colic artery.

The attachment of the mesosigmoid to the body wall shows much variation.99 In most individuals, the attachment starts in the left iliac fossa and extends diagonally downward and to the right. In others, the attachment is sinuous (shaped like a C, S, or inverted U). According to Vaez-Zadeh and Dutz,100 the average length of the attachment in 140 autopsies was 7.9 cm (Fig. 18-37A). These authors found the breadth of the mesentery ranging from an average of 5.6 cm in 100 autopsies in New York to 15.2 cm in 40 autopsies in Iran. Whether this difference is genetic or dietary is not clear. The left ureter passes through the base of the sigmoid mesocolon through the intersigmoid recess (Fig. 18-37B).

Fig. 18-37.

Sigmoid mesocolon. A, Average measurements. B, The relation of the base of the sigmoid mesocolon to the left ureter. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983. Data in A from Vaez-Zadeh K, Dutz W. Ileosigmoid knotting. Ann Surg 1970;172:1027; with permission.)

Surgical Anatomy of the Rectum and Anal Canal

The junction between the sigmoid colon and the rectum has been variously described:


A point opposite the left sacroiliac joint

Level of the 3rd sacral vertebra

Level at which the sigmoid mesentery disappears

Level at which sacculations and epiploic appendages disappear and taeniae broaden to form a complete muscle layer (long transition)

Level at which the superior rectal artery divides into right and left branches

Construction with anterior angulation (proctoscopy)

Level of superior rectal fold (inconstant)

Transition between rugose mucosa of colon and smooth mucosa of rectum (cadaver)

These levels are neither consistent with each other nor constant from one individual to another. Some are useful to the surgeon, some to the anatomist, and some to the proctoscopist. The reader need not be discouraged. There is, fortunately, no compelling surgical reason for establishing a definite boundary between the sigmoid colon and the rectum.

The lower boundary of the rectum is no more agreed upon than the upper boundary. Part of the disagreement between the surgeon and the anatomist lies in differences between the living patient and the cadaver, but much of it is due to a terminology rich in both synonyms and ambiguities.

Some anatomy textbooks describe the anal canal as the region lying distal to the pectinate (dentate) line, while to the surgeon, all the region distal to the insertion of the levator ani muscles is the anal canal. The surgical anal canal (the anorectum of Harkins) includes the anatomic anal canal and the distal 2 cm of the rectum above the pectinate line. In addition to the changes in anatomy and physiology at the pectinate line, pathology unique to the distal 4 cm of the intestinal tract (2 cm above and 2 cm below the line) distinguishes a unit, the “surgical anal canal” (Table 18-7).6

Table 18-7. The Pectinate Line and Changes in the Surgical Anal Canal

  Below the Pectinate Line Above the Pectinate Line
Embryonic origin Ectoderm Endoderm
  Lining Stratified squamous Simple columnar
  Arterial supply Inferior rectal artery Superior rectal artery
  Venous drainage Systemic, by way of inferior rectal vein Portal, by way of superior rectal vein
  Lymphatic drainage To inguinal nodes To pelvic and lumbar nodes
  Nerve supply Inferior rectal nerves (somatic) Autonomic fibers (visceral)
Physiology Excellent sensation Sensation quickly diminishes
  Cancer Squamous cell carcinoma Adenocarcinoma
  Varices External hemorrhoids Internal hemorrhoids

Source: Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.

Peritoneal Reflections

The entire upper one-third of the rectum is covered by peritoneum (Fig. 18-38). As the rectum passes deeper into the pelvis, more and more fat is interposed between the rectal musculature and the peritoneum. The mesorectum, which suspends the rectum from the posterior body wall, comes off more laterally, leaving bare progressively more of the posterior rectal wall. The peritoneum finally leaves the rectum and passes anteriorly and superiorly over the posterior vaginal fornix and the uterus in females or over the superior ends of the seminal vesicles and the bladder in males. This creates a depression, the rectouterine or rectovesical pouch; with infection, this may become filled with pus.

Fig. 18-38.

The line of peritoneal reflection on the rectum; lateral view in the male. More of the rectum is covered anteriorly than posteriorly. The measurements of the anal canal and lower rectum from the anal verge are approximate. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

In summary, the relations of the rectum are as follows:


Male (extraperitoneal)


– Anterior: prostate, seminal vesicles, ductus (vas) deferens, ureters, urinary bladder

– Posterior: sacrum and fascia of Waldeyer, coccyx and its muscles, levator ani, median sacral vessels, roots of the sacral nerve plexus


a. Extraperitoneal


– Anterior: posterior vaginal wall

– Lateral: intestinal loops

b. Intraperitoneal


– Anterior: posterior vaginal wall, upper uterus, uterine (fallopian) tubes, ovaries

– Lateral: pelvic wall

– Posterior: as in the male

Some authors like to divide the anorectum into 3 parts. This division also is acceptable if the student of rectal surgery remembers that the upper one-third is surrounded by peritoneum, the middle one-third is covered only anteriorly by peritoneum (being located nearly retroperitoneally) and the lower one-third is extraperitoneal (retroperitoneal).

Pelvic Diaphragm and Continence

The floor of the pelvis is the pelvic diaphragm, through which the rectum passes. The diaphragm is composed of two paired muscles, the levator ani and the coccygeus (Fig. 18-39). The levator ani may be considered to be made up of three muscles: the iliococcygeus, the pubococcygeus, and the puborectalis. The puborectalis is essential to maintaining rectal continence, and is considered by some authors to be part of the external sphincter and not a part of levator ani. The visible, intrapelvic border of the levator hiatus is formed by the medial borders of the pubococcygeus, not the puborectalis. The puborectalis is attached to the lower back surface of the symphysis pubis and the superior layer of the deep perineal pouch (urogenital diaphragm). Fibers from each side of the muscle pass posteriorly and then join posterior to the rectum, forming a well-defined sling (Fig. 18-39).

Fig. 18-39.

Diagram of the pelvic diaphragm from below. Note that the levator ani is composed of three muscles: puborectalis, pubococcygeus, and iliococcygeus. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The puborectalis with the superficial and deep parts of the external sphincter and the proximal part of the internal sphincter form the so-called anorectal ring. This ring can be palpated; and since cutting through it will produce incontinence, it must be identified and protected during surgical procedures. The details of the external sphincter will be discussed with the morphology of the anal canal (see “Surgical Anatomy of the Rectum and Anal Canal”).



The anal verge is the junction of skin around the anal opening and the anal mucosa.

The dentate or pectinate line is located 2 cm above the anal verge and represents the junction between the anal transition zone below and the anal nonkeratinizing squamous mucosa.

In the upper end of the surgical anal canal (which is located 4 cm from the anal verge and 2 cm from the dentate line) is the palpable anorectal ring. This is the level of the sling provided by the puborectalis.

The divisions of the anorectum can be remembered as 4 and its multiples. The surgical anal canal is 4 cm long; the rectum, 12 cm long; the total anorectum is approximately 16 cm.

Fibrofatty tissue covers the surgical anal canal anteriorly, laterally, and posteriorly. The distal 8 cm of the anorectum is mostly (or partially) retroperitoneal; posteriorly it is related to the sacrum and coccyx.

The posterior rectal wall is related to two fasciae: a thin one close to the rectal wall, and a thick one (the fascia of Waldeyer) covering the anterior aspects of the sacrococcygeal area.

Only the thin fascia should be removed with the rectum. Removal of the sacral fascia will produce venous bleeding which is extremely difficult to control.

Careful cleaning of the distal 8 cm of the anorectum will produce enough length for a sphincter-saving procedure.

Anorectal anastomosis after good mobilization of the distal 8 cm of the anorectum can be safely done at the dentate line (endo-anal anastomosis).

Some anatomic guidelines vary depending on the peritoneal reflection, sex of the patient, peculiarity of pelvic anatomy, and weight of the patient.

The rectal mucosa has no role in the act of defecation.

Today’s acceptable “save” rule is a resection line of 2.5-3 cm distal to the lower margin of the tumor mass.

The surgeon should always remember the magic number 4 for the segmental anatomy of the anorectum, as well as an empiric 2.5-3 cm margin below the tumor as the line of resection. Frozen specimen of the distal end, appropriately marked, is essential. If the tumor is located 2-3 cm above the dentate line, abdominoperineal resection is the procedure of choice. Discussion with the patient prior to surgery about the possibility of permanent colostomy is imperative.

During a low anterior resection, the length of the rectum after good mobilization may be increased by 4 cm to 5 cm.

Fascial Relations and Tissue Spaces

The parietal fascia of the pelvic basin is continuous with the transversalis fascia of the abdominal cavity. In the pelvis this fascia would include that which covers the obturator internus laterally and the piriformis posterolaterally. It also includes the fascial layers of the pelvic diaphragm; that is, the fascia of the levator ani (Fig. 18-40) and coccygeus muscles.

Fig. 18-40.

Diagram of some of the fasciae of the pelvis seen in coronal section. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Six potential spaces around the rectum (Fig. 18-41) are recognized by Shafik.101 They are important because they may become sites of infection. The fascial layers that bound these spaces help limit the spread both of infection and of neoplastic disease, although all are potentially confluent with one another.

Fig. 18-41.

The spaces of the anus and rectum. 1. Pelvirectal space. 2, Ischioanal (ischiorectal) space. 3, Intersphincteric spaces. 4, Subcutaneous space. 5, Central space. 6, Submucous space. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Subcutaneous Space

The subcutaneous space corresponds to the perianal space of Milligan and associates.102 It is bounded above by the lowest muscular loop of the external sphincter, below by the perianal skin, and medially by the epithelium of the anal verge. It is filled with fat and by fibers of the corrugator ani cutis. The subcutaneous space is in communication above with the central space and laterally with the ischioanal (ischiorectal) space. Medially, the medial central septum separates it from the submucous space. The subcutaneous space is often considered to be a part of the ischioanal space.

Central Space

The central space is considered by Shafik101 to be the main perianal space; it is in communication with each of the others. It surrounds the anal canal and is bounded by the termination of the longitudinal muscles above and the lowest muscular loop of the external sphincter below. Within it lie the tendon fibers of the longitudinal muscles.

Intersphincteric Spaces

The intersphincteric spaces are four upward extensions of the central space. They are the fascial planes between the longitudinal intersphincteric muscles that form the upper boundary of the central space. Proceeding from lateral to medial, the first and third spaces open into the ischioanal space, the second space opens into the pelvirectal space, and the most medial space communicates with the submucous space. These “spaces” are potential pathways of infection and are potential spaces only.

Ischioanal (Ischiorectal) Fossa

The ischioanal (ischiorectal) fossa is a pyramidal space on either side of the anal canal and lower rectum, posterior to the base of the urogenital diaphragm. Its base is at the perianal skin; its medial wall is the external anal sphincter and levator; its lateral wall is the internal obturator fascia; and its apex is where the levator muscles join the obturator internus muscle. The two spaces communicate posteriorly through the retrosphincteric space.103

The lower portion of the ischioanal (ischiorectal) space has been termed the perianal space by Milligan.104 Laterally it communicates with the gluteal fat and the subcutaneous space. It is usually considered a part of the ischioanal fossa.

Pelvirectal Spaces

The pelvirectal spaces lie above the levators and are bounded superiorly by the pelvic peritoneum, laterally by the pubococcygeus muscle, and medially by the rectum. The spaces are filled with fibroadipose tissue. The fibrous elements of the tissue, known as the lateral ligaments of the rectum, are part of the pelvic fascia and connect the parietal pelvic fascia with the walls of the rectum and pelvis. These ligaments form a triangle, with its base on the side wall of the pelvis and the apex joining the rectum. The lateral ligaments conduct the middle rectal vessels and nerves.

The lateral pelvirectal spaces on either side of the rectum communicate behind the rectum, in front of the sacrum, and above the levators. This communication is separated from the rectal wall by the fascia propria of the rectum and from the sacrum by a thickened parietal pelvic fascia called the fascia of Waldeyer. The middle sacral vessels lie within this fascia, which is often considered to be a separate space (the rectorectal or presacral space).

The extraperitoneal anterior part of the rectum is covered with a bilaminar fascial layer (Denonvilliers’ fascia) that extends from the anterior peritoneal reflection above to the perineal body below. Posterolaterally, this connective tissue septum is continuous with connective tissue of the sacrogenital or uterosacral folds and the lateral pillar of the rectum. This bilaminar fascial layer, also called the rectovaginal or rectovesical septum, separates the rectum from the prostate and seminal vesicles in males or from the vagina in females. This fascia forms a barrier to the spread of cancer or infection either anteriorly or posteriorly.

Submucous Space

The submucous space of Shafik105 lies beneath the anal mucosa and the internal sphincter. It is the most distal portion of the submucosa of the digestive tract. While it is certainly a possible pathway for infection, it does not contribute to the longitudinal spread of cancer.106 The authors of this chapter consider it a definite and rather stout layer of the enteric wall; however, it can be considered a space.

A mucosal ligament described by Parks107 as connecting the anal mucosa and the internal sphincter would form the lower limit of the submucous space. The presence of such a ligament has not been confirmed by Shafik105 or by Goligher.41

Retrorectal Space

The retrorectal space, according to Jackman et al.,108 has the following boundaries:


Anterior: Fascia propria of the posterior rectal wall

Posterior: Presacral fascia

Lateral: Iliac vessels, ureters, lateral rectal ligaments

Superior: Peritoneum

Inferior: Retrosacral fascia


In Embryology for Surgeons,13 which was co-written by the senior author of this chapter, the mesorectum was neither mentioned nor discussed. Gray’s Anatomy26 refers to the “dorsal mesorectum…which does not form a true mesentery, however, but. . . a woven fibroreolar sheet with patterned variations in thickness and fibre orientation.” The rectum is said to differ from the sigmoid colon by having “no sacculations, appendices epiploicae or mesentery“26 (italics ours).

Heald and colleagues109 have published multiple excellent articles providing a complete education about the mesorectum. Heald110 defines the mesorectum as “the integral visceral mesentery surrounding the rectum . . .covered by a layer of visceral fascia providing a relatively bloodless plane, the so-called ‘holy plane’ (Fig. 18-42A & B). The dorsal mesentery is embryologically responsible for the genesis of the mesorectum.

Fig. 18-42.

The “holy plane.” A, Diagrammatic representation. B, Suggested plane of excision shown by dashed line. (A, Modified from Heald RJ. The “holy plane” of rectal surgery. J R Soc Med 1988;81:503-508. B, Modified from Heald RJ, Husband EM, Ryall RDH. The mesorectum in rectal cancer surgery – the clue to pelvic recurrence? Br J Surg 1982;69:613-616; with permission.)

In the lab, the senior author of this chapter has found this ‘holy plane,’ and he agrees with Heald about its several relations. Posteriorly, the plane is located between the visceral fascia which surrounds the mesorectum and the parietal presacral fascia (fascia of Waldeyer). In the male, Denonvilliers’ fascia (Fig. 18-43) constitutes the anterior surface of the mesorectum, which is fused with the posterior surface of the fascia of Denonvilliers’. Inferiorly, the mesorectum and the fascia of Waldeyer condense to form the rectosacral ligament in the vicinity of S4.

Fig. 18-43.

Schematic representation of the relationships of the mesorectum to the anatomic structures in the male. The neurovascular bundle contains the nerves responsible for erection and ejaculation, and aspects of bladder function. (Modified from Heald RJ, Moran BJ. Embryology and anatomy of the rectum. Semin Surg Oncol 1998;15:66-71; with permission.)

The senior author of this chapter is sure that the readers of the beautiful publications of Heald would have no objections to renaming the ‘holy plane,’ the Plane of Heald.

We quote from Konerding et al.:111

The perirectal tissue gives rise to the rectal fascia or adventitia, also known as mesorectum. The connective tissue space between rectal and parietal pelvic fascia can be dissected as a plane free of vessels and nerves. Surgical dissection along this plane with complete mesorectum excision results in reliable excision of all relevant lymphatic pathways with extensive preservation of continence and sexual function.

Enker et al.112 presented four surgical planes for anatomic dissection of the rectum for the following oncologic surgeries (see the section “Anorectal Cancer Surgery” later in this chapter for details):


Total mesorectal excision

Total mesorectal excision with autonomic nerve preservation

Mesorectal excision

Extrafascial excision of the rectum

After low anterior resection with total mesorectal excision, Law et al.113 recommend routine creation of a stoma for males and selective use of diversion for females for the avoidance of lower anastomotic leakage.

Total or partial mesorectal excision for rectal carcinoma has been advocated by several authors.114-119 Maurer et al.120 stated that continence-preserving surgery on patients with rectal carcinoma may be performed in over 80% of patients by partial or total removal of the mesorectum with positive results.

Arterial Supply, Venous Drainage

The following is a detailed presentation of the blood supply of the anorectum. Some of this material has already been presented in less detailed form in other sections of “Surgical Anatomy of the Colon”; because it is a complicated topic, some repetition is worthwhile.

The arteries of the rectum and anal canal are the unpaired superior rectal artery, the paired middle and inferior rectal arteries (Fig. 18-44), and the median sacral arteries. The superior rectal (hemorrhoidal) artery arises from the inferior mesenteric artery and descends to the posterior wall of the upper rectum. Supplying the posterior wall, it divides and sends right and left branches to the lateral walls of the middle portion of the rectum down to the pectinate (dentate) line.

Fig. 18-44.

Diagram of the arterial supply to the rectum and anus. The median sacral artery supplying a few small branches to the posterior wall of the rectum is not shown. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Many surgeons are under the impression that the middle rectal (hemorrhoidal) arteries are always present in the lateral rectal stalks. Some authors30,121 have found these arteries to be inconstant; Michels122 found them to be constant, but varying in number and in origin. All originated (directly, or often indirectly) from the internal iliac artery. The number varied from three to nine and the diameter from 1.0 mm to 2.5 mm. In 58 percent of subjects, there was a grossly visible anastomosis between the middle and superior rectal arteries.

Boxall et al.123 found that the vessel in the lateral ligaments of the rectum, called the middle rectal artery by surgeons, was actually an “accessory” middle rectal artery that is present in about 25 percent of individuals. The main trunk of the middle rectal artery was inferior to the rectal stalk and could be endangered when the rectum is separated from the seminal vesicle, prostate, or vagina. In 12 cadavers, the arteries entered the rectal wall with the levator ani muscle; in 9 it was 2-4 cm higher. These findings may explain why some surgeons feel that the lateral ligaments may be cut with impunity.

In our experience, the middle rectal artery is usually absent in the female. It is probably replaced by the uterine artery. In the male, the chief beneficiaries of the artery are the rectal musculature and the prostate gland. Last124 agrees.

Vogel and Klosterhalfen125 reported that the middle rectal artery supplies the rectum accessorily and stated that this is the reason for suture leaks at the dorsocaudal area of the profunda. However, Goligher126 reported that the rectum and anus can survive divisions of the superior and middle rectal arteries.

The inferior rectal (hemorrhoidal) arteries arise from the internal pudendal arteries and proceed ventrally and medially to supply the anal canal distal to the pectinate line.

The median sacral artery arises just above the bifurcation of the aorta and descends beneath the peritoneum on the anterior surface of the lower lumbar vertebrae, the sacrum, and the coccyx. It sends several very small branches to the posterior wall of the rectum.

Venous drainage of the rectum and anal canal is discussed in “Vascular Supply” of the large intestine.


Lymph channels of the rectum and anal canal form two extramural plexuses, one above and one below the pectinate line (see Fig. 18-24). The upper plexus drains through posterior rectal nodes to a chain of nodes along the superior rectal artery to the pelvic nodes. Some drainage follows the middle and inferior rectal arteries to hypogastric nodes. Below the pectinate line, the plexus drains to the inguinal nodes.

There is considerable disagreement about connections between the two plexuses across the pectinate line, but if such connections exist, they are small. Regardless of this, drainage above the pectinate line from any part of the rectum is upward to the pelvic nodes; drainage below the line is to inguinal nodes. The importance of this line is that 85 percent of pathology is located in this area. External drainage to inguinal nodes appears to be limited to lesions involving the skin of the anal or perianal region.41

The watershed of the extramural lymphatic vessels is at the pectinate line (Fig. 18-44). The watershed for the intramural lymphatics is higher, at the level of the middle rectal valve (Fig. 18-45). These two landmarks may be kept in mind by the mnemonic “two, four, eight,” meaning:

2 cm = anal verge to pectinate line

4 cm = surgical anal canal (above and below the pectinate line)

8 cm = anal verge to middle rectal valve

Fig. 18-45.

Diagram of lymph drainage of the anus and rectum. The watershed for extramural drainage is at the pectinate line. The watershed for intramural drainage is at the level of the middle rectal valve, about 8 cm above the anal verge. IRA, Inferior rectal artery. MRA, Middle rectal artery. SRA, Superior rectal artery. LRV, Lower rectal valve. MRV, Middle rectal valve. SRV, Superior rectal valve. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The anatomy of metastasis of malignant tumors of the colon and rectum is perhaps as follows (Fig. 18-46):


1. Intramural stage. Cancer begins in the epithelium of the colon wall. Longitudinal spread in the submucosa is not common and, when present, extends only a few centimeters.106 Until the tumor has penetrated the mucosa and submucosa and involved the muscular and serosal layers, no metastasis occurs.

2. Direct extension. The pericolic fat is usually the first of the neighboring structures to be involved.

3. Venous drainage. Metastasis to the liver and lungs by way of the inferior mesenteric vein and the portal veins is an obvious pathway. A second pathway is from pelvic veins to the vertebral veins.127 This explains metastases to the vertebral column.

We quote from Koch et al.128:

Metastatic disease in colorectal cancer results from hematogenic dissemination of tumor cells…The significantly higher detection rate in mesenteric venous blood emphasizes the importance of the filter function of the liver for circulating tumor cells in the portal venous blood. Tumor cell detection in central and peripheral venous blood, however, shows that this filtering process is limited and indicates early systemic hematogenic tumor cell dissemination in colorectal cancer.


4. Additional pathways by which cancer spreads. Cancer spreads: (a) by lymphatics from nearby epicolic to paracolic to intermediate to principal lymph nodes; (b) by means of the peritoneal cavity with implants on the serosal surfaces of other viscera. Ueno et al.129 stated that indirect cancer involvement of the extrarectal autonomic nerves and/or the surrounding tissue occurred in direct proportion to the extent of cancer spread to the mesorectum. Nerve plexus involvement had an unfavorable prognosis.

Fig. 18-46.

Spread from a primary carcinoma of the colonic epithelium. Spread may be by extension in the submucosa, or by peritoneal seeding from extension into the subserous fat. More commonly, metastases travel through lymphatics to principal lymph nodes at the root of the mesentery, or by veins to the portal system. Thickness of arrows indicates the frequency with which spread occurs. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Lateral lymphatic spread is uncommon and is limited to lesions less than 4 cm from the pectinate line. Lesions above this level spread upward along the superior rectal/ inferior mesenteric artery system.130 The fact that recurrence of tumors of the lower one-third of the rectum is much higher than that of tumors of the upper two-thirds suggests that lateral spread may be more frequent than has been suspected.131

Downward spread of lesions of the rectum is similarly rare, probably only about 2 percent.132 A margin of 2-3 cm distal to the tumor should be allowed in anterior resection.

Williams and Beart133 stated that despite the fact that the Dukes 1932 system is considered the gold standard and is still used because of its simplicity and accuracy, the following should be considered predictors of survival and will play a role in the equation of staging: the number of positive lymph nodes, the depth of invasion, nucleopathology, flow cytometric characteristics, histological grade, and vascular or lymphatic invasion. The same authors recommend the use of the TNM (tumor, nodes, metastasis) system.

Granfield et al.134 reported that routine CT of the abdomen and pelvis will show the distribution of regional lymph node metastases in carcinoma of the left side of the colon, rectum, and anus.

Lymphatic Drainage to Adjacent Organs

Madden and McVeigh135 emphasized the importance of lymphatic communication between tumors of the colon at the splenic flexure and lymph nodes in the hilum of the spleen. To perform an adequate resection including the lymphatic drainage, they recommended removing en bloc the distal one-half of the transverse colon, the flexure, the whole of the descending and the proximal sigmoid colon with the mesentery, the distal one-half of the greater omentum, the proximal two-thirds of the gastrocolic ligament, the spleen, and the tail of the pancreas.

Block and Enquist136 found a large number of lymphatic channels passing from the lower one-third of the rectum to the posterior vaginal wall, the cul-de-sac, broad ligaments, and lateral cervical (cardinal) ligaments. For this reason, they recommended that resection for carcinoma of the lower two-thirds of the rectum in women include the rectum, the cul-de-sac, the uterus, tubes, ovaries, and posterior vaginal wall, the lateral cervical and broad ligaments, the levator ani muscles, and the ischioanal (ischiorectal) fat in continuity.


The possible autonomic innervation of the anorectum is as follows. The internal rectal sphincter’s motor innervation is supplied by sympathetic fibers that cause contraction. The pelvic splanchnic nerve (parasympathetic) and the hypogastric nerve (sympathetic) supply the lower rectal wall. Together these two nerves serve to form the rectal plexus. The levator ani muscles are controlled by the third and the fourth sacral nerves.

Davies137 reported the following about the innervation of the rectum, bladder, and internal genitalia in anorectal dysgenesis in the male.

Using a posterior sagittal approach to expose retroperitoneal viscera and nerves, the anatomy of the pelvic autonomic nerve plexus was studied in normal and abnormal male cadaver specimens. This plexus is found on the anterolateral surface of the lower rectum surrounded by endopelvic fascia. The autonomic nerves that supply the plexus reach it from posterior, lateral to the midline by passing over the surface of the rectum. The nerves of this plexus are distributed with the terminal branches of the internal iliac arteries, mainly with the vessels of the inferior vesical plexus. The rectum receives its autonomic nerves with its arterial blood supply, the superior rectal artery. The nerves of the pelvic plexus supply the genitourinary viscera that lie anterior to the rectum and in front of the fascia of Denonvilliers. The named fascial layers of the pelvis play a major role in determining the anatomic plane of these structures. In anorectal agenesis the plexus becomes a more midline structure. Because the pelvic fascia is often deficient in these cases these nerves lie vulnerable to inappropriate midline dissection (Figs. 18-47, 18-48, and 18-49).

Fig. 18-47.

The dissection planes in front and behind the normal rectum. (Modified from Davies MRQ. Anatomy of the nerve supply of the rectum, bladder, and internal genitalia in anorectal dysgenesis in the male. J Pediatr Surg 32:536-541, 1997; with permission.)

Fig. 18-48.

The anatomy of the pelvic nerve plexus in the presence of a normal rectum and anal canal. (Modified from Davies MRQ. Anatomy of the nerve supply of the rectum, bladder, and internal genitalia in anorectal dysgenesis in the male. J Pediatr Surg 32:536-541, 1997; with permission.)

Fig. 18-49.

The anatomy of the pelvic nerve plexus in anorectal agenesis with a rectovesical fistula. (Modified from Davies MRQ. Anatomy of the nerve supply of the rectum, bladder, and internal genitalia in anorectal dysgenesis in the male. J Pediatr Surg 32:536-541, 1997; with permission.)

Motor innervation of the internal rectal sphincter is supplied by sympathetic fibers that cause contraction and by parasympathetic fibers that inhibit contraction. The parasympathetic fibers are carried by pelvic splanchnic nerves which also convey the afferent nerve fibers that mediate the sensation of rectal distention. The external rectal sphincter is innervated by the inferior rectal branch of the internal pudendal nerve and by the perineal branch of the fourth sacral nerve.

The pelvic splanchnic nerves (parasympathetic and sensory) and the hypogastric nerve (sympathetic) supply the lower rectal wall. These two sources together form the rectal plexus. The levator ani muscles are supplied by the nerve to the levator ani, usually a branch from S4, with variant contributions from S3 and S5.

The inferior rectal branches of the internal pudendal nerve follow the inferior rectal arteries and supply the sensory innervation of the perianal skin.138

Remember that the pudendal nerve innervates the external sphincter and possibly the puborectalis muscle. The sympathetic nerves have no influence on the muscular wall of the rectum. Evacuation is accomplished by the pelvic splanchnic nerves; continence is maintained by the pudendal and the pelvic splanchnic nerves.

Since the pelvic parasympathetic nerves are responsible for erection and the sympathetic nerves of the hypogastric plexus are responsible for ejaculation, the surgeon should be familiar with the pathway of these nerves and dissect the posterior rectal wall from the sacrum, the prostate, and the lateral pelvic wall as close to the posterior rectal wall as possible.

The topographic anatomy of the nervi erigentes was studied by Stelzner et al.,139 who reported that the nerves are located along the diaphragmatic part of the urethra before entering the cavernous bodies. During proctocolectomy, in order to preserve sexual function, these authors advised leaving a piece of the rectal muscle that covers the diaphragmatic part of the urethra.

Williams and Slack140 studied the sexual function of males and females after rectosigmoid and rectum excision by examining the specimens for the presence of nerve tissue. Their results suggest that as the amount of nerve tissue in the specimen increases, impaired sexual activity increases.

For a conceptual picture of the “flow” of pelvic autonomic nerves which cannot be seen during mobilization of the rectosigmoid, and to help surgeons avoid inadvertent injury to these nerves, we recommend reading a brief review by Pearl et al.141 of the structure and function of these nerves.

Defecation and Continence

Distention of the rectum is the initial stimulus for defecation. Distention (Fig. 18-50A), with a rise in pressure, acting on mural receptors, produces reflex contraction of the rectal musculature. At the same time, the internal sphincter relaxes. This portion of the process is mediated by the intrinsic nerves only, with no contribution from extrinsic nerves.

Fig. 18-50.

Neural pathways involved in defecation. A, Rectal distension initiates relaxation of the internal sphincter and effective voluntary closure by the external sphincter. B, Defecation with relaxation of both sphincters, contraction of muscles in the rectal wall, and increased intraabdominal pressure. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The external (voluntary) sphincter (Figs. 18-50, 18-51) is normally in a state of contraction by a reflex from muscle spindles by means of the sacral spinal cord. Continence is obtained by a second reflex from the distended rectal wall to the sacral cord increasing contraction of the external sphincter and relaxing the rectal wall, reducing the urge to defecate. This reflex can be reinforced by voluntary effort if defecation is inhibited.

Fig. 18-51.

The three loops of the external anal sphincter: subcutaneous (C), superficial (B), and deep (A). Continence depends on the preservation of at least one of the three. Some subcutaneous muscle fibers encircle the anus; some attach to the perianal skin anteriorly at D. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

If defecation is to proceed, facilitating impulses arise from the cerebrum, pass to the sacral cord, then to the external sphincter to relax it142 (Fig. 18-50B). Contraction of the longitudinal muscles, together with peristalsis starting in the sigmoid colon, results in extrusion of the stool. The process is aided by voluntary straining (Valsalva maneuver).

Gunterberg et al.143 stated that damage to the parasympathetic nerves can almost completely abolish activity of the internal sphincter. Speakman et al.144 speculated that incontinence in some patients is from as yet unidentified defects in the innervation of the internal anal sphincter.

Is damage to the puborectalis during vaginal delivery the most significant cause of incontinence? Sunderland145 and Henry et al.146 support this concept. But recent studies indicate that a great proportion of incontinence is from other types of injuries during delivery. Fornell and her colleagues147 reported a high incidence of fecal incontinence after anal sphincter rupture at childbirth.

Surgical Considerations


Digital examination of the anus and lower rectum is essential. No patient should leave the office without rectal examination. In the male both lobes of the prostate gland should be palpated, as well as the right and left seminal vesicles and the retrorectal or presacral space, the ischioanal space, and the cystorectal space. The sphincteric mechanism should be evaluated. In the female palpate the cervix, posterior uterine wall, space of Douglas, ischioanal space, and retrorectal space. Try to palpate both ovaries and both tubes.

In the female, the cul-de-sac (space of Douglas) is lower in comparison to the male; a pelvic abscess formation can be drained by aspiration or incision of the postvaginal vault. In the male, the only way to drain such an abscess is through the posterior rectal wall after performing a careful digital examination to find a better location for the incision or insertion of the trocar.

Chifflet148 observed that there is no fascial barrier between rectum and vagina in women as there is in men between the rectum and the prostate (Denonvilliers’ fascia). Two of the authors of this chapter (GLC and JES) have routinely noted the presence of the septum in dissections of female cadavers, however. Transverse or vertical breaks in the rectovaginal septum often presage the occurence of enteroceles and rectoceles in women. Surgical correction of these problems is often performed by repair of the septum or the insertion of a prosthetic mesh to replace its function. Watson et al.149 advised operative repair of the rectovaginal septum (transperineal with Marlex mesh) in patients with large rectoceles.

Merad et al.150 stated that “prophylactic drainage of the pelvic space does not improve outcome or influence the severity of complications.” Ross et al.151 advised mesorectal excision and radiotherapy for patients with local recurrence of rectal cancer if these tumors are identified with endorectal ultrasound or pathology.

The superior rectal artery (which is the downward continuation of the inferior mesenteric artery) can be ligated with impunity and must be ligated during rectosigmoid resections. The needs of this area for arterial blood will be taken care of by the middle and inferior rectal arteries.

The controversy regarding ligation of the middle rectal artery continues. We quote from Siddharth and Ravo:152

The superior rectal artery is the main blood supply of the rectum. Its branching on the rectum is varied, but it has a rich anastomosis with the other rectal arteries, namely, the middle rectal and inferior rectal arteries. Sudeck’s point is not critical. The middle rectal artery varies in number and origin and is not essential provided the inferior rectal artery is intact.

At a 1986 meeting of the American Association of Clinical Anatomists attended by the senior author of this chapter (JES), the position of the Mayo Clinic (as represented by Oliver Beahrs) was that ligation of the middle rectal artery is not necessary. The Vanderbilt surgeons (represented by H. William Scott, Jr.) disagreed. Anatomically Beahrs was right when he expressed his belief that the middle rectal artery is, for all practical purposes, a prostatic small artery. However Scott, also, was correct when he advised ligation to avoid a possible hematoma.

Our advice is to ligate all tissues on the superior surface of the levator ani at its insertion to the rectal wall; among them should be the middle rectal artery.


Remember that a presacral plexus of veins is located under the endopelvic fascia that covers the sacral periosteum; this is the No. 1 danger zone during low anterior resection. Do not remove the fascia. Work close to the posterior rectal wall. If bleeding occurs use thumbtacks, bone wax, Oxycel (oxidized cellulose), or apply pressure with a surgical pad and lap packs. Leave the endopelvic fascia in situ and always dissect close to the posterior rectal wall. This is the correct plane for avoiding anatomic complications. Another source of bleeding is the median sacral artery. This can be controlled without too much difficulty.

The lateral ligaments of the rectum should be divided using Hemoclips, which work better than suture ligations. These poorly defined anatomic entities are controversial. Are they related to the middle rectal artery or to the nerves of the rectum? From a surgical standpoint it makes no difference whether the ligaments are posterolateral, the middle rectal artery is anterolateral, and the nerves are closer to the ligaments. All these perirectal components should be divided between hemoclips.

To avoid impotence, protect the nerves of the right and left pelvic plexuses. The parasympathetic fibers responsible for erection arise as the nervi erigens (pelvic splanchnic nerves) from the second, third and fourth sacral ventral primary rami. To avoid anatomic complications of neurologic origin, dissection close to the posterior rectal wall is mandatory.

During the perineal part of abdominoperineal resection, remember the location of the following anatomic entities (which are located within the anal triangle) from below upward. There are two surgical zones: the posterior (safe) and the anterior (dangerous).


Posterior: Cut the strong rectococcygeal fascia (anococcygeal ligament) close to and anterior to the coccyx. Cut the puborectalis muscle close to the anorectal wall using an anterior incision.

Anterior: The transverse perineal muscle should be identified and used as a landmark. Always dissect posterior to the muscle to avoid urethral injury. A Foley catheter prior to surgery is essential.

Protect the pudendal nerve by dissecting close to the anorectal wall. Between the surgical anal canal and the prostate in the male or vagina in the female is the perineal body in the general surgeon’s terms, or the area called the perineum by the gynecologist. The anatomic landmarks in this area are the superficial transverse perineal muscle and the urogenital diaphragm. The membranous urethra penetrates the muscle and is protected only by not extending any anterior dissection. The fascia of Denonvilliers will be found by posterior traction and downward pushing of the specimen. The prostate will be found in this area; protect it.

Laterally, of course, ligate the inferior rectal vessels at the medial wall of the ischioanal fossa.

Remember the double fascia of Denonvilliers, the rectovesical or rectovaginal septum. Infiltration of the rectovesical septum by prostatic or rectal carcinoma is well known. Since the anterior leaflet of the fascia is related to the prostate and the posterior leaflet is related to the rectum, ideally the surgeon should explore the space (of Proust) between. But as Healey and Hodge153 stated in their beautiful book Surgical Anatomy, “it is not always easy to find the passage between ‘wind and water.'”


Remember that external hemorrhoids are covered by skin; internal hemorrhoids are covered by anal mucosa.

Anal fissure is a longitudinal crack from the anal verge very near the dentate line. In both males and females it is most common at the posterior anal wall. An anterior anal fissure may be present in females; it may be single or multiple.

Characteristically the fissure has two pockets: one close to the dentate line (anal crypt) and one (subcutaneous pocket) at the anal verge.

Therefore, the fissure is anatomically associated with the subcutaneous ring of the external sphincter.


Anal fistulas (Figs. 18-52, 18-53, and 18-54) extend from an anal crypt in the vicinity of the dentate line (internal or primary opening) to a possible pathway along the lymphatics to the skin, forming the external or secondary opening. Remember Goodsall’s rule on location and pathway of fistula tracts (Fig. 18-54): A posterior external opening has a curved pathway to the internal primary opening in the midline and posteriorly; an anterior external opening has a straight pathway to an opposite internal opening.

With ischioanal (ischiorectal) abscess, early draining with skin incision as close to the anus as possible is mandatory to avoid supralevator space inflammatory process and external fistula far away from the anus.

Fig. 18-52.

Diagrammatic representation of the pathogenesis of anal fistula. A, Infected material from the bowel invades one or more of the anal crypts and the tiny vestigial anal glands. This “primary” process is at the dentate line. B and C, The infection spreads to the perianal and perirectal tissue indirectly by way of the lymphatics or directly by breaks in the continuity of the gland duct structure. D, Abscess formation. E, The abscess spontaneously ruptures or is incised on the perianal skin surface, and the fistulous tract is complete. The skin opening (at “b”) is a “secondary” opening. If the abscess had drained into the rectum (at “a”), the secondary opening would have been at that point. F, Collapse of the abscess leaves the commonly seen narrow fistulous tract. Editor’s note: For instructional purposes, Nesselrod divides anal infection into 3 stages. Stage I is entry of infectious material into anal crypts that funnel the material into anal ducts and glands. In Stage II the perianal tissues, and possibly perirectal tissues, are invaded by infection. Stage III consists of manifestations of infection, including abscess and fistula. (Modified from Nesselrod JP. Proctology in General Practice. Philadelphia: WB Saunders, 1950; with permission.)

Fig. 18-53.

Common locations of anal abscesses and fistulous openings. The ischioanal (ischiorectal) fossae (infralevator spaces) are much more commonly involved than the pelvirectal (supralevator) spaces. From the latter an abscess may extend through the intervening pelvic diaphragm to the lower ischio-anal space. A supralevator abscess may drain spontaneously into the rectum, as at 2′. Subsequent shrinkage of the abscess leads to a chronic fistulous process similar to the tract from 4 to 4′. Occasionally an abscess will drain through its primary opening, and is then a sinus. An ordinary anal fistula is 3 to 3′; 1, 2, 3 and 4 are primary openings; 1′, 2′, 3′ and 4′ are secondary openings. Ab, Abscess. (Modified from Nesselrod JP. Proctology in General Practice. Philadelphia: WB Saunder, 1950; with permission.)

Fig. 18-54.

Perianal lymphatic drainage and Goodsall’s rule. A, Direction of perianal lymphatic plexus. B, Goodsall’s rule: Fistulas with an external (secondary) opening situated posterior to an imaginary line passing transversely through the center of the anus usually have the internal (primary) opening in the midline and posteriorly, so that the tract is curved. When the external (secondary) opening is anterior to the transverse line, the internal (primary) opening is immediately opposite; hence the tract is straight. (Modified from Nesselrod JP. Anal, perianal, perineal and sacrococcygeal sinuses. Am J Surg 1942;56:154-165; with permission.)

Histology of the Colonic Wall

The layers of the wall of the large intestine are essentially similar to those of the wall of the small intestine. The chief differences are: (1) the absence of mucosal villi; (2) the longitudinal muscularis externa in three discrete bands (taeniae) rather than in a continuous cylinder; (3) the presence of epiploic appendices (appendages); and (4) the presence of haustra or sacculations.

The colonic wall has 5 layers:



muscularis externa


muscularis mucosa


Serosa is the visceral peritoneum that covers the colon, abdominal pelvic viscera, and the mesenteries; it does not cover the posterior attachments of the colon. The appendix, cecum, and transverse and sigmoid colons are covered totally by peritoneum.

The muscularis externa layer is composed of the inner circular and outer longitudinal layers, between which lies the myenteric plexus (of Auerbach). The outer layer is not as complete as the inner; it is responsible for the formation of the taeniae coli. Both layers form a network of smooth muscle.

The submucosa is areolar tissue containing veins, lymphatics, the terminal portion of small arteries, and the submucosal plexus (of Meissner).

The muscularis mucosa is a thin network of circular and longitudinal smooth muscle, fixed and interwoven together.

The mucosa is formed by a simple columnar epithelium containing goblet cells.

Dieulafoy’s lesion (cirsoid aneurysm) may cause massive lower gastrointestinal bleeding from a minute submucosal arteriole that bleeds through a punctate erosion in an otherwise normal mucosa. Dieulafoy’s lesions of the colon, rectum, and anal canal have been reported.154

The colon is characterized by three longitudinal muscular bands, the taeniae coli. Between these bands, the longitudinal muscle layer is highly attenuated, reduced to a thickness of less than half that of the circular coat. The bands begin their divergence from one another at the base of the appendix and, as a result, provide a useful guide to the position of the appendix. The taenia libera or free taenia (see Fig. 18-18) is located on the ventral surface of the cecum and the ascending and descending colons. The taenia omentalis (omental taenia) is found posterolaterally. Posteromedially, at the attachment of the mesocolon, is the taenia mesocolica (mesenteric taenia).

Between the taeniae are saccular formations, the haustra, which are characteristic of the colon. The haustra have an enigmatic origin. It is presumed that the length of the taeniae is less than that of the bowel itself, resulting in the formation of the haustra. Indeed, if the taeniae are cut, the colon increases in length.

The taeniae coli are about 1 cm wide. They are most completely developed in the right colon. In the rectum, their architecture changes to broad bands. These are located anteriorly and posteriorly in the rectal wall, and occasionally completely cover the rectosigmoid area and the rectum. Distally, on the rectal ampulla, some anterior longitudinal fibers leave the bowel as the rectourethralis muscle to insert upon the perineal body.

The epiploic appendages (appendices epiploica) (see Fig. 18-18) are another characteristic of the external surface of the colon. They are sessile or pedunculated adipose masses enveloped by the serosa of the colon. They are absent from the rectum. The epiploic appendages can become infarcted or gangrenous, thereby producing severe epigastric or left abdominal pain. These fat-filled pouches can also serve to conceal true diverticula from the wall of the bowel.

The appendages are small pouches of peritoneum that arise from the external surface of the colon. They range from 3 cm to as much as 15 cm in length, depending on the obesity of the patient. The epiploic appendices arise in two rows, chiefly from the lateral and medial intertaenial areas of the transverse and descending colons. They may be present anywhere from the cecum to the proximal rectum.

The surgeon is interested in the epiploic appendages because they may be the sites of diverticula (see Fig. 18-15). Fat may conceal the presence of the diverticulum on inspection, but fecoliths in the diverticula are frequently palpable.41 The appendices are also subject to infarction and torsion; both produce symptoms of an acute abdomen. Goligher41 warned that epiploic appendages should be ligated without traction (see Fig. 18-15). This prevents unintentional pulling of a loop of a long colic artery into the appendiceal neck, and its accidental inclusion in the ligation.

Occasionally the epiploic appendix becomes inflamed secondary to venous thrombosis or torsion, producing a clinical picture similar to diverticulitis; but primary epiploic appendagitis (PEA) is rarely diagnosed. Rao et al.155 reported significant potential to accurately differentiate PEA from diverticulitis and appendicitis at CT scanning if the correct CT techniques are used and radiologists become aware of the CT features of PEA.

Kuganeswaran and Fisher156 reported a giant sigmoid diverticulum of 33 cm, the largest recorded in the literature.



The large bowel epithelium is columnar, without any villi. Characteristically, the flat mucosa of the large bowel is penetrated by tubular glands, forming the crypts of Lieberkühn which open into its lumen. Mucus-secreting goblet cells and some endocrine cells are found within the epithelium of the crypts.

Welton157 proposed that the “human colonic microvascular endothelial cell in culture is a legitimate model for the study of the human colon in the normal and diseased states.”

The muscularis externa of the colonic wall is formed by smooth muscles (longitudinal and circular). These are responsible for peristaltic contraction.

The arrangement of the muscular network of the colonic wall from mucosa to serosa is as follows: the muscularis mucosa just below the colonic epithelium is formed by an inner circular and outer longitudinal layer which are heavily fixed and interwoven together.

The circular layer surrounds the colonic wall; the longitudinal layer covers the circular layer (responsible for formation of the taeniae).

Histology of the Rectum

The upper rectum contains one to four crescentic plicae, the rectal folds or valves of Houston. Typically there are three folds: left superior, right middle, and left inferior. They are encountered by the sigmoidoscope at 4-7 cm, 8-10 cm, and 10-12 cm from the anal verge.

These folds contain mucosa, submucosa, and some muscle. Their positions are marked by a groove on the outer wall, and they do not disappear on rectal distention.

Gordon and Nivatvongs158 state that the Houston valves do not contain all the layers of the rectal wall and biopsy of the valves has minimal risk of perforation.

Histology of the Anal Canal

Musculature of the Wall of the Anal Canal

Two layers of smooth muscle surround the anal canal. The innermost layer is formed by a greatly thickened circular coat which is continuous with the circular muscularis externa of the colon. This is the internal sphincter of the anal canal (see Fig. 18-41). The second smooth muscle layer is composed of longitudinal fibers continuous with the fibers of the taeniae coli.

The downward continuation of the longitudinal muscle is the so-called conjoined longitudinal coat which forms fibers that penetrate both internal and external sphincters and extends to the perianal connective tissue and perhaps to the skin and the anal mucosa. Therefore, the conjoined longitudinal coat is considered the anchor of the surgical anal canal. However, the anatomy of the perianal area is still very controversial. O’Kelly et al.159 stated that in the anal canal both the conjoined longitudinal coat and the internal anal sphincter are specialized sphincteric smooth muscle which relax under the influence of beta-adrenoceptor stimulation.

Shafik101 would divide the longitudinal muscle fibers into three layers that lie between the internal and external sphincters. These layers, the medial, intermediate, and lateral longitudinal muscles, are separated by fibrous connective tissue septa that join to form a “central tendon.” Fibers from this pierce the external sphincter to form the corrugator cutis ani.

The longitudinal muscle fibers prevent separation of the sphincteric elements from each other and also permit a telescopic movement between internal and external sphincters. We witness this in the operating room when the external sphincter rolls back and the internal sphincter rolls forward. Goligher41 suggested that this is why, in the past, surgeons performing sphincterotomy sometimes thought they were cutting the external sphincter but were, in reality, cutting the internal sphincter. They were unknowingly performing the correct operation.

Delancey et al.160 performed cadaveric evaluation of the length of the internal and external anal sphincters. The internal sphincter, located between the anal mucosa and the external sphincter, extends more than a centimeter above the cranial margin of the external sphincter. This region is damaged in fourth-degree obstetric lacerations.

Haas and Fox,161 emphasizing the importance of the perianal connective tissue, state that the anus is anchored by fibers of the conjoined longitudinal coat which bundle the internal and external sphincters, penetrate the perianal fat, reach the pelvic wall and the lower levator fascia, and end at the perianal skin. This peculiar complicated network acts in an antagonistic action to the muscles of the anal canal which are able to overcome such action and close the anus.

The same authors suggest that stretching (transecting the sphincters) may cause loss of elasticity or mobility in the sphincteric anal apparatus with some impairment of function. Maybe so, but we have our doubts; therefore we agree with the anatomy described by Haas and Fox, but we perhaps disagree with the possible pathology.

We quote from Sangwan and Solla162:

The internal anal sphincter, the smooth muscle component of the anal sphincter complex, has an ambiguous role in maintaining anal continence. Despite its significant contribution to resting anal canal pressures, even total division of the internal anal sphincter in surgery for anal fistulas may fail to compromise continence in otherwise healthy subjects. However, recently reported abnormalities of the innervation and reflex response of the internal anal sphincter in patients with fecal incontinence indicate its significance in maintaining continence. The advent of sphincter-saving surgery and restorative proctocolectomy has re-emphasized the major contribution of the internal anal sphincter to resting pressure and its significance in preventing fecal leakage. The variable effect of rectal excision on rectoanal inhibitory reflex has led to a reappraisal of the significance of this reflex in discrimination of rectal contents and its impact on anal continence. Electromyographic, manometric, and ultrasonographic evaluation of the internal anal sphincter has provided new insights into its pathophysiology.

Unlike the longitudinal and circular muscles of the anal canal which are smooth and arise from splanchnic mesoderm, the external sphincter is striated muscle and arises from somatic mesoderm.

The external sphincter is often described as having three separate fiber bundles or loops (see Fig. 18-51): subcutaneous, superficial, and deep. Although in normal individuals these bundles are continuous and show no gross or histologic evidence of separation, it is useful to consider the three parts separately. Shafik163 believes that the three loops together form an efficient anal closure. Any single one of the loops is capable of maintaining continence to solid stools, but not to fluid or gas. The subcutaneous portion surrounds the outlet of the anus, attaching to the perianal skin anteriorly. Some fibers completely encircle the anus.

The superficial portion surrounds the anus and continues within the anococcygeal ligament, which attaches posteriorly to the coccyx. This creates the small triangular space of Minor behind the anus. It is worth noting that the contraction of this part of the sphincteric mechanism pulls the anus posteriorly toward the coccyx, perhaps serving to augment the posterior angulation of the anal canal and thus enhancing the action of the puborectalis. Anteriorly, some fibers insert into the transverse perineal muscles at the perineal body, creating a potential space toward which anterior midline fistulas may point.

The deep portion, like the subcutaneous portion, surrounds the canal, with no obvious anterior or posterior attachments. In Shafik’s view, the deep portion and the puborectalis muscle are a single unit.164

The degree to which these portions, together with their anterior and posterior attachments, are separated from one another has been a source of controversy. We agree that no real separation exists between deep and superficial portions of the sphincter, although an intersphincteric groove is palpable. We question, however, some of the anterior and posterior attachments described by Oh and Kark.165 Our view is shown in Figure 18-55.

Fig. 18-55.

Diagram of the extrinsic muscles of the surgical anal canal. 1, Coccyx. 2, Pubis. 3, Levator ani muscle. 4, Puborectalis muscle. 5, Deep external sphincter. 6, Superficial external sphincter. 7, Subcutaneous external sphincter. 8, Anococcygeal ligament. 9, Anal verge. 10, Rectum. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Lining of the Surgical Anal Canal

There are three histologic regions of the anal canal. The cutaneous zone, up to the anal verge (anocutaneous line), is covered by pigmented skin that has hair follicles and sebaceous glands. Above the anal verge is the transitional zone, which consists of modified skin that has sebaceous glands without hair. It extends to the pectinate line defined by the free edges of the anal valves. Above the line begins the true mucosa of the anal canal (Fig. 18-56).

Fig. 18-56.

The interior of the anal canal showing the rectal columns, anal valves, and anal sinuses (crypts). They form the pectinate line. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The pectinate line is formed by the margins of the anal valves, small mucosal pockets between the 5-10 vertical folds of the mucosa known as the anal columns of Morgagni. These columns extend upward from the pectinate line to the upper end of the surgical anal canal, at the level of the puborectalis sling. They are formed by underlying parallel bundles of the muscularis mucosae. Hollinshead166 noted that the actual junction of stratified squamous and columnar epithelia is usually just above the pectinate line; hence the mucocutaneous line is not precisely equivalent to the pectinate line.

The pectinate line is the most important landmark in the anal canal. It marks the transition between the visceral area above and the somatic area below. The arterial supply, the venous and lymphatic drainage, the nerve supply, and the character of the lining all change at or very near the pectinate line (see Table 18-7).

In spite of all the changes that occur at the pectinate line, pathology, unique to a region 2 cm above and 2 cm below the line, makes a unit of the “surgical anal canal.”6

Histology of the Anal Glands and Anal Papillae

The pockets formed by the anal valves are termed anal sinuses or crypts (Fig. 18-56). In 4-8 of these, especially those on the posterior rectal wall, ducts lead downward and outward, reaching the intersphincteric longitudinal muscle and occasionally penetrating the internal sphincter. Such anal ducts are present in about three-fourths of fetuses121 and in about one-half of adults.167 When present, these ducts appear to be vestigial structures; most lack mucus-secreting gland cells.121 Anal ducts may become infected and provide pathways for anal fistulas. We quote from Klosterhalfen et al.:168 “[A]nal sinuses and anal intramuscular glands are separate anatomic entities. . . For idiopathic, chronic anal diseases anal sinuses have little surgical significance. Anal intramuscular glands should be the anatomic correlate of anal fistulas.”

Where the margins of the anal valves join the anal column, some individuals have small projections, the anal papillae. Usually they cause no symptoms. In a few patients, one or more of these may become hypertrophied and, when large enough, may prolapse.169,170 Schutte and Tolentino171 found papillae in 13 percent of newborns and 46 percent of adults. Most were at the apices of the anal column. They varied from 2 mm or less to a “fibrous polyp” 2 cm in length.

The surgical anal canal is lined by squamous epithelium along the 2 cm below the dentate line and by columnar epithelium above the dentate line. The anal transitional zone has a length 0.5 cm to 2 cm; it is located just above the anal valves. The peculiar innervation of the anal canal is shown in Figure 18-57.

Fig. 18-57.

A diagrammatic representation of the innervation of the anal canal and perianal skin. The nerve endings characteristic of the transitional zone and the anal valve region are shown. (Modified from Keighley MRB, Williams NS. Surgery of the Anus, Rectum and Colon. Philadelphia: Saunders, 1993; with permission.)


Colonic motility is enigmatic. Each segment, part, or region functions in a seemingly independent manner. Practically speaking, the right colon accepts the ileal contents. Multiple phenomena take place: mixing, stirring, kneading, some absorption, and propulsion. These phenomena are manifestations of colonic contraction.

The left colon is a storage reservoir, with further propulsion activities directed toward the anorectum.

Colonic and anorectal function are altered after posterior rectopexy. Mollen and colleagues172 reported that total and segmental transit times doubled, while the effects on anorectal function were not statistically significant. Division of the lateral ligaments did not significantly influence postoperative functional outcome.

Surgery of the Colon

Abdominoperineal Resection

The abdominal phase of an abdominoperineal resection is an extended left colectomy with a presacral dissection to mobilize the rectum, which will be removed in the perineal phase.

The left colon should be mobilized down to the rectovesical or rectouterine fossa by medial and lateral incision of the peritoneal ligaments of the sigmoid colon. The left ureter and left gonadal vessels are visualized; the inferior mesenteric artery and its downward continuation as the superior rectal artery are identified, the duodenum is pushed upward, and the surgeon is ready to enter the pre-sacral area by bloody presacral dissection.

Potential bleeding may occur from the presacral veins which lie beneath the endopelvic fascia. The fascia should not be removed. Use clips for hemostasis; warm packs with Gelfoam will reduce bleeding.

Use long scissors for the perirectal tissues and the fascia of Waldeyer which bridges the sacrum and coccyx to the lower rectum. Blunt dissection with the surgeon’s hand may be used to complete the procedure. Remember the complicated anatomy of the mesorectum.

The tip of the prostate, with Denonvilliers’ fascia, or the tip of the uterine cervix as well as the tip of the coccyx may now be palpated; the hypogastric nerve and hypogastric (pelvic) plexus must be preserved to avoid problems with ejaculation or a neurogenic bladder.

We quote from Havenga et al.173:

Between the rectum and the sacrum a retrorectal space can be developed, lined anteriorly by the visceral leaf and posteriorly by the parietal leaf of the pelvic fascia. The hypogastric nerve runs anterior to the visceral fascia, from the sacral promontory in a laterocaudad direction. The splanchnic sacral nerves originate from the sacral foramina, posterior to the parietal fascia, and run caudad, laterally and anteriorly. After piercing the parietal layer of the pelvic fascia, approximately 4 cm from the midline, the sacral nerves run between a double layer of the visceral part of the pelvic fascia. The relationship between the hypogastric nerves, the splanchnic nerves and the pelvic fascia was comparable in all six specimens examined.

The perineal phase of the abdominoperineal resection encounters the following structures: pudendal vessels (which should be ligated), levator sling (which should be excised widely), and membranous urethra of the male (in which a Foley catheter has been placed prior to surgery). Use sharp dissection to separate the prostate from the lower rectum. The perineum should be partially or completely closed; a suction drain is advisable.

Gorski et al.,174 in a review of the literature, reported 276 cases of retrorectal malignancy. Table 18-8 indicates the tumor types.

Table 18-8. Malignant Retrorectal Tumors

  Total (%)
  Chordoma 140 (50.7)
  Teratocarcinoma 12 (4.3)
  Neurofibrosarcoma 12 (4.3)
  Neuroblastoma 9 (3.3)
  Ependymoma 7 (2.5)
  Osteosarcoma 7 (2.5)
  Ewing’s tumor 6 (2.2)
  Chondrosarcoma 9 (3.3)
  Myeloma 6 (2.2)
  Metastatic carcinoma 21 (7.6)
  Liposarcoma 3 (1.1)
  Hemangiosarcoma 6 (2.2)
  Fibrosarcoma 5 (1.8)
  Leiomyosarcoma 9 (3.3)
  Unknown origin 9 (3.3)
  Lymphoma 8 (2.9)
  Pericytoma 4 (1.4)
  Carcinoid 1 (0.4)
  Other 2 (0.7)
Total 276 (100)

Source: Modified from Gorski T, Khubchandani IT, Stasik JJ, Riether R. Retrorectal carcinoid tumor. South Med J 1999; 92:417-420; with permission.

Surgery of Colonic Trauma

Brasel et al.175 stated that “simple suture or resection and anastomosis at the time of initial exploration is the dominant management method for penetrating colonic trauma.”

Curran and Borzotta,176 after evaluating 5,400 cases of colon injury, advised primary repair of colon injury, but in selected cases advised colostomy.

 Read an Editorial Comment

Colon Resection: Open Method

Specific knowledge of normal and abnormal anatomy is mandatory for the uncomplicated resection of part or all of the large bowel. The surgeon must be familiar with the peritoneal attachments, the blood supply, and the lymphatic pathways in this area where cancer is so common. Lack of knowledge of the peritoneal reflections and the mesenteries will produce technical problems in the operating room, such as poor mobilization or tension at the anastomosis, with postoperative leakage and peritonitis. Inadequate knowledge of the blood supply and its variations can be catastrophic. Ignorance of the pathways and patterns of lymphatic distribution may result in a fatal disseminated disease. Therefore, a brief summary with emphasis on specific anatomy and some of the anatomic entities involved with techniques of colon surgery is necessary.

We agree with the statement of McDaniel et al.177:

The distribution of regional lymph node metastases in carcinomas of the cecum, ascending colon, and transverse colon follows the vascular distribution in the ileocolic mesentery, ascending mesocolon, and transverse mesocolon. The location of these metastatic nodes can be recognized on CT scans when the anatomy of the vessels in the ileocolic mesentery and mesocolon is well understood. This knowledge is important in the preoperative staging of carcinomas of the colon for curative surgery and in the early detection of recurrent nodal disease after curative surgery.

Lane et al.178 advised the following procedures for cecal diverticulitis:


If solitary diverticulum is present, only diverticulectomy

With multiple diverticuli and cecal phlegmon or if neoplastic disease cannot be ruled out, right hemicolectomy is recommended even if the colon is not prepared

McDonald et al.179 reported that the size the colonic polyp plus the presence of high-grade dysplasia are key factors to be considered in formulating a treatment plan.

 Read an Editorial Comment

Opinions vary on the value of total mesorectal excision (TME). Lopez-Kostner et al.180 reported that total mesorectal excision is not necessary for cancers of the upper rectum. Heald et al.181 concluded that “precise total mesorectal excision from above appears oncologically superior to abdominoperineal resection.” Havenga et al.182 stated that TME and extended lymphadenectomy gives better recurrence and survival results than conventional surgery in the treatment of primary rectal cancer.

Very rarely colonic adenocarcinoma may arise without a gross mucosal lesion. Such a case was presented by Wimmer et al.183 when the tumor developed in the underlying submucosa with negative colonoscopy.

Hayashi et al.184 reported the no-touch isolation technique without surgical manipulation may be useful in preventing cancer cells from being shed into the portal vein during resection of colorectal cancer.

 Read an Editorial Comment

Anatomy of Exposure and Mobilization

Condon and Lamphier185 suggested the concept of three layers of structures in the abdominal cavity. The first layer is the digestive tube with its nerves and vessels. The second layer contains the kidneys, adrenals, ureter, aorta, and inferior vena cava. The third layer is the transversalis fascia lining the parietal muscles.

Figure 18-23 shows the extent of colectomy recommended for cancer at various sites in the colon. A standard right colectomy is essentially a midline resection; it includes a few centimeters of the terminal ileum, the cecum, the right colon, and the proximal half of the transverse colon. These are the segments served by the ileocolic, right colic, and right branches of the middle colic arteries (Fig. 18-23B). Following mobilization of the right colon, expose the kidney, ureter, spermatic vessels, inferior vena cava, aorta, iliac vessels, duodenum, pancreas, and retroperitoneal muscles.186

There are four reasons for choosing the standard right colectomy rather than a lesser resection:

(1) the lymphatic drainage makes a lesser resection inadequate

(2) the proximal colon is more difficult to use for anastomosis because it is not completely covered by peritoneum and may have attached fat or membranous veils

(3) the ileum has a good blood supply and is less subject to suture-line necrosis than is the colon

(4) the operation is easier, since the hepatic flexure is easy to mobilize

An exception may be made for benign tumors. In these cases, as much of the terminal ileum as possible should be preserved to prevent diarrhea resulting from a lack of resorption of bile salts.

Koea et al.187 reported that extended resection by right colectomy and en bloc duodenectomy or by en bloc pancreaticoduodenectomy for localized primary colon carcinoma invading the pancreas or duodenum are safe procedures associated with prolonged survival time.

In left colectomy, the lowest sigmoid artery may be too short for adequate sigmoid resection. Dixon188 suggested a method to overcome this problem (Fig. 18-58): the last sigmoid artery, together with a small portion of the proximal stump, may be sectioned to permit an anastomosis free from tension. The sigmoid artery and the problem of the short transverse mesocolon should be evaluated before resection is started.

Fig. 18-58.

A, Following resection of the sigmoid colon, the last sigmoid artery may be too short to permit the required anastomosis. B, The last sigmoid artery together with a small portion of the proximal stump may be sectioned to permit an anastomosis free from tension. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Mobilization of the left colon for a successful pull-through operation was discussed by Barnes.189

Sharma and Klaasen190 reported that the procedure of choice for elderly patients with colonic carcinoma adherent to the duodenum is duodenal seromyectomy.

Eisen et al.191 made the following recommendations about intussusception in adults:


Colectomy should be performed without reduction of intussusception because malignant tumor is often the cause of intussusception in adults

Small bowel intussusception should be reduced only if there is benign disease or if resection is likely to result in short bowel syndrome

Total proctocolectomy with pouch ileoanal anastomosis is used to treat patients with ulcerative colitis and familial polyposis. Michelassi and Hurst192 reported that the procedure allows patients to live without a permanent stoma and experience a high degree of continence and an acceptable number of daily bowel movements.


Below we list several types of colostomy; some are included entirely for historical interest. The procedures in italics are those that are currently being performed, and here we will describe only those. End colostomies with Hartmann’s pouch or mucous fistula have limited application, but ReMine and Dozois193 and Bell194 have reaffirmed their usefulness.


1. Cecostomy


a. Cecal exteriorization

b. Tube cecostomy

2. Loop colostomy

3. Double-barreled colostomy (Bloch-Paul-Mikulicz)

4. Tube and marsupialization colostomy

5. Interrupted colostomy


a. End colostomy with Hartmann’s pouch

b. End colostomy with mucous fistula

We quote from Bardoel et al.195 on their cadaveric studies:

[T]he rectus abdominis muscle is ideally suited for the construction of a stoma sphincter. The muscle is located in the appropriate anatomical location for stoma creation; it has a long vascular pedicle; and the preserved, segmental intercostal innervation pattern allows the muscle to be tailored and mobilized so as to completely wrap a fecal stoma without significant muscle denervation.

They plan future “functional” studies to see if the rectus abdominis muscle can be successfully trained to become fatigue-resistant.

Tube Cecostomy

A double purse-string absorbable suture at the anterior taenia is the first step in a tube cecostomy (Fig. 18-59A). The cecum must be fixed to the abdominal wall without leaving a dead space between the stoma and the wall. Removal of the appendix is desirable. The appendiceal stump may be used to insert the Foley catheter before closing it. A three-way, large Foley catheter can be used for irrigation and suction as well as for injection of antibiotics when necessary.

Fig. 18-59.

A, Tube cecostomy. The dashed line surrounded by the purse string suture indicates the incision for the insertion of the Foley catheter. B, Loop colostomy. The loop of colon has been brought through the incision, held in place by a glass or plastic rod with rubber tubing connecting its ends. C, End colostomy. The limbs of the colostomy should be brought to the surface through separate incisions. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Although cecostomy is less popular than it once was, we have used the procedure successfully several times for decompression and for suture-line protection. With a short transverse mesocolon in an obese patient, cecostomy is the ideal operation.

Tschmelitsch et al.196 compared loop colostomy/ ileostomy (C/I) and tube cecostomy (TC) in patients who had undergone low anterior resections for rectal cancer. They reported that the rate of anastomotic leaks, fecal peritonitis, reoperation for anastomotic leak/fistulas, permanent colostomies, and postoperative mortality was not significantly different between the groups, and concluded, “a C/I for protection of the anastomosis did not improve outcome significantly as compared with a TC. With a properly fashioned TC and adequate postoperative management a second operation (for colostomy closure) can be avoided and the overall hospital stay is significantly reduced.”

Loop Colostomy

A loop colostomy (Fig. 18-59B) is feasible only in the transverse or sigmoid colon because a mesentery is required. If the transverse mesocolon is short, mobilization of the hepatic and splenic flexures will provide a more mobile loop.

At the hepatic flexure, the colon is attached to the second part of the duodenum and the lower pole of the right kidney. There is no right phrenocolic ligament. At the splenic flexure, the colon is attached to the lower pole of the left kidney and to the diaphragm by the phrenocolic ligament. Careful division of the ligament is important to avoid injury to the spleen.

With obstruction of the left colon, the transverse colon is dilated; unless adhesions hold it in the pelvis, the transverse colon is high. Identification of dilated intestine can be a problem. The presence of taeniae distinguishes the colon from the small intestine, but with dilatation, the taeniae tend to become less visible. The left colon has many more epiploic appendages than does the transverse colon. Transilluminate the mesocolon if possible. The middle colic and marginal arteries should be identified and preserved. Keep in mind that the blood supply of the colon is not as rich as that of the small intestine.

Sigmoid loop colostomy is, for practical purposes, left colon colostomy. The stoma should be located at the junction of the descending and sigmoid colons so that the peritoneal fixation of the descending colon will protect the proximal stoma from prolapse.

Sakai et al.197 stated that transverse colostomy and loop ileostomy are equally safe procedures for temporary fecal diversion, and recommend further comparative study of these procedures for patient satisfaction and quality of life.

End Colostomy

In end colostomy (Fig 18-59C), the stoma should be in the most proximal sigmoid loop, very close to the distal end of the descending colon. Externally, the stoma should be in the left lower quadrant somewhere between the umbilicus and the left anterior superior iliac spine. A site which is the mirror image of McBurney’s point is usually ideal. The considerations are: (1) mechanical, for optimum placement of the colostomy bag, and (2) anatomic, for closure of the left paracolic gutter to avoid obstructing the small intestine.

The proximal cut end of the sigmoid colon should be brought 5 to 8 cm above the surface of the skin and observed for color. The colostomy should be loose enough to permit the insertion of one finger. Be sure to close the space between the superficial fascia and the aponeuroses of the external oblique muscle and the anterior rectus sheath. If this space is not closed, it will become the site of a postoperative hernia.

The distal rectal stump can be closed to form Hartmann’s pouch,198 or it may be brought to the surface to form a “mucous” fistula. With modern knowledge of colon anatomy and with good colonic preparation, Hartmann’s procedure is performed less frequently today. ReMine and Dozois193 defended the use of this procedure, pointing out its safety and good long-term survival as well as its advantage in permitting subsequent restoration of bowel continuity, especially if the resection is above the line of peritoneal reflection. Another advocate of the Hartmann procedure was Bell,194 who performed a typical Hartmann procedure or used a distal mucous fistula.

However Belmonte et al.199 reported that colostomy closure after Hartmann’s procedure is associated with significant length of hospitalization and morbidity, and leaves one third of patients with permanent stomas.

Rosoff200 introduced the question of the value of the critical point of Sudeck (see “Critical Point of Sudeck” earlier in this chapter). Ligation of the inferior mesenteric or superior rectal artery almost never results in necrosis of the left colon.35,41

The surgeon should perform the colostomy he or she judges to be best. The critical point of Sudeck may be ignored; however, the surgeon must observe the color, bleeding, and arterial pulsation in the segment to be anastomosed or to be brought outside. No one need be critical if the surgeon chooses to do a loop colostomy or an end colostomy with a pouch or mucous fistula.

Laparoscopic Colectomy

Liu et al.201 concluded that laparoscopic intestinal surgery is both feasible and safe in selected patients with inflammatory bowel disease. Technical advances have enabled laparoscopic lower anterior resection to transect the lower rectum in the same way as is done with laparotomy.202 Reissman et al.203 concluded that the feasibility of laparoscopic and laparoscopy-assisted colorectal surgery has been well established. Leung et al.204 stated that the immediate and medium-term results of laparoscopy-assisted resection of rectosigmoid carcinoma are promising.

There is controversy as to the safety of laparoscopic colectomy for malignancy. Bouvet et al.205 advocate laparoscopic colectomy for carcinoma of the colon, as a sound oncologic procedure.

Targarona et al.206 reported that laparoscopy can help disseminate aggressive tumors and therefore should be reserved for diagnostic and staging procedures or for treatment of low-grade malignant tumors.

Tomita et al.207 considered laparoscopic surgery for treatment of colorectal cancer, including indications and contraindications. They stated that more research on surgery for colorectal carcinoma, as well as advances in laparoscopic technology, will be needed for future successful application of laparoscopic surgery to the treatment of colorectal carcinoma.

Stocchi and Nelson208 recommended worldwide trials for laparoscopic colectomy to treat carcinoma of the colon. We agree.

We quote the wise counsel of Marubashi et al.209:

[W]hile laparoscopy-assisted colectomy (LAC) is a safe and minimally invasive form of surgery from which faster recovery can be expected, it should only be performed in carefully selected patients with advanced colorectal carcinomas because most of the major complications associated with LAC using current devices and techniques may be prevented by performing traditional open surgery.

The authors of this chapter consider laparoscopic procedures for the colon to be evolutionary. Therapeutic laparoscopic resection of carcinoma of the colon should be restricted to prospective and randomized trials until there is enough hard data to indicate its safety.

The Colon as a Reconstructive Entity

Bussi et al.210 advised that pharyngocoloplasty after total pharyngolaryngoesophagectomy for carcinoma of the hypopharyngoesophageal junction is a reliable treatment option when gastric pull-up is considered risky or is contraindicated.

Surgery of the Anorectum

Cloacal Exstrophy

Soffer et al.211 reported on the surgical treatment of cloacal exstrophy:

During neonatal repair, a colostomy should be formed incorporating all pieces of colon, no matter how small. With time, most patients will be able to form solid stool, and a pull-through should be undertaken if that ability exists. Decisions regarding genitourinary reconstruction should be made only after the gastrointestinal plan is established to achieve the optimal use of available bowel.

Howell et al.212 and Ricketts et al.213 advised the following about the surgery and treatment of exstrophy of the cloaca:


1. Preservation of the entire hindgut, including the terminal colon

2. End colostomy using the tailgut

3. Primary closure of omphalocele, if possible at once or in stages

4. Reapproximation of the bladder halves

5. Closure of the exstrophic bladder

6. Bladder augmentation at a later stage, using either bowel or stomach

7. Conversion to female gender

8. Staged abdominoperineal pull-through of the fecal stoma, or conversion to a continent ileostomy, depending on the anatomy and available bowel

Anatomic Guidelines

No anorectal procedure should be undertaken without digital and sigmoidoscopic examination. The following is the anatomy as encountered by the examiner’s finger or as seen in the sigmoidoscope. Digital examination should always precede sigmoidoscopy. It relaxes the sphincters and reveals any obstruction that might be injured by the sigmoidoscope.

The anal verge (Fig. 18-60) separates the pigmented perianal skin from the pink transition zone. The verge is the reference line for the position of all other structures encountered.

Fig. 18-60.

Diagram of anorectal landmarks for sigmoidoscopic examination. Patient in knee-chest or knee-elbow position. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

When the gloved and lubricated index finger is inserted so that the distal interphalangeal joint is at the anal verge (Fig. 18-61A), the subcutaneous portion of the external (voluntary) sphincter is felt as a tight ring around the distal half of the distal phalanx. The fingertip should detect the pectinate line of anal valves that lies about 2 cm above the anal verge. The anal columns (Morgagni) above valves also may be felt. External hemorrhoids, polyps, and hypertrophied anal papillae in this region are readily detected.

Fig. 18-61.

Digital examination. A, Distal interphalangeal joint at the anal verge. Hemorrhoids can be detected at this stage. B, Middle interphalangeal joint at the anal verge. C, Metacarpophalangeal joint at the anal verge. The tip of the finger is at or just above the inferior rectal valve. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

Further insertion of the finger to the level of the middle interphalangeal joint (Fig. 18-61B) brings the first joint to the anorectal ring formed by the deep component of the external sphincter, the puborectalis loop, and the upper margin of the internal sphincter. The ring is felt posteriorly and laterally, but not anteriorly.

Still further penetration of the finger to the level of the metacarpophalangeal joint (Fig. 18-61C) allows the distal phalanx to enter the rectum. The left lower rectal fold may often be touched. At this point the pelvirectal space lies lateral and the rectovesical or rectovaginal space lies anterior. Further anterior to the rectum one can palpate the prostate gland in men and the upper vagina and cervix in women.

The sigmoidoscope should be inserted, aimed at the patient’s umbilicus (Fig. 18-62A). At 4 cm from the anal verge, the tip will be at the anorectal ring. With the obturator removed (Fig. 18-62B), the left lower rectal fold should be visible. At about 8 cm from the verge, the middle rectal fold may be seen. This is the level of the peritoneal reflection. The superior rectal fold is reached at 10 to 12 cm, and beyond this, passage of the instrument is easy.

Fig. 18-62.

Sigmoidoscopic examination. A, The instrument is directed toward the umbilicus. The tip is just past the anorectal ring. B, With obturator removed, the instrument is passed by direct observation. The tip shown here is almost up to the middle rectal valve. (Modified from Skandalakis JE, Gray SW, Rowe JS Jr. Anatomical Complications in General Surgery. New York: McGraw-Hill, 1983; with permission.)

The rectum, despite its name (which means “straight”), is not straight. The perineal flexure is a posterior bend just above the anorectal ring. Beyond this is the sacral flexure as the rectum follows the curve of the sacrum anteriorly. Lateral flexure to the left occurs at the first and third (inferior and superior) rectal folds and to the right at the middle rectal fold. The most dangerous area is between the middle and superior rectal folds, just above the peritoneal reflection. This is the area in which perforation by the sigmoidoscope may occur.

Saunders et al.214 investigated the thesis that colonoscopy is a more difficult procedure in Western than Oriental patients by comparing anatomic findings at laparotomy of 115 Western (Caucasian) and 114 Oriental patients:

Sigmoid adhesions were found more frequently in Western (17%) compared to Oriental (8%) patients, P = 0.047. A descending mesocolon of 10 cm occurred in 10 (8%) Western patients but only 1 (0.9%) Oriental patient, P = 0.01. The splenic flexure was more frequently mobile in Western patients (20%) compared to Oriental (9%) patients, P = 0.016. In 29% of Western patients the mid-transverse colon reached the symphysis pubis, or lower when pulled downward, in contrast to 10% of Oriental patients, P <0.001. There was no significant difference in total colonic length comparing Western (median = 114 cm, range 68-159 cm) to Oriental (median = 111 cm, range 78-161 cm) patients. Western patients have a higher incidence of sigmoid colon adhesions and increased colonic mobility when compared to Orientals.

Anorectal Fistulas

Most anal fistulas are complications of anorectal abscesses, and most such abscesses arise in the anal glands that open into the base of the anal valves. Incision and drainage of an abscess predisposes to fistula formation.107,215,216

Conservative treatment is called for here. Fistulectomy (radical removal of the fistula) with extensive section of the sphincters will produce incontinence. Fistulectomy is safe only where preservation of the anal ring can be ensured; otherwise fistulotomy is the procedure of choice.

Anorectal Abscesses

With anorectal abscesses, early drainage by a long, radial incision is the correct technique to avoid recurrence. An excellent presentation by Nomikos217 advised accurate anatomic localization of anorectal abscesses.

Anal Fissure

An anal fissure, which has a boatlike shape, is a painful ulcer in the anal canal, and may be superficial or deep. In children it is usually secondary to constipation; in adults it is most likely idiopathic (except in cases of preexisting disease such as Crohn’s disease, AIDS, or anal tuberculosis, or predisposing factors such as anal intercourse) or iatrogenic.

The accepted treatment is partial internal sphincterotomy (lateral or midline) or pharmacologic sphincterotomy (topical application of glyceryl trinitrate or intrasphincteric injection of botulinum toxin). A superficial vertical mucosal tear secondary to severe and tight spasm of the anal sphincters may be treated by four finger dilatation and very rarely by lateral internal sphincterotomy. Argov and Levandovsky218 state that open lateral sphincterotomy performed in an ambulatory setting is the gold standard treatment for chronic anal fissure due to the very high long-term success rate and negligible occurrence of complications. Fecal or flatus incontinence is reported with lateral sphincterotomy; according to Sharp219 the incidence is 0-35%.


Rule out epidermoid carcinoma as well as venereal diseases prior to any dilatation.

Pilonidal Disease

Pilonidal disease produces a cyst or sinus with hair. While it may be of congenital origin, it is most likely that this enigmatic condition is acquired. The treatment is also controversial. If abscessed, incise and drain. Incision and curettage is favored by da Silva220 with regard to morbidity, healing, recurrence, and cure. If a persistent sinus remains, then excision of the sinus in toto, leaving the wound open, is perhaps the procedure of choice.

Malignant degeneration of pilonidal cysts was reported by Davis and colleagues.221

 Read an Editorial Comment


Hemorrhoids may be internal or external.

Good knowledge of technique and surgical anatomy is essential when treating hemorrhoids.

Rectal Prolapse

Rectal prolapse (Fig. 18-63) is a pathologic entity in which some or all layers of the rectum protrude mucosal side out from the surgical anal canal.

Fig. 18-63.

Development of rectal prolapse. A, Sagittal section through the normal pelvis. B, Start of rectal prolapse. The proximal segment (intussusceptum) is telescoping into the distal segment (intussuscipiens). C, Intussusceptum at the pectinate line. D, Intussusceptum protrudes through the anus. (Modified from Skandalakis LJ, Gadacz TR, Mansberger AR Jr, Mitchell WE Jr, Colborn GL, Skandalakis JE. Modern Hernia Repair. New York: Parthenon, 1996; with permission.)

There are several approaches and procedures which may be employed:


Perineal approach with amputation of the sigmoid colon and rectosigmoid anastomosis

Thiersch perineal approach

Rectopexy by abdominal approach or low anterior resection. Herold and Bruch222 reported good results with laparoscopic retropexy in patients with rectal prolapse and morphologic outlet constipation.

Rectal sling of Ripstein. Schultz et al.223 reported low mortality and recurrence, but high complication rate, with Ripstein rectopexy. Increased constipation occurred with improved continence in some patients, especially those with internal rectal intussusception.

Ivalon sponge wrap of Wells

Transanal excision. Transanal excision for low rectal cancer in early stage disease (carcinoma in situ and T1 lesion) with favorable histology was reported safe and effective by Blair and Ellenhorn.224

Agachan et al.225 advised perineal rectosigmoidectomy with levatoroplasty for the treatment of rectal prolapse.

Pikarsky and colleagues226 stated that the outcome of surgery is similar in cases of primary or recurrent prolapse, with the same surgical options valid in both scenarios. They reported an overall success rate of 85.2 percent.

For further information and techniques, please consult Modern Hernia Repair.227

Anorectal Cancer Surgery

We quote from the beautiful anatomical presentation of Enker et al.112:

In the posterior visceral compartment of the pelvis, four anatomic layers determine which surgical planes are available for sharp dissection within the pelvis. These layers are 1) the visceral layer of the pelvic fascia; 2) the parietal layer(s) of the pelvic fascia; 3) the medial layer of the internal iliac vascular adventitia; and 4) the lateral layers of the internal iliac vascular adventitia and the obturator spaces. The bilateral paravesical spaces may also provide ventral access to the deep pelvis in rare instances.

The circumstances defining the use of each of the planes include:


Primary rectal cancers without extension to the visceral layer of the pelvic fascia.

Primary rectal cancers with extension to the visceral layer of the pelvic fascia.

Primary rectal cancers with extension beyond the visceral layer of the pelvic fascia, but resectable with negative (uninvolved) margins. Examples include 1) attachment to or invasion of the so-called lateral ligament (which actually contains no ligamentous tissue), 2) localized involvement of the middle rectal artery or the internal iliac vessels, and 3) middle rectal or internal iliac node involvement.

Recurrent rectal cancer due to previously unresected internal iliac or mid-rectal artery node involvement.

Obturator space involvement by either primary or recurrent disease, or by nodal disease presenting as a resectable mass.

Direct spread to the levator ani by a laterally located, low-lying rectal cancer.

Under these differing circumstances, the knowledge and use of these planes can make a difference between successful resection of a primary or recurrent rectal cancer and a negative outcome, such as uninvolved circumferential margins or the death of the patient due to unresectable pelvic disease.

Approaches and procedures include:


Anterior resection if margin distal to the tumor is 5 cm

Abdominoperineal rectal excision

Sphincter-saving resection


The surgical treatment of anorectal cancer is controversial; most controversial are the sphincter-saving procedures. As Dehni and colleagues228 wisely stated, “Conservation of the sphincter mechanism should never compromise the oncologic outcome of surgery and the method of neorectum construction must provide acceptable function for patients.”

Several factors influence the choice of procedure:


Anatomic location of tumors (above or below dentate line)

Pathology (epithelial or nonepithelial)



Involvement of other anatomic entities related to the anorectum such as parts of the urogenital system (including lymph nodes)

Age and general condition of patient

Distal metastasis

Read and Kodner229 reported that restoration of intestinal continuity following proctectomy for cancer is now possible due to better understanding of the spread of rectal cancer as well as of pelvic physiology. Occasionally a diverting loop ileostomy may be necessary. Di Matteo et al.230 reported better anal sphincteric continence with high or low colorectal anastomosis than with coloanal anastomosis. Madoff et al.231 reported that sphincter replacement by electrically stimulated skeletal muscle neosphincter and artificial anal sphincter provide a continent option for patients with end-stage fecal incontinence and those requiring abdominoperineal resection.

Locally excised T1 rectal cancer with negative surgical margins should receive adjuvant chemoradiation.232 Adequate shielding of the anal sphincter during chemoradiation is essential when a sphincter-preserving procedure is performed for low rectal cancer.233

This is the era of the sentinel lymph node in cancer surgery not only in the breast, but also in other anatomic entities such as the colorectum.234 Saha et al.235 reported that sentinel lymph node mapping can easily be performed in colorectal cancer patients, with an accuracy rate of more than 95%.

Anatomic Complications of the Colon

Colon Resection

Anastomotic Complications

Schrock and colleagues236 investigated leakage in 1,703 colonic anastomoses. Disruption of the anastomosis occurred in 4.5 percent. The mortality among patients with leakage was 36.8 percent in enterocolic anastomoses, 36.8 percent in left colocolostomies, and 26.5 percent in left colorectal anastomoses. The incidence of anastomotic dehiscence increased with the age of the patient, anemia, prior radiation therapy, infection, hypotension, intraoperative transfusion, and carcinoma at the line of resection. Anastomotic dehiscence occurred in 1.7 percent of colonic anastomoses performed under ideal circumstances and in 6.7 percent of anastomoses performed with one or more unfavorable conditions present. Neither the experience of the surgeon nor the details of the technique showed significant differences.

One- or two-layered anastomosis, omental wrapping, proximal colostomy, preoperative cleansing, and antibiotics all reduce leakage, but they do not eliminate it entirely. Welch et al.237 stated: “It could be assumed that if barium enemas were given almost immediately after every colonic anastomosis, 100 percent of them would show some leakage.” While we consider this view too pessimistic, we recognize that leakage of colonic anastomoses is a much more frequent problem than leakage of small intestinal anastomoses. Tocchi et al.238 reported that omentoplasty by means of intact omentum reduced the severity of anastomotic leakage following rectal resection, although the incidence of anastomotic disruption was not affected by the use of omentoplasty. Moran and Heald239 stated that with optimum technique and the use of a combination of linear and circular staplers, leakage is a problem largely confined to anastomoses within 6 cm of the anal verge.

Vogel and Klosterhalfen125 stated that the dorsocaudal sector of the rectal ampulla is an area of the rectum deficient in blood supply. This is due to the separation of the inferior and superior rectal arteries by the pelvic diaphragm.

Patricio et al.240 reported that in patients older than 50, the hypogastric (internal iliac) arteries provide less blood to the anorectum. These authors hypothesize that a deficient blood supply is one of several factors responsible for leakage at a colorectal anastomosis. Because the inferior mesenteric artery is ligated, this hypothesis is very sound.

Jansen et al.241 reported an ingenious method to close the anastomosis. Two ring magnets were used to produce end-to-end approximation of the layers of the gut wall. Following healing of the resection, the magnets pass out of the body. The report covered 21 patients in whom the magnetic anastomosis was employed. In two (9.5 percent), dehiscence took place.

In spite of the fact that leakage cannot be entirely prevented, principles of good technique must be observed:


Anastomose only healthy colon

Avoid tension on the anastomosis

Ensure a good blood supply at the cut ends of both bowels. Good color, bleeding at the cut edge, and pulsation of mesenteric vessels are indicators of a good blood supply.

Avoid intramural hematoma. If a hematoma appears to be spreading, do not hesitate to resect more colon. Good hemostasis is absolutely mandatory; ligate all vessels.

Clean fat, epiploic appendices, and mesentery from both proximal and distal edges of the anastomosis. Clean not more than 1 to 1.5 cm.

Preserve an adequate lumen

Close mesenteric gaps if possible. If they cannot be closed, open them as wide as possible.

Avoid sepsis by good preparation of the colon prior to surgery. Careful selection of appropriate antibiotics is mandatory before and after surgery.

If there is doubt about the anastomosis, perform a proximal colostomy. This is a life-saving procedure.

In addition to postoperative leakage, colon anastomoses are subject to bleeding and hematoma formation as well as ischemia, necrosis, and fistula formation.

Early obstruction at the anastomotic site can result from edema or excessive inversion; late obstruction can be caused by recurrent carcinoma. Hernia through a mesenteric defect is a possible complication. Mesenteric defects should be closed or enlarged to avoid the possibility of strangulated transmesenteric hernia.



The circular muscle is probably the strongest stroma of the intestinal wall.

The single Cheatle slit (Fig. 18-64) is made by incising longitudinally the antimesenteric border of the narrow intestinal segment.

The double Cheatle slit (Fig. 18-65) is made by incising both antimesenteric borders of the proximal and distal segments. The more mobile bowel end is rotated approximately 180°. The end of the Cheatle incision is sutured to the cutting edge (rim) of the bowel. The same technique is used to approximate the remaining Cheatle incision.

Fig. 18-64.

The single Cheatle slit.

Fig. 18-65.

The double Cheatle slit.

Vascular Injury

Several vessels are subject to injury with consequent hemorrhage.


In right colectomy: the superior mesenteric artery, superior mesenteric vein, branches of the middle colic vein extending to the first part of the duodenum, and marginal artery

In left colectomy: the left gastroepiploic artery, left splenic polar artery, vessels of the renocolic ligament that may connect Gerota’s fascia of the kidney, left gonadal vein, marginal artery, and middle colic artery

In low anterior resection: most of the vessels endangered by left colectomy (left iliac vein, veins of the left pelvic wall, and presacral venous plexus which is beneath the endopelvic fascia)

In addition, hematomas, abscesses, or fistulas are possible.

Organ Injury


Graham and Goligher242 reviewed 1,605 patients, 14 of whom suffered ureteric injury (one bilateral) during rectal excision. A higher frequency was reported by Andersson and Bergdahl,243 who found 30 ureteric injuries among 801 patients. The two sides are almost equally affected. Injuries occur either at the level of ligation of the inferior mesenteric vessels or, more often, just above the level of the lateral rectal ligaments. Graham and Goligher242 observed that the left ureter lies 1.25-5 cm or more lateral to the inferior mesenteric artery. Where the ureter is close to the artery, it can inadvertently be hooked with the vessel and ligated. This must be avoided by pushing the left ureter laterally and backward following incision of the left lateral peritoneum.

If the tumor is not adherent, the ureter can be retracted blindly from the lateral ligament. However, the only reliable method for sparing the ureter is by exposing it from the brim of the pelvis almost to the bladder. If the ureter is infiltrated by the tumor, it will, of course, have to be resected.

The ureter may survive ligation or crushing injury if the ligation or clamp is removed at once. If the ureter does not appear normal, the injured portion must be resected. Where resection has occurred, the choice is between end-to-end anastomosis of the ureter or reimplantation of the ureter into the bladder. The latter procedure is the easier of the two.244

Duodenum, Liver, Pancreas, and Spleen

In right colectomy, mobilization of the hepatic flexure and the right transverse mesocolon may accidentally injure the duodenum, liver, or pancreas. In left colectomy, the duodenum, spleen, and tail of the pancreas may be injured during mobilization of the splenic flexure and the left transverse colon.

Prostate, Seminal Vesicle, Vagina, Bladder, and Rectum

The prostate, seminal vesicle, vagina, and bladder can be injured during low left colon resections. In 11 of 801 patients reviewed by Andersson and Bergdahl,243 trauma to the bladder appeared following operation. Voiding dysfunction was described by Kirkegaard and associates245 in 20 patients who underwent low anterior resection of midrectal carcinoma. These conditions were the result of autonomic nerve injury, and did not arise from direct trauma to the bladder.

Autonomic nerve preservation in association with total mesorectal excision for cancer of the rectum minimizes sexual and urinary dysfunction, according to Havenga et al.246

We quote Mancini et al.247:

Traditional rectal cancer surgery has been burdened with a high rate of sexual and urinary dysfunctions due to intraoperative injury or the cutting of the sympathetic and/or parasympathetic nerves…The parasympathetic nerve trunks were those most often damaged because of perineural tumor spreading. Partial to complete sexual impotence was observed in 44% of the patients and surprisingly, preoperative dysfunctions were detected by means of the multidisciplinary approach in one third of these patients. Therefore, only 30.5% of the patients presented with strictly postoperative sexual impotency, above all, those who had undergone high-dose preoperative chemoradiation for T3 or T4 middle to low rectal cancer.

Inadequate Procedure

A leaking anastomosis is the most obvious sign of an inadequate procedure, but, as we have seen, it is unrealistic to expect every anastomosis to be free from leakage. Goligher and colleagues248 reported radiological leakage in 51 percent of low and high anterior resections. Chassin249 had only two patients with clinical leakage in 62 anastomoses. Many radiological leaks do not present with clinical symptoms.

Other evidence of inadequate procedure includes:


Obstruction resulting from turning in too much tissue at the anastomosis

Improper preparation of the ends to be anastomosed

Failure to close (or open widely) a mesenteric defect

Excessive tension on the anastomosis from a short transverse mesocolon

Poor lymphadenectomy with future recurrence

The following are several anatomic points to be remembered to ensure performance of a complete lymphadenectomy (and, therefore, adequate cancer surgery):

Right Colectomy


Ligate the ileocolic artery at its point of origin from the superior mesenteric artery, or as high as possible. Preserve ileal branches.

If an extended right colectomy is necessary, ligate the middle colic artery in the above fashion.

Left Colectomy


High ligation of the inferior mesenteric artery must be carried out. To accomplish this, elevate the third part of the duodenum, which is notorious for covering the origin of the inferior mesenteric artery from the aorta.

Perform high ligation of the inferior mesenteric vein employing slight, gentle elevation of the pancreas, in order to explore the entrance of this vein into the splenic vein.

Avoid traction of the spleen (see surgical anatomy section of the spleen chapter for details).


Anatomic complications of cecostomy include:


Intraperitoneal abscess secondary to leakage at the stoma; poor purse-string suture is a contributing factor

Abdominal wall abscess resulting from failure to completely fix the cecum to the abdominal wall

Poor or nonfunctioning cecostomy due to a Foley catheter that is too small, has blockage, or its catheter tip in the cecum is located poorly, especially when the appendiceal stump is used


Most of the anatomic complications of colostomy result from inadequate procedures. Many sources of error are possible.

Among a series of 181 consecutive colostomies,250 one patient died, and 50 suffered complications. Of the 50 complications, 12 resulted from an opening that was too tight; 24 resulted from an opening that was too loose. An excessively tight opening leads to edema, ischemia, and stenosis; an excessively loose opening leads to prolapse, redundant mucosa, peristomal hernia, retraction, evisceration, and small intestinal obstruction. Wound infection and tumor recurrence each occurred in 3 patients.

In 100 of the patients in Hines and Harris’ series,250 the colostomy was subsequently closed. There was 1 death; 17 patients had complications. Incisional hernia occurred in 7 patients, obstruction in 4, and infection in 2. A much higher rate (43.5 percent) of complications at closure has been reported by Varnell and Pemberton.251 The Hartmann procedure had significantly more complications than did loop colostomy. Nearly two-thirds of the complications were the result of infection.

In a series reported by Stothert and colleagues,252 41 patients underwent emergency colonic stoma formation. Six patients died and 19 had major or minor complications. Among the major complications were intraabdominal abscesses, peristomal abscesses, necrosis of the stoma, and peristomal hernia. The researchers advised using better technique to avoid complications.

Establishment of an end colostomy requires a good blood supply to the distal stump. If the end of the colon is ischemic, the tightness of the stoma will be irrelevant. If the terminal blood supply is good, strangulation from an abdominal aperture that is too small or is stenosed by skin may produce ischemia and necrosis in spite of a well-prepared colonic end.

Remember that retraction of a colostomy is due to peristaltic contractions trying to move feces from the obstructed colon. These contractions can be powerful.

We quote Kasperk et al.253:

Parastomal herniation is a very frequent complication in enterostomy. The therapeutic strategy consists of three approaches: local fascial repair, relocation of the stoma, and a variety of more elaborate procedures, many of which also involve the use of nonabsorbable meshes. Despite this multitude of available techniques, recurrence rates are high, and long-term complications, especially after mesh implantation, are frequent.

Abdominoperineal Resection

The complications of abdominoperineal resection can be summarized as: (1) vascular injury, (2) organ injury, and (3) nerve injury.

Presacral veins, the left iliac vein, the middle rectal artery (if present), and inferior hemorrhoidal vessels are subject to injury.

Trauma to the left ureter, duodenum, urinary bladder, or male urethra must be avoided. Other complications of abdominoperineal resection are inadequate reconstruction of the peritoneal floor, small bowel obstruction, colostomy complications, rupture of the rectum, and contamination.

Injury to the sympathetic or parasympathetic nerves may result in bladder dysfunction, failure of ejaculation, and severance of the nervi erigentes with impotence and retention of urine.41 The surgeon should remember that the nervi erigentes penetrate the fascia of Waldeyer; the fascia of Waldeyer should be divided at the tip of the coccyx. Nakai et al.254 reported that unilateral sacrifice of sacral nerves results in little bladder or anorectal dysfunction.

The ureter (and the lateral ligaments of the rectum) must be traced deep into the pelvis by careful dissection without elevation. Division of the colon and formation of the colostomy can then be done. The pelvic peritoneum should be closed to avoid herniation and obstruction of the small bowel.

According to Porter et al.,255 62% of inadvertent rectal perforations during abdominoperineal resection take place during the perineal part of surgery. The authors advised special care.

Iatrogenic Perforations with Barium Enema or Colonoscopy

Gedebou et al.256 stated that in iatrogenic colon perforations due to barium enema or colonoscopy without significant contamination, either primary repair or resection and anastomosis can be performed with acceptable morbidity.

Anatomic Complications of the Anorectum

Anorectal Fistulas

Among 133 patients in the series of Adams and Kovalcik,216 there were 80 fistulectomies, with 1 patient in whom section of the sphincter was necessary. The other 53 patients had fistulotomy. Complications occurred in 14 patients: recurrence in 5, spinal headache in 3, postoperative bleeding in 2, and temporary rectal incontinence in 1. Other complications are stenosis and stricture. Avoid complete division of the external sphincter. Avoid complete anterior division of the external sphincter in females.

Church et al.257 reported that hemorrhage from presacral veins, perforation of the rectum, damage to the pelvic autonomic nerves, and inadequate clearance of rectal cancer are major complications of rectal surgery.

Anorectal Abscesses

Fistulas may form due to poor drainage of abscesses. Initial fistulotomy reduces the number of recurrences requiring surgery. Knoefel et al.258 state that the risk of developing incontinence increases with recurrent anorectal disease, not with careful fistulotomy.

Anal Fissure

Fecal incontinence in elderly patients may occur secondary to dilatation or sphincterotomy.


We quote Liberman and Thorson259:

Anal stenosis may be anatomic (stricture) or functional (muscular). Anal stricture is most often a preventable complication. It is most commonly seen after overzealous surgical hemorrhoidectomy. A well-performed hemorrhoidectomy is the best way to avoid anal stricture. Symptomatic mild functional stenosis and stricture may be managed conservatively with diet, fiber supplements, and stool softeners. A program of gradual manual or mechanical dilation may be required. Sphincterotomy and various techniques of anoplasty have been used successfully in the treatment of symptomatic moderate to severe functional anal stenosis and stricture, respectively.

Anal or distal rectal strictures can occur secondary to nonepithelialization of the mucosa of the surgical anal canal and distal rectum when the mucosa is removed in toto. Anal fissure is a rare complication of hemorrhoids. Incontinence is very rare.

Timaran et al.260 reported a case of unexpected anal adenocarcinoma in a hemorrhoidectomy specimen, but they advised that the rarity of this finding does not support routine histopathological examination.

Rectal Prolapse

The complications of the surgical treatment of rectal prolapse depend on the procedure chosen by the surgeon:


Anastomotic complications


– Leakage

– Disruption of anastomosis

– Obstruction

Vascular injury. The following vessels should be protected to prevent bleeding:


– Left gastrohepatic artery and vein

– Left splenic polar artery

– Renocolic vessels

– Left gonadal vein

– Marginal artery

– Middle colic artery

– Vein of the left pelvic wall

– Presacral venous plexus

– Superior, middle, and inferior rectal veins

Organ injury


– Ureters

– Spleen

– Prostate

– Seminal vesicles

– Vagina

– Urinary bladder

Nerve injury


– Autonomic nerve injury with bladder dysfunction and impotence

Inadequate procedures


– Failure to close mesenteric defect

Anorectal Cancer Surgery

Complications of anorectal cancer surgery may include impotence, anastomotic leak with local abscess formation or generalized peritonitis, bleeding of venous origin, or urinary retention.

Bissett and Hill261 present the following complications which may occur during rectal mobilization for carcinoma:


Ureteric injury

Rectal perforation


Injury to the autonomic nerves

Recurrence of cancer in the pelvis

Ho et al.262 stated that transanally introduced stapling technique which involves anal manipulation may result in postoperative anal sphincter defects and impaired anal pressures.

Ahmed Shafik

Large Intestine


As promised earlier in this chapter, we present here some selected topics on the large intestine and extensive research on the anorectum by Dr. Ahmed Shafik.

Dr. Shafik wishes to acknowledge his great appreciation to Margot Yehia and Waltraut Reichelt for their invaluable assistance in preparing this material.

Right Colon Flexure Syndrome

The right colon flexure syndrome, which was described by Shafik,263 presented with right-sided abdominal pain, distension and constipation. At operation a triangular web (Fig. 18-66) with an underlying narrow segment at the hepatic colonic flexure was found responsible for colonic obstruction. The stenotic segment had complete longitudinal muscle coverage due to fusion of the 3 taenia coli. Colo-colic anastomosis relieves the symptoms. The condition seems to have a developmental origin.

Fig. 18-66.

Operative findings at the hepatic flexure. Web is divided to demonstrate the stenotic segment. (Modified from Shafik A. Right colon flexure syndrome. Report of two cases. Coloproctology 1981;3:105-106; with permission.)

Rectosigmoid Junction

The physioanatomic aspects of the rectosigmoid junction (RSJ) are rarely considered in the literature. Herein is a concise review of the surgical anatomy and related physiology of the RSJ.

Sigmoidorectal Junction Sphincter and Reflex

A recent study has recognized the presence of a high pressure zone in the RSJ at rest.98 It exists between two lower pressure zones: the sigmoid colon above and the rectum below. The segment of the colon at the RSJ seems to act as a physiologic sphincter preventing feces from crossing automatically from the sigmoid colon to the rectum during a cycle of mass colonic contraction.

Upon colonic contraction, the colonic contents travel downward through the descending colon to the sigmoid colon. The factors preventing the colonic contents from continuing directly to the rectum during colonic contraction are believed to be:


The S-curve of the sigmoid colon

A flap-valve action induced by the angle that links the sigmoid with the rectum

The high pressure zone at the RSJ

During the time of fecal collection in the sigmoid colon, the pressures in the RSJ and rectum showed no significant change against the resting state suggesting that the rectum and RSJ did not exhibit mechanical activity. This state was maintained until the fecal mass attained a certain volume, at which point the sigmoid colon contracts and the RSJ relaxes, dispelling the sigmoid contents to the rectum.98 A reflex relationship exists between the sigmoid colon and the RSJ which is mediated through what Shafik calls: “sigmoidorectal junction inhibitory reflex.”98 The reflex functions to relax the RSJ upon sigmoid contraction, thereby allowing the sigmoid contents to cross into the rectum. It seems to also regulate the process of sigmoid stool delivery to the rectum: it does not relax until the colonic mass reaches a certain volume that would initiate sigmoid contraction.

Studies have demonstrated that the RSJ pressure increases upon rectal contraction. This reflex relationship, which Shafik termed “sigmoidorectal junction excitatory reflex,”98 closes the RSJ upon rectal contraction, resulting in the rectal contents being dispelled distally. This in turn prevents reflux of rectal contents back to the sigmoid colon. The sigmoido-rectal junction excitatory reflex seems to function only upon rectal contraction leading to RSJ closure.

Thus, a physiologic sphincter appears to exist at the RSJ. Researchers failed to detect such a sphincter anatomically. The same state of affairs is found with respect to the lower esophageal sphincter: it exists functionally but not anatomically.

Functional Activity of Sigmoid Colon and Rectum during Fecal Storage in

The Sigmoid

Another study264 revealed that slow sigmoid colon distension did not effect the increase of sigmoid pressure, indicating that the sigmoid colon adapts as it receives new material from the colon. This adaptation continued with increasing balloon distension of the sigmoid, until, at a certain volume, the sigmoid colon contracted and pushed the balloon to the rectum (Fig. 18-67). It seems that the stretch receptors in the sigmoid colon wall, which initiate sigmoid contraction, are not stimulated before a certain level of distension is reached.

Fig. 18-67.

Pressure response of the sigmoid colon, rectosigmoid junction (RSJ), rectum, and rectal neck (anal canal) to slow-rate balloon distension of the sigmoid colon. (Modified from Shafik A. Functional activity of the sigmoid colon and rectum. A study during fecal storage in the sigmoid. Coloproctology 1997;19:236-241; with permission.)

This is in contrast to rapid sigmoid distension which effected sigmoid contraction with approximately half the volume required for sigmoid contraction by the slow distension test (Fig. 18-68). There might be 2 types of stretch receptors in the sigmoid wall: one responding to slow distension and the other to rapid distension. Under normal physiologic conditions, the sigmoid receives the feces from the colon and stores them until a volume has accumulated which is big enough to stimulate the stretch receptors and initiate sigmoid colon contraction which would push the stools to the rectum. Shafik speculates that rapid sigmoid distension occurs in pathologic conditions, such as diarrhea, and triggers sigmoid contraction with smaller amounts of fecal material than normal.

Fig. 18-68.

Pressure response of the sigmoid colon, rectosigmoid junction (RSJ), rectum, and rectal neck (anal canal) to rapid-rate balloon distension of the sigmoid colon. (Modified from Shafik A. Functional activity of the sigmoid colon and rectum. A study during fecal storage in the sigmoid. Coloproctology 1997;19:236-241; with permission.)

Hypertonic Rectosigmoid Junction Syndrome

The presence of a new clinicopathological entity causing constipation which Shafik termed the ‘hypertonic rectosigmoid junction’265 could be demonstrated in 6 women and 2 men (mean age 44.2 ± 10.3 years) who complained of chronic constipation. They had normal bowel habits before that time. Intestinal transit was delayed with accumulation of pellets in the sigmoid colon. Defecography and electromyography of the external anal sphincter and levator ani muscle were normal. The resting pressure was normal in the sigmoid colon, rectum, and rectal neck, but elevated in the rectosigmoid sphincter (RSS). The sigmoidorectal inhibitory and excitatory reflexes were absent. Histologic findings in biopsies from the sigmoid colon and rectum were normal, but from the RSS were aganglionic (Fig. 18-69). RSS “achalasia” was diagnosed. Endoscopic dilatation of the RSS effected improvement of 5 patients. The remaining 3 patients underwent sigmoidomyotomy (Fig. 18-70A, B). The 8 patients are now 9 to 38 months without recurrence of the constipation. RSS achalasia constitutes a clinicopathologic entity which should be considered in the etiology of constipation.

Fig. 18-69.

Photomicrograph of a biopsy from the rectosigmoid sphincter showing aganglionosis. Hematoxylin eosin x 1000. (From Shafik A. The hypertonic rectosigmoid junction: description of a new clinicopathologic entity causing constipation. Surg Laparosc Endosc 1997;7:116-120; with permission.)

Fig. 18-70.

Technique of sigmoidomyotomy. A, Muscle coat of the rectosigmoid junction is incised down to the mucosa. B, Mucosa bulges throughout the length of the incision. (Modified from Shafik A. The hypertonic rectosigmoid junction: description of a new clinicopathologic entity causing constipation. Surg Laparosc Endosc 1997;7:116-120; with permission.)

Electrorectogram and Rectosigmoid Pacemaker

The electrical activity of the rectal detrusor was studied experimentally and in humans both intrarectally and transcutaneously.266-268 Pacesetter potentials (PPs) or slow wave activity was recorded in the rectum. PPs start in an area that corresponds anatomically to the RSJ. The PPs propagate distally in a caudad direction with the same frequency and velocity in the individual subject.

Action potentials (APs) followed or were superimposed on the PPs. They were accompanied simultaneously with increased rectal pressure, which denotes rectal contraction. They occurred less frequently than the PPs, and their frequency was inconsistent in each individual subject. The frequency and amplitude of PPs and APs increased with increasing rectal distension until the rectal balloon was dispelled. The occurrence of APs in relation with PPs was usually random. This points to irregular or nonrhythmic segmental rectal contraction. However, once they occurred they were conducted distally in a regular fashion like PPs. After rectal myotomy, the orderly distal propagation of APs disappeared because myotomy had interrupted their pathway. The APs cause a contractile sweep along the rectum. Yet, this contractile activity induces a rectal pressure increase which is too small to be perceived as rectal sensation.

The aforementioned findings would suggest that the RSJ is the site of a pacemaker.266-268 It triggers the PPs which are propagated distally along the rectum. The PPs seem to initiate the APs which are accompanied with episodes of rectal pressure elevation representing contractile activity. This postulates that rectal contractions result from APs and not from PPs. The PPs seem to pace the rectal contractile activity in terms of direction and frequency.

Rectosigmoid Pacemaker Evidence

It seems that there are specialized receptors at the RSJ that trigger rectal detrusor contraction on stimulation by distension of the RSJ. These receptors act as a ‘pacemaker’ regulating the stool delivery from the sigmoid colon through the rectum to the rectal neck.269,270 The concept of a ‘recto-sigmoid pacemaker’ is evidenced269-271 by: (a) rectal detrusor contraction on RSJ distension, (b) propagation of the rectal electric waves from the RSJ proximodistally, (c) absence of rectal detrusor response to distension of the anesthetized RSJ, (d) absence of detrusor contraction on distension of the colon in anterior resection operation for rectal cancer, and (e) reproducibility. The mechanism of action of the rectosigmoid pacemaker in rectal motility needs to be discussed.

Rectosigmoid Pacemaker and Rectal Motility

When the sigmoid colon contracts, it delivers its stool contents to the rectum. As the stools pass across the RSJ, they stimulate the rectosigmoid pacemaker leading to rectal detrusor contraction.98,264 A contraction wave seems to start at the RSJ and to spread into the rectal detrusor, serving a dual action: a) it closes the RSJ, and b) it amputates the fecal column delivered by the sigmoid colon. The upper part remains in the sigmoid colon, while the lower part is delivered to the rectal neck by a wave of ‘mass-squeeze’ contraction of the rectal detrusor98,272,273 (Fig. 18-71). The contraction wave travels caudad across the rectal detrusor pushing the fecal column into the rectal neck.

Fig. 18-71.

“Mass-squeeze” contraction of the rectal detrusor. A, Stool distends the rectum. B, Rectosigmoid junction closed by means of the rectosigmoid junction excitatory reflex. C, Rectal contractile wave travels caudad pushing the fecal column to the rectal neck. (Modified from Shafik A. Study of the rectal detrusor motility in normal and constipated subjects. Proc 2nd Int Mtg Coloproct, Ivrea/Italy 1992:78-90; with permission.)

After the rectum evacuates its contents, it relaxes. The fecal portion that remained in the sigmoid colon is then delivered to the rectum, stimulating the recto-sigmoid pacemaker into evoking another rectal contraction wave that evacuates the rectum. The process of rectosigmoid pacemaker stimulation is repeated until the sigmoid reservoir becomes empty.98,272,273 The stimulus to sigmoid colon contraction is unknown. Relaxation, following the contraction of the rectal detrusor, seems to trigger sigmoid contraction to deliver further stools into the rectum.

If, after the rectal contraction wave has been initiated, evacuation is inopportune, the external anal sphincter is voluntarily contracted, hence preventing internal sphincter relaxation and reflexively aborting rectal detrusor contraction; this occurs through the voluntary inhibition reflex.163 The rectum relaxes and accommodates for the new contents. It is under these conditions that stools can be palpated in the rectum on digital examination, though under normal physiologic conditions the rectum is empty. When conditions are favorable, the loaded rectum is emptied either by straining with resulting mechanical compression of the detrusor or by a rectal contraction wave initiated when the RSJ is stimulated by another fecal mass delivered from the sigmoid colon to the rectum.

Artificial Pacemaker

The role of an artificial pacemaker in initiating rectal contractions was assessed both experimentally and in humans.


An artificial pacemaker was applied to the rectum of 18 mongrel dogs aiming at assessing its effectiveness in inducing rectal contraction.271 The pacemaker consisted of a hooked needle, a metal plate, a battery, and a telegrapher’s key. The needle was hooked into the dog’s rectal muscle coat at the RSJ and the metal piece was applied to the skin. Upon electric pulsing of the pacemaker, the rectal pressure showed significant increase while the rectal neck pressure was significantly decreased. Electric pacing succeeded in expelling the balloon in all the dogs.


In 46 patients, the effectiveness of an artificial pacemaker in the treatment of chronic idiopathic constipation was studied: 26 had chronic idiopathic constipation and 20 were normal controls.274 The results demonstrated the effectiveness of the artificial pacemaker in inducing rectal contraction in both the normal controls and constipated subjects. This is evident from the significant rectal pressure increase as well as from balloon expulsion upon electric pulsing of the pacemaker.

Relationship between Electrosigmoidogram and Electrorectogram

In the sigmoid colon, slow waves or PPs were monophasic with a large negative deflection, while the rectal PPs were triphasic with a small positive, large negative, and another small positive deflection275 (Figs. 18-72, 18-73). The frequency and amplitude of the PPs in the rectum were significantly higher than those of the sigmoid colon. The fact that shape, frequency, amplitude, and velocity of conduction of the electric waves differed in the electrosigmoidogram (ESG) from those in the electrorectogram (ERG) suggests that the rectal electromechanical activity is not a continuation of the sigmoid colon activity and seems to be initiated in the rectum.275 This agrees with previous studies which have shown that the rectal motility may be regulated by a “pacemaker” at the rectosigmoid junction.266,269 The increased intrasigmoid and rectal pressures associated with APs indicate that they possess contractile activity that continued to increase with increasing balloon distension until, at a certain volume, the balloon is dispelled outside the sigmoid colon or rectum.

Fig. 18-72.

Electrorectogram showing triphasic pacesetter potentials followed randomly by action potentials. (From Shafik A. Electrosigmoidogram, electrorectogram and their relation. Front Biosci 1997;2:12-16 Pub Med No. 9294095; with permission.)

Fig. 18-73.

Electrosigmoidogram showing monophasic pacesetter potentials followed randomly by action potentials. (From Shafik A. Electrosigmoidogram, electrorectogram and their relation. Front Biosci 1997;2:12-16 Pub Med No. 9294095; with permission.)

Motility of Sigmoid Colon and Rectum

The sigmoid colon receives the stools from the colon by a mass action.276 The new contents distend the sigmoid colon and augment its electromechanical activity.275 The latter increases with growing accumulation of sigmoid contents until the stools are expelled to the rectum. The movements of the sigmoid colon appear to be of the “mass type.” Shafik’s study showed that the proximal contraction of the sigmoid colon and its relaxation distal to the simulated stool occurred synchronously in one comprehensive mass action and not segmentally.264,275 This mechanism pushes the contents of the sigmoid colon en masse to the rectum. The results showed also that the distension of the sigmoid colon did not affect the electromechanical activity of the rectum. Therefore, the mass action of the sigmoid colon is independent from the rectal mass action.

Tube and Marsupialization Colostomy

The technique for tube and marsupialization colostomy consists of creating a small opening into the colon (1-2 cm in length) and suturing its edges to the abdominal wall.277 The procedure is referred to as a tube colostomy when a tube is put into the colostomy (Fig. 18-74), and a marsupialization colostomy (Fig. 18-75) when no tube is used. The technique is especially indicated in seriously ill patients with colonic obstruction.

Fig. 18-74.

“Tube colostomy.” A, Purse-string suture applied to the colon. B, Incision into the colon. C, Tube introduced into the bowel lumen. (Modified from Shafik A. “Tube” and “marsupialization” colostomy: a simplified technique for colostomy. Am J Proct 1982;33:12-16; with permission.)

Fig. 18-75.

“Marsupialization colostomy.” A, Incision into the colon. B, Edges of the colonic opening sutured to the skin. C, A wide colonic opening showing eversion of posterior colonic wall forming a spur. (Modified from Shafik A. “Tube” and “marsupialization” colostomy: a simplified technique for colostomy. Am J Proct 1982;33:12-16; with permission.)

“Tube” colostomy is a deflating colostomy whereas “marsupialization” colostomy is both deflating and defunctioning. Both colostomies have advantages over the formal colostomies. The operations are easily and rapidly performed in any part of the colon, and do not require colonic mobilization or exteriorization. They are performed under local anesthesia, and preserve colonic length. They require a small parietal opening, thus avoiding colostomy prolapse or hernia. Tube colostomy closes spontaneously and marsupialization colostomy requires simple closure.


Intense research by Shafik and others into the anorectal region has been provoked both by unsatisfactory results from diverse clinical procedures and by some key references in the medical literature that are still taken for granted, even though they date back about 400 years to the time of Vesalius (1514-1564).278 As a result of research in recent years, ample new anatomic and physiologic data have become available which have shed brighter light on the anatomy and function of the anorectal region.

Some of the new data are consistent with our classical knowledge; some negate older observations and conclusions. The majority of findings are intriguing additions to our perception of the physioanatomic relationships of this area. They have had a direct bearing on the adaptation of the basic concepts of the anorectal region. The newer concepts have taken a significant place in our understanding of the etiologies of hitherto unresolved rectal pathologies, including some that had been earlier considered idiopathic.

The findings obtained by intensive investigations into the embryologic aspects of the anorectum, in particular, have provided missing links which have contributed in large measure to our understanding of the functions, the interdependence of the different elements contained in the area, and the disorders that affect them. Hence, by drawing from and applying modern knowledge and understanding, modern surgery stands an incomparably better chance of improving the results of treatment.


Embryologic investigations have supplied the essential bits of mosaic which are necessary for the understanding of the principles governing the anorectum through its development from the hindgut. Studies have conclusively revealed that the rectum extends down to the perineal skin and constitutes one single embryologic, anatomic, and functional unit, together with its neck (the anal canal). The anal canal consequently exists neither as an embryologic entity, nor as an anatomically separate entity, but only as the epithelium that lines the terminal part of the hindgut and forms an opening for the course of the hindgut to the exterior. Failure to comprehend these facts has contributed to the controversy about the embryologic bases of the different congenital anomalies of the area.

Anatomic Evidence Related to the Anal Musculature

The anal musculature consists of two components: visceral and somatic, which are separated by the intersphincteric space.

Visceral Component

The visceral component of the anal musculature comprises the internal sphincter and the longitudinal muscle, both of which are formed of smooth muscle bundles and are autonomically innervated. Contrary to the views of some investigators41,279 who hold that these muscles belong to the anal canal, all evidence suggests that they are part of the rectum. Accordingly, the internal sphincter is the downward extension of the circular muscle coat of the rectum which is thickened at the rectal neck inlet, and proceeds down to the perineal skin, from which it is separated by the subcutaneous space101 (Fig. 18-76A). The longitudinal muscle is but a continuation of the rectal longitudinal muscle coat. It stretches along the rectal neck to stop short of the lower end of the internal sphincter, where it contributes to the central tendon which inserts into the perineal skin by multiple tendinous fibers (Fig. 18-76B).

Fig. 18-76.

Coronal sections of the rectal muscle coats. A, Cadaveric specimen demonstrates that the circular and longitudinal rectal muscle coats extend down to the perineal skin, and that the longitudinal muscle inserts into the perineal skin after base loop penetration. Verhoeff-van Gieson x 7. a, Rectal circular muscle coat and internal sphincter. b, Rectal longitudinal muscle coat. c, Central tendon. d, Rectal longitudinal muscle insertion in perineal skin. e, Levator plate. f, Suspensory sling. g, Fascia covering levator plate. h, Tunnel septum. i, Top loop of external sphincter (conjoined deep external sphincter and puborectalis). j and k, Intermediate and base loops of external sphincter. l, Hiatal ligament. B, Diagrammatic coronal section of the muscle coats. (A, From Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. II. Anatomy of the levator ani muscle with special reference to puborectalis. Invest Urol 1975;13:175-182; with permission. B, Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. V. The rectal neck: Anatomy and function. Chir Gastroenterol 1977:11:319; with permission.)

Somatic Component

The somatic component of the anal musculature consists of striated muscles which include parts of the levator ani, including the puborectalis, and the external sphincter. The somatic component forms the levator tunnel (Fig. 18-77) which embraces the lower end of the hindgut. It develops in the perineum independently of the visceral component.280 It functions under voluntary control, and serves to control the involuntary visceral component.

Fig. 18-77.

The levator tunnel. (Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. VIII. Levator hiatus and tunnel. Anatomy and function. Dis Colon Rectum 1979;22: 539; with permission.)

The two components, visceral and somatic, although separate anatomically, have been found to be functionally interrelated so that the voluntary component acts through the involuntary, and vice versa. Thus, the “voluntary inhibition reflex” action of the external sphincter is effected through the involuntary internal sphincter.163 Furthermore, the suspensory sling of the levator ani (somatic) and the longitudinal muscle (visceral) act jointly to open the rectal neck at defecation.281 For these reasons, the somatic component is an integral part of the hindgut visceral component.

Intersphincteric Space

The intersphincteric space is an embryologically developed space which acts to separate the visceral and somatic components of the hindgut, and is occupied by areolar tissue. It lies between the longitudinal muscle and the suspensory sling of the levator ani muscle; it extends along the lower part of the hindgut to the subcutaneous space, and is continuous above with the pelvirectal space.101 Through the intersphincteric space, the lower part of the rectum, including its visceral component, can be mobilized from within the levator tunnel (somatic component) for removal of otherwise inaccessible rectal lesions, with preservation of the voluntary somatic component282 (Fig. 18-78).

Fig. 18-78.

The technique of rectal mobilization for local removal of a rectal lesion. A, Plane of dissection between suspensory sling and longitudinal muscle. B, Rectal neck and lower rectum proper mobilized and pulled down outside anal wound. (Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XII. Anorectal mobilization. A new surgical access to rectal lesions. Preliminary report. Am J Surg 1981;142:625-635; with permission.)

Embryologic Evidence

The rectum develops from the hindgut (Fig. 18-79A, B, C), which extends down to the perineal skin as can be deduced from the extension of its circular and longitudinal muscle coats to this level. The blind end of the hindgut, reaching the perineal skin, is invaginated by the proctodeal dimple (Fig. 18-79D). The invagination is effected by the ectoderm only, without muscular elements.282 In the course of invagination, an anorectal sinus forms, but is usually obliterated (Fig. 18-79E, F); and the proctodeal skin is remodeled282,283 (Fig. 18-79G). It seems that the hindgut is guided to its normal location in the perineum by tracking through the levator tunnel (Fig. 18-79B, C).

Fig. 18-79.

The embryologic development of the lower end of the hindgut. A, Hindgut migrating toward perineal skin. B, Levator tunnel formation. Hindgut migrates through and is guided by levator tunnel to the perineal skin. C, Hindgut reaches and is fixed to perineal skin by its longitudinal muscle coat. D, Proctodeal dimple invaginates hindgut lower end with resultant anorectal sinus formation. E, Rectal membrane rupture with a resulting hindgut opening to the exterior and paraanal space formation. F, Anorectal sinus obliteration with persistence of paraanal space. G, Rectal neck remodeling resulting in paraanal space obliteration. a, Hindgut longitudinal muscle coat. b, Hindgut circular muscle coat. c, Levator plate. d, Suspensory sling. e, Top loop of external sphincter (conjoined deep external sphincter and puborectalis). f,g, Intermediate and base loops of external sphincter. h, Internal sphincter (hindgut circular muscle coat). i, Hindgut longitudinal muscle attachment to perineal skin. j, Hindgut membrane (rectal membrane). k, Anorectal sinus. l, Pectinate line. m, Paraanal space. n, Obliterated anorectal sinus. o, Skin lining hindgut. (Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. X. Anorectal sinus and band: anatomic nature and surgical significance. Dis Colon Rectum 1980;23:170; with permission.)

The stimulus to proctodeal dimpling has not been identified. It may be triggered either by the fixation of longitudinal hindgut fibers to the perineal skin or in the course of formation of the levator tunnel.282 The latter possibility seems more likely, because proctodeal dimpling occurs also without the hindgut having entered the tunnel or being fixed to the perineal skin. The formation of the levator tunnel may therefore stimulate not only normal hindgut migration to the perineum, but also proctodeal dimpling. As an embryologic result of this process, proctodeum-hindgut intussusception achieves both rupture of the rectal membrane (Fig. 18-79E) to open the hindgut to the exterior, and fusion of the lining of the endodermal hindgut with the perineal ectoderm within the lower end of the hindgut at the pectinate line.

Vascular Supply

The hindgut artery is the inferior mesenteric artery. Its continuation, beyond the last sigmoid branch, is the superior rectal artery which normally accompanies the hindgut to its termination at the perineal skin. In postnatal life, however, the artery normally ends approximately at the pectinate line – probably because invagination of the hindgut by the proctodeum involves its lower part during the obliteration of the anorectal sinus. The hindgut below this level acquires blood supply from the inferior rectal branch of the internal pudendal artery.

Rectal Neck

The rectum is the terminal part of the hindgut, between the third sacral vertebra and the perineal skin. Its lower segment, which can be termed “rectal neck” (Fig. 18-80) due to the narrowing imposed on it by the surrounding sphincters, extends downward from the level of the levator plate to the perineal skin, whereas the part above is the “rectum proper.” The rectal neck therefore represents that portion of the bowel usually referred to as the “anal canal.”282 The only role the proctodeum has in the rectal neck is as the lining of its terminal end, which develops in the course of fusion of the hindgut mucosa with the perineal skin, as described above.

Fig. 18-80.

The new anatomic features of the rectum. a, Internal rectal sphincter. b, Longitudinal rectal muscle. (Courtesy of Prof. Ahmed Shafik, MD.)

This simulates the case of the gastric columnar mucosa: although it normally extends upward into the distal 3 cm of the esophagus to fuse with its squamous epithelium,284 this columnar-lined esophageal segment is neither considered a separate entity of the esophagus nor given a separate name. The junction of the rectum proper with its neck is the “rectal neck inlet” (Fig. 18-80), whereas the rectal neck opening to the exterior is the “rectal neck outlet.” The skin lining the lower rectal neck is the “intrarectal skin,” while the skin surrounding the rectal neck outlet is the “perirectal skin.”

Structural-Functional Adaptation of the Rectal Neck

The rectal neck has undergone certain anatomic changes in adaptation to its function as a regulator of the mechanisms of continence and defecation.285 Thus, the presence of the rectal angle induced by the puborectalis sling and levator plate (Fig. 18-81), at the junction of the rectum proper with its neck, seems to play a role in rectal continence. The thickening of the rectal circular muscle coat around the rectal neck to form the internal sphincter keeps the rectal neck closed involuntarily. Also the longitudinal muscle attachments to the perineal skin, after penetrating the base loop, are of functional significance: they fix the rectum to the perineal skin in such a way that the longitudinal muscle, on contraction at stool, not only shortens and opens up the rectal neck, but also stretches open and pulls up the base loop to unseal the rectal neck outlet.105,281 The intrarectal skin lining is functionally important in that the rich somatic ectodermal innervation supplies the lower rectal neck with sensitivity to stimuli, and thus constitutes an essential component in the mechanisms of continence and defecation.

Fig. 18-81.

Defecation mechanism. A, At rest. B, At defecation. Observe opening of both rectal neck inlet and outlet as well as opening and shortening of rectal neck. (Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. V. The rectal neck: Anatomy and function. Chir Gastroenterol 1977;11:319; with permission.)

Furthermore, the rectal neck sphincter apparatus is arranged to keep each involuntary muscle under the control of a voluntary one.163 Hence, the external sphincter voluntarily opposes the involuntary action both of the internal sphincter and the longitudinal muscle. In order to resist the desire to pass stool or flatus, external sphincter contraction inhibits the reflex internal sphincter relaxation on detrusor contraction; this is called the “voluntary inhibition reflex” action.163 At the same time, base loop contraction stretches tight, and inhibits contraction of, the longitudinal muscle which passes through the base loop substance.163 More information on this subject follows under “Mechanism of Continence.”

Congenital Rectal Anomalies

A New Concept of Genesis

The conclusion that the hindgut migrates to the perineal skin, guided by the levator tunnel, to be invaginated and perforated by the proctodeal dimple could explain the different congenital rectal anomalies.283,285 The rectum may develop completely, enter the levator tunnel and reach the perineal skin, but not be invaginated by the proctodeum (Fig. 18-82A, B); or it may be invaginated without rupture of the rectal membrane (Fig. 18-82C, D). In both conditions, the end of the hindgut may remain blind or result in a fistula in the perineum or genitourinary tract (Fig. 18-82B, D). Failure of rectal neck remodeling could result in rectal neck stenosis (Fig. 18-82E). The hindgut might fail to enter the levator tunnel because of delayed formation of, or poor development of, the levator tunnel and hence not be guided to its normal location in the perineum. In such cases then, the end of the hindgut lies above the levator tunnel and either remains blind (Fig. 18-82F) or seeks a way to the exterior by development of a fistular connection to the genitourinary tract (Fig. 18-82G).

Fig. 18-82.

Congenital anomalies of the rectum. A, B, Hindgut migrates through levator tunnel to perineal skin. No proctodeum-hindgut invagination. Hindgut either (A) ends blindly or (B) fistulizes to perineal skin or genital tract. C, D, Hindgut migrates through levator tunnel to perineal skin. Proctodeum-hindgut invagination occurs but without rectal membrane rupture. Hindgut either (C) ends blindly or (D) fistulizes to perineal skin or genital tract. E, Rectal neck stenosis due to failure of rectal neck remodeling. F, G, Failure of hindgut to enter levator tunnel. Hindgut (F) ends blindly or (G) fistulizes to genitourinary tract. G, Proctodeum may dimple or not. a, Levator plate. b, Suspensory sling. c,d,e, Top, intermediate, and base loops of external sphincter. f, Longitudinal rectal muscle. g, Internal rectal sphincter. h, Fistula. (Courtesy of Prof. Ahmed Shafik, MD.)

Anorectal Sinus and Band

The proctodeal dimple, by invaginating the lower end of the hindgut (Fig. 18-79A, B) as it approaches the perineal skin, pushes up and tightly stretches the rectal membrane which separates the hindgut from the proctodeum; meanwhile it enfolds the lower end of the hindgut mucosa285 (Fig. 18-79C, D). Invagination continues until the rectal membrane ultimately ruptures, leaving the anal valves to mark its site (Fig. 18-79E, F).

Two spaces result from proctodeal “invagination” of the hindgut: (1) an outer space which can be termed the “anorectal sinus,” and (2) an inner space which can be called the “paraanal space”283,285 (Fig. 18-79E). The anorectal sinus represents the enfolded part of the hindgut mucosa and seems to include the structures which investigators refer to as the anal glands.

Histopathology of the Anorectal Sinus and Band and Epithelial Debris

The anorectal sinus (Fig. 18-83) is a diverticulum-like process which extends downward to a variable extent in the lower rectal neck submucosa from the level of the anal valves.285 The sinus may be deep and narrow and extend along the internal sphincter, or it may be small and shallow. Deep sinuses are commonly found in deceased neonates and children, and are complete circumferentially around the lower rectal neck.285 In adult specimens the anorectal sinus is shallow and divided longitudinally into 4 to 9 compartments. It is lined by columnar epithelium in some areas, and stratified columnar in others.

Fig. 18-83.

Paracoronal section of rectal neck showing an anorectal sinus extending down to the central space (hematoxylin and eosin; magnification x 8). a, Internal sphincter. b, Anorectal sinus extending down to (c) central space. d, Longitudinal muscles. e, f, g, Top, intermediate, and base loops of external sphincter. (Courtesy of Prof. Ahmed Shafik, MD.)

The anorectal band (Fig. 18-84) is a fibroepithelial tube which lies in the submucosa of the rectal neck below the anal valves.285 Histologically, it consists of collagen fibers impregnated with islands of epithelial cells. Clumps of epithelial cells found scattered in the lower rectal neck submucosa constitute epithelial debris285 (Fig. 18-85). These clumps are composed of round, oval, or columnar cells and are arranged in masses, sheets, or in the form of pseudoacini. The cells have been found to occur at the site of the absent anorectal sinus, and probably represent remnants of the sinus.

Fig. 18-84.

Coronal section of rectal neck showing anorectal band in anal submucosa (Verhoeff-van Gieson; magnification x 27). a, Anal lining. b, Anorectal band. c, Internal sphincter. d, Longitudinal muscle. e, External sphincter. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-85.

Photomicrograph showing epithelial debris of the anorectal sinus in the anal submucosa (hematoxylin and eosin; magnification x 110). a, Anal lining. b, Epithelial debris. (Courtesy of Prof. Ahmed Shafik, MD.)

Fate and Anomalies of the Anorectal Sinus

Normally the anorectal sinus is obliterated. The process of obliteration simply represents the completion of the embryonic fusion of the proctodeum with the hindgut. Sinus obliteration leads to the formation of the anorectal band (Fig. 18-79F) which normally disappears.285 This is followed by rectal neck remodeling in such a way that the proctodeal wall recedes laterally to come into alignment with the rectal wall, with the obliteration of the paraanal space and widening of the lower rectal neck (Fig. 18-79G).

The sinus has been found divided into 4 to 9 compartments, however, indicating irregularity in the process of closure.285 This is consistent with Shafik’s findings that the anorectal sinus may persist or be only partially obliterated and form epithelium-lined structures in the submucosa of the lower rectal neck. On the other hand, obliteration of the sinus may be complete, yet leave the anorectal band or epithelial debris that marks the position of the anorectal sinus285 (Fig. 18-79F). Failure of rectal neck remodeling, with persistence of the paraanal space, results in narrowing of the lower rectal neck (Fig. 18-79E).

Role of the Anorectal Sinus, Anorectal Band, and Epithelial Debris in the Genesis of Rectal


Persistence of the anorectal sinus or its epithelial elements, as represented in the anorectal band or epithelial debris, could explain the pathogenesis of some idiopathic rectal neck lesions such as perirectal abscess and fistula, chronic fissure, pruritus ani, hemorrhoids, cysts, and rectal neck adenocarcinoma.285

Perirectal Suppurations and Fistulas

A study of the pathology of perirectal fistula286,287 revealed a significantly higher rectal neck pressure (128 cm H2O) than normal (78 cm H2O). Epithelial cells were detected in the vicinity of the fistular track in 90% of the cases studied. These cells were arranged in acini or masses (Fig. 18-86). Multiple intraepithelial microabscesses were detected in these masses. Presumably, these epithelial cells are but the epithelial debris of the anorectal sinus. Chronic perirectal suppuration is believed to be due to the epithelial debris which maintains the infective process.287 The cure of perirectal suppuration of a nonchronic form would be related to absence of epithelial debris, or its destruction in the course of infection. The cause of fistular recurrence following properly performed operations could be due to the infection of other islands of epithelial debris.287 The rectal neck pressure becomes normal after correction of the fistula.

Fig. 18-86.

Photomicrograph showing epithelial debris in the vicinity of the fistulous track. The cells are arranged in masses lined by hyperplastic nonkeratinized stratified squamous epithelium which contains multiple microabscesses (hematoxylin and eosin). A, Magnification x 5. B, Magnification x 60. a, Fistulous track. b, Squamous epithelial mass. c, Microabscess. d, Rectal muscle bundles. (Courtesy of Prof. Ahmed Shafik, MD.)

Perianal Spaces and Anorectal Suppuration

Studies by Shafik101 have identified 6 perianal spaces: subcutaneous, central, intersphincteric, ischiorectal (ischioanal), submucosal, and pelvirectal (Fig. 18-87). Infection was found to start as a central abscess in the central space, which, if neglected, spread to the above mentioned spaces, and led to central, intersphincteric, ischiorectal, and pelvirectal abscess and fistula formation286 (Fig. 18-88).

Fig. 18-87.

The anatomic structure of the anal sphincter mechanism and the perianal spaces. Paracoronal section of rectal neck (hematoxylin and eosin; magnification x 19). a, Submucous space. b, Internal sphincter. c, e, j, l, Intersphincteric space. d, Medial layer of longitudinal muscle. f, Intermediate layer of longitudinal muscle, anal suspensory sling. g, Central space and central tendon. h, Base loop of external sphincter. i, Subcutaneous space containing corrugator cutis. k, Lateral layer of longitudinal muscle. m, Intermediate loop of external sphincter. n, External anal septum. o, Top loop of external sphincter. p, Ischiorectal (ischioanal) space. q, Levator plate. r, Pelvirectal space. (From Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. IV. Anatomy of perineal spaces. Invest Urol 1976;13:424; with permission.)

Fig. 18-88.

Diagrammatic representation of the mode of spread of central space infection and the resultant anorectal abscess. a, Central abscess. b, Subcutaneous abscess. c, Ischiorectal (ischioanal) abscess. d, Intersphincteric abscess. e, Pelvirectal abscess. (Courtesy of Prof. Ahmed Shafik, MD.)

Fistula cauterization was performed with a cautery probe. By destroying the epithelial debris, this treatment alternative achieves satisfactory results.288

Role of the Anorectal Sinus in Chronic Anal Fissure

Assessment of pathologic changes in chronic fissure289 resulted in identification of epithelial cells in the floor of the fissure, superficial to the internal sphincter. The cells were round, oval, or columnar, and were arranged in clumps or in the form of pseudoacini (Fig. 18-89). It is believed that these epithelial cells are just anorectal sinus remnants that exist in the rectal neck submucosa either as epithelial debris or an anorectal band.289 In a few cases, the anorectal sinus was detected in the floor of the fissure (Fig. 18-90), whereas in others no epithelial cells could be found, as occurs in heavily infected fissures. The mean maximal rectal neck pressure was significantly above normal (124 cm H2O).

Fig. 18-89.

Photomicrographs showing the epithelial debris in the floor of a chronic anal fissure. (Hematoxylin-eosin stain). A, Magnification x 40. B, Magnification x 150, reduced 17 percent. a, Squamous epithelium at the edge of the fissure. b, Epithelial cells. c, Inflammatory cells. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-90.

Chronic anal fissure showing an anorectal sinus at its floor. a, Squamous epithelium. b, Anorectal sinus. c, Granulations of the fissure. (Hematoxylin-eosin stain; magnification x 18.5, reduced 13 percent). (Courtesy of Prof. Ahmed Shafik, MD.)

The mechanism of fissure formation (Fig. 18-91) includes disruption of the lower rectal neck lining and the resulting exposure of the epithelial cells or the anorectal sinus to repeated infection in the floor of the wound.289 Epithelial cells act as multiple sites (sequestra) for harboring the infection and are responsible for fissure chronicity. Where epithelial debris could not be detected in the fissure, the death of the epithelial cells over the course of infection was assumed –which would explain the spontaneous cure of some fissures.289

Fig. 18-91.

Mechanism of fissure formation. Epithelial debris in the submucosa of the rectal neck (A) exposed in the wound floor due to anal lining disruption (B). Anorectal band in the submucosa of the rectal neck (C) exposed in the wound floor due to anal lining traumatization (D). Anorectal sinus in the submucosa of the rectal neck (E) exposed in the wound floor due to anal lining traumatization (F). (Courtesy of Prof. Ahmed Shafik, MD.)

The presence of high pressure in the rectal neck predisposes it to trauma by feces, due to the partial stenosis of the rectal neck. The stenosis can be induced either by the constricting effect of the anorectal band or failure of rectal neck remodeling.285,289 Evidence of partial stenosis is provided both by the significant high rectal neck pressure and the necessity of straining at stool, even with soft and bulky stools. The location of the fissure exclusively in the lower, and not in the upper rectal neck, is due to the anorectal sinus remnants contained therein; whereas, on the other hand, the posterior and (rarely seen) anterior median fissure is ascribed to the existence of 2 weak areas opposite the intermediate and base loop ellipses of the external sphincter289 (Fig. 18-92). Tears are more frequent posteriorly because the posterior rectal neck wall is less supported than the anterior wall.

Fig. 18-92.

The two weak areas in the rectal neck. A, Posterior area corresponds to the intermediate loop ellipse. B, Anterior area corresponds to the base loop ellipse. (Courtesy of Prof. Ahmed Shafik, MD.)

Treatment of chronic fissure is directed at ablation of epithelial debris in the fissure. Both internal sphincterotomy and anal dilatation aim at sectioning the rigid, fibrous, tube-shaped anorectal band (“anorectal bandotomy”),290 and provide the expansion of the rectal neck on defecation that prevents the repeated trauma from fecal material. These procedures do not eradicate the epithelial remnants in the fissure floor, however; fissure excision does remove these remnants.289

Role of the Anorectal Sinus in Hemorrhoids

In the presence of hemorrhoids, the mean rectal neck pressure is significantly higher than normal, 118 cm H2O ± 24 (s) versus 76 cm H2O ± 18 (s).290 There is, however, no significant difference in the rectal neck pressures between large and small hemorrhoids. Fibrous bands (Fig. 18-93) were detected in the lower rectal neck submucosa of all studied hemorrhoid patients, but in only 1 of the control subjects.290 The bands may extend from the bottom of an incompletely obliterated anorectal sinus. The hemorrhoid neck overhangs the top end of the fibrous band. Evidence suggests that this fibrous band is the anorectal band, an embryonic vestige.290

Fig. 18-93.

Photomicrograph of biopsy from the lower rectal neck of hemorrhoid patient, showing a fibrous tissue band extending in the submucosa from the level of pectinate line downward. (Verhoeff-van Gieson stain, magnification x 13). a, Squamous epithelium. b, Fibrous band. c, Internal sphincter. (Courtesy of Prof. Ahmed Shafik, MD.)

The invariable detection of fibrous bands (anorectal bands) with hemorrhoids postulates a relationship. High rectal neck pressure indicates a degree of stenosis which commonly accompanies the anorectal band.285 The presence of the fibrous band and the accompanying stenosis of the lower rectal neck hinder full rectal neck expansion at defecation, with a resulting partial obstruction to the descending fecal mass (Fig. 18-94). An excessive straining effort is necessary to achieve dilatation sufficient for evacuation. This would explain the cause of straining, even with soft and bulky stools, which precedes and accompanies hemorrhoid formation. The fecal mass, passing through the stenosed lower rectal neck, repeatedly traumatizes the mucosa at the upper end of the anorectal band, which would slacken and ultimately prolapse290 (Fig. 18-94).

Fig. 18-94.

Mechanism of hemorrhoid formation. A, Descending fecal mass partially obstructed at lower rectal neck with extra straining due to forcing of the fecal mass through the stenosed rectal neck with resultant mucosal slackening. B, Gradual mucosal prolapse. Prolapsing mucosa overhangs top end of anorectal band. Arrow points to loose mucosa, which glides to share in hemorrhoid formation. Pectinate line is in its place. C, Firm attachment of the cutaneous lining of the rectal neck (c) to the anorectal band (b) is loosened by the hemorrhoid mass. The cutaneous lining is dragged downward, displacing the pectinate line (a) outside the rectal neck orifice. a, Pectinate line. b, Anorectal band. c, Cutaneous lining of rectal neck. (A, B, Modified from Shafik A, Mohi-el-Din M. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defaecation. XXII. The pathogenesis of hemorrhoids and their treatment by anorectal bandotomy. J Clin Gastroent 1984;6:129-137; with permission; C, Courtesy of Prof. Ahmed Shafik, MD.)

The 25% incidence of hemorrhoids in our population above 45 years corresponds with the 20% incidence of persisting anorectal bands in a series of normal cadaveric specimens.285 This suggests an individual susceptibility to hemorrhoids related to the presence of the anorectal band. It appears that the constricting nature of the anorectal band not only leads to mucosal prolapse and hemorrhoid formation, but is also responsible for the higher fissure incidence in patients with hemorrhoids, in comparison to patients without hemorrhoids.290 Moreover, the constricting effect of the anorectal band could explain the manifestations in the syndrome of “urethral discharge, constipation, and hemorrhoids.”291 In this syndrome, the hemorrhoid patient experiences urethral discharge only when he is constipated. The discharge results from the squeezing effect of the descending hard fecal mass on the prostate, seminal vesicles, and vasal ampullae which are compressed, consequent to lower rectal neck congestion induced by hemorrhoids. In these patients, the anorectal band seems not only to cause hemorrhoid formation, but also to augment the squeezing effect of the hard feces.291

Shafik’s studies have shown that hemorrhoids are a “disease” in which the rectal masses represent a “manifestation,” but not the cause.285,290 Both the high rectal neck pressure and hemorrhoid masses are attributable to lower rectal neck stenosis due to the constricting action of the anorectal band and the failure of rectal neck remodeling with a resulting mucosal prolapse and venous congestion.290 In light of these findings, a new technique has been devised for treatment of hemorrhoids consisting of 2 main steps:292 (1) anorectal bandotomy (Fig. 18-95) involving division of the anorectal band and (2) ligation of the hemorrhoid mass at its base (Fig. 18-96). The achieved results are satisfactory. Preoperative high rectal neck pressures normalized postoperatively.292 The procedure is easy, and prevents the otherwise not infrequent recurrence by dealing directly with the primary etiologic factor.

Fig. 18-95.

Anorectal bandotomy of the hemorrhoids. A, Site. B, Anorectal band in lower rectal neck. C, Shaded area showing the extent of division. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-96.

Steps in hemorrhoid ligation. A, Tip of hemorrhoid mass is seized in an artery forceps. B, Curved forceps applied to hemorrhoid pedicle and a ligature placed proximal to it. C and D, Ligatures are tied. (Courtesy of Prof. Ahmed Shafik, MD.)

Role of the Anorectal Sinus in Pruritus Ani

The cause of pruritus ani was unknown in more than 50% of the cases until the detection of epithelial cells in the dermis of the lower rectal neck offered new possibilities regarding its etiology.293 The cells were oval and arranged in sheets, showed acinar arrangements of columnar cells, or were arranged in sheets of squamous epithelium (Fig. 18-97). They were swollen and vacuolated, with the nucleus displaced and edematous. Lymphocytic aggregations were found in close proximity to the ectopic epithelial cells. Occasionally the pruritic areas were found heavily infected, and epithelial debris could not be detected.

Fig. 18-97.

Photomicrographs showing epithelial debris in the deep dermis of rectal neck of pruritus ani. The cells, arranged in sheets, are oval, swollen and vacuolated with the nucleus pushed aside. Lymphocytic aggregations are in the neighborhood of the epithelial debris. Haematoxylin and eosin. A, Magnification x 50. B, Magnification x 220. a, Epidermis. b, Epithelial cells. c, Lymphocytic aggregations. (Courtesy of Prof. Ahmed Shafik, MD.)

Epidermal hyperplasia, prickle cell hydrops, dermal edema, and lymphocytic aggregations, as well as absence of leucocytes, are changes which are characteristic of dermatitis due to “sterile” irritation. The irritant factor seems to exist locally, because pruritus affects no other parts of the body. The local sterile irritative changes are probably produced by the epithelial elements, which are believed to be anorectal sinus remnants.293 These changes are initiated by the antigenic action of either the epithelial cells or their secretions, as evidenced by epidermal hydropic degeneration, as well as by edema and infiltration of the dermis with plasma cells and lymphocytes.293 Where epithelial debris could not be identified it was assumed that it was destroyed by infection. Pruritus ani seems to be an “autoimmune reaction” resulting from cell antigenicity induced by buried “epithelial remnants” which probably originate in the anorectal sinus.

The treatment of pruritus ani aims at destruction of the epithelial debris in the submucosa of the lower rectal neck as the primary cause of pruritus ani. Injection of a 5% solution of phenol in almond oil into the submucosa of the lower rectal neck under short duration general anesthesia achieves a cure.294 Relapse, if any, might be ascribed to incomplete epithelial debris destruction, as a result either of insufficient dosage or improper placement of the solution; it is easily cured with a second injection.

Surgical Anatomy and Physiology

Anatomy of Anal Sphincters and Pelvic Floor

Smooth Musculature

In the past our conventional textbooks described the anatomic concept of this area with the classical information: The internal anal sphincter starts at the anorectal junction by thickening of the rectal circular muscle coat (Fig. 18-98). The sphincter is up to 5 mm thick. It has a rounded lower border which lies in the lowest part of the external anal sphincter. The sphincter consists of smooth muscle bundles which run obliquely at the proximal and distal ends of the muscle and horizontally in its middle part.

Fig. 18-98.

The anorectal musculature. (Courtesy of Prof. Ahmed Shafik, MD.)

At the anorectal junction, the rectal longitudinal muscle coat fuses with inferiorly directed fibers of the pubococcygeus to form a conjoined longitudinal muscle layer (Fig. 18-98). Strands from this muscle layer penetrate the internal anal sphincter and the lower part of the external anal sphincter; some fibers reach the ischiorectal (ischioanal) fossa, the perianal skin and the mucous membrane of the intersphincteric groove. Puckering of the perianal skin was ascribed to some of these fibroelastic strands, while other investigators attributed it to the corrugator cutis ani muscle which is considered part of the panniculus carnosus.

This picture differed after Shafik’s studies105,282 revealed that the longitudinal anal muscle consists of 3 layers: medial, intermediate, and lateral (see Figs. 18-76, 18-81, 18-87), each having a different origin and being separated from the other by a fascial septum. The medial layer is a continuation of the longitudinal rectal muscle coat, and the intermediate layer represents the downward prolongation of the levator plate, whereas the lateral muscle is the longitudinal extension of the top loop. The longitudinal muscle ends at the level of the lower border of the internal sphincter by a fibrous condensation called the central tendon (Figs. 18-76, 18-87). The tendon splits into multiple small fibrous septa which include a medial and a lateral septum as well as intermediate septa (see Fig. 18-81). The intermediate septa penetrate the base loop of the external anal sphincter, then they split and decussate to form the corrugator cutis ani which inserts into the perianal skin. The medial septum proceeds medially to attach to the anal lining between the internal sphincter and base loop of the external sphincter. The lateral septum passes laterally into the ischiorectal fossa.

Striated Musculature

Classically, the external anal sphincter has been described as having deep, superficial, and subcutaneous parts. The deep part consists of circular fibers. The superficial part is elliptical, being attached to the tip of the coccyx posteriorly and to the perineal body anteriorly. The subcutaneous part is a circular ring of fibers.

This description did not comply with the findings of Shafik’s anatomic investigations,163,282 which could demonstrate that the external anal sphincter is a triple loop system (Figs. 18-76, 18-92). The top loop consists of the deep portion of the external anal sphincter fused with the puborectalis muscle. It sends a downward prolongation which shares in the formation of the longitudinal muscle (Figs. 18-76, 18-92). The top loop forms a U-shaped ribbon around the upper rectal neck and is attached to the symphysis pubis. The intermediate loop is attached to the coccyx, and the base loop to the perianal skin anteriorly. Each of the 3 loops is enclosed in a separate fascial compartment (Fig. 18-76).

The levator ani has been defined as consisting of iliococcygeus, pubococcygeus, and puborectalis; their fibers are in continuity from the ischial spine to the body of the pubis, across the white line of the obturator fascia. The iliococcygeus portion arises from the posterior half of the white line and the ischial spine and inserts into the coccyx and the anococcygeal raphe. The pubococcygeus arises from the anterior half of the white line and the body of pubis and is inserted into the coccyx and anococcygeal raphe. The puborectalis is attached anteriorly to the pubic body, passes backwards around the anorectal junction and meets with fibers of the opposite side to form a U-shaped sling.

Shafik’s studies282 have confirmed that the levator ani consists of the pubococcygeus and iliococcygeus, but have revealed that the puborectalis belongs to the external anal sphincter, not to the levator ani:282 the puborectalis fuses with the deep external anal sphincter to form the top loop of the external anal sphincter. The pubococcygeus is funnel-shaped (Fig. 18-76); it consists of a transverse and a vertical portion.282 The transverse portion on both sides forms a platelike structure called the levator plate (Figs. 18-76, 18-77). At the anorectal junction, the levator plate bends sharply downwards to form the vertical portion called the anal suspensory sling.282 The latter constitutes the middle layer of the longitudinal anal muscle. The levator plate is connected to the anorectal junction by a fibrous condensation called the hiatal ligament.

Mechanism of Continence

A great deal of controversy has arisen as to the relative importance of the external and internal anal sphincters in the mechanism of anal continence. Thus, Goodsall295 and Allingham296 believed that the internal sphincter is a more important muscle than the external one, whereas Tuttle297 and Lockhart-Mummery298 claimed the reverse. This conflict was the result of inaccurate anatomic information gained at that time.299 Milligan and Morgan’s account300 contributed somewhat to the clarity of the anatomy of the anal musculature.

Anal continence is a joint function of the external and internal anal sphincters; the former is responsible for voluntary continence, whereas the latter maintains involuntary continence.142 Contrary to the views of investigators holding that the levator ani has a sphincteric action on the anorectal junction,299,300 it acts as the rectal neck dilator.301 Functioning mainly during defecation, the levator ani plays no role in anal continence. On contraction, the levator plate is elevated and retracted laterally, pulling the hiatal ligament which opens the rectal neck inlet301 (Fig. 18-81). Furthermore, contraction of the anal suspensory sling of the levator muscle opens up and shortens the rectal neck.

The part played by the longitudinal muscle in anal continence is minimal. The corrugations formed by the corrugator cutis, which result from the insertion of the longitudinal muscle into the perianal skin, help the base loop of the external anal sphincter induce an airtight anal occlusion.105 The muscle’s major role occurs during defecation. Its contraction results in shortening and opening of the rectal neck105 (Fig. 18-81). Furthermore, it fixes the rectal neck during fecal descent to prevent anal prolapse.

The internal anal sphincter is responsible for involuntary continence. Being unstriated, it maintains continence for long periods. However, it cannot resist a call to stool or flatus; that is, it cannot maintain voluntary continence, because it is reflexively relaxed when the rectal detrusor contracts on entrance of stools or flatus. There is a reciprocal action between the detrusor and the internal sphincter: when one contracts, the other relaxes reflexively.142,302 Accordingly, it is the reflex action that makes the internal sphincter relax and prevents it from opposing defecation. This is shown clinically by the fact that a patient with rectal inertia is continent, despite the rectum being full of stool and the external sphincter being relaxed. The explanation for this could well be that no reflex internal sphincter relaxation occurs, because the detrusor contraction is lost by inertia. Continence in such cases is maintained by the internal sphincter.

The external anal sphincter is responsible for voluntary continence. It contracts to resist a call to stool, or to interrupt or terminate the act of defecation.163 It seems that the muscle induces continence by a double action:163 (1) prevention of internal sphincter relaxation on detrusor contraction –an action effected by what Shafik refers to as the “voluntary inhibition reflex”– and (2) direct compression of the rectal neck, or the “mechanical action.”

Voluntary Inhibition Reflex

As stool enters the rectum, the detrusor contracts and the internal sphincter reflexively relaxes to unseal the rectal neck (Fig. 18-99). However, the neck does not open unless the external sphincter has voluntarily relaxed. If there is no desire to defecate, the external sphincter contracts, thereby mechanically preventing relaxation of the internal sphincter (Fig. 18-100). Failure of the internal sphincter to relax reflexively inhibits contraction of the detrusor which relaxes and dilates to accommodate the new contents.163 Commonly the individual, while opposing defecation, senses the sudden detrusor relaxation which occurs due to failure of the internal sphincter to relax (a characteristic sensation, especially if the rectum contains gases); this is followed by loss of the desire to defecate.

Fig. 18-99.

External and internal sphincters at rest and during defecation. A, At rest: detrusor relaxed and internal sphincter involuntarily contracted. B, During defecation: detrusor contracted and external and internal sphincters relaxed. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-100.

Mechanism of voluntary inhibition reflex action to oppose a call to stool. A, Detrusor contraction with failure of internal sphincter relaxation due to voluntary external sphincter contraction. B, Reflex detrusor relaxation due to failure of internal sphincter relaxation (voluntary inhibition reflex action). (Courtesy of Prof. Ahmed Shafik, MD.)

Voluntary external sphincter contraction to prevent reflex internal sphincter relaxation, with subsequent inhibition of detrusor contraction, is what could be called “voluntary inhibition reflex.”163 The latter seems to be the main action responsible for voluntary continence; it leads to immediate detrusor relaxation and loss of the desire to defecate. For this reason, internal sphincter integrity is necessary not only for involuntary, but also for voluntary continence since the action of the voluntary inhibition reflex is mediated through it. Voluntary inhibition is further potentiated by the base loop contraction acting on the longitudinal muscle.163 Thus, on external sphincter contraction, the base loop glides medially, stretching the longitudinal muscle tight by pulling on the fibrous prolongations penetrating the base loop substance. The tightened longitudinal muscle limits internal sphincter relaxation and thus augments the voluntary inhibition action.

Voluntary Mechanical Action

In addition to its voluntary inhibition reflex action, external sphincter contraction firmly seals the rectal neck by mechanical compression.163 Being striated, the external sphincter cannot contract for a long period to maintain continence mechanically. The mechanical compression action is thus brief, and serves to occlude the rectal neck by the time the detrusor relaxes as a result of the action of the voluntary inhibition reflex. If the detrusor does not relax, the external sphincter tires out and relaxes, and defecation should occur. Once the detrusor relaxes and the desire for defecation wanes, the external sphincter relaxes and continence is then maintained by the internal sphincter. However, after some time, the loaded detrusor recontracts; if the desire to defecate is still opposed, the external sphincter contracts again. This process is repeated until either defecation is acceded to or a prolonged contraction of the unstriated detrusor tires out the striated — and briefly contracting — external sphincter which involuntarily relaxes, leading to internal sphincter relaxation and opening of the rectal neck.163

Stress Defecation

In conditions of internal sphincter damage and resulting loss of the voluntary inhibition reflex, voluntary continence is induced only by the mechanical action of the external sphincter. Being striated, the external sphincter cannot contract long enough to withstand the uninhibited prolonged contraction of the loaded detrusor. Thus, detrusor contraction continues despite external sphincter contraction until the latter sphincter fatigues and relaxes and the detrusor forces evacuation163 (Fig. 18-101). Accordingly, in cases of internal sphincter damage, once the desire to defecate is initiated, evacuation should occur. This condition, which is referred to as “stress defecation,” is observed in patients after internal sphincterotomy for anal fissure.163 This not only underscores the importance of the internal sphincter in the mechanism of voluntary continence, but could also offer an explanation for the impaired control of flatus and liquid feces which occurs after internal sphincterotomy, as reported by Bennett and Goligher.303

Fig. 18-101.

Mechanism of stress defecation. A, Detrusor contraction with external sphincter contraction; internal sphincter is damaged. B, Detrusor continues contraction, uninhibited by the damaged internal sphincter; external sphincter fatigues, relaxes, and defecation occurs. (Courtesy of Prof. Ahmed Shafik, MD.)

Single Loop Continence Theory

The external sphincter consists of three loops (Fig. 18-102): top, intermediate, and base. Each has its own attachments, direction of muscle bundles, and innervation, which are separate and different from the others.164 Each loop also has its own fascial investment (Fig. 18-103). The top loop is directed forward, with an upward inclination to be attached to the symphysis pubis and the adjoining pubic bone; it is innervated by the inferior rectal branch of the pudendal nerve. The intermediate loop proceeds horizontally backward to gain attachment to the coccyx, and is supplied by the perineal branch of the fourth sacral nerve. The base loop is directed forward and slightly downward, and is attached anteriorly to the perianal skin in, and close to the midline; it is innervated by the inferior rectal nerve.

Fig. 18-102.

Triple loop system of external sphincter. (Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. V. The rectal neck: Anatomy and function. Chir Gastroenterol 1977;11:319; with permission.)

Fig. 18-103.

Coronal section of rectal neck (Verhoeff-van Gieson; magnification x 6). Each of the three loops of the external sphincter has its own fascial covering. a, Top loop. b, Intermediate loop. c, Base loop. d, Fascial septa. e, Longitudinal muscle. f, Internal sphincter. (From Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. I. The external anal sphincter. A triple-loop system. Invest Urol 1975;12:412-419; with permission.)

As a result of the arrangement of these parts of the external sphincter, and also because of the fact that each loop has its own innervation, any single loop can function as a sphincter separately.163 Accordingly, the external sphincter action in continence can be achieved by the contraction of one singular element and not necessarily the three loops. This constitutes the basis of the “single loop continence” theory.163 On contraction, a single loop induces continence by the action of the voluntary inhibition reflex and by mechanical occlusion. The latter action is significantly tight, as loop contraction is effected not only by direct compression but also by anal kinking.164 The understanding of the “single loop continence” theory is of paramount surgical significance in planning for the treatment of anorectal abnormality, especially fistulas and incontinence. Thus, “unless all three of the loops are destroyed, continence can be maintained by any single loop.”163

Fallacy of the Anorectal Ring

The anorectal ring is the name given by Milligan and Morgan300 to the muscular ring which encircles the anorectal junction, and is composed of the upper borders of the internal and external sphincters and puborectalis. It is claimed that this ring is responsible for anal continence and that its complete division results in incontinence.299,300 In the light of recent studies, this ring represents only the upper border of the top loop of the external sphincter, and plays no specific role in the mechanism of anal continence in comparison to the intermediate and base loops.164,301

As has already been mentioned, anal continence is the result of the action of the voluntary inhibition reflex as well as of mechanical anal compression. These two actions could be achieved by any one of the three loops of the external sphincter, not necessarily the top, although the action of the top loop is most advantageous because of the natural kink at the rectal neck inlet which shares, although minimally, in the actual mechanism of anal continence.

The present view does not conflict with the opinion of other investigators299,300 however, who hold that division of the anorectal ring in the treatment of high anal fistula leads to incontinence. The explanation could well be that in such cases the intermediate and base loops have also been divided with the result that the integrity of all three of the loops has been destroyed, and incontinence is inevitable. Failure to realize this fact has led investigators to overestimate the role of the anorectal ring in anal continence. If one single loop, intermediate or base, could be spared in the treatment of high anal fistula, the patient will be continent in spite of division of the top loop.163

Role of Internal Sphincter in Management of Anal Incontinence

The significant role of the internal sphincter in voluntary continence should be considered in the repair of anal incontinence after traumatic damage of the external sphincter because frequently, in such cases, the internal sphincter is destroyed as well. A successful external sphincter repair should aim, however, at effecting not only mechanical occlusion but also voluntary inhibitory reflex action.163 This necessitates repair of both the external and internal sphincters.

It seems that neglecting repair of the internal anal sphincter would explain the unsatisfactory operative results that are commonly obtained in such cases. In traumatic incontinence, the external and internal sphincters are commonly amalgamated at the injured site, in a mass of fibrous tissue. Internal sphincter injury would cause the rectal detrusor to lose its inhibitory component. Once the detrusor contracts, it immediately evacuates through the unguarded rectal neck. The operation practiced conventionally for repair of anal incontinence calls for suturing of the external sphincter;304 whereas the internal sphincter is either ignored or included in the repair of the external sphincter. In both of these procedures the internal sphincter loses its voluntary inhibition reflex action as well as the ability to effect involuntary rectal neck closure.

A sound repair should free the internal sphincter from the external and suture each muscle separately. This allows the external sphincter to induce not only mechanical occlusion but also voluntary inhibition reflex action –mediated through the internal sphincter as described earlier.163 Plastic operations for anal incontinence to supplement the external sphincter with muscle grafts or fascial slings305 should also be planned to induce voluntary inhibition reflex action, rather than mechanical compression. Accordingly, the graft should be applied to surround the internal sphincter. Furthermore, in patients in whom tone of the anal sphincter is to be restored by electronic implants,306 it is better to sew the electrodes to the internal sphincter than to the external. In the former, the voluntary inhibition reflex action is induced by direct stimulation and contraction of the internal sphincter.163 This is more effective than applying the electrodes to the damaged external sphincter which lacks the efficient contraction to produce either mechanical compression or voluntary inhibition reflex action.

Involuntary Action of the External Anal Sphincter

Studies of the histologic structure of the striated external anal sphincter following internal anal sphincter excision in dogs have revealed a distinct preponderance of smooth muscle fibers around the 10th postoperative month; this was confirmed by manometric and electromyographic experiments.307,308 The increased nonstriated element in the external anal sphincter seems to be a structural-functional adaptation, meaning that the external anal sphincter, now a “compound muscle,” takes on the otherwise lost involuntary function of the excised internal anal sphincter muscle.

Artificial Rectal Neck Pressure to Assist Anal Sphincteric Action

In partial fecal incontinence when (for any of a variety of reasons) the rectal neck pressure has suffered a significant drop below normal, “filler”-injections into the anal submucosa produce an effect that can assist the diminished anal sphincteric activity by increasing, and eventually normalizing, the rectal neck pressure, thereby restoring fecal continence. There are two types of injections: perianal injection of polytetrafluoroethylene and perianal injection of autologous fat.

Perianal Injection of Polytetrafluoroethylene

In the treatment of partial fecal incontinence,309 the etiology of which may be idiopathic or following internal sphincterotomy, Teflon or polytef paste can be used. Five milliliters of polytef paste are injected, without anesthesia, into the rectal neck submucosa above the pectinate line, at both the 3 and 9 o’clock positions. Cure is usually achieved after the 1st injection or, in some cases, after repeat injections. Improvement is believed to be due to the increase in rectal neck pressure produced by the cushion effect of the polytef submucosal injection. The technique is simple, easy, and without complications. It is performed on an outpatient basis.309

Perianal Injection of Autologous Fat

For the second type of injection, 50-60 ml of fat are harvested from the abdominal wall, thoroughly washed, and injected submucosally into the rectal neck at 3 and 9 o’clock positions.310 The treatment is effective and achieves normalization of the pre-injection low rectal neck pressure. Occasional failures are due to insufficient washing of the fat or improper positioning of the needle; two to three repetitions of the injections achieves full fecal continence. The body fat is easily obtainable, biocompatible, and inexpensive. The technique is simple, easy, cost-effective, and performed on an outpatient basis.

Mechanism of Defecation

The complexity of the mechanism of defecation seems to require a brief description before we look into the details of its anatomic, physiologic, and pathologic aspects. Then, having acquired a more profound understanding of the essential interrelations, we can probe the treatment modalities that have been developed.

The levator ani, which is the muscle of defecation,282 is funnel-shaped, with a transverse part called the levator plate, and a vertical part called the suspensory sling282 (see Fig. 18-76). The hiatal ligament connects the levator plate to the rectal neck.301 The levator ani muscle is the rectal neck dilator. As a striated muscle, it contracts voluntarily at defecation to open the rectal neck to allow the stool to pass externally311 (see Fig. 18-81). The puborectalis muscle is a U-shaped flat muscle that embraces the upper part of the rectal neck and acts as a constrictor muscle. It contracts to oppose, interrupt, or terminate the act of defecation.282,301

The concerted functions of the anorectal musculature at defecation are initiated and harmonized by voluntary impulses and reflex actions. When the rectal detrusor is distended with fecal mass and the stretch receptors are stimulated, the rectoanal-inhibitory reflex is initiated, and thus the rectal detrusor contracts and the internal sphincter relaxes. Detrusor contraction triggers two other reflexes: the rectopuborectalis reflex312 and the rectolevator reflex.313 Although acting simultaneously, these two reflexes have opposite functions: the rectolevator reflex effects a reflex levator contraction that opens the rectal neck, whereas at the same time, the puborectalis reflex actuates the reflex puborectalis contraction. Puborectalis contraction functions to seal the rectal neck, or keep it closed as impulses reach the conscious level to probe the circumstances for defecation. If time or circumstance is inopportune, the puborectalis continues voluntary contraction.

Voluntary puborectalis contraction evokes two reflex actions: the reflex levator relaxation through the levator-puborectalis reflex,314 and the reflex detrusor relaxation by means of the voluntary inhibition reflex.163 Furthermore, it aborts the rectoanal-inhibitory reflex which relaxes the internal sphincter. Hence, voluntary puborectalis contraction, through the voluntary inhibition reflex, prevents internal sphincter relaxation which results in reflex detrusor relaxation and waning of the urge to defecate (Fig. 18-100). However, as soon as the sensation of desire is felt, and circumstances are appropriate for defecation, the puborectalis muscle relaxes voluntarily and the detrusor evacuates its contents. This demonstrates that the act of defecation is under voluntary control despite the presence of reflex actions sharing the mechanism of defecation.

Thus, although the rectoanal-inhibitory and rectolevator reflexes function to open the rectal neck, the rectopuborectalis and the levator puborectalis reflexes keep the rectal neck closed until the decision for defecation has been taken. Straining at the start of defecation is a normal physiologic process and as such is part of the mechanism of defecation. By elevating the intraabdominal pressure, straining triggers the straining-levator reflex315 which effects levator contraction and the opening of the rectal neck for spontaneous evacuation of the stools.

Deflation Reflex: Role in Defecation

Contraction of the external anal sphincter on rectal “inflation” was reported in 1935 by Denny-Brown and Robertson315 and was confirmed by other investigators.316,317 Rectal inflation is accompanied by internal sphincter relaxation and momentary external sphincter contraction.316,318,319 Shafik could demonstrate that the external sphincter contracts also upon rectal “deflation” through a reflex he calls “deflation reflex.”320

The recto-anal inhibitory reflex comprises internal sphincter relaxation simultaneously with external sphincter contraction. It seems that the latter serves to close, momentarily, the rectal neck by the time the impulses reach the conscious decision whether to evacuate or not. If conditions are favorable, the external anal sphincter relaxes and defecation occurs. If the conditions are inopportune, external sphincter contraction continues preventing internal anal sphincter relaxation which leads to reflex rectal detrusor relaxation by means of the voluntary inhibition reflex.163

Reflex external anal sphincter contraction on rectal “deflation” is believed to play a role at defecation as well.320 It functions to “interrupt” or “terminate” the act of defecation. Thus, at stool, the rectum contracts upon receiving fecal material from the sigmoid colon. The rectal neck opens and the rectal contents are delivered to the outside. Once the rectum is empty and becomes deflated, the external sphincter reflexively contracts.320 To “interrupt” defecation, the base loop contraction of the external sphincter amputates the fecal column, expelling the lower portion. Following this, the rectal neck closes progressively upward by successive intermediate and top loop contractions, returning to the rectum any fecal material trapped in the rectal neck.163 This process is repeated as long as the rectum receives fecal material from the sigmoid colon. At the end of defecation, reflex external sphincter contraction helps to “terminate” the act. The last fecal portion is pushed down the rectal neck to the outside by a process of “vermicular” contractions (Fig. 18-104) induced by the external sphincter triple loop system.164

Fig. 18-104.

Vermicular contractions of the external sphincter triple-loop. A, Fecal mass lies within the relaxed external sphincter. B, Top loop contraction and intermediate loop relaxation. C, Intermediate loop contraction and base loop relaxation. D, Base loop contraction. (Modified from Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. 1. The external anal sphincter. A triple-loop system. Invest Urol 1975;12:412-419; with permission.)

Voluntary Inhibition Reflex and the Deflation Reflex

The rectal detrusor and internal anal sphincter have a reciprocal action;317,319 when one contracts, the other reflexively relaxes. Thus, at defecation, rectal detrusor contraction to evacuate its contents is associated with reflex internal sphincter relaxation and opening of the rectal neck. Detrusor contraction is involuntary and can be maintained for long periods without being exhausted. The internal sphincter is kept relaxed as long as the rectal detrusor remains contracted. However, at the end of detrusor contraction, the deflation reflex comes into action. Upon detrusor contraction and internal sphincter relaxation, and after the rectum is completely evacuated and deflated, the external anal sphincter reflexively contracts. External sphincter contraction impedes relaxation of the internal sphincter which leads to reflex detrusor relaxation.163 Thus, the “deflation reflex” functions to terminate internal sphincter relaxation and detrusor contraction. Consequently, the detrusor relaxes to receive fresh fecal material or to terminate defecation.

Role of the Deflation Reflex as a Diagnostic Tool

The deflation reflex may prove of diagnostic significance in defecation disorders. Detectable changes in latency and amplitude of the evoked response would indicate a defect in the reflex pathway such as a muscle or nerve damage from a disease of the spinal cord, spinal nerve roots, or peripheral nerves, or from a central lesion. The reflex may thus be incorporated as an investigative tool in the study of patients with anorectal disorders.

Anatomic Components of the Mechanism of Defecation

Levator Plate

The levator plate muscle bundles are arranged into two main groups, each with its own attachments, direction, and function: a) the lateral bundles, which form the “lateral mass,” and b) the medial bundles which form two “crura”282 (Fig. 18-105).

Fig. 18-105.

Diagram illustrating the levator plate with its two lateral masses and two crura. (Courtesy of Prof. Ahmed Shafik, MD.)

The lateral mass is triangular in shape and arises with a wide base from the obturator internus fascia. The muscle bundles converge as they pass backward, with medial and downward inclination to be attached to the coccyx (Fig. 18-105). The lateral mass has a free lateral border separated from the iliococcygeus by a triangular gap.

The most medial fibers of the levator plate form two crura which bound the levator hiatus (Fig. 18-105). Each crus has its separate origin from the back of the lower part of the pubic body, about 2 cm above its lower border, and proceeds backward, with an upward inclination as a horizontal strip of fleshy bundles. In the midline posteriorly, the fleshy bundles of the two crura become tendinous and decussate in a crisscross pattern, forming the anococcygeal raphe. Laterally, each crus merges with the corresponding lateral mass. Medially, the inner borders of the two crura are connected to the intrahiatal structures along the hiatal ligament. The muscle bundles of the two crura are not directly attached to any of the pelvic viscera. Three patterns can be identified for the origin of the crura from the pubic body (Fig. 18-106): the classic pattern, crural overlap, and crural crossing.282

Fig. 18-106.

Diagram illustrating crural patterns. A, Classic pattern. B, C, Crural overlap. D, Crural crossing. (Courtesy of Prof. Ahmed Shafik, MD.)

The classic pattern of origin of the levator crura (Fig. 18-106A) is the most common. The two crura arise from the body of the pubic bone almost side by side, without overlap or crossing. The gap between the two crura at their origin is occupied by the puboprostatic or pubovesical ligaments.

In the crural overlap pattern (Figs. 18-106B,C), the proximal ends of the two crura overlap at their origin from the symphysis pubis. The crura may arise one above the other, or originate from a common tendinous origin at the back of the symphysis pubis and adjoining pubic bone, and then split into two slips.

In the crural “scissor” (crossing) (Fig. 18-106D), the right crus arises from the left pubic body and the left crus from the right one. They cross each other by overlapping one limb over the other or by interdigitation of the fibers from both limbs.

Levator Tunnel

The intrahiatal structures, namely the rectal neck and the prostate in the male, and the rectal neck with vagina and urethra in the female, are ensheathed along their way down from the levator hiatus to the perineum in a muscular tube to which Shafik has applied the name “levator tunnel”282 (Figs. 18-76, 18-77, 18-107, 18-108). The tunnel’s posterior wall is longer than its anterior, due to the obliqueness of the levator plate. The length of the posterior wall in adults varies from 3 to 4 cm, and that of the anterior wall from 2.5 to 3 cm. The tunnel wall consists of two muscular coats of striated bundles (Figs. 18-76, 18-77, 18-109): an inner longitudinal coat formed by the suspensory sling, and an outer loop formed by the puborectalis.

Fig. 18-107.

Paracoronal section of vesical detrusor and neck in female cadaver (Verhoeff-Van Gieson stain, magnification x 5). a, Urinary bladder. b, Levator plate. c, Hiatal ligament. d, Longitudinal muscle. e, Internal sphincter. f, Suspensory sling. g, Puborectalis. h, External urethral sphincter. i, Tunnel septum. (From Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. VIII. Levator hiatus and tunnel. Anatomy and function. Dis Colon Rectum 1979;22:539-549; with permission.)

Fig. 18-108.

Coronal section of the rectal neck shows the levator tunnel. a, Levator plate. b, Suspensory sling. c, Top loop (fused puborectalis and deep external anal sphincter). d, Intermediate loop of external anal sphincter. e, Base loop of external anal sphincter. f, Internal anal sphincter. g, Longitudinal anal muscle. k, Pelvirectal space. l, Ischiorectal (ischioanal) space (hematoxylin and eosin; magnification x 7). (From Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. VIII. Levator hiatus and tunnel. Anatomy and function. Dis Colon Rectum 1979;22:539; with permission.)

Fig. 18-109.

Diagram illustrating the “individual” sphincters arising from the puborectalis which acts as a “common” sphincter for the intrahiatal structures. (Courtesy of Prof. Ahmed Shafik, MD.)

The suspensory sling forms the inner coat of the levator tunnel. At the level of the levator hiatus, the levator plate bends sharply downward to form a vertical muscular cuff around the intrahiatal organs called the “suspensory sling”301 (Figs. 18-76, 18-77, 18-107, 18-108). Detailed study of the latter sling has shown that it consists of longitudinally arranged striated muscle bundles impregnated with collagen. It is separated from the intrahiatal structures by a fascial septum which can be called the “tunnel septum.”282 From a descriptive viewpoint, the suspensory sling can be divided into two parts: anal and urethral.

The anal suspensory sling extends along the back and sides of the rectal neck. In the lower part, it fuses with the longitudinal muscle coat of the rectal neck, and then splits into multiple fibrous septa which penetrate the base loop of the external anal sphincter and insert into the perianal skin (see Fig. 18-76).

The urethral suspensory sling constitutes the anterior portion of the suspensory sling. It extends alongside the urethra and vagina in females, and the prostate in males. It ends by splitting into multiple septa which penetrate the substance of the external urethral sphincter. In females, it inserts in the skin around the external urethral meatus, whereas in males it merges with the intermuscular septa of the external urethral sphincter282 (Fig. 18-107).

The outer loop of the tunnel wall consists of the puborectalis (Figs. 18-77, 18-109). As noted above, the muscle arises as a vertically oriented band by means of a tendinous attachment to the lower part of the symphysis pubis and adjoining pubic bone below the levator crural attachment. The aponeurotic origin gives rise to fleshy bundles, and the muscle proceeds backward with a downward inclination, underlying the inner border of the levator crura, and loops around the back of the rectal neck.282 Behind the rectal neck, it becomes continuous with the corresponding muscle of the opposite side to form a U-shaped muscle, without intervention of the anococcygeal raphe. In no part along its extent is the muscle attached to the levator plate.

As the puborectalis proceeds backward from its origin, it gives off muscle bundles to each intrahiatal organ forming the “individual” voluntary sphincters for these organs282 (Fig. 18-109). In the male, the isolated puborectalis fibers surround the membranous urethra with a sheath of striated muscle bundles forming the external urethral sphincter (Figs. 18-110, 18-111). In the female, they give rise to the external urethral sphincter and, further back, provide the vaginal tube with another muscular sheath, the vaginal sphincter. Posteriorly, the puborectalis, in both sexes, gives off loop fibers to embrace the rectal neck and to form the deep part of the external anal sphincter. However, the puborectalis and deep external anal sphincter were found fused together to such an extent that they could not be differentiated either morphologically or histologically; the conjoined muscle was termed “top loop.”169,301 The origin of the individual sphincters from the puborectalis muscle (PRM) could be demonstrated not only anatomically but also physiologically.

Fig. 18-110.

Coronal section of the urinary bladder and prostate shows the levator tunnel. a, Urinary bladder. b, Prostate. c, Prostatic urethra. d, Membranous urethra. e, External urethral sphincter. f, Puborectalis. g, Levator plate. h, Suspensory sling inserted into external urethral sphincter. i, Rectal neck (opened). (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-111.

Coronal section of the urinary bladder and prostate shows that the puborectalis gives rise to the external urethral sphincter. a, Urinary bladder. b, Prostate. c, Prostatic urethra. d, Membranous urethra. e, Puborectalis. f, External urethral sphincter. g, Levator plate. h, Hiatal ligament. i, Suspensory sling. (Courtesy of Prof. Ahmed Shafik, MD.)

Origin of External Anal, Urethral, Vaginal, and Prostatic Sphincters

An anatomic study282 has demonstrated that the puborectalis muscle gives origin to the following sphincters: external anal (EAS), external urethral (EUS), prostatic (PS), and vaginal (VS). The response of EAS, EUS, PS, and VS to stimulation of the PRM was studied with the aim of physiologic validation of their anatomic origin from the PRM.321 Twenty-eight healthy volunteers were examined (16 men, 12 women, mean age 40.6 ± 8.3 SD years).

The PRM was stimulated by a needle electrode. The response of the EAS and EUS was recorded by needle electrode, whereas the PS and VS were evaluated by manometric measuring of the urethral and vaginal pressures. Upon stimulation, the electromyogram (EMG) activity of the EAS and EUS increased, as did the prostatic urethral and vaginal pressures. There was no response to stimulation of the anesthetized PRM; this might indicate that the sphincters contract in response to PRM stimulation. The EMG recorded no latency, suggesting that the motor units of the EAS and EUS were simultaneously activated with those of the PRM and also that their muscle fibers seem to be directly derived from the PRM.

Mass Contraction of the Pelvic Floor Muscles

The above-mentioned studies demonstrated both anatomically and physiologically that the EAS, EUS, PS, and VS and the bulbocavernosus muscle (BC) originate from the PRM. It is hypothesized that stimulation of any of these muscles would lead to contraction of all the others. Because the levator ani (pubococcygeus) muscle (LA) also has the same innervation as the previously discussed muscles, it is further suggested that it, too, contracts reflexively upon stimulation of any of these muscles. Shafik tested this hypothesis in 18 healthy volunteers.322 EAS was stimulated and the response of the EUS, PRM, LA, and BC was determined. Each muscle was thereafter stimulated separately and the response of the other pelvic floor muscles was registered. Stimulation of any of the pelvic floor muscles effected an increased EMG activity of the rest of the muscles.322 The muscle contraction was instantaneous with no latency in any of the muscles except the LA EMG activity which showed a mean latency of 21.3 ± 6.6 ms. The response of the pelvic floor muscles seems to be attributable to muscle stimulation both directly and indirectly through activation of pudendal nerve fibers in the muscles. The study demonstrated that the pelvic floor muscles behave as one muscle; they contract or relax en masse. The “mass contraction” of the pelvic floor muscles might explain some of the physiologic phenomena that occur during pelvic organ evacuation. However, in addition to this mass contraction, a voluntary “selective” individual muscle activity exists by which each individual muscle would act independently of the others.323

Hiatal Ligament

The levator tunnel is connected to the intrahiatal organs by a fascial condensation called the “hiatal ligament”301 (Figs. 18-76, 18-77, 18-107). A detailed study282 of the ligament has demonstrated that this structure arises circumferentially around the hiatal margin as a continuation of the fascia on the pelvic surface of the levator plate. The ligament has its origin slightly lateral to the inner edge of the levator plate over which it passes to fill the space between the levator plate and intrahiatal structures. In this space the pelvic fascia is condensed to form the hiatal ligament. The ligament splits fan-wise into multiple septa which insert into the vesical neck and rectal neck inlet in both sexes as well as into the upper end of the vagina in females (Fig. 18-109). The inserting fibers penetrate the visceral fascia covering the intrahiatal organs, and become continuous with the fibers of their intermuscular fascia; some fibers attach directly to the adventitia of these organs. This arrangement gives firmness to the ligamentous insertion.

Histologically, the hiatal ligament consists of elastic fibers intermingled with collagen. No muscle bundles were detected within the ligament. Anteriorly, the hiatal ligament fills the gap between the two levator crura at their origin, forming the puboprostatic or pubovesical ligament.282 After giving rise to the hiatal ligament, the pelvic fascia over the levator plate continues downward as the “tunnel septum” (Figs. 18-76, 18-107) which lines the inner aspect of the levator tunnel.282

The pubovesical ligament represents only the anterior portion of the hiatal ligament.282 It extends backward from the lower part of the symphysis pubis, between the two crural origins, to insert, in both sexes, into the vesical neck in a way similar to that of the hiatal ligament. Contrary to the view of investigators that the ligament in males is attached to the prostate, it was observed that the main insertion is into the vesical neck, with only a few fibers attaching to the prostate. Accordingly, the term “pubovesical” is more appropriate for the ligament in both sexes. In the classic crural pattern, the ligament is well formed as it stretches between the two separated crural origins. However, in crural overlap or scissors, the ligament was smaller and shorter. The pubovesical ligament, being a part of the hiatal ligament, shares with the latter its functional activity as will be mentioned later.

The tunnel septum (Figs. 18-76, 18-107) is the downward continuation of the fascia on the pelvic surface of the levator plate.282 It lines the inner aspect of the levator tunnel and separates it from the fascia propria of the intrahiatal organs. It is loosely connected to these structures, being separated by a potential space filled with areolar tissue. It consists of collagen impregnated with a few elastic fibers. The septum not only separates the tunnel from its contents but demarcates the voluntary from the involuntary components of these contents. The understanding of this anatomic fact appears to be of major surgical importance because, with the line of cleavage along the tunnel septum, the intrahiatal structures could be mobilized from within the levator tunnel without injury to their voluntary sphincteric mechanism.282,324

Physiologic Considerations of the Mechanism of Defecation

The levator plate (Fig. 18-105) consists of two zones which seem to be functionally separate: a lateral “visceral support” zone represented by the two lateral masses, and a medial “dilator” zone represented by the two crura.282 The two lateral masses are fixed muscular sheets which, on contraction, together with the iliococcygeus, become elevated and function to “support” the viscera. The two crura represent the functionally active components of the levator plate.

A previous study has shown a free passage, across the anococcygeal raphe, of the tendinous fibers of the two levators, denoting the existence of a “digastric pattern.”301 Such a pattern allows the two crura to contract simultaneously as a single sheet. On levator contraction, the two crura become elevated and laterally retracted, resulting in “dilatation” of the levator hiatus (Fig. 18-112). The dilator crural action seems to be an effect of the crisscross arrangement of the anococcygeal raphe, the presence of which prevents the crural constrictive action on the intrahiatal contents.282 Thus, on crural contraction the anococcygeal raphe is shortened and broadened by the narrowing of the meshes between its fibers, with a resulting dilatation of the hiatus (Fig. 18-112).

Fig. 18-112.

Diagram illustrating the “dilator” action of the anococcygeal raphe (ACR). A, At rest: long and narrow. B, At crural contraction, anococcygeal raphe becomes short and broad (hiatal dilatation). (Courtesy of Prof. Ahmed Shafik, MD.)

If the crural muscle bundles were continuous on both sides without the intervention of the anococcygeal raphe, as is the case in the puborectalis, their contraction would constrict the intrahiatal structures. In the classical crural pattern, the separation of the two crura at their origin seems to provide them with the free mobility necessary for proper functioning. Deviation from the classical pattern occurred in 28%.282 Crural overlap, or crossing, may interfere with the mechanism of hiatal action such that full hiatal dilatation required during the act of defecation or urination may not be achieved. Nevertheless, a good overlap seems to keep the intrahiatal structures in position more firmly, thus reducing the risk of prolapse.282

Double Sphincteric Control

Each intrahiatal structure is provided with a double voluntary sphincteric apparatus: a) an “individual” organ sphincter derived from the puborectalis, which is specific for the organ, and b) a “common” tunnel sphincter, the bulk of the puborectalis, which acts on the intrahiatal organs collectively.282,323 Thus, nature’s way of providing separate sphincteric activity for the individual organs under the control of a common continent muscle secures not only an individual sphincteric function for the organ, but a harmonized action among the structures enclosed within the tunnel. Furthermore, the double sphincteric mechanism provided to each organ could be a guarantee of functional maintenance in case either of the two sphincters is damaged.323 It is suggested that injury of either sphincter alone is not sufficient to induce incontinence of the concerned organ. Hence, unless both sphincters, the “individual” and “common,” are destroyed, continence could be maintained by either.282,323

Levator Complex

The levator tunnel seems to occupy an important role in the mechanisms of defecation and urination. It is provided with two striated muscle coats which differ in their morphologic structure and function: an inner longitudinal (the suspensory sling) which is a tunnel “dilator,” and an outer loop (the puborectalis) which is a tunnel “constrictor.”282,301 The two coats act reciprocally: when one contracts, the other relaxes. During active contraction, the inner tunnel coat shares in opening the rectal or bladder necks, while the outer coat occludes them.

The levator crura (the functionally active components of the levator plate, as previously described) together with the levator tunnel and hiatal ligament form a “levator complex.” This is structurally and functionally adapted to serve a dual function: a) to fix, and prevent herniation of, the intrahiatal organs, and b) to share in the mechanisms of defecation and urination.282,301

Role of the Levator Complex in Intrahiatal Fixation

The levator tunnel is connected to the intrahiatal structures through the hiatal ligament and along the attachment of its inner coat (the suspensory sling) to these structures. These two links not only bind the intrahiatal structures firmly to the levator tunnel, but also harmonize their action.282,301 As the suspensory sling is a direct continuation of the levator plate, it slings the intrahiatal structures from the side wall of the pelvis. Thus, on levator contraction at stool or urination, the suspensory sling contracts, with a resultant shortening and widening of the levator tunnel. At the same time, it slings up and fixes the intrahiatal structures during evacuation of their contents: hence its name, the “suspensory sling.”301

While the suspensory sling fixes the intrahiatal structures in a “vertical” plane, the hiatal ligament provides a “horizontal” suspension.282,301 The rectal and vesical necks are continuously exposed to variations of intraabdominal tension due to respiratory movements and/or straining. The hiatal ligament constitutes a flexible connection between the levator tunnel and the intrahiatal structures. It allows for a certain degree of rectal and vesical neck mobility during respiration, defecation, and urination. Meanwhile, it seals the levator hiatus and prevents the intra-abdominal pressure from leaking through the hiatus to the intrahiatal organs, thus interfering with their functional activity. The hiatal ligament, thus, keeps the tunnel “pressure tight.”282 This mechanism is maintained as long as the intra-abdominal pressure is within physiologic limits. An increase beyond these limits, as in conditions of chronic straining, would tend to throw its load on the levator plate and tunnel, and eventually lead to prolapse of the intrahiatal structures.282,301

Thus, the anococcygeal raphe, being tendinous, becomes overstretched and subluxated under the effect of chronically increased tension. Consequently, the levator plate sags down, leading to pull and stretch of the hiatal ligament, as well as subluxation of the suspensory sling. The levator hiatus becomes widened and lowered such that the rectal and vesical necks lie above it and are directly exposed to the intraabdominal pressure action, which interferes with their stability and function. Sagging of the levator plate, and hiatal widening as well as hiatal ligament and suspensory sling subluxation result eventually in loss of support, slackening, and prolapse of the intrahiatal structures.282,325

Role of the Levator Complex in the Mechanism of Defecation and Urination

The defecation and urination mechanism has two components, intrinsic and extrinsic.282 The intrinsic component involves the detrusor and its sphincter; it acts involuntarily, being composed of smooth muscle fibers. It is under the voluntary control of the extrinsic component which consists of the levator complex, the fibers of which are striated. The intrinsic component functions reflexively: as the detrusor, rectal, or vesical muscle contracts, the internal sphincter relaxes. However, continuation of the acts of defecation or urination depend essentially on the extrinsic component.

When defecation or urination is desired, the “common” tunnel sphincter (the puborectalis) and the “individual” sphincter specific for the organ to be evacuated are voluntarily relaxed. Straining is necessary to maintain the act of evacuation. The Valsalva maneuver elevates the intraabdominal pressure, thereby serving two purposes: a) it compresses and helps evacuation of the detrusor, and (b) it evokes the straining-levator315 and straining-puborectalis reflexes.326 The former tends to open the rectal or vesical neck while the latter closes it by the time signals reach the level of consciousness and the decision is made whether to evacuate or not. Although the intraabdominal pressure compresses the detrusor, it spares the rectal neck and urethra, inasmuch as they are located in the pressure-tight levator tunnel.

The levator plate is connected to the rectal and vesical necks along the hiatal ligament, and to the external sphincters (anal and urethral) through the suspensory sling.282,301 On straining at stool or urination, the levator crura contract and are elevated and laterally retracted. In doing so, they pull the hiatal ligament, effecting traction to open the rectal neck inlet and bladder neck. Meanwhile the suspensory sling, a downward prolongation of the levator plate, contracts simultaneously with it. The contraction of the sling and plate has a twofold action: a) it shortens and widens the levator tunnel and, in turn, the rectal neck and urethra, and b) it pulls open the external sphincters (anal and urethral) by the action of the suspensory sling’s terminal fibers which pass through the external urethral sphincter and base loop of the external anal sphincter.282,301 The final result of levator contraction is that the rectal neck and/or vesical neck and urethra are opened for the rectum and/or bladder to evacuate their contents.

Electromechanical Activity of the Rectum and Rectosigmoid Pacemaker

Rectal motility is complex, and is regulated and coordinated by motor and reflex actions.273,312-315,326 Disturbances, of these mechanisms can lead to constipation. The “mass squeeze” contraction theory offers an explanation for rectal motility at defecation.273 A single contraction wave, starting at the rectosigmoid junction, pushes the fecal matter from the rectum into the opened rectal neck; it is repeated until the rectum is completely evacuated. In dogs, the passage of an inflated balloon through the rectosigmoid junction caused a significant increase in rectal pressure; this response is absent in the anesthetized rectosigmoid junction, as well as in inertia-type constipation, whereas it is present in obstructive constipation.266,271 A “pacemaker” has been suggested to exist in the rectosigmoid junction, organizing the motor activity of the rectum and defining its maximal contractile frequency.269 This pacemaker may trigger rectal contraction when stimulated by stools traversing the rectosigmoid junction. In inertia constipation, the pacemaker seems to be disordered.269,271

Rectal myoelectrical activity is characterized by regular pacesetter potentials266 (Fig. 18-113). Evidence supports the view that these potentials originate at the rectosigmoid junction and propagate distally.266,271 Action potentials follow the pacesetter potentials, with rectal pressure increasing simultaneously, representing contractile activity. This provides additional evidence that the rectosigmoid junction could be the site of a pacemaker that evokes the pacesetter potentials. Animal studies have demonstrated that electrical stimulation applied to the rectosigmoid junction by an artificial pacemaker increases rectal pressure, but decreases rectal neck pressure.266,271 An artificial pacemaker in the treatment of chronic constipation has been proven to be beneficial and effective.269,271

Fig. 18-113.

Electrorectogram from a normal subject showing pacesetter potentials followed randomly by action potentials. The frequency was calculated as cycles per minute. (Courtesy of Prof. Ahmed Shafik, MD.)

Anatomy-Related Pathologies and Treatment

The importance of the anatomic and physiologic findings described above and the resulting conclusions became apparent when their impact on pelvic pathology —in particular, anorectal pathologic conditions— was probed. The newly acquired knowledge has replaced inadequate concepts of the etiologies of pathology with new ones applicable to many pelvic disorders. This knowledge has made essential contributions toward solving a variety of longstanding surgical problems.

Intrahiatal Organ Mobilization from Within the Levator

With the understanding of the precise anatomy of the levator hiatus and tunnel, mobilization of intrahiatal organs can be performed safely, with preservation of voluntary sphincteric mechanism.282,324 Thus, high rectal neck and lower rectal lesions can be approached per perineum by mobilizing the rectal neck and lower rectum from within the levator tunnel. Dissection extends between the longitudinal anal muscle and the suspensory sling and through the intersphincteric space, up to the pelvirectal space, the plane of cleavage being along the tunnel septum (see Fig. 18-78). The rectal neck and lower rectum can thus be cored out of the tunnel, pulled down, and brought outside the anal orifice so that any lesion within them can be managed easily, with sphincter preservation.282,324

Likewise, middle- or lower-third malignant rectal tumors can be excised radically by a combined abdominoperineal approach with sphincter preservation. The abdominal portion of the procedure is accomplished as usual, mobilizing the sigmoid and rectum down to the levator plate. The perineal part calls for coring out the rectal neck from within the levator tunnel in the way mentioned above.324 The rectal neck, rectum, and sigmoid so mobilized are pulled outside the anal orifice and excised, and the colonic end is fixed to the perianal skin so as to be under the sphincteric control of the levator tunnel.324

It is suggested that by following the above-mentioned surgical principles, other intrahiatal organs (e.g., urethra, bladder, vagina, or uterus) can be mobilized and dealt with per perineum from within the levator tunnel, again with preservation of the sphincters.

Reversion to Normal Defecation after End Colostomy

Comprehension of the interrelations described above has an important, and immediate, bearing on another aspect of sphincter “preservation.” Electromyographic (EMG) studies of the perineal muscles after abdominoperineal operation with end colostomy for rectal cancer have revealed in 50% of the cases that the levator ani and puborectalis muscles have been spared – as was also confirmed by gross anatomic and histologic examination of excised specimens.327 Hence, advantage is taken of the presence of the muscles to reverse the otherwise permanent end colostomy by mobilizing and suturing the colonic stump to the perineal skin to lie under sphincteric control. Thus the fecal stream is diverted back to the perineal route to restore normal defecation and full fecal continence.


Excessive and exhaustive straining in defecation, even when the stools are soft and bulky, is a pathologic manifestation that Shafik refers to as “strainodynia.”328 Four types of strainodynia could be identified in his studies: “band,” sphincter, levator, and detrusor.

Band Strainodynia

‘Band’ strainodynia is more common in men. It presents with a stool of normal character, and is associated with elevated rectal neck pressure. Rectal neck biopsy reveals a fibrous band. Treatment by bandotomy results in adequate expansion of the rectal neck at defecation and thus relieves the problem.

Sphincter Strainodynia

‘Sphincter’ strainodynia is a condition in which internal and external sphincters fail to relax (anismus), or their contraction (dyssynergia) on rectal contraction results in strainodynia. In constipation, the puborectalis and external anal sphincter muscles may contract, rather than relax, upon attempted defecation.329-331 Contraction of the external anal sphincter with rectal distension is a condition Shafik has termed “detrusor-sphincter dyssynergia syndrome.”330 It has to be differentiated from the detrusor-rectal neck dyssynergia syndrome in which the internal anal sphincter contracts instead of relaxing on rectal distension.332 Differentiation is made by recording EMG activity both of the external and internal sphincter on rectal distension. It is worth mentioning that the stools in both syndromes commonly are soft and bulky, although the main complaint is constipation. Fiber diet and laxatives do not improve either of the 2 syndromes. External sphincter myotomy in detrusor-sphincter dyssynergia syndrome and internal sphincter myectomy in detrusor-rectal neck dyssynergia syndrome relieve straining at defecation.330,332

Levator Strainodynia

‘Levator’ strainodynia is due to levator ani muscle dysfunction which includes: levator dysfunction syndrome,333 detrusor-levator dyssynergia syndrome,334 and levator paradoxical syndrome.335 Levator strainodynia occurs with repeatedly obstructing stools at defecation even though they have a normal character (i.e. soft and bulky). EMG of the levator ani muscle is diagnostic.

In the levator dysfunction syndrome, the levator ani muscle is subluxated and sags down; it shows no EMG activity at rest or on contraction.333 Rectal prolapse may follow. Levatorplasty has satisfactory results.333

In the detrusor-levator dyssynergia syndrome, the levator muscle relaxes instead of exhibiting normal contraction on detrusor distension. EMG resting activity disappears, instead of increasing as occurs under normal physiologic conditions. There is no response to pharmacologic therapy, although biofeedback treatment may be beneficial.

With the levator ‘paradoxical’ syndrome,335 the rectal neck pressure on straining is elevated instead of diminished. EMG here, too, is diagnostic. The levator muscle shows increased EMG activity on voluntary squeezing, and little or no activity on straining; whereas the normal levator EMG exhibits no change in resting activity on squeezing, and increased activity on straining. These results suggest that levator action is paradoxical. Normally, the levator muscle contracts to open the rectal neck at defecation; the paradoxical action results in failure of the rectal neck to open on straining at stool. Biofeedback treatment may also be useful for this condition.

Detrusor Strainodynia

‘Detrusor’ strainodynia has a higher frequency of occurrence in women.336 Pressure and EMG studies of anal sphincters and levator ani muscle are normal. Rectal detrusor retropulsion is evident. One hypothesis relates rectal retropulsion to an ectopic pacemaker in the rectal detrusor that generates abnormal impulses.336 The pacemaker that regulates the rectal function seems to be located normally at the rectosigmoid junction.269,271 Pharmacologic therapy fails to cure the condition while antegrade balloon expulsion training is successful.


There are many known causes for oligofecorrhea, a term which can be applied to infrequent defecation, such as two or fewer weekly bowel movements.337 “Idiopathic oligofecorrhea” is the category without a traceable cause. According to clinical and investigative data, oligofecorrhea presents in 3 different stages, which share major abnormal findings, such as high rectal neck resting pressure, reduced or absent rectoanal-inhibitory reflex, internal sphincter hypertrophy, and degenerated nerve plexus of the internal anal sphincter.337 The abnormal innervation of the internal sphincter seems to interfere with the rectoanal-inhibitory reflex action, with a resulting failure of relaxation of the internal sphincter on rectal distension. The degenerative changes of the nerve plexus seem to affect mainly the parasympathetic supply, thereby producing predominantly sympathetic activity resulting in abnormal internal sphincter contraction, with an eventual result of muscle hypertrophy.

The 3 stages have criteria by which they are differentiated. In the deep intersphincteric groove category (stage 1), the lower end of the internal anal sphincter is thick and the intersphincteric groove is deeper than normal. Internal sphincterotomy has gratifying results.

In the everted intersphincteric groove category (stage 2), the intersphincteric groove lies outside the rectal neck orifice and is deeper than normal. The lower edge of the internal sphincter is thick and prominent and lies at the level of the external anal sphincter. It descends on straining to project as a cone outside the rectal neck outlet, producing the “cone sign.”337 Internal sphincterotomy results in improvement.

In cone anus (stage 3), the lower end of the internal sphincter protrudes as a cone without straining, outside the rectal neck orifice and below the lower edge of the external sphincter. On palpation, the sphincter is thicker and firmer than normal. The internal sphincter cone elongates on straining. Internal sphincter myotomy has resulted in improvement of all patients in this group.

Physioanatomy of the Hemorrhoidal Venous Plexus

And Related Pathologic and Therapeutic Potentials

The presence of a hemorrhoidal venous plexus at the level of both the submucosa and adventitia was demonstrated to extend over the entire length of the rectum and its neck.338 This plexus constitutes the means of direct communication between the 3 rectal veins. The submucosal plexus (Fig. 18-114) consists of characteristic transverse venous rings, whereas the adventitial plexus (Fig. 18-115) is formed of intercommunicating oblique veins. Neither the fusiform, saccular, nor serpiginous dilatations of the anal submucosal plexus as described by Thomson339 to be a regular feature of normal anatomy, nor the so-called corpus cavernosum recti by Stelzner340 could be demonstrated. Contrary to views of some investigators341,342 that the “collecting hemorrhoidal veins” (Fig. 18-116) lie in the columns of Morgagni, the study by Shafik and Mohi-El-Din338 has revealed that they exist in the rectal adventitia. The columns are simply plicated mucosal folds that result from both the fusion of the wide hindgut with the narrow proctodeum and the tonic action of the rectal neck sphincters.

Fig. 18-114.

Cadaveric specimen showing barium sulfate solution injected into inferior mesenteric vein. It demonstrates that the rectal submucosal plexus extends along the whole of the rectum including its neck and is arranged in transverse venous rings. Arrows indicate the hemorrhoidogenital veins connecting the hemorrhoidal with the vesicoprostatic venous plexus. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-115.

Barium sulfate injected into inferior mesenteric vein. The oblique large veins are those of the adventitial plexus, whereas the upper small transverse veins belong to the submucosal plexus. The pectinate area shows the radiological blush. The bladder wall is opacified through the hemorrhoidogenital veins. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-116.

Barium sulfate injected into inferior mesenteric vein. Rectum and urinary bladder were inflated, frozen, and bisected. Artery forceps points to urethra. Specimen shows “collecting veins.” Upper arrows point to hemorrhoidogenital veins through which bladder wall is opacified. Lower arrows point to inferior hemorrhoidal plexus. (Courtesy of Prof. Ahmed Shafik, MD.)

Two sites of portosystemic communication could be identified in the rectum: interhemorrhoidal and hemorrhoidogenital.338 The first of these occurs between the three rectal veins, both submucosally and adventitially. The communication site was identified in the adventitia but not in the submucosa. The portal blood is shunted through this communication to the internal iliac vein. The second communication is through rectogenital collateral channels (Fig. 18-117) which connect the rectal venous plexus with the vesicoprostatic or vaginal plexuses. It seems that this portosystemic connection is extensive, because the urinary bladder, vagina, and uterus were opacified each time the inferior mesenteric vein was injected with barium sulfate.338

Fig. 18-117.

Inferior mesenteric vein injected with barium sulfate. Rectum and urinary bladder inflated, frozen, and cut-sectioned. Specimen shows hemorrhoidogenital veins passing from rectal neck to vesical plexus. (Courtesy of Prof. Ahmed Shafik, MD.)

In contrast to investigators341,342 who mention that the hemorrhoidal venous plexus is located in the lower rectum and anal canal, and is only submucosal in position, studies by Shafik and Mohi-El-Din338 have demonstrated that it not only extends along the whole rectum and its neck, but is both submucosal and adventitial. Being extensive, the plexus can absorb excess venous congestion along its entire length before it becomes varicose; likewise, the varicosity involves the entirety of the venous plexus and not only its lower submucous part. It simulates, in this respect, the diffuse congestion and varicosity of the pampiniform plexus in varicocele. This fact and the understanding that the portal hemorrhoidal blood can work its way to the systemic circulation through two portosystemic shunts (interrectal and rectogenital) tend to negate the theory of venous congestion in the lower part of the hemorrhoidal plexus as being the primary event in hemorrhoidogenesis.

Findings of Shafik and Mohi-El-Din338 support the conclusion that hemorrhoids are a mucosal prolapse resulting primarily from the constricting effect of the anorectal band upon the rectal neck.338 The observations explain the rarity of hemorrhoids in portal hypertension, and are consistent with the incidence of hemorrhoids in bilharzial liver cirrhosis patients in Egypt, which does not differ from the incidence in subjects without bilharzial liver cirrhosis.290,291 It also explains the rarity of rectal bleeding contrasted with esophageal bleeding in these patients.

Porto-Systemic Circulation in the Rectum

Under normal physiologic conditions, the submucosal hemorrhoidal plexus drains into the adventitial plexus, which then drains into the 3 paired rectal veins: superior, middle, and inferior. Because of the submucosal and adventitial anastomoses of the three veins in the rectal neck, and due to the presence of the rectogenital veins, portal blood can drain into the systemic circulation – particularly when the rectum contracts at defecation. This was proved when contrast medium was injected into the rectal neck submucosa of normal living subjects: the dye appeared in the vesicoprostatic and vesicovaginal venous plexuses343,344 (Fig. 18-118A & B).

Fig. 18-118.

Anal cystography in a living subject. A, Dye injected into lower rectal neck submucosa. X-ray, ten minutes after injection, showing dye outlining prostatic and lower vesical plexus. B, Whole urinary bladder opacified 30 minutes after injection. (From Shafik A. Anal cystography. New technique of cystography. Preliminary report. Urology 1984;23:313-316; with permission.)

Systemic blood cannot drain into the portal venous system, however. When either barium sulfate or blue plastic was injected into the deep dorsal vein of the penis, it could not be recovered in the hemorrhoidal plexus.338 This is probably due to the presence of valves in the middle and inferior rectal veins which direct the blood to the systemic circulation, and not vice versa. Unlike elsewhere, portal blood shunted from the rectum and left colon to the systemic circulation seems to cause no toxicity problems from metabolic products, because this blood carries no nutrient materials.

Rectourinary Syndrome

The venous channels that communicate between the rectum and genitourinary organs may provide explanation for some of the previously unexplained pathologic lesions affecting these organs, such as idiopathic prostatitis, cystourethritis, vaginitis and cervicitis, recurrent bacteriuria, and others.291,344-346 Portal blood is known to carry various organisms derived from the colon and rectum, notably E. coli. It appears that these organisms, under certain conditions, find their way through the communicating veins to the genitourinary organs and cause infection there. This is especially so if there is portal backflow, as occurs in portal hypertension, abdominal tumors, or pregnancy. Similarly, the anorectal congestion which is encountered with lesions like hemorrhoids, fissures, or abscesses is liable to result in congestion of genitourinary organs, which become readily infected from the organisms in the portal blood.

Rectal congestion from hemorrhoids also has been found to cause prostato-vesicular congestion, which in turn leads to idiopathic urethral discharge,291 recurrent bacteriuria, and cervicitis.345,346 Recurrent bacteriuria, cystourethritis, and recurrent cervicitis have Escherichia coli as their causative agent in the majority of cases. Bacteriologic examination of the organisms recovered from urine, cervix, and feces has demonstrated that they belong to the same serogroup.345 Hemorrhoids, fissures, ulcers, and fistulas act as E. coli reservoirs, as well as leading to anorectal and genitourinary organ congestion. The E. coli-laden blood from the congested anal bacterial reservoirs passes through the communicating veins to the congested pelvic organs, causing their infection.345,346 Treatment of the anal lesion, with or without obliteration of the rectogenital veins by sclerotherapy, is curative.345

Communicating veins also seem to play a vital role in genitourinary bilharziasis: schistosomal worms are either urinary (haematobium) or intestinal (mansoni). They mature in the intrahepatic portal venules. The male then carries the female worm in the gynecophoral canal and migrates down the portal vein and its mesenteric tributaries. Schistosoma mansoni worms, being bigger than the haematobium type, cannot proceed further and lay down their eggs in the colonic and rectal wall. The smaller schistosoma haematobium worms seem to continue their journey through the rectogenital veins to the vesico-prostatic or vaginal plexus where the eggs are deposited in the pelvic genitourinary organs338 – which would explain the mystery of the route adopted by this type of schistosoma to reach the genitourinary organs from the liver.

Role of Pelvic Organ Venous Plexuses in Diagnostics

X-ray imaging can be used to demonstrate the configuration of the venous plexuses of the rectum and urogenital organs and the venous interconnections across the pelvic floor. Contrast medium injected into the anal submucosa diffuses from the site of injection through the vesico-prostatic plexus, outlining the prostate and the wall of the urinary bladder, but not its lumen. An anal cystography is obtained343 (Fig. 18-118). By the same procedure, the dye reaches the vaginovesical plexus so that cystovaginohysterography can be performed344 (Fig. 18-119).

Fig. 18-119.

Inferior mesenteric vein of female cadaveric specimen injected with barium sulfate. Vagina, uterus, and urinary bladder were opacified through hemorrhoidogenital veins. (Courtesy of Prof. Ahmed Shafik, MD.)

Role of the Rectogenital Veins for Chemotherapy in Pelvic Malignancies

Detection of the 2-6 small rectogenital veins described which communicate unidirectionally between the rectal veins and the vesicoprostatic or rectovaginal plexuses called for a method that would adequately exploit the prodigious potential of this anatomic route. The submucosal anal injection technique338,347 was created for a direct anal or perineal approach.348 It has developed into a landmark contribution from the point of view of nonsurgical and minimally invasive measures in the therapy of pelvic disorders, including advanced and metastatic stages of malignancies. The procedure is simple, easy, and of marked efficacy. For most of its applications it can be considered an office procedure.

Animal experiments had earlier shown that the anal route is adequate for administration of chemotherapy. The distribution of 14C-labeled misonidazole, a radiation sensitizer, was studied in the serum and tissues of rats by comparison of the submucosal anal route with the oral route.349 The drug concentration in the bladder tissue, relative to the serum, was highest by the anal route with a level of 8, and 5 times that of the serum 15 and 30 minutes, respectively, after administration. When the drug was administered orally, however, the drug concentration in the bladder tissue reached only one-fourth of, or a level equal to, the serum level in the same periods. In a similar experiment assessing the drug concentration in the uterus and vagina relative to serum, the submucosal anal route achieved a concentration of 10 and 8 times, respectively, of the serum level after 15 minutes. Oral administration in the same period of time produced a drug concentration of one-fifth and one-quarter the serum level.350

In view of these results, submucosal anal injection of chemotherapeutic agents was used with satisfactory results in the treatment of pelvic malignancies such as advanced vesical, prostatic, cervical, and rectal cancer.347,351-353 No side effects were encountered. An essential advantage of this treatment modality is that there are no systemic effects of the therapeutic chemical agent, because a high drug concentration is achieved in the tumor while the serum drug concentration remains low. This was confirmed experimentally and by studies of methotrexate concentration in vesical and rectal tumor tissue and serum after anal injection compared to parenteral injection.349,350,354 Chronic prostatitis is effectively treated via the submucosal anal route by injecting the appropriate antimicrobial agent as defined by sensitivity tests.355

Arterial Pattern of the Rectum and Its Clinical Application

The arterial supply of the rectum was studied in 32 cadavers by Shafik and Moustafa.356 The superior rectal artery (SRA) (Fig. 18-120) and vein were found to be enclosed in a fibrous sheath which was connected to the posterior rectal surface by an anterior mesorectum containing the “transverse rectal branches,” and to the sacrum by avascular posterior mesorectum. Small lymph nodes were scattered alongside the anterior mesorectum. The SRA gave rise to 4 branches: transverse rectal, descending rectal, rectosigmoid, and terminal. The transverse rectal arteries (Fig. 18-121) arose from the SRA in 24 specimens and from the descending rectal artery in 8. They were distributed to the upper half of the rectum. The rectosigmoid artery was distributed to the descending limb of the sigmoid colon and rectosigmoid junction. Examining terminal branches of the SRA in 32 cadavers, the authors found 2 branches in 21, and 3 branches in 11. The branches communicated in the lower half of the rectum. Inferior rectal arteries were present in all the dissected cadavers while middle rectal arteries could be identified in only 50% of the cadavers. Two arterial patterns were recognized: anular in the upper rectal half provided by the transverse rectal arteries, and plexiform in the lower half supplied by the SRA terminal branches.

Fig. 18-120.

Diagrammatic illustration of the superior rectal artery giving rise to the transverse rectal arteries. (Modified from Shafik A, Moustafa H. Study of the arterial pattern of the rectum and its clinical application. Acta Anat 1996;157:80-86; with permission.)

Fig. 18-121.

Angiogram showing superior rectal artery (A) giving rise to transverse rectal branches (B). a, Lateral view. b, Anteroposterior view. Terminal branches of the transverse rectal arteries do not anastomose anteriorly and this results in the formation of a “bloodless line.” The superior rectal artery terminates at the mid-rectum giving rise to 3 branches. (From Shafik A, Moustafa H. Study of the arterial pattern of the rectum and its clinical application. Acta Anat 1996;157: 80-86; with permission.)

The anterior mesorectum was vascular as it contained the transverse rectal arteries, whereas the posterior mesorectum was avascular.356 These findings should be considered during rectal mobilization in operations for rectal prolapse or cancer to avoid excessive bleeding. In rectopexy for rectal prolapse,357 the rectum is mobilized and this is preferably done through the posterior avascular mesorectum. Also mesorectum removal with rectal excision for rectal cancer reduces the recurrence rate.358 To avoid bleeding, dissection is best performed in the posterior avascular plane. The pararectal lymph nodes existed alongside the anterior mesorectum. Their dissection necessitates excision of the superior rectal vessels and their mesenteries. This might explain the higher 5-year survival rate in rectal cancer when the mesorectum is included in the radical operation.358 Removal of the mesorectum includes removal of the pararectal lymph nodes as well.

There were fewer branches of the SRA on the anterior rectal wall than on the posterior. The vascular branching diminished gradually from the posterior to the anterior aspect to the extent that a “bloodless line” is suggested to exist in the midline anteriorly (Figs. 18-120, 18-121). It is assumed that for this reason incisions in the posterior rectal wall heal faster than anterior wall incisions and, in addition, are less likely to leak. Incisions in the anterior rectal wall should have bleeding edges before being closed in order to avoid disruption and leak; they may be covered by a colostomy.

The rectosigmoid branch of the SRA constitutes a communication between the SRA and inferior mesenteric artery. The arterial pattern in the sigmoid colon and upper half of the rectum is segmental, and is provided by the transverse arterial rings which seem to constitute end arteries. The rectosigmoid artery might act to shunt the blood from the inferior mesenteric artery to the SRA in case of arterial interruption in the rectosigmoid area.

The middle rectal arteries seem to play a minor role in the arterial supply of the rectum as they are inconsistent. Their low incidence has to be considered during rectal dissection in operations for rectal prolapse or cancer. This is in contrast to the inferior rectal arteries which were consistent in all the dissected specimens. They were distributed to the rectal neck. Anastomosis occurred between the 3 rectal arteries at the submucosal level.

Rectum: Conduit or Storage Organ?

Most standard texts state that the rectum is usually empty.359-362 However, the evidence upon which this statement is based is deficient. In a study by Shafik et al.,363 stools were palpable in the lower rectum in 64.5% (31 of 48 subjects) and were demonstrated radiologically in 75% (12 of 16) (Fig. 18-122). In 3 of 16 subjects the stools were located in the rectum too high to be detected digitally, although they were demonstrated radiologically (Fig. 18-123). Stools situated more than 7.5 cm above the anal verge are likely to be impalpable. The high location of the stools in the rectum seems to be peculiar. It is believed that the upper rectum can retain stools owing to the presence of the mucosal folds which may shelve the stools and keep them in the upper rectum. The higher stool frequency per week in subjects with an empty rectum as opposed to those with a full rectum is believed to be due to many factors including the type of diet and the response to the defecation reflex. Meanwhile the authors could not find a relation between the time elapsed since the last defecation and a full or an empty rectum.

Fig. 18-122.

Barium enema showing stools (arrow) in rectum. (From Shafik A, Ali YA, Afifi R. Is the rectum a conduit or storage organ? Int Surg 1997;82:194-197; with permission.)

Fig. 18-123.

Air insufflated into the rectum. Stools (arrows) are located above the upper rectal valve. (From Shafik A, Ali YA, Afifi R. Is the rectum a conduit or storage organ? Int Surg 1997;82:194-197; with permission.)

The above study363 demonstrated that the rectum may contain stools under normal physiologic conditions. This is in contrast to investigators who hold the view that the rectum is normally empty of feces and that, as the stools pass to the rectum from the sigmoid colon, the recto-anal inhibitory reflex is initiated and defecation occurs.359-362 It is believed that the rectum receives the stools and stores them until they reach a certain volume which would stimulate the rectal stretch receptors and the defecation reflex. In such a case, digital rectal examination or radiography can detect stools in the rectum under normal physiologic conditions.

Another explanation for the presence of stools in the rectum under normal physiologic conditions could be that the stools enter the rectum in a volume sufficient to initiate the defecation reflex. Should the circumstances be unfavourable for defecation, the external sphincter voluntarily contracts evoking the voluntary inhibition reflex which results in rectal relaxation and waning of the desire to defecate.163 The defecation reflex might not return until an additional amount of feces enters the rectum; this can take several hours to occur, during which time the stools may be palpable in the rectum.

Pudendal Nerve

The pudendal nerve is an important motor and sensory nerve to the pelvic organs and perineum. It provides supply for the anal and urethral sphincters, as well as cutaneous and skeletal motor innervation of the penis and clitoris. Pudendal neuropathy or nerve injury leads to pathologic changes of these structures.333,364-367 Pudendal nerve compression in the pudendal canal causes the pudendal canal syndrome (an entrapment syndrome) with a resulting incompetence of anal and urethral sphincters.368-370 Electrode application to the pudendal nerve or to its roots or branches has been used for treatment of anal and urethral sphincter insufficiency.371-376

Sound knowledge of the precise anatomic location and pattern of the pudendal nerve and its roots and branches is essential to facilitate their detection. The localization of the nerve by its surgical anatomy has important diagnostic and therapeutic advantages in procedures such as confirmation of the diagnosis of pudendal canal syndrome or corroboration of indications for pudendal canal decompression operation.368-370,377 Adequate anatomic knowledge and thorough screening studies of the pudendal nerve are imperative to:


Expose the nerve in patients with pudendal canal syndrome wherein the roof of the canal is slit open to decompress the nerve368-370,377-383

Provide exposure for the positioning of electrodes for the performance of an accurate nerve stimulation

Stimulate the nerve roots or branches for fecal and urinary incontinence or erectile dysfunction371-377

Stimulate the nerve to assess its functional integrity and that of the pelvic floor musculature

Determine pudendal nerve terminal motor latency364

The pudendal nerve arises from the anterior rami of S2-4, and is formed of three roots and two cords (Fig. 18-124). It may take a contribution from S1 and S5 (Fig. 18-125). The first root continues as the upper cord while the 2nd and 3rd roots unite to form the lower cord. The 2 cords fuse to form the pudendal nerve.384 In neurostimulation of the pudendal nerve, identification of the cords helps in applying the neuroprosthesis to the appropriate cords. The upper cord is longer and thinner than the lower; it was thick and shorter, however, when it received a contribution from S1. In all dissected cadavers, the pudendal nerve was formed by the merging of the two cords above the sacrospinous ligament before it crossed the pelvic surface of the ligament a short distance medial to the ischial spine.384 In no specimen did the nerve cross the bony ischial spine. This is in contrast to the view of other investigators who state that the pudendal nerve passes behind the ischial spine.385,386 For pudendal nerve block or stimulation, the needle or electrode has to be moved medially after locating the ischial spine, allowing for close apposition to the pudendal nerve.

Fig. 18-124.

Photograph of an adult cadaver showing the first root (S2) that passes downward and laterally to form the upper cord (UC). The second root (S3) passes downward to fuse with the third root (S4) forming the lower cord (LC). The two cords fuse together just above the sacrospinous ligament (L) to form the pudendal nerve (PN). IR, Inferior rectal nerve. P, Perineal nerve. D, Dorsal nerve of penis. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-125.

Photograph of an adult cadaver showing the three roots (S2, S3, and S4) and the two cords (UC, upper cord, and LC, lower cord) of the pudendal nerve above the sacrospinous ligament (L). The upper cord receives a contribution from the first sacral root (S1). (Courtesy of Prof. Ahmed Shafik, MD.)

The inferior rectal nerve originated from the pudendal nerve in the pudendal canal in 18 of 20 dissected cadavers (7 male, 13 female). In one, it arose from the pudendal nerve in the lesser sciatic foramen before entering the pudendal canal, and in another it arose directly from S3.384 The inferior rectal nerve supplies the external anal sphincter and levator ani muscle. In pudendal canal syndrome, pudendal nerve compression in the pudendal canal leads to anal or scrotal pain, fecal or urinary incontinence and/or erectile dysfunction. Decompression of the pudendal canal relieves the symptoms.368-370,377 It is highly probable that subjects in whom the inferior rectal nerve arises from the pudendal nerve proximal to the pudendal canal or from the pudendal nerve roots are spared the anorectal manifestations of the pudendal canal syndrome.384

In seven male cadavers the pudendal nerve gave off an “accessory rectal nerve” behind the sacrospinous ligament (Fig. 18-126), which innervated the levator ani muscle and perineal and perianal skin.384 Juenemann et al.385 dissected three male cadavers and reported a branch arising from the pudendal nerve, but described the nerve as innervating the transversus perinei and ischiocavernosus muscles. Shafik et al. failed to demonstrate the innervation of these muscles by the accessory rectal nerve, and it seems that the two branches are different. The finding of the accessory rectal nerve exclusively in male cadavers and not in the female is mysterious. The levator ani muscle in such cadavers was doubly innervated on its perineal surface by the inferior rectal nerve and the accessory rectal nerve. While the inferior rectal nerve arose in the pudendal canal, the accessory rectal nerve arose above it and therefore might be spared the neuropathy of the other branches of the pudendal nerve in the pudendal canal syndrome.384

Fig. 18-126.

Photograph of male cadaver showing the three branches of the right pudendal nerve (i) arising in the pudendal canal: inferior rectal nerve (e), perineal nerve (j), and dorsal nerve of penis (c). They show also the accessory rectal branch (a) which arises behind the sacrospinous ligament. (Courtesy of Prof. Ahmed Shafik, MD.)

The pudendal nerve can be used to study the integrity of pelvic floor muscles, in biofeedback training, nerve blocks, stimulation trials to treat chronic incontinence, and in nerve conduction studies or evoked potential recordings.384

Pudendal Canal

The anatomy of the pudendal canal (PC) (Fig. 18-127) was studied in 26 cadavers: 10 stillborn and 16 adult (mean age 48.2 years).387 Two approaches were used to expose the PC: gluteal and perineal. The PC was an obliquely lying tube with a mean length of 0.8 cm in the stillborn and 1.6 cm in the adult cadavers. It started at a mean distance of 0.8 cm from the ischial spine in the stillborn and of 1.6 cm in the adult cadavers, and ended at a mean distance of 0.7 cm and 2.6 cm, respectively, from the lower border of the symphysis pubis. The PC wall was formed by splitting of the obturator fascia and not by the lunate fascia. The PC contained the pudendal nerve and vessels embedded in loose areolar tissue. The 3 branches of the neurovascular bundle arose inside the canal in all but 3 cadavers. The wall of the PC consisted of collagen and elastic fibers while that of the obturator fascia consisted of collagen only. The PC seems to be structurally adapted to serve certain functions.

Fig. 18-127.

Dissected cadaveric specimen showing the pudendal canal (p) with the pudendal nerve (i) entering it posteriorly and its branches leaving it anteriorly. m, Sacrospinous ligament. g, Obturator internus muscle. e, Inferior rectal nerve. b, Dorsal nerve of clitoris. j, Perineal nerve. (Courtesy of Prof. Ahmed Shafik, MD.)

Structural-Functional Adaptation of the Pudendal Canal

The anatomic structure of the PC seems to provide it with maximum functional performance.387 Thus, the loose areolar tissue in which the neurovascular bundle is embedded allows for changes in the vessel diameter in response to pelvic organ activities, especially during sexual arousal and erection, without blood supply embarrassment. The functional adaptation of the PC is further provided by its histologic structure. The crisscross plywood arrangement gives the PC wall a textile nature that allows the canal to change its shape in adaptation to pressure changes in the pudendal vessels. The PC can thus expand, giving the pudendal vessels a space to engorge during sexual arousal and intercourse. Meanwhile, the elastic fibers included in the wall effect spontaneous return of the PC to its original size by their elastic recoil. With such structure, the PC is suggested to be acting as a “pump” that assists venous return in the pudendal veins. Furthermore, the elastic fibers seem to prevent collagen overstretch and PC subluxation as a result of its continuous distension by vessel engorgement.

Pulley Action of the Pudendal Canal

In the meantime, it appears that the PC acts as a “pulley” for the pudendal neurovascular bundle.387 The latter, arising in the pelvis, passes through the PC on its journey to the ischiorectal fossa. It seems that the pulley action of the PC not only fixes the bundle during its travel to the pelvic floor muscles but also prevents it from being traumatized by the continuous movement of these muscles.

Pudendal Canal Syndrome

The pulley action of the PC may be disrupted by dysfunctioning pelvic floor muscles or disordered defecation. Thus in cases of levator dysfunction syndrome,388 the levator muscle becomes subluxated and sags. This would lead to continuous pull on the neurovascular bundle with eventual occurrence of pudendal neuropathy and pudendal arteritis, a condition we call “pudendal canal syndrome.”368-370,377-383,389-391 This syndrome presents with the sensory and motor manifestations of the pudendal nerve such as fecal incontinence,369,381,389 anal pain,368 stress urinary incontinence,390,391 erectile dysfunction,377 and perineal hypoesthesia.368-370,377-379,381-383,389-391 The PC syndrome is treated by PC decompression.368-370,377-379,381-384, 390,391 The canal is slit open to release the pudendal vessels and nerve from their entrapment.

Pudendal Canal Decompression

Surgeons exposing the PC must know its precise anatomy so they can avoid injury to the neurovascular bundle or its branches. The PC lies obliquely with its proximal end close to the ischial spine and its distal end lateral to, and level with, the inferior border of the symphysis pubis;387 it is located a few centimeters above the ischial tuberosity. In view of these anatomic landmarks, the PC could be exposed through 2 routes: perineal,368 and posterior.378 In the perineal route (Fig. 18-128), the PC is exposed through a para-anal incision. The inferior rectal nerve is identified in the ischiorectal (ischioanal) fossa and taken as a guide to the pudendal nerve in the PC. The posterior approach to the PC (Fig. 18-129) is performed through a parasacral incision, dividing the underlying gluteus maximus muscle, exposing the pudendal nerve, and following it to the PC. However, the perineal approach was more satisfactory than the posterior because it is easier, simpler, and less time-consuming.

Fig. 18-128.

Pudendal canal decompression operation (anterior approach). A, B, Incision. C, Inferior rectal nerve crossing ischiorectal (ischioanal) fossa. D, Inferior rectal nerve hooked with index finger. E, Inferior rectal nerve followed to pudendal nerve. Inset, Pudendal nerve in and outside pudendal canal. (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-129.

Pudendal canal decompression operation (posterior approach). A, Parasacral incision. B, Division of gluteus maximus muscle. C, Pudendal nerve and vessels exposed. (Modified from Shafik A. The posterior approach in the treatment of pudendal canal syndrome. Coloproctology 1992; 14:310-315; with permission.)

In conclusion, knowledge of the surgical anatomy of the PC is necessary not only for diagnostic purposes but also for designing a proper therapeutic approach.

New Methods of Investigation

Although we have a great variety of diagnostic methods at our disposal, the need for qualitative improvement remains constant. This improvement need not necessarily be effected by more sophisticated instrumentation, however. With regard to the assessment of disorders of the anorectum, Shafik et al. have introduced approaches that are less sophisticated in that they simply put to use known anatomic or physiologic principles; yet they give excellent objective results in highly reliable and detailed ways.


With respect to anorectal complaints, conventional examinations reveal little correlation between subjective symptoms and investigative results. This may be partly due to the fact that patients’ assessment of their own defecation is frequently misleading. The main reason, however, is the inadequacy of these types of investigations to reveal exactly what occurs during defecation. Neither defecometry392 nor the balloon expulsion test393 are representative of the procedure of defecation in view of the minimal –if any– increase in voluntary abdominal pressure required with normal defecation as opposed to the significant rectal expulsion pressure needed to expel the balloons.393 In search of a method simulating natural rectal evacuation as closely as possible, Shafik and colleagues394,395 introduced the fecoflowmeter, which produces an authentic and comprehensive record of the act of defecation in all its details.

The fecoflowmeter apparatus is based on weighing the defecated fluid; it consists of a weight transducer, amplifier, and oscillograph. The technique of fecoflowmetry is simple and noninvasive.395 The patient is instructed to evacuate the bowel prior to examination by either defecation or enema. Then, with the patient in the left lateral position, a 1 L water enema is given slowly. The water is incubated at 37°C and instilled under gravity through a 14F catheter placed in the rectum 8-12 cm from the rectal neck orifice. The catheter is removed after enema administration. The subject is then asked to walk around, retain the enema as long as possible, and to sit on the commode of the fecoflowmeter when feeling the urge to defecate. The fecal flow rate is the product of rectal detrusor action against outlet resistance and measures the defecated volume passed per time unit.395 Fecoflowmetry provides quantitative as well as qualitative data concerning the act of defecation. All objective parameters are assessed in one test. The defecated volume, the flow time, the time to maximum flow, as well as the mean and maximum flow rates are calculated from the defecation flow curve.395 The shape of the normal curve reflects the dominance of an active detrusor and the modulation of a passive rectal neck. The ascending limb illustrates the rectal detrusor contraction, while the descending limb displays the function of the rectal neck.

The results of contrasting the fecoflowmetric parameters of chronically constipated patients to those of normal controls394,395 have shown that the time to maximum flow is longer in the constipated group than in the normal group, while the mean and maximum flow rates are lower. The volume expelled is smaller in the constipated patients. Furthermore, the time elapsing from the administration of the enema to the sensation of urgency is prolonged in the constipated group. The flow curve of the controls (Fig. 18-130) is characteristic:395 It is obelisk-shaped. The ascending limb rises steeply and is commonly smooth. The descending limb has a tendency to be drawn out, but is also smooth. In contrast, the flow curve of constipated patients (Fig. 18-131) presents with a less steeply rising ascending limb and with fluctuation in flow intensity.395 Also, preevacuation fluctuations are common. As the maximum flow is reached, a plateau may form. The descending limb commonly shows fluctuations.

Fig. 18-130.

Normal defecation flow curve. a-b, Defecation time (seconds). c-d, Maximum flow rate (ml/sec). a-d, Time to maximum flow (seconds). (Courtesy of Prof. Ahmed Shafik, MD.)

Fig. 18-131.

Defecation flow curve in patients with chronic constipation. The ascending limb rises less steeply than normal and shows fluctuations. The descending limb also shows fluctuations. (Courtesy of Prof. Ahmed Shafik, MD.)

Reflex Reactions

The reflexes which are involved in the mechanisms of defecation and continence, namely recto-puborectalis reflex,312 recto-levator reflex,313 levator-puborectalis reflex,314 straining-levator reflex,315 straining-puborectalis reflex,326 and levator-sphincter reflex396 are additional investigative tools. They are reliable, objective, and substantial sources of information on the physiologic state of the pelvic floor muscles and of the nerves supplying them. Evaluation of the reflex actions of the pertinent musculature contributes to establishing a diagnosis in anorectal disorders with yet more accuracy, in a more comprehensive manner, and in less time because the approach is direct. The technique of evaluating the interrelationships between reflexes of the rectal area and the responses of the involved musculature follows an easily understood principle. A balloon-tipped catheter is introduced into the rectum and is attached to a pressure transducer. An electromyographic concentric needle electrode is inserted into the muscle that is to be investigated for its response to rectal balloon inflation. Detectable changes in the latency, duration, or motor unit action potentials of the evoked response may indicate a defect in the reflex pathway that could be an indication of muscular or nerve damage.

Water Enema Test

Another simple office test, the so-called water enema test, was introduced for anorectal investigation.397 At 37°C, 1.5 L of water is instilled under gravity into the rectum through a 16-gauge Nélaton catheter at a rate of 150 ml/min. The subject is asked to report the first rectal sensation as well as desire and urge to evacuate. The volume of water infused at the time of these occurrences is determined, and compared with standard values from controls. Defecation disorders can be identified with this easy and noninvasive office test.

Inguinal Pelviscopy

Inguinal pelviscopy (Fig. 18-132) examines structures in the extra- and intraperitoneal pelvic cavity.398 Through a 1 cm incision above the symphysis pubis, a Veress needle is inserted into the retropubic space, and gas is insufflated; the pelviscope is introduced through a second 1 cm incision in the area overlying the superficial inguinal ring and advanced in the inguinal canal to cross the deep inguinal ring into the extraperitoneal pelvic cavity. It is used to visualize and biopsy questionable masses in the pelvis. The approach is safe, because the entry is through a natural pathway, the inguinal canal; the examination is extraperitoneal.398 It is direct, as the deep inguinal ring is near and represents a natural gateway to the pelvic cavity. Furthermore, extraperitoneal masses are difficult and hazardous to approach by the standard intraperitoneal laparoscopic technique. Inguinal pelviscopy can also deal with intraperitoneal pelvic and abdominal lesions, when the instrument is made to pierce the peritoneum at the deep inguinal ring or close to the fallopian tube. It is most suitable for visualizing the ovaries, tubes, and uterus. These organs are close to the deep inguinal ring and can be approached either extra- or intraperitoneally or both. Inguinal pelviscopy is used not only for diagnostic purposes but also for some minor therapeutic procedures. Pelvic adhesions can be easily located and managed. Tubal sterilization is safely done. Small, simple cysts are aspirated and biopsied. With such a simple and safe technique, repeated inguinal pelviscopy allows pelvic lesions to be followed and surgery to be properly timed.

Fig. 18-132.

Route to enter pelvic cavity (dotted line). (Courtesy of Prof. Ahmed Shafik, MD.)


Rectometry is a simple, noninvasive way of assessing rectal volume, pressure, and compliance at the first rectal and urge sensations in one test.399 Carbon dioxide is infused into a balloon introduced into the rectum and connected to a pressure transducer. Simultaneous measurement of the rectal and intra-abdominal pressure is done, and a rectometrogram (Fig. 18-133) is obtained. It reads the volume of carbon dioxide infused and the intrarectal and detrusor pressure at both the first rectal and urge sensations. In addition to supplying information on the quantitative values, the curve configuration itself differentiates not only between normal and constipated subjects, but also between the obstructive and inertia types of constipation.399 Apart from its diagnostic value, rectometry can be used to follow the effectiveness of drugs on detrusor function.

Fig. 18-133.

A, Rectometrogram of normal evacuation. B, Intraabdominal pressure curve, showing slight increase of pressure at evacuation. First dash below tone limb denotes first rectal sensation, second dash denotes urge sensation. (From Shafik A, Moneim KA. Rectometry: A new method assessing rectal function. Coloproctology 1991;13:237-243; with permission.)

Electrorectogram in Normal Subjects and in Rectal Pathologic Conditions

Electrorectography is a noninvasive and nonradiologic technique by which the electric activity of the rectum is recorded, using a silver/silver chloride electrode situated 1 cm from the tip of a 6F catheter which is applied to the rectal mucosa by suction.266,267 The procedure could be included as a valuable method in assessing rectal detrusor efficiency and in diagnosing disorders and pathologic rectal conditions. In normal subjects, regular and reproducible pacesetter potentials are recorded with a mean frequency of 2.6 ± 0.4 cycle/min, amplitude of 1.9 ± 0.6 mV and velocity of 4.2 ± 0.9 cm/sec. They are followed randomly by action potentials267 (see Fig. 18-113). In inertia constipation, pacesetter potentials are infrequent (bradyrectia) and action potentials are registered only occasionally. In obstructive constipation, regular and reproducible pacesetter potentials record a higher frequency and velocity than normal (tachyrectia).400 In chronic proctitis, the pacesetter potentials frequency is higher, while amplitude and velocity are lower than normal. Action potentials have a higher frequency and amplitude.401 In Hirschsprung’s disease, pacesetter potentials or action potentials are not recordable, and a “silent” reproducible electrorectogram is obtained.402 Recently an electrorectogram has been recorded percutaneously;268 likewise, percutaneous electrovesicography and electrovasography studies in normal and pathologic conditions have been carried out.403,404

Anatomic Complications of Surgery of Anorectum

Chronic Anal Fissure and Hemorrhoids

Manual dilatation of the anus has been advocated for treatment of chronic anal fissure and hemorrhoids.405,406 It is likely that it acts by disrupting the anorectal band described by Shafik285 with a resulting release of the anal stenosis which seems to be responsible for the high rectal neck pressure that accompanies chronic anal fissure and hemorrhoids.289,292

The eight-finger dilatation may be accompanied by unpredictable sphincter damage.342 Anorectal bandotomy292 divides the band more accurately than manual dilatation and causes no sphincteric damage.

Internal anal sphincterotomy is widely used in the treatment of chronic anal fissure. Partial fecal incontinence, i.e. incontinence to fluid stools or flatus, occurs in approximately 25%. Fecal soiling may occur due to the “key-hole” deformity created by a posterior internal sphincterotomy, while lateral sphincterotomy might cause a less prominent groove and fewer control disorders than a posterior one. “Stress defecation”163 or inability to hold stools for a long period might occur after internal sphincterotomy due to loss of the voluntary inhibition reflex as already mentioned.

Complications related to open or closed hemorrhoidectomy are few. Bleeding occurs in 4% of cases and is commonly of a secondary nature.407

Anal stenosis may be encountered due to excessive removal of anoderm. Fecal incontinence results from the destruction of sphincters during excision of the hemorrhoidal masses. It occurs in 0.4% of cases. Recurrence may occur from missed daughter or mother hemorrhoids.


The commonest complications of fistula operations are recurrence and functional disorders. Recurrence may be a true recurrence due to incomplete excision of the track or a “de novo” fistula due to infection of new areas of epithelial debris.408 Functional disorders occur in the form of complete or partial fecal incontinence as a result of injury of the anal sphincters. Incontinence for solid stools was encountered in 3.4 % and for loose stools or flatus in 17% and 25%, respectively.409 Fistulotomy may be associated with partial fecal incontinence since the internal anal sphincter is always divided in such operations. Fistulectomy injures the external anal sphincter more commonly than fistulotomy and thus has a higher incidence of incontinence. For this reason, fistulotomy is now more commonly practiced than fistulectomy.

Rectal Prolapse

Both constipation with excessive straining (strainodynia) and fecal incontinence are frequently associated with rectal procidentia. Preceding studies have shown that procidentia is due to levator subluxation and sagging that might result from pregnancy trauma, strainodynia, or senility.325 Levator sagging exposes the rectal neck to the direct action of the intraabdominal pressure. At stool, the increased intraabdominal pressure, acting on the rectal neck, obstructs the fecal stream with a resulting constipation.333 In cases with advanced levator subluxation and sagging, the pudendal nerve is pulled down and stretched. This might lead to pudendal neuropathy and pudendal canal syndrome368 which presents with fecal incontinence.369

Operations for cure of rectal prolapse are either perineal or abdominal. The anatomic complications of the abdominal rectopexy with or without sigmoidectomy are constipation which is attributable to the constricting action of the graft applied to the rectum. Furthermore, a preexisting constipation or fecal incontinence may not improve after rectopexy as the latter cures the rectal prolapse by mechanical fixation of the rectum without managing the disordered levator ani muscle.410 Therefore, repair of the levator muscle should be an integral part in any operation for the cure of rectal prolapse.410

Rectal Cancer

Today operations for rectal cancer are directed to sphincter-saving procedures. In tumors of the upper or middle third of the rectum, there is no problem with continence. Tumors of the lower third may compromise continence due to encroachment on the control mechanism, whether sphincteric or sensory. These patients complain commonly of fecal soiling.

Other complications that occur during abdominal dissection of the rectum are injury to the pelvic nerve leading to urinary and sexual disorders in the form of urinary retention, impotence, and ejaculatory failure. To avoid these complications, nerve-sparing operations are now adopted in many centers. Other anatomic complications include ureteric injuries and median sacral artery injury.

Abdominoperineal Resection

The anatomic complications of abdominoperineal resection can be summarized as: (1) vascular injury, (2) organ injury, and (3) nerve injury.

Presacral veins, the left iliac vein, the middle rectal artery (if present), and the inferior hemorrhoidal vessels are subject to injury.

Trauma to the left ureter, duodenum, urinary bladder, and male urethra must be avoided. Other complications of abdominoperineal resection are: inadequate reconstruction of the peritoneal floor, small bowel obstruction, colostomy complications, rupture of the rectum, and contamination.

Injury to the sympathetic or parasympathetic nerves may result in bladder dysfunction, failure of ejaculation, impotence, and retention of urine.41 The surgeon should remember that the pelvic splanchnics (nervi erigentes) penetrate the fascia of Waldeyer; therefore, the fascia of Waldeyer should be divided at the tip of the coccyx.

The ureter (and the lateral ligaments of the rectum) must be traced deep into the pelvis by careful dissection without elevation; after this, division of the colon and formation of the colostomy can be done. The pelvic peritoneum should be closed to avoid herniation and obstruction of the small bowel.

The perineal phase of the abdominoperineal dissection encounters the following structures: the pudendal vessels (which should be ligated), the levator plate (which should be excised widely), and the membranous urethra of the male (into which a Foley catheter is placed prior to surgery). Use sharp dissection to separate the prostate from the lower rectum. The perineum should be partially or completely closed; a suction drain is advisable.


1. Van Damme JPJ. Behavioral anatomy of the abdominal arteries. Surg Clin North Am 1993;73:699-725.

2. Ferguson JA, Parks AG. Cited in Rains AJH, Ritchie HD (eds). Bailey & Love’s Short Practice of Surgery (18th ed). London: HK Lewis & Co, 1981, p. 1095.

3. Langman JM, Rowland R. The number and distribution of lymphoid follicles in the human large intestine. J Anat 1986;149: 189-194. [PubMed: 3693106]

4. Okamota E, Ueda T. Embryogenesis of intramural ganglia of the gut and its relation to Hirschsprung’s disease. J Pediatr Surg 1967;2: 437.

5. O’Rahilly R, Müller F. Human Embryology & Teratology (2nd ed). New York: Wiley-Liss, 1996, p. 234.

6. Rowe JS Jr, Skandalakis JE, Gray SW, Olafson RP, Steinmann RJ. The surgical anal canal. Contemp Surg 5:107-116, 1974.

7. Gross RE. The Surgery of Infancy and Childhood. Philadelphia: Saunders, 1958.

8. Davis DL, Poynter CWM. Congenital occlusions of the intestine. Surg Gynecol Obstet 1922;34:35.

9. Sturim HS, Ternberg JL. Congenital atresia of the colon. Surgery 1966;59:458. [PubMed: 5906239]

10. Dalla Vecchia LK, Grosfeld JL, West KW, Rescorla FJ, Scherer LR, Engum SA. Intestinal atresia and stenosis: a 25-year experience with 277 cases. Arch Surg 1998;133:490-497.

11. Lambrecht W, Kluth D. Hereditary multiple atresias of the gastrointestinal tract: report of a case and review of the literature. J Pediatr Surg 1998;33:794-797. [PubMed: 9607506]

12. Stone WD, Hendrix TR, Schuster MM. Aganglionosis of the entire colon in an adolescent. Gastroenterology 48:636-641, 1965. [PubMed: 14277359]

13. Skandalakis JE, Gray SW. Embryology for Surgeons (2nd ed). Baltimore: Williams & Wilkins, 1994.

14. Lohlun J, Margolis M, Gorecki P, Schein M. Fecal impaction causing megarectum-producing colorectal catastrophes. A report of two cases. Dig Surg 2000;17:196-198. [PubMed: 10781993]

15. Wulkan ML, Georgeson KE. Primary laparoscopic endorectal pull-through for Hirschsprung’s disease in infants and children. Semin Laparosc Surg 1998;5:9-13. [PubMed: 9516554]

16. Balthazar EJ. Congenital positional anomalies of the colon: Radiographic diagnosis and clinical implications. Gastrointest Radiol 1977;2:41-47. [PubMed: 615801]

17. DePrima SJ, Hardy DC, Brant WE. Reversed intestinal rotation. Radiology 1985;157:603-604. [PubMed: 4059546]

18. Prassopoulos P, Raissaki M, Daskalogiannaki M, Gourtsoyiannis N. Retropsoas positioned bowel: incidence and clinical relevance. J Comput Assisted Tomogr 1998;22:304-307. [PubMed: 9530399]

19. Herman TE, Coplen D, Skinner M. Congenital short colon with imperforate anus. Pediatr Radiol 2000;30:243-246. [PubMed: 10789902]

20. Wakhlu AK, Wakhlu A, Pandey A, Agarwal R, Tandon RK, Kureel SN. Congenital short colon. World J Surg 20:107-114, 1996. [PubMed: 8588400]

21. Kim HL, Gow KW, Penner JG, Blair GK, Murphy JJ, Webber EM. Presentation of low anorectal malformations beyond the neonatal period. Pediatrics 2000;105:E68.

22. Connaughton JC, Poletti L, Broderick T, Sugerman HJ. Rectal duplication cyst with a large perineal hernia presenting as recurrent perineal abscesses. Surgery 124:926-928, 1998. [PubMed: 9823409]

23. Susuki K, Baba M, Sasaki K. Two cases of non-rotation in adults: Its anatomical analysis. Clin Anat 1993;6:111-115.

24. Peveretos P, Krespis E, Polydorou A, Golematis B. Incontinence anale idiopathique: Cure chirurgicale [French]. J Chir (Paris) 1987; 124:399-402.

25. Gordon PH. Retrorectal tumors. In: Gordon PH, Nivatvongs S (eds.). Principles and Practice of Surgery for the Colon, Rectum, and Anus, St. Louis, MO: Quality Medical Publishing, Inc. 1992, pp. 383-400.

26. Williams PL, Warwick R, Dyson M, Bannister LH (eds). Gray’s Anatomy (37th ed). Edinburgh: Churchill Livingstone, 1989, pp. 1774-1787, p. 194.

27. Goligher JC. Surgical anatomy of the colon, rectum and anal canal. In: Turell R (ed). Diseases of the Colon and Anorectum (2nd ed). Philadelphia: WB Saunders, 1969.

28. Saunders BP, Phillips RK, Williams CB. Intraoperative measurement of colonic anatomy and attachments with relevance to colonoscopy. Br J Surg 1995;82:1491-1493. [PubMed: 8535800]

29. Sadahiro S, Ohmura T, Yamada Y, Saito T, Taki Y. Analysis of length and surface area of each segment of the large intestine according to age, sex, and physique. Surg Radiol Anat 1992;14: 251-257. [PubMed: 1440190]

30. Steward J, Rankin FW. Blood supply of the large intestine: Its surgical considerations. Arch Surg 1933;26:843.

31. George R. Topography of the unpaired visceral branches of the abdominal aorta. J Anat 1935;69:196.

32. Michels NA, Siddharth P, Kornblith PL, Parke WW. The variant blood supply to the small and large intestine: its import in regional resections. J Int Coll Surg 1963;39:127.

33. Cheng B, Chen K, Gao S, Tu Z. Colon interposition. Recent Results Cancer Res 2000;155:151-160. [PubMed: 10693249]

34. Furst H, Hartl WH, Lohe F, Schildberg FW. Colon interposition for esophageal replacement: an alternative technique based on the use of the right colon. Ann Surg 2000;231:173-178. [PubMed: 10674607]

35. Griffiths JD. Surgical anatomy of the blood supply of the distal colon. Ann R Coll Surg 1956;19:241-256. [PubMed: 13363265]

36. Van Damme JPJ, Bonte J. Vascular Anatomy in Abdominal Surgery. New York: Thieme Verlag, 1990.

37. Dworkin MJ, Allen-Mersh TG. Effect of inferior mesenteric artery ligation on blood flow in the marginal artery – dependent sigmoid colon. J Am Coll Surg 1996;183:357-360. [PubMed: 8843265]

38. Adachi Y, Kakisako K, Sato K, Shiraishi N, Miyahara M, Kitano S. Factors influencing bowel function after low anterior resection and sigmoid colectomy. Hepatogastroenterology 2000;47:155-158. [PubMed: 10690600]

39. Haigh PI, Temple WJ. The gargantuan marginal artery sign: a case report of averting total necrosis of the small intestine. Surgery 123; 362-364, 1998.

40. Sudeck P. Ueber die Gefässversorgung des Mastdarmes in Hinsicht auf die operative Gangrän. Med Wschr (Muenchen) 1907;54:1314.

41. Goligher JC. Surgery of the Anus, Rectum and Colon (4th ed). London: Baillière Tindall, 1980.

42. Bertelli L, Lorenzini L, Bertelli E. The arterial vascularization of the large intestine: anatomical and radiological study. Surg Radiol Anat 1996;18(Suppl I):S1-S59.

43. Voiglio EJ, Boutillier du Retail C, Neidhardt JPH, Caillot JL, Barale F, Mertens P. Gastrocolic vein: Definition and report of two cases of avulsion. Surg Radiol Anat 1998;20:197-201. [PubMed: 9706679]

44. Fishman SJ, Shamberger RC, Fox VL, Burrows PE. Endorectal pull-through abates gastrointestinal hemorrhage from colorectal venous malformations. J Pediatr Surg 2000;35:982-984. [PubMed: 10873049]

45. Yada H, Sawai K, Taniguchi H, Hoshima M, Katoh M, Takahashi T. Analysis of vascular anatomy and lymph node metastases warrants radical segmental bowel resection for colon cancer. World J Surg 21:109-115, 1997. [PubMed: 8943187]

46. Guyton AC, Hall JE. Textbook of Medical Physiology (9th ed). Philadelphia: WB Saunders Co, 1996.

47. Anson BJ, McVay CB. Surgical Anatomy (5th ed). Philadelphia: Saunders, 1971.

48. Pavlov S, Pétrov V. Sur l’anse sous-clavièr de l’artère sous-clavièr droite rétro-oesophagienne. Folia Med (Plovdiv) 1968;10:73-78. [PubMed: 4972026]

49. Skandalakis JE. Volvulus of the cecum: clinical, anatomical and experimental work (thesis). Athens (Greece): University of Athens Medical School, 1949.

50. Treves F. The anatomy of the intestinal canal and peritoneum in man. Br Med J 1885;1:415.

51. Sappey PC. Traité d’Anatomie Descriptive (3rd ed). Paris: VA Delahaye et Cie., 1879. Cited in: DiDio LJA, Anderson MC. The “Sphincters” of the Digestive System: Anatomical, Functional and Surgical Considerations. Baltimore: Williams & Wilkins, 1968, p. 152.

52. DiDio LJA, Anderson MC. The “Sphincters” of the Digestive System: Anatomical, Functional and Surgical Considerations. Baltimore: Williams & Wilkins, 1968.

53. Kumar D, Phillips SF. The contribution of external ligamentous attachments to function of the ileocecal junction. Dis Colon Rectum 1987;30:410-416. [PubMed: 3595357]

54. Bogers JJ, Van Marck E. The ileocecal junction. Histol Histopath 1993;8:561-566. [PubMed: 8358165]

55. Gedgaudas-McClees, R Kristin. (Clinical Professor of Radiology, Saint Joseph’s Hospital, Atlanta GA) Letter to: John E. Skandalakis. 1995 Dec 20.

56. Silverman PM, Kelvin FM, Baker ME, Cooper C. Computed tomography of the ileocecal region. Comput Med Imaging Graph 1988; 12:293-303.

57. Khope S, Rao PL. Cecal volvulus in a 2-month-old baby. J Pediatr Surg 1988;23:1038. [PubMed: 3244083]

58. Shoop SA, Sackier JM. Laparoscopic cecopexy for cecal volvulus. Case report and a review of the literature [Review]. Surg Endosc 1993;7:450-454. [PubMed: 8211629]

59. Frank AJ, Goffner LB, Fruauff AA, Losada RA. Cecal volvulus: The CT whirl sign. Abdom Imaging 1993;18:288-289. [PubMed: 8508096]

60. Moore JH, Cintron Jr, Duarte B, Espinosa G, Abcarian H. Synchronous cecal and sigmoid volvulus: Report of a case (Review]. Dis Colon Rectum 1992;35:803-805. [PubMed: 1644007]

61. Theuer C, Cheadle WG. Volvulus of the colon. Am Surg 1991;57: 145-150. [PubMed: 2003701]

62. VanderKolk W, Snyder CA, Figg DM. Cecal-colic adult intussusception as a cause of intestinal obstruction in Central Africa. World J Surg 20:341-344, 1996. [PubMed: 8661842]

63. Lear TF, Gray SW, Milsap JH, Skandalakis JE. Benign lesions of the right colon. J Med Assoc Ga 60:271-275, 1971. [PubMed: 5561498]

64. Astler VB, Miller EB, Snyder RS, McIntyre CH, Lillie RH. Benign surgical lesions of the cecum. Arch Surg 86:435, 1963. [PubMed: 13965296]

65. Hatch KF, Blanchard DK, Hatch GF III, Wertheimer-Hatch L, Davis GB, Foster RS Jr, Skandalakis JE. Tumors of the appendix and colon. World J Surg 24:430-436, 2000. [PubMed: 10706915]

66. Hatch KF, Blanchard DK, Hatch GF III, Wertheimer-Hatch L, Davis GB, Foster RS Jr, Skandalakis JE. Tumors of the rectum and anal canal. World J Surg 24:437-443, 2000. [PubMed: 10706916]

67. Simon AM, Birnbaum BA, Jacobs JE. Isolated infarction of the cecum. Radiology 2000;214:513-516. [PubMed: 10671602]

68. Nelson RL. Division of the colorectum into anatomic subsites: Why and where? J Surg Oncol 1998;69:1-3. [PubMed: 9762883]

69. Poon RT-P, Chu K-W. Inflammatory cecal masses in patients presenting with appendicitis. World J Surg 1999;23:713-716. [PubMed: 10390592]

70. Mann GN, Scoggins CR, Adkins B. Perforated cecal adenocarcinoma presenting as a thigh abscess. South Med J 90(9):949-951, 1997.

71. Skandalakis JE, Colborn GL, Skandalakis LJ. The embryology of the inguinofemoral area: an overview. Hernia 1:45-54, 1997.

72. Wolfer JA, Beaton LE, Anson BJ. Volvulus of the cecum: anatomical factors in its etiology. Report of a case. Surg Gynecol Obstet 1942;74:882.

73. Symington J. The relations of the peritoneum to the descending colon in the human subject. J Anat Physiol 1892;26:530.

74. Haskin PH, Teplick SK, Teplick JG, Haskin ME. Volvulus of the cecum and right colon. JAMA 1981;245:2433. [PubMed: 7230481]

75. Rogers RL, Harford FJ. Mobile cecum syndrome. Dis Colon Rectum 1984;27:399-402. [PubMed: 6734364]

76. Wright TP, Max MH. Cecal volvulus: review of 12 cases. South Med J 1988;81:1233-1235. [PubMed: 3175730]

77. Ballantyne GH, Brandner MD, Beart RW Jr, Ilstrup DM. Volvulus of the colon: Incidence and mortality. Ann Surg 1985;202:83-92. [PubMed: 4015215]

78. Rabinovici R, Simansky DA, Kaplan O, Mavor E, Manny J. Cecal volvulus [Review]. Dis Colon Rectum 1990;33:765-769. [PubMed: 2202566]

79. Jones IT, Fazio VW. Colonic volvulus: Etiology and management [Review]. Dig Dis 1989;7:203-209. [PubMed: 2663257]

80. Benacci JC, Wolff BG. Cecostomy. Therapeutic indications and results. Dis Colon Rectum 1995;38:530-534. [PubMed: 7736886]

81. Marinella MA. Acute colonic pseudo-obstruction complicated by cecal perforation in a patient with Parkinson’s disease. South Med J 1997;90(10):1023-1026.

82. Woodburne RT, Burkel WE. Essentials of Human Anatomy (9th ed). New York: Oxford University Press, 1994.

83. Charnsangavej C, DuBrow RA, Varma DGK, Herron DH, Robinson TJ, Whitley NO. CT of the mesocolon. Part 1. Anatomic considerations. Radiographics 1993;13:1035-1045. [PubMed: 8210588]

84. Charnsangavej C, DuBrow RA, Varma DGK, Herron DH, Robinson TJ, Whitley NO. CT of the mesocolon. Part 2. Pathologic considerations. Radiographics 1993;13:1309-1322. [PubMed: 8290726]

85. Robillard GL, Shapiro AL. Variational anatomy of the middle colic artery. J Int Coll Surg 1947;10:157.

86. Hollinshead WH. Anatomy for Surgeons (2nd ed). New York: Harper & Row, 1968, Volume 2, p. 497.

87. Horiuchi T, Shimomatsuya T, Uchinami M, Yoshida M, Amaya H, Aotake T, Chiba Y, Imamura Y. Retroperitoneoscopic excision of a mesenteric cyst. J Laparoendosc Adv Surg Tech (Part A) 2000;10;59-61.

88. Dixon AK, Henderson PJ, Oliver JL, Matthewson MH. Posteriorly situated retroperitoneal colon: A study using CT and MRI. Clin Anat 1993;6:269-274.

89. Hadar H, Gadoth N. Positional relations of colon and kidney determined by perirenal fat. Am J Roentgenol 1984;143:773-776. [PubMed: 6332483]

90. Sherman JL, Hopper KD, Green AJ, Johns TT. The retrorenal colon on computed tomography: A normal variant. J Comput Assist Tomogr 1985;9:339-341. [PubMed: 3973159]

91. Hopper KD, Maj MC, Sherman JL, Luethke JM. The retrorenal colon in the supine and prone patient. Radiology 1987;162:443-446. [PubMed: 3797658]

92. Helms CA, Munk PL, Witt WS, Davis GW, Morris J, Onik G. Retrorenal colon: Implications for percutaneous diskectomy. Radiology 1989;171:864-865. [PubMed: 2717766]

93. Prassopoulos PN, Gourtsoyiannis N, Cavouras D, Pantelidis N. A study of the variation of colonic positioning in the pararenal space as shown by computed tomography. Eur J Radiol 1990;10:44-47. [PubMed: 2311605]

94. LeRoy AJ, Williams HJ, Bender CE, Segura JW, Patterson DE, Benson RC. Colon perforation following percutaneous nephrostomy and renal calculus removal. Radiology 1985;155:83-85. [PubMed: 3975424]

95. Bonaldi G, Belloni G, Prosetti D, Moschini L. Percutaneous discectomy using Onik’s method: 3 years’ experience. Neuroradiology 1991;33:516-519. [PubMed: 1780054]

96. Grant JCB, Basmajian JV. Grant’s Method of Anatomy (7th ed). Baltimore: Williams & Wilkins, 1965, p. 258.

97. Manasse P. Die arterielle Gefässversorgung des S. Romanum in ihrer Bedetung für die operative Verlagerung desselben. Arch Klin Chir 1907;83:999.

98. Shafik A. Sigmoido-rectal junction reflex: role in the defecation mechanism. Clin Anat 9:391-394, 1996. [PubMed: 8915619]

99. Anson BJ. An Atlas of Human Anatomy (2nd ed). Philadelphia: Saunders, 1963.

100. Vaez-Zadeh K, Dutz W. Ileosigmoid knotting. Ann Surg 1970; 172:1027. [PubMed: 5496474]

101. Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation: IV. Anatomy of the perianal spaces. Invest Urol 13:424, 1976. [PubMed: 1270238]

102. Milligan ETC, Morgan CN, Jones LE, Officer R. Surgical anatomy of the anal canal, and the operative treatment of haemorrhoids. Lancet 1937:2:1119.

103. Courtney H. Anatomy of the pelvic diaphragm and anorectal musculature as related to sphincter preservation in anorectal surgery. Am J Surg 1950;79:155. [PubMed: 15399368]

104. Milligan ETC. The surgical anatomy and disorders of the perianal space. Proc R Soc Med 1943;36:365.

105. Shafik JA. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation: III. The longitudinal anal muscle: Anatomy and role in anal sphincter mechanism. Invest Urol 1976;13:271. [PubMed: 1245390]

106. Haagensen CD, Feind CR, Herter FP, Slanetz CA, Weinberg JA. The Lymphatics in Cancer. Philadelphia: Saunders, 1972.

107. Parks AG. Pathogenesis and treatment of fistula-in-ano. Br Med J 1961;1:463.

108. Jackman RJ, Clark PL III, Smith ND. Retrorectal tumors. JAMA 1951;145:956-962. [PubMed: 14803290]

109. Heald RJ, Moran BJ. Embryology and anatomy of the rectum. Semin Surg Oncol 1998;15:66-71. [PubMed: 9730411]

110. Heald RJ. The “Holy Plane” of rectal surgery. J R Soc Med 1988;81:503-508. [PubMed: 3184105]

111. Konerding MA, Heintz A, Huhn P, Junginger T. [Rectal carcinoma. Optimizing therapy by knowledge of anatomy with special reference to the mesorectum]. [German] Zentralbl Chir 1999;124; 413-417.

112. Enker WE, Kafka NJ, Martz J. Planes of sharp pelvic dissection for primary, locally advanced, or recurrent rectal cancer. Semin Surg Oncol 2000;18:199-206. [PubMed: 10757885]

113. Law WL, Chu KW, Ho JWC, Chan CW. Risk factors for anastomotic leakage after low anterior resection with total mesorectal excision. Am J Surg 2000;179:92-96. [PubMed: 10773140]

114. Aeberhard P, Fasolini F, del Monte G. Anatomical basis and rationale of total mesorectal excision. Swiss Surg 1997;3:243-247. [PubMed: 9427862]

115. Hohenberger W, Schick CH, Gohl J. Mesorectal lymph node dissection: is it beneficial? Langenbecks Arch Surg 1998;386;402-408.

116. Tiret E, Pocard M. [Total excision of the mesorectum and preservation of the genitourinary innervation in surgery of rectal cancer]. [French] Ann Chir 1999;53:507-514. [PubMed: 10427843]

117. Leo E, Belli F, Andreola S, Gallino G, Bonfanti G, Ferro F, Zingaro E, Sirizotti G, Civelli E, Valvo F, Gios M, Brunelli C. Total rectal resection and complete mesorectum excision followed by coloendoanal anastomosis as the optimal treatment for low rectal cancer: the experience of the National Cancer Institute of Milano. Ann Surg Oncol 2000;7: 85-86 and 125-132.

118. Leong F. Selective total mesorectal excision for rectal cancer. Dis Colon Rectum 2000;43:1237-1240. [PubMed: 11005489]

119. Tocchi A, Mazzoni G, Lepre L, Liotta G, Costa G, Agostini N, Miccini M, Scucchi L, Frati G, Tagliacozzo S. Total mesorectal excision and low rectal anastomosis for the treatment of rectal cancer and prevention of pelvic recurrences. Arch Surg 2001; 136:216-220.

120. Maurer CA, Renzulli P, Meyer JD, Buchler MW. [Rectal carcinoma. Optimizing therapy by partial or total mesorectum removal]. [German] Zentralbl Chir 1999;124:428-435. [PubMed: 10420530]

121. Burke RM, Zavela D, Kaump DH. Significance of the anal gland. Am J Surg 1951;82:659. [PubMed: 14885618]

122. Michels NA. Blood Supply and Anatomy of the Upper Abdominal Organs with a Descriptive Atlas. Philadelphia: Lippincott, 1955.

123. Boxall TA, Smart PJG, Griffiths JD. The blood supply of the distal segment of the rectum in anterior resection. Br J Surg 1962;50:399.

124. Last RJ. Anatomy: Regional and Applied (5th ed). Baltimore: Williams & Wilkins, 1972.

125. Vogel P, Klosterhalfen B. The surgical anatomy of the rectal and anal blood vessels [German]. Langenbecks Arch Chir 1988;373: 264-269. [PubMed: 3268710]

126. Goligher JC. The blood supply to the sigmoid colon and rectum. Br J Surg 1949;37:157-162. [PubMed: 15395084]

127. Batson OV. The function of the vertebral veins and their role in the spread of metastases. Ann Surg 1940;112:138.

128. Koch M, Weitz J, Kienle P, Benner A, Willeke F, Lehnert T, Herfarth C, von Knebel Doeberitz M. Comparative analysis of tumor cell dissemination in mesenteric, central, and peripheral venous blood in patients with colorectal cancer. Arch Surg 2001;136:85-89. [PubMed: 11146784]

129. Ueno H, Mochizuki H, Fujimoto H, Hase K, Ichikura T. Autonomic nerve plexus involvement and prognosis in patients with rectal cancer. Br J Surg 2000;87:92-96. [PubMed: 10606917]

130. Grinnell RS. Lymphatic metastases of carcinoma of the colon and rectum. Ann Surg 1950;131:494.

131. Morgan CN. Treatment of cancer of the rectum. Am J Surg 1968; 115:442. [PubMed: 4172220]

132. Morgan CN. Carcinoma of the rectum. Ann R Coll Surg Engl 1965;36:73. [PubMed: 14259236]

133. Williams ST, Beart RW Jr. Staging of colorectal cancer. Semin Surg Oncol 1992;8:89-93. [PubMed: 1615269]

134. Granfield CA, Charnsangavej C, Dubrow RA, Varma DG, Curley SA, Whitley NO, Wallace S. Regional lymph node metastases in carcinoma of the left side of the colon and rectum: CT demonstration. AJR Am J Roentgenol 1992;159:757-761. [PubMed: 1529837]

135. Madden JL, McVeigh GJ. The extension of operation in the treatment of carcinoma in the region of the splenic flexure. Surg Clin North Am 1954;34:523.

136. Block IR, Enquist IF. A more radical resection for carcinoma of the rectum in the female. Surg Gynecol Obstet 1964;119:1328. [PubMed: 14227763]

137. Davies MRQ. Anatomy of the nerve supply of the rectum, bladder, and internal genitalia in anorectal dysgenesis in the male. J Pediatr Surg 32(4):536-541, 1997.

138. Duthie HL, Gairns FW. Sensory nerve endings and sensation in the axial region of man. Br J Surg 1960;47:38.

139. Stelzner F, Fritsch H, Fleischhauer K. The surgical anatomy of the genital nerves of the male and their preservation in excision of the rectum [German]. Chirurg 1989;60:228-234. [PubMed: 2721301]

140. Williams JT, Slack WW. A prospective study of sexual function after major colorectal surgery. Br J Surg 1980;67:772-774. [PubMed: 7427033]

141. Pearl RK, Monsen H, Abcarian H. Surgical anatomy of the pelvic autonomic nerves: A practical approach. Am Surg 1986;52(5): 236-237.

142. Gaston EA. The physiology of fecal continence. Surg Gynecol Obstet 1948;87:280.

143. Gunterberg B, Kewenter J, Petersen I, Stener B. Anorectal function after major resections of the sacrum with bilateral or unilateral sacrifice of sacral nerves. Br J Surg 1976;63:546-554. [PubMed: 953450]

144. Speakman CTM, Hoyle CHV, Kamm MA, Henry MM, Nicholls RJ, Burnstock G. Decreased sensitivity of muscarinic but not 5-hydroxytryptamine receptors of the internal anal sphincter in neurogenic faecal incontinence. Br J Surg 1992;79:829-832. [PubMed: 1393486]

145. Sunderland S. Nerves and Nerve Injuries (2nd ed). Edinburgh: Churchill Livingstone, 1978, pp. 82-86.

146. Henry MM, Parks AG, Swash M. The pelvic floor musculature in the descending perineum syndrome. Br. J Surg 1982;69:470-472. [PubMed: 7104636]

147. Fornell EKU, Berg G, Hallböök O, Matthiesen LS, Sjödahl R. Clinical consequences of anal sphincter rupture during vaginal delivery. J Am Coll Surg 1996;183:553-558. [PubMed: 8957456]

148. Chifflet A. Surgery for cancer of the lower rectum: The perirectal fascia with reference to conservative surgery and technic. Dis Colon Rectum 1964;7:493.

149. Watson SJ, Loder PB, Halligan S, Bartram CI, Kamm MA, Phillips RKS. Transperineal repair of symptomatic rectocele with Marlex mesh: a clinical, physiological and radiologic assessment of treatment. J Am Coll Surg 183:257-261, 1996. [PubMed: 8784320]

150. Merad F, Hay J-M, Fingerhut A, Yahchouchi E, Laborde Y, Pélissier E, Msika S, Flamant Y, The French Association for Surgical Research. Is prophylactic pelvic drainage useful after elective rectal or anal anastomosis? A multicenter controlled randomized trial. Surgery 1999;125:529-535. [PubMed: 10330942]

151. Ross A, Rusnak C, Weinerman B, Kuechler P, Hayashi A, Mac Lachlan G, Frew E, Dunlop W. Recurrence and survival after surgical management of rectal cancer. Am J Surg 1999;177:392-395. [PubMed: 10365877]

152. Siddharth P, Ravo B. Colorectal neurovasculature and anal sphincter. Surg Clin North Am 1988;68:1185-2000. [PubMed: 3194816]

153. Healey JE Jr, Hodge J. Surgical Anatomy (2nd ed). Toronto: BC Decker, 1990, p. 262.

154. Azimuddin K, Stasik JJ, Rosen L, Riether RD, Khubchandani IT. Dieulafoy’s lesion of the anal canal: a new clinical entity. Report of two cases. Dis Colon Rectum 2000;43:423-426. [PubMed: 10733128]

155. Rao PM, Rhea JT, Wittenberg J, Warshaw AL. Misdiagnosis of primary epiploic appendagitis. Am J Surg 176:81-85, 1998. [PubMed: 9683140]

156. Kuganeswaran E, Fisher JK. Giant sigmoid diverticulum: a rare manifestation of diverticular disease. South Med J 91(10):952-955, 1998.

157. Welton ML. Human colonic microvascular endothelial cells as a model of inflammatory bowel disease. Am J Surg 174:247-250, 1997. [PubMed: 9324131]

158. Gordon PH, Nivatvongs S. Principles and Practice of Surgery for the Colon, Rectum, and Anus. St. Louis: Quality Medical Publishing, 1992, p. 7.

159. O’Kelly TJ, Brading A, Mortensen NJ. In vitro response of the human anal canal longitudinal muscle layer to cholinergic and adrenergic stimulation: evidence of sphincter specialization. Br J Surg 1993;80:1337-1341. [PubMed: 16456908]

160. Delancey JO, Toglia MR, Perucchini D. Internal and external anal sphincter anatomy as it relates to midline obstetric lacerations. Obstet Gynecol 1997;90:924-927. [PubMed: 9397104]

161. Haas PA, Fox TA Jr. The importance of the perianal connective tissue in the surgical anatomy and function of the anus. Dis Colon Rectum 1977;20:303-313. [PubMed: 862491]

162. Sangwan YP, Solla JA. Internal anal sphincter. Advances and insights. Dis Colon Rectum 1998;41:1297-1311. [PubMed: 9788395]

163. Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation: IX. Single loop continence: A new theory of the mechanism of anal continence. Dis Colon Rectum 1980;23:37. [PubMed: 7379650]

164. Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. The external anal sphincter: A triple loop system. Invest Urol 1975;12:412 [PubMed: 1112669]

165. Oh C, Kark AE. Anatomy of the external anal sphincter. Br J Surg 1972;59:717. [PubMed: 5070135]

166. Hollinshead WH. Anatomy for Surgeons. New York: Hoeber-Harper, 1956.

167. Kratzer GL, Dockerty MB. Histopathology of the anal ducts. Surg Gynecol Obstet 1947;84:333.

168. Klosterhalfen B. Offner F, Vogel P, Kirkpatrick CJ. Anatomic nature and surgical significance of anal sinus and anal intramuscular glands. Dis Colon Rectum 1991;34:156-160. [PubMed: 1704306]

169. Heiken JP, Zuckerman GR, Balfe DM. The hypertrophied anal papilla: recognition on air-contrast barium enema examinations. Radiology 1984;151:315-318. [PubMed: 6709897]

170. Kusunoki M, Horai T, Sakanoue Y, Yanagi H, Yamamura T, Utsunomiya J. Giant hypertrophied anal papilla. Case report. Eur J Surg 1991;157:491-492. [PubMed: 1681940]

171. Schutte AG, Tolentino MG. A second study of anal papillae. Dis Colon Rectum 1971;14:435. [PubMed: 5165593]

172. Mollen RM, Kuijpers JH, van Hoek F. Effects of rectal mobilization and lateral ligament division on colonic and anorectal function. Dis Colon Rectum 2000;43:1283-1287. [PubMed: 11005498]

173. Havenga K, DeRuiter MC, Enker WE, Welvaart K. Anatomical basic of autonomic nerve-preserving total mesorectal excision for rectal cancer. Br J Surg 1996;83:384-388. [PubMed: 8665201]

174. Gorski T, Khubchandani IT, Stasik JJ, Riether R. Retrorectal carcinoid tumor. South Med J 1999;92:417-420. [PubMed: 10219364]

175. Brasel KJ, Borgstrom DC, Weigelt JA. Management of penetrating colon trauma: A cost-utility analysis. Surgery 1999;125:471-479. [PubMed: 10330934]

176. Curran TJ, Borzotta AP. Complications of primary repair of colon injury: literature review of 2,964 cases. Am J Surg 1999; 177:42-47.

177. McDaniel KP, Charnsangavej C, DuBrow RA, Varma DG, Granfield CA, Curley SA. Pathways of nodal metastasis in carcinomas of the cecum, ascending colon, and transverse colon: CT demonstration. AJR Am J Roentgenol 1993;161:61-64. [PubMed: 8517322]

178. Lane JS, Sarkar R, Schmit PJ, Chandler CF, Thompson JE. Surgical approach to cecal diverticulitis. J Am Coll Surg 1999;188:629-634.

179. McDonald JM, Moonka R, Bell RH. Pathologic risk factors of occult malignancy in endoscopically unresectable colonic adenomas. Am J Surg 1999;177:384-387. [PubMed: 10365875]

180. Lopez-Kostner F, Lavery IC, Hool GR, Rybicki LA, Fazio VW. Total mesorectal excision is not necessary for cancers of the upper rectum. Surgery 124:612-618, 1998. [PubMed: 9780979]

181. Heald RJ, Smedh RK, Kald A, Sexton R, Moran BJ. Abdominoperineal excision of the rectum—an endangered operation. Dis Colon Rectum 1997;40:747-751. [PubMed: 9221846]

182. Havenga K, Enker WE, Norstein J, Moriya Y, Heald RJ, van Houwelingen HC, van de Velde CJ. Improved survival and local control after total mesorectal excision or D3 lymphadenectomy in the treatment of primary rectal cancer: an international analysis of 1411 patients. Eur J Surg Oncol 1999;25:368-374. [PubMed: 10419706]

183. Wimmer AP, Bouffard JP, Storms PR, Pilcher JA, Liang CY, DeGuide JJ. Primary colon cancer without gross mucosal tumor: unusual presentation of a common malignancy. South Med J 91(12):1173-1176, 1998.

184. Hayashi N, Egami H, Kai M, Kurusu Y, Takano S, Ogawa M. No-touch isolation technique reduces intraoperative shedding of tumor cells into the portal vein during resection of colorectal cancer. Surgery 1999;125:369-374. [PubMed: 10216526]

185. Condon RV, Lamphier TA. Surgical considerations in mobilizing the hepatic flexure. Surg Gynecol Obstet 1950;90:623. [PubMed: 15418432]

186. Phelan JT, Nadler SH. A technique of hemicolectomy for carcinoma of the right colon. Surg Gynecol Obstet 1968;126:355.

187. Koea JB, Conlon K, Paty PB, Guillem JG, Cohen AM. Pancreatic or duodenal resection or both for advanced carcinoma of the right colon: is it justified? Dis Colon Rectum 2000;43:460-465. [PubMed: 10789739]

188. Dixon CF. Anterior resection for carcinoma low in the sigmoid and the rectosigmoid. Surgery 1944;15:367.

189. Barnes JP. A pull-through technique for rectal resection. Surg Gynecol Obstet 1966;123:357. [PubMed: 5913489]

190. Sharma P, Klaasen H. Duodenal seromyectomy in the management of adherent colonic carcinoma in elderly patients. Can J Surg 40(4):289-293, 1997.

191. Eisen LK, Cunningham JD, Aufses AH. Intussusception in adults: institutional review. J Am Coll Surg 1999;188:390-395. [PubMed: 10195723]

192. Michelassi F, Hurst R. Restorative proctocolectomy with J-pouch ileoanal anastomosis. Arch Surg 2000;135:347-353. [PubMed: 10722040]

193. ReMine SG, Dozois RR. Hartmann’s procedure: Its use with complicated carcinomas of sigmoid colon and rectum. Arch Surg 1981; 116:630. [PubMed: 7235957]

194. Bell GA. Closure of colostomy following sigmoid colon resection for perforated diverticulitis. Surg Gynecol Obstet 1980;150:85. [PubMed: 7350709]

195. Bardoel JWJM, Stadelmann WK, Tobin GR, Werker PMN, Stremel RW, Kon M, Barker JH. Use of the rectus abdominis muscle for abdominal stoma sphincter construction: an anatomical feasibility study. Plast Reconstr Surg 2000;105:589-595. [PubMed: 10697165]

196. Tschmelitsch J, Wykypiel H, Promegger R, Bodner E. Colostomy vs tube cecostomy for protection of a low anastomosis in rectal cancer. Arch Surg 1999;134:1385-1388. [PubMed: 10593339]

197. Sakai Y, Nelson H, Larson D, Maidl L, Young-Fadok T, Ilstrup D. Temporary transverse colostomy vs loop ileostomy in diversion. Arch Surg 2001;136:338-342. [PubMed: 11231858]

198. Hartmann H. Nouveau procede d’ablation des cancers de la partie terminale du colon pelvien. Strasbourg (France): Trentieme Congres de Chirurgie, 1921.

199. Belmonte C, Klas JV, Perez JJ, Wong WD, Rothenberger DA, Goldberg SM, Madoff RD. The Hartmann procedure: first choice or last resort in diverticular disease? Arch Surg 131:612-617, 1996. [PubMed: 8645067]

200. Rosoff, L. Discussion. ReMine SG, Dozois RR. Hartmann’s procedure: Its use with complicated carcinomas of sigmoid colon and rectum. Arch Surg 1981;116:630.

201. Liu CD, Rolandelli R, Ashley SW, Evans B, Shin M, McFadden DW. Laparoscopic surgery for inflammatory bowel disease. Am Surg 61(12):1054-1056, 1995.

202. Ichihara T, Nagahata Y, Nomura H, Fukumoto S, Urakawa T, Aoyama N, Kuroda Y. Laparoscopic lower anterior resection is equivalent to laparotomy for lower rectal cancer at the distal line of resection. Am J Surg 2000;179:97-98. [PubMed: 10773141]

203. Reissman P, Cohen S, Weiss EG, Wexner SD. Laparoscopic colorectal surgery: ascending the learning curve. World J Surg 20: 277-282, 1996. [PubMed: 8661831]

204. Leung KL, Kwok SPY, Lau WY, Meng WCS, Lam TY, Kwong KH, Chung CC, Li AKC. Laparoscopic-assisted resection of rectosigmoid carcinoma. Arch Surg 132:761-765, 1997. [PubMed: 9230862]

205. Bouvet M, Mansfield PF, Skibber JM, Curley SA, Ellis LM, Giacco GG, Madary AR, Ota DM, Feig BW. Clinical, pathologic, and economic parameters of laparoscopic colon resection for cancer. Am J Surg 1998;176:554-558. [PubMed: 9926789]

206. Targarona EM, Martínez J, Nadal A, Balagué C, Cardesa A, Pascual S, Trias M. Cancer dissemination during laparoscopic surgery: tubes, gas, and cells. World J Surg 22:55-61, 1998. [PubMed: 9465762]

207. Tomita H, Marcello PW, Milsom JW. Laparoscopic surgery of the colon and rectum. World J Surg 1999;23:397-405. [PubMed: 10030864]

208. Stocchi L, Nelson H. Laparoscopic colectomy for colon cancer: trial update. J Surg Oncol 1998;68:255-267. [PubMed: 9721713]

209. Marubashi S, Hiroshi Y, Monden T, Hata T, Takahashi H, Fujita S, Kanoh T, Iwazawa T, Matsui S, Nakano Y, Tateishi H, Kinuta M, Takiguchi S, Okamura J. The usefulness, indications, and complications of laparoscopy-assisted colectomy in comparison with those of open colectomy for colorectal carcinoma. Surg Today 2000;30:491-496. [PubMed: 10883457]

210. Bussi M, Ferrero V, Riontino E, Gasparri G, Camandona M, Cortesina G. Problems in reconstructive surgery in the treatment of carcinoma of the hypopharyngoesophageal junction. J Surg Oncol 2000;74:130-133. [PubMed: 10914822]

211. Soffer SZ, Rosen NG, Hong AR, Alexianu M, Pena A. Cloacal exstrophy: a unified management plan. J Pediatr Surg 2000;35:932-937. [PubMed: 10873038]

212. Howell G, Caldamone A, Snyder H, Ziegler M, Ducket J. Optimal management of cloacal exstrophy. J Pediatr Surg 1983; 18:365-369.

213. Ricketts RR, Woodard JR, Zwiren GT, Andrews HG, Broecker BH. Modern treatment of cloacal exstrophy. J Pediatr Surg 1991;26:444-450. [PubMed: 2056406]

214. Saunders BP, Masaki T, Sawada T, Halligan S, Phillips RK, Muto T, Williams CB. A peroperative comparison of Western and Oriental colonic anatomy and mesenteric attachments. Int J Colorectal Dis 1995;10:216-221. [PubMed: 8568407]

215. Scoma JA. Salvati EP, Rubin RJ. Incidence of fistulas subsequent to anal abscesses. Dis Colon Rectum 1974;17:357. [PubMed: 4830804]

216. Adams D, Kovalcik PJ. Fistula in ano. Surg Gynecol Obstet 1981; 153:731. [PubMed: 7292274]

217. Nomikos IN. Anorectal abscesses: need for accurate anatomical localization of the disease. Clin Anat 10:239-244, 1997. [PubMed: 9213040]

218. Argov S, Levandovsky O. Open lateral sphincterotomy is still the best treatment for chronic anal fissure. Am J Surg 2000;179:201-202. [PubMed: 10827320]

219. Sharp FR. Patient selection and treatment modalities for chronic anal fissure. Am J Surg 171:512-515, 1996. [PubMed: 8651398]

220. da Silva JH. Pilonidal cyst: cause and treatment. Dis Colon Rectum 2000;43:1146-1156.

221. Davis KA, Mock CN, Verasaci A, Lentrichia P. Malignant degeneration of pilonidal cysts. Am Surg 1994;60:200-204. [PubMed: 8116982]

222. Herold A, Bruch HP. [Laparoscopic therapy of functional disorders of the rectum and pelvic floor]. [German] Langensbecks Arch Chir (Suppl Kongressband) 1997;114:905-908. [PubMed: 9574296]

223. Schultz I, Mellgren A, Dolk A, Johansson C, Holmstrom B. Long-term results and functional outcome after Ripstein retropexy. Dis Colon Rectum 2000;43:35-43. [PubMed: 10813121]

224. Blair S, Ellenhorn JD. Transanal excision for low rectal cancers is curative in early-stage disease with favorable histology. Am Surg 2000;66:817-820. [PubMed: 10993607]

225. Agachan F, Reissman P, Pfeifer J, Weiss EG, Nogueras JJ, Wexner SD. Comparison of three perineal procedures for the treatment of rectal prolapse. South Med J 90(9):925-932, 1997.

226. Pikarsky AJ, Joo JS, Wexner SD, Weiss EG, Nogueras JJ, Agachan F, Iroatulam A. Recurrent rectal prolapse: what is the next good option? Dis Colon Rectum 2000;43:1273-1276. [PubMed: 11005496]

227. Skandalakis LJ, Gadacz TR, Mansberger AR Jr, Mitchell WE Jr, Colborn GL, Skandalakis JE. Modern Hernia Repair. New York: Parthenon, 1996.

228. Dehni N, Cunningham C, Sarkis R, Parc R. Results of coloanal anastomosis for rectal cancer. Hepatogastroenterology 2000;47: 323-326. [PubMed: 10791180]

229. Read TE, Kodner IJ. Proctectomy and coloanal anastomosis for rectal cancer. Arch Surg 1999;134:670-677. [PubMed: 10367879]

230. Di Matteo G, Mascagni D, Zeri KP, Torretta A, Di Matteo FM, Maturo A, Peparini N. Evaluation of anal function after surgery for rectal cancer. J Surg Oncol 2000;74:11-14.

231. Madoff RD, Baeten CG, Christiansen J, Rosen HR, Williams NS, Heine JA, Lehur PA, Lowry AC, Lubowski DZ, Matzel KE, Nicholls RJ, Seccia M, Thorson AG, Wexner SD, Wong WD. Standards for anal sphincter replacement. Dis Colon Rectum 2000;43:135-141. [PubMed: 10696884]

232. Lamont JP, McCarty TM, Digan RD, Jacobson R, Tulanon P, Lichliter WE. Should locally excised T1 rectal cancer receive adjuvant chemoradiation? Am J Surg 2000;180:402-406. [PubMed: 11182387]

233. Gervaz P, Rotholtz N, Pisano M, Kaplan E, Secic M, Coucke P, Pikarsky A, Efron J, Weiss E, Wexner S. Quantitative short-term study of anal sphincter function after chemoradiation for rectal cancer. Arch Surg 2001;136:192-196. [PubMed: 11177140]

234. Borgstein P, Meijer S. Historical perspective of lymphatic tumour spread and the emergence of the sentinel node concept. Eur J Surg Oncol 1998;24:85-89. [PubMed: 9591019]

235. Saha S, Wiese D, Badin J, Beutler T, Nora D, Ganatra BK, Desai D, Kaushal S, Nagaraju M, Arora M, Singh T. Technical details of sentinel lymph node mapping in colorectal cancer and its impact on staging. Ann Surg Oncol 2000;7:82-84 and 120-124.

236. Schrock TR, DeVeney CW, Dunphy JE. Factors contributing to leakage of colonic anastomoses. Ann Surg 1973;177:513. [PubMed: 4540874]

237. Welch CE, Ottinger LW, Welch JP: Manual of Lower Gastrointestinal Surgery. New York: Springer-Verlag, 1980.

238. Tocchi A, Mazzoni G, Lepre L, Costa G, Liotta G, Agostini N, Miccini M. Prospective evaluation of omentoplasty in preventing leakage of colorectal anastomosis. Dis Colon Rectum 2000;43: 951-955. [PubMed: 10910241]

239. Moran B, Heald R. Anastomotic leakage after colorectal anastomosis. Semin Surg Oncol 2000;18:244-248. [PubMed: 10757890]

240. Patricio J, Bernades A, Nuno D, Falcao F, Silveira L. Surgical anatomy of the arterial blood supply of the human rectum. Surg Radiol Anat 1988;10:71-75. [PubMed: 3131901]

241. Jansen A, Brummelkamp WH, Davies GAG, Klopper PJ, Keeman JN. Clinical application of magnetic rings in colorectal anastomosis. Surg Gynecol Obstet 1981;153:537. [PubMed: 7280943]

242. Graham JW, Goligher JC. The management of accidental injuries and deliberate resections of the ureter during excision of the rectum. Br J Surg 1954;42:151. [PubMed: 13209038]

243. Andersson A, Bergdahl L. Urologic complications following abdominoperineal resection of the rectum. Arch Surg 1976;111: 969. [PubMed: 949259]

244. Béland G. The abdominal surgeon and the ureter. Can J Surg 1979;22:540. [PubMed: 16456907]

245. Kirkegaard P, Hjortrup A, Sanders S. Bladder dysfunction after low anterior resection for mid-rectal cancer. Am J Surg 1981;141: 266. [PubMed: 7457746]

246. Havenga K, Enker WE, McDermott K, Cohen AM, Minsky BD, Guillem J. Male and female sexual and urinary function after total mesorectal excision with autonomic nerve preservation for carcinoma of the rectum. J Am Coll Surg 182:495-502, 1996. [PubMed: 8646349]

247. Mancini R, Cosimelli M, Filippini A, Tedesco M, Pugliese P, Marcellini M, Pietrangeli A, Lepiane P, Mascagni D, Cavaliere R, Di Matteo G. Nerve-sparing surgery in rectal cancer: feasibility and functional results. J Exper Clin Cancer Res 2000;19:35-40. [PubMed: 10840933]

248. Goligher JC, Graham NG, DeDombal FT. Anastomotic dehiscence after anterior resection of rectum and sigmoid. Br J Surg 1970;57:109. [PubMed: 5467147]

249. Chassin JL. Operative Strategy in General Surgery. New York: Springer-Verlag 1980.

250. Hines JR, Harris GD. Colostomy and colostomy closure. Surg Clin North Am 1977;57:1379. [PubMed: 74099]

251. Varnell J, Pemberton LB. Risk factors in colostomy closure. Surgery 1981;89:683. [PubMed: 7245028]

252. Stothert JC, Brubacher L, Simonowitz DA. Complications of emergency stoma formation. Arch Surg 1982;117:307. [PubMed: 7065872]

253. Kasperk R, Klinge U, Schumpelick V. The repair of large parastomal hernias using a midline approach and a prosthetic mesh in the sublay position. Am J Surg 2000;179:186-188. [PubMed: 10827315]

254. Nakai S, Yoshizawa H, Kobayashi S, Maeda K, Okumura Y. Anorectal and bladder function after sacrifice of the sacral nerves. Spine 2000;25:2234-2239. [PubMed: 10973408]

255. Porter GA, O’Keefe GE, Yakimets WW. Inadvertent perforation of the rectum during abdominoperineal resection. Am J Surg 172:324-327, 1996. [PubMed: 8873522]

256. Gedebou TM, Wong RA, Rappaport WD, Jaffe P, Kahsai D, Hunter GC. Clinical presentation and management of iatrogenic colon perforations. Am J Surg 172:454-458, 1996. [PubMed: 8942543]

257. Church JM, Raudkivi PJ, Hill GL. The surgical anatomy of the rectum—a review with particular relevance to the hazards of rectal mobilisation. [Review]. Int J Colorectal Dis 1987;2:158-166. [PubMed: 3309101]

258. Knoefel WT, Hosch SB, Hoyer B, Izbicki JR. The initial approach to anorectal abscesses: fistulotomy is safe and reduces the chance of recurrences. Dig Surg 2000;17:274-278. [PubMed: 10867462]

259. Liberman H, Thorson AG. Anal stenosis. Am J Surg 2000;179: 325-329. [PubMed: 10875995]

260. Timaran CH, Sangwan YP, Solla JA. Adenocarcinoma in a hemorrhoidectomy specimen: case report and review of the literature. Am Surg 2000;66:789-792. [PubMed: 10966042]

261. Bissett IP, Hill GL. Extrafascial excision of the rectum for cancer: a technique for the avoidance of the complications of rectal mobilization. Semin Surg Oncol 2000;18:207-215. [PubMed: 10757886]

262. Ho YH, Tsang C, Tang CL, Nyam D, Eu KW, Seow-Choen F. Anal sphincter injuries from stapling instruments introduced transanally: randomized, controlled study with endoanal ultrasound and anorectal manometry. Dis Colon Rectum 2000;43: 169-173. [PubMed: 10696889]

263. Shafik A. Right colon flexure syndrome. Report of two cases. Coloproctology 1981;3:105-106.

264. Shafik A. Functional activity of the sigmoid colon and rectum: a study during fecal storage in the sigmoid. Coloproctology 1997; 19:236-241.

265. Shafik A. The hypertonic rectosigmoid junction: description of a new clinicopathologic entity causing constipation. Surg Laparosc Endosc 1997;7:116-120. [PubMed: 9109239]

266. Shafik A. Study of the electrical and mechanical activity of the rectum: an experimental study. Eur Surg Res 1994;26:87-93. [PubMed: 8005171]

267. Shafik A. Study of the electromechanical activity of the rectum. II. Human study. Coloproctology 1993;15:215-221.

268. Shafik A, Nour A, Abdel-Fattah A. Transcutaneous electrorectography. Human electrorectogram from surface electrodes. Digestion 1995;56:479-482. [PubMed: 8536817]

269. Shafik A. Rectosigmoid pacemaker: role in defecation mechanism and constipation. Dig Surg 1993;10:95-100.

270. Shafik A. Rectosigmoid pacemaker: role in defecation mechanism and constipation. (abstract) Dis Colon Rectum 1992;35:29-30.

271. Shafik A. Artificial pacemaker for rectal evacuation. Coloproctology 1992;14:96-98.

272. Shafik A, Moneim KA. Dynamic study of the rectal detrusor activity at defecation. Digestion 1991;49:167-174. [PubMed: 1769432]

273. Shafik A, El-Sibai O, Moustafa R, Shafik I. Study of the mechanism of rectal motility. The “mass squeeze contraction.” Arch Physiol Biochem 2002;109:418-423.

274. Shafik A. Artificial pacemaker for rectal evacuation. II. Role in patients with chronic idiopathic constipation. Proc 3rd Int Mtg Coloproct, Ivrea/Italy 1994, pp. 31-34.

275. Shafik A. Electrosigmoidogram, electrorectogram and their relation. Front Biosci 2 (1997), b12-16; PubMed No. 9294095.

276. Guyton AC, Hall JE. The gastrointestinal tract: nervous control, movement of food through the tract and blood flow. In: Guyton AC, Hall JE. (eds), Human Physiology and Mechanisms of Disease (6th ed). WB Saunders, Philadelphia 1997, pp. 511-523.

277. Shafik A. “Tube” and “marsupialization” colostomy: a simplified technique for colostomy. Am J Proct 1982;33:12-16.

278. Vesalius (1555). Quoted by Gabriel WB. The Principles and Practice of Rectal Surgery (5th ed). London: HK Lewis and Company, 1963, p. 15.

279. Last RJ: Anatomy, Regional and Applied. 5th Edition, reprint. Edinburgh & London: The English Language Book Society and Churchill Livingstone, 1973, p. 531.

280. Bill AH, Jr, Johnson RJ: Failure of migration of the rectal opening as a cause for most cases of imperforate anus. Surg Gynecol Obstet 1958;106:643. [PubMed: 13556488]

281. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. V. The rectal neck: Anatomy and function. Chir Gastroenterol 1977;11:319-336.

282. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. VIII. Levator hiatus and tunnel. Anatomy and function. Dis Colon Rectum 1979;22:539-549. [PubMed: 527446]

283. Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XIX. The malformed rectum. A new theory of pathogenesis with simplified classification. Am J Proct Gastroent Col Rect Surg 1983;34:16-20.

284. Harding Rains AJ, David Ritchie H. In: Bailey and Love’s Short Practice of Surgery. The esophagus (17th ed). The English Language Book Society & H.K. Lewis & Co. Ltd., 1979, p. 759.

285. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. X. Anorectal sinus and band: anatomical nature and surgical significance. Dis Colon Rectum 1980;23:170-179. [PubMed: 7379670]

286. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. VII. Anal fistula: A simplified classification. Dis Colon Rectum 1979;22:408-414. [PubMed: 498899]

287. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XXVI. Fistula-in-ano. A new theory of pathogenesis. Coloproctology 1988;10:148-156.

288. Shafik A, Abdel-Wahab E-S, El-Sibai O, Khalil A: Anorectal fistulae: Results of treatment with cauterization. Dig Surg 1994; 11:16-19.

289. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XV. Chronic anal fissure. A new theory of pathogenesis. Am J Surg 1982;144:262-268. [PubMed: 7102937]

290. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XXII. The pathogenesis of hemorrhoids and their treatment by anorectal bandotomy. J Clin Gastroent 1984;6:129-137. [PubMed: 6715850]

291. Shafik A. Urethral discharge, constipation and hemorrhoids. New syndrome with report of 7 cases. Urology 1981;18:155-160. [PubMed: 7269018]

292. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XXVII. Treatment of hemorrhoids. Report of a technique. Am J Surg 1984;148:393-398. [PubMed: 6476232]

293. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XVI. Pruritus ani. A new theory of pathogenesis. Coloproctology 1981;3:239-243.

294. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XXIII. An injection technique for the treatment of idiopathic pruritus ani. Int Surg 1990;75:43-46. [PubMed: 2318572]

295. Goodsall DH, Miles WE: Diseases of the Anus and Rectum. London: Longmans, Green & Co. Part I, 1900-1905.

296. Allingham W, Allingham HW: The Diagnosis and Treatment of Diseases of the Rectum; Being a Practical Treatise on Fistula, Piles, Fissure and Painful Ulcer, Procidentia, Polypus, Stricture, Cancer, etc. London: Bailliere, Tindall and Cox, 1901.

297. Tuttle JP: A Treatise on Diseases of the Anus, Rectum and Pelvic Colon. New York: Appleton, 1902.

298. Lockhart-Mummery JP: Diseases of the Rectum and Colon and Their Surgical Treatment (2nd ed). Baltimore: William Wood, 1934.

299. Goligher JC: Surgery of the Anus, Rectum and Colon (3rd ed). London: Baillière Tindall, 1975, pp. 15, 17, 225.

300. Milligan ET, Morgan CN: Surgical anatomy of the anal canal with special reference to anorectal fistulae. Lancet 1934;2:1150, 1213.

301. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. II. Anatomy of the levator ani muscle with special reference to puborectalis. Invest Urol 1975;13:175-182. [PubMed: 1193814]

302. Garry RC: The responses to stimulation of the caudal end of the large bowel in the cat. J Physiol 1933;78:208.

303. Bennett RC: Goligher JC. Results of internal sphincterotomy for anal fissure. Br Med J 1962;2:1500.

304. Parks AG, McPartlin JF: Late repair of injuries of the anal sphincter. Proc R Soc Med 1971;64:1187. [PubMed: 5131253]

305. Mann A: Gracilis anoplasty: report of a successful case. Aust NZ J Surg 1970;39:405. [PubMed: 5269352]

306. Caldwell KP: A new treatment of rectal prolapse. Proc R Soc Med 1965;58:792. [PubMed: 5836734]

307. Shafik A, Gamal El-Din MA, El-Bagoury EM, Abdel Hamid Z, El-Said B, Metwalli S, Olfat E: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. The involuntary action of the external anal sphincter. Histologic study. Acta Anat 1990;138:359-363. [PubMed: 1699384]

308. Shafik A, Gamal El-Din MA, El-Sibaei O, Abdel Hamid Z, El-Said B: Involuntary action of the external anal sphincter. Manometric and electromyographic studies. Eur Surg Res 1992;24: 188-196. [PubMed: 1499607]

309. Shafik A: Polytetrafluoroethylene injection for the treatment of partial fecal incontinence. Int Surg 1993;78:159-161. [PubMed: 8354617]

310. Shafik A: Perianal injection of autologous fat for treatment of sphincteric incontinence. Dis Colon Rectum 1995;38:583-587. [PubMed: 7774467]

311. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. 17. Mechanism of defecation. Coloproctology 1982;4:49-54.

312. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. 42. Recto-puborectalis reflex. Coloproctology 1990;12:170-172.

313. Shafik A. Recto-levator reflex. The description of a new reflex and its clinical application. Preliminary report. Clin Physiol Biochem 1993;10:13-17. [PubMed: 8339517]

314. Shafik A: Levator-puborectalis reflex. The description of a new reflex and its clinical significance. Pract Gastroenterol 1991;15: 28-35.

315. Shafik A: Straining-levator reflex: the description of a new reflex and its clinical significance. Coloproctology 1991;13:314-319.

316. Denny-Brown D, Roberston EG. An investigation of the nervous control of defecation. Brain 1935;58: 256-310.

317. Schuster MM, Hookman P, Hendrix TR, Mendeloff AI. Simultaneous manometric recording of internal and external anal sphincter reflexes. Bull Johns Hopkins Hosp 1965;116:79-88. [PubMed: 14261147]

318. Arhan P, Faverdin C, Persoz B, Devroede G, Dubois F, Dornic C, Pellerin D. Relationship between viscoelastic properties of the rectum and anal pressure in man. J Appl Physiol 1976;41:677-682. [PubMed: 993154]

319. Gowers WR. The autonomic action of the sphincter ani. Proc R Soc (London) 1877;26:77.

320. Shafik A. Deflation reflex. Description and clinical significance. Anat Rec 1997; 29: 405-408.

321. Shafik A. A study on the origin of the external anal, urethral, vaginal and prostatic sphincters. Int Urogynecol J 1997;8:126-129. [PubMed: 9449582]

322. Shafik A. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation: Mass contraction of the pelvic floor muscles. Int Urogynecol J 1998;9:28-32. [PubMed: 9657175]

323. Shafik A. Pelvic double-sphincter control complex: Theory of pelvic organ continence with clinical application. Urology 1984; 23:611-618.

324. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XII. Anorectal mobilization. A new surgical access to rectal lesions. Preliminary report. Am J Surg 1981;142:625-635.

325. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XIII. Rectal prolapse. A concept of pathogenesis. Am J Proct Gastroent Col Rect Surg 1981;32:6. [PubMed: 7304730]

326. Shafik A: Straining-puborectalis reflex: description and significance of a new reflex. Anat Rec 1991;229:281-284. [PubMed: 2012316]

327. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. Reversion to normal defecation after combined operation and end colostomy for rectal cancer. Am J Surg 1986;15:278-284.

328. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. 31. Strainodynia: an etiopathologic study. J Clin Gastroenterol 1988;10:179-184. [PubMed: 3418076]

329. Preston DM, Lennard-Jones JE: Anismus in chronic constipation. Dig Dis Sci 1985;30:413-418. [PubMed: 3987474]

330. Shafik A: Detrusor-sphincter dyssynergia syndrome. A new syndrome and its treatment by external sphincter myotomy. Eur Surg Res 1990;22:243-248. [PubMed: 2265659]

331. Turnbull GK, Lennard-Jones JE, Batram CL: Failure of rectal expulsion as a cause of constipation: why fiber and laxative sometimes fail. Lancet 1986;1:767-769. [PubMed: 2870270]

332. Shafik A: Detrusor-rectal neck dyssynergia syndrome: a new syndrome with report of 9 cases. Int Surg 1991;76:241-244. [PubMed: 1778722]

333. Shafik A: A new concept of the anatomy of the anal sphincter mechanism and the physiology of defecation. XVIII. The levator dysfunction syndrome: a new syndrome with report of 7 cases. Coloproctology 1983;5:159-165.

334. Shafik A: Detrusor-levator dyssynergia syndrome: a new syndrome with report of 8 cases. Coloproctology 1990;12:369-373.

335. Shafik A: Levator paradoxical syndrome. A new syndrome. Proc 3rd Int Mtg Coloproctology (Ivrea, Italy) 1994, pp. 48-54.

336. Shafik A: Rectal detrusor retropulsion syndrome: report of 5 cases. Proc 2nd Int Mtg Coloproctology (Ivrea, Italy) 1992, pp. 69-77.

337. Shafik A: Idiopathic oligofecorrhea: a clinicopathologic entity. Pathogenesis and treatment. Digestion 1991;48:51-58. [PubMed: 1868968]

338. Shafik A, Mohi-el-Din M. A new concept of the anatomy of the anal sphincter mechanism and the physiology of defaecation. XXIV. Haemorrhoidal venous plexuses: anatomy and role in haemorrhoids. Coloproctology 1985;7:291-296.

339. Thomson WAF: The nature of haemorrhoids. Br J Surg 1975; 62:542-552.

340. Stelzner F: Haemorrhoids and other diseases of the corpus cavernosum recti and the anal canal. Ger Med Mon 1963;8:177-182.

341. Goligher JC: Surgery of the Anus, Rectum and Colon. Baillière Tindall, London (5th ed). 1984, pp. 98-149.

342. Goldberg SM, Gordon PH, Nivatvongs S: Essentials of Anorectal Surgery. Philadelphia: Lippincott, 1980, pp. 69-85.

343. Shafik A: Anal cystography. New technique of cystography. Preliminary report. Urology 1984;23:313-316. [PubMed: 6702049]

344. Shafik A, Mohi M: Pelvic organ venous communications. Anatomy and role in urogenital diseases. A new technique of cysto-vagino-hysterography. Am J Obstet Gynecol 1988;159:347-351. [PubMed: 3407691]

345. Shafik A: Role of hemorrhoids in the pathogenesis of recurrent bacteriuria with a new approach for treatment. Eur Urol 1985;11: 392-396. [PubMed: 3910441]

346. Shafik A, Saleh F, Twefik O, Abdel-Azim S, Saad R, Olfat E, Sharkawi A: The role of hemorrhoids in urinary tract infection: A clinicobacteriologic study. Pract Gastroenterol 1989;13:21-29.

347. Shafik A, Haddad S, Elwan F, El-Metnawi W, Olfat E: Anal submucosal injection: A new route for drug administration in pelvic malignancies. II. Methotrexate anal injection in the treatment of advanced bladder cancer (preliminary study). J Urol 1988;140:501. [PubMed: 3045340]

348. Shafik A: Perineal versus anal submucosal injection of chemotherapeutic drugs. Eur Urol 1992;21:256. (Letter to the Editor).

349. Shafik A, El-Merzabani MM, El-Aaser AA, et al: Anal submucosal injection. A new route for drug administration in pelvic malignancies. I. Experimental study of misonidazole distribution in serum and tissues, with special reference to urinary bladder. Invest Radiol 1986;21:278-281. [PubMed: 3957602]

350. Shafik A, El-Desouky G: Anal submucosal injection: A new route for drug administration in pelvic malignancies. III. Misonidazole distribution in serum, uterus and vagina: An experimental study. Gynecol Obstet Invest 1990;29:219-223. [PubMed: 2358197]

351. Shafik A: Anal submucosal injection: A new route for drug administration in pelvic malignancies. V. Advanced prostatic cancer: results of methotrexate treatment using the anal route. Preliminary study. Eur Urol 1990;18:132-136. [PubMed: 2152339]

352. Shafik A: Anal submucosal injection: A new route for drug administration in pelvic malignancies. IV. Submucosal anal injection in treatment of cancer of uterine cervix. Preliminary study. Am J Obstet Gynecol 1989;161: 69-72. [PubMed: 2750823]

353. Shafik A, El-Metnawi W, El-Sibai O. Treatment of advanced rectal cancer by anal submucosal injection. Eur J Surg Oncol 1999;25:76-81.

354. Shafik A, El-Dawi M, El-Metnawy W. Anal submucosal injection: methotrexate concentration in rectal tumor tissue and serum after anal compared with parenteral injection. Anti-Cancer Drugs 1994;5:650-654. [PubMed: 7888703]

355. Shafik A: Anal submucosal injection: A new route for drug administration. VI. Chronic prostatitis: A new modality of treatment with report of eleven cases. Urology 1991;37:61. [PubMed: 1986477]

356. Shafik A, Moustafa H. Study of the arterial pattern of the rectum and its clinical application. Acta Anat 1996;157:80-86. [PubMed: 9096745]

357. Ripstein CB. Procidentia: definitive corrective surgery. Dis Colon Rectum 1972;15:334-336. [PubMed: 5072253]

358. Heald RJ. Ryall RDH. Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1986;i:1479-1482.

359. Guyton AC. Movement of food through the alimentary tract. In: Guyton AC (ed). Human Physiology and Mechanics of Disease (4th ed). Philadelphia: WB Saunders, 1987, pp. 486-496.

360. Connell AM. The motility of the pelvic colon. I. Motility in normals and in patients with asymptomatic duodenal ulcer. Gut 1961;2:175-186. [PubMed: 13695051]

361. Truelove SC. Movements of the large intestine. Physiol Rev 1966; 46:457-512.

362. Edwards DA, Beck ER. Movement of radiopacified feces during defecation. Am J Dig Dis 1971;16:709-711. [PubMed: 5096786]

363. Shafik A, Ali YA, Afifi R. Is the rectum a conduit or storage organ? Int Surg 1997;82:194-197. [PubMed: 9331852]

364. Kiff ES, Swash M: Slow conduction in pudendal nerves in idiopathic fecal (neurogenic) incontinence. Br J Surg 1984;71: 614-616. [PubMed: 6743983]

365. Neil ME, Parks AG, Swash M: Physiological studies of the pelvic floor in idiopathic fecal incontinence and rectal prolapse. Br J Surg 1981;68:531-536.

366. Jones PN, Lubowski DZ, Swash M, Henry MM: Relationship between perineal descent and pudendal nerve incontinence. Int J Colorect Dis 1987;9:3-7.

367. Womack NR, Morrison JFB, Williams NS: The role of pelvic floor denervation in etiology of idiopathic fecal incontinence. Br J Surg 1986;73:404-408. [PubMed: 3708299]

368. Shafik A: Pudendal canal syndrome. Description of a new syndrome and its treatment: Report of 7 cases. Coloproctology 1991; 13:102-110.

369. Shafik A: Pudendal canal decompression in the treatment of idiopathic fecal incontinence. Dig Surg 1992;9:265-271.

370. Shafik A: Chronic scrotalgia: Report of four cases with successful treatment. Pain Digest 1993;3:252-256.

371. Thuroff JW, Bazeed MA, Schmidt RA, Wiggin DM, Tanagho EA: Functional pattern of sacral root stimulation in dogs. I. Micturition. J Urol 1982;127:1031-1033.

372. Thuroff JW, Bazeed MA, Schmidt RA, Wiggin DM, Tanagho EA: Functional pattern of sacral root stimulation in dogs. II. Urethral closure. J Urol 1982;127:1034-1038. [PubMed: 7086981]

373. Hohenfellner M, Paick JS, Trigo Rocha F, Schmidt RA, Kaula NF, Thuroff JW, Tanagho EA. Site of deafferentation and electrode placement for bladder stimulation: Clinical implication. J Urol 1992;147:1665-1669. [PubMed: 1593717]

374. Shafik A: Pudendal nerve stimulation for anal and urethral sphincter control. Experimental study. Eur J Gastroenterol Hepatol 1994;6:345-349.

375. Shafik A: Perineal nerve stimulation for urinary sphincter control. Experimental study. Urol Res 1994;22:151-155. [PubMed: 7992459]

376. Shafik A: Sacral root stimulation for controlled defecation. Eur Surg Res 1995;27:63-68. [PubMed: 7890007]

377. Shafik A: Pudendal canal decompression in the treatment of erectile dysfunction. Arch Androl 1994;32:141-149. [PubMed: 8166577]

378. Shafik A. The posterior approach in the treatment of pudendal canal syndrome. Coloproctology 1992;14:310-315.

379. Shafik A. Role of pudendal canal syndrome in the etiology of fecal incontinence in rectal prolapse. Digestion 1997;58:489-493. [PubMed: 9383642]

380. Shafik A. Pudendal artery syndrome presenting as ischemic proctitis. Report of 3 cases. Dig Surg 1996;13:53-58.

381. Shafik A. Pudendal canal decompression for the treatment of fecal incontinence in complete rectal prolapse. Am Surg 1996; 62:339-343.

382. Shafik A. Pudendal canal syndrome: a new etiological factor in prostatodynia and its treatment by pudendal canal decompression. Pain Dig 1998;8:32-36.

383. Shafik A. Pudendal canal syndrome as a cause of vulvodynia and its treatment by pudendal nerve decompression. Eur J Obstet Gynecol Reprod Biol 1998;80:215-220. [PubMed: 9846672]

384. Shafik A, El-Sherif M, Youssef A, El-Sibai O: Surgical anatomy of the pudendal nerve and its clinical implications. Clin Anat 1995;8:110-115. [PubMed: 7712320]

385. Juenemann KP, Lue TF, Schmidt RA, Tanagho EA: Clinical significance of sacral and pudendal nerve anatomy. J Urol 1988; 139:74-77.

386. Schmidt RA: Technique of pudendal nerve localization for block or stimulation. J Urol 1989;142:1528-1531. [PubMed: 2585630]

387. Shafik A, Doss S. Pudendal canal: surgical anatomy and clinical implication. Am Surg 1999;65:176-180. [PubMed: 9926755]

388. Shafik A. The levator dysfunction syndrome. A new syndrome with report of seven cases. Coloproctology 1983;5:159-165.

389. Shafik A. Pudendal canal decompression in the treatment of idiopathic fecal incontinence. Dis Colon Rectum 1993;36:17 (abstract).

390. Shafik A. Stress urinary incontinence: an alternative concept of pathogenesis. Int Urogynecol J 1994;5:3-11.

391. Shafik A. Pudendal canal decompression in the treatment of urinary stress incontinence. Int Urogynecol J 1994;5:215-220.

392. Lestar B, Penninckx FM, Kerremans RP: Defecometry: A new method for determining the parameters of rectal evacuation. Dis Colon Rectum 1989;32:197-201. [PubMed: 2920625]

393. Barnes PR, Lennard-Jones JE: Patients with constipation of different types have difficulty in expelling a balloon from the rectum. Gut 1984;25:562-563.

394. Shafik A, Khalid A: Fecoflowmetry in defecation disorders. Pract Gastroenterol 1990;14:46-52.

395. Shafik A, Abdel-Moneim K: Fecoflowmetry: A new parameter assessing rectal function in normal and constipated subjects. Dis Colon Rectum 1993;36:35-42. [PubMed: 8416777]

396. Shafik A: Levator-sphincter reflex. Description of a new reflex and its clinical significance. Coloproctology 1992;14:172-175.

397. Shafik A. Water enema test: a means of assessing rectal function. Pract Gastroenterol 1992;16:24J-24P.

398. Shafik A: Inguinal pelviscopy. A new approach for examining the pelvic organs. Gynecol Obstet Invest 1990;30:159-161. [PubMed: 2265801]

399. Shafik A, Moneim KA. Rectometry: A new method assessing rectal function. Coloproctology 1991;13:237-243.

400. Shafik A: Electrorectography in chronic constipation. World J Surg 1995;19:772-775. [PubMed: 7571680]

401. Shafik A: Electrorectogram in chronic proctitis. World J Surg 1993;17:675-679. [PubMed: 8273392]

402. Shafik A: The electrorectogram in Hirschsprung’s disease. A new diagnostic tool. Preliminary report. Pediatr Surg Int 1995;10: 478-480.

403. Shafik A, Abdel-Fattah A: Transcutaneous electrovesicography. Urologia 1995;62:371-374.

404. Shafik A: Electrovasography in normal and vasectomized men before and after vasectomy reversal. Int J Androl 1996;19:33-38. [PubMed: 8698536]

405. Récamier M. Quoted by Maisonneure JG. Du traitment de la fissure a l’anus par la dilatation forcée. Gaz d’Hop (3rd series) 1849; 1:220. [PubMed: 16456907]

406. Lord PH. Conservative management of hemorrhoids. Part II: Dilatation treatment. Clin Gastroenterol 1975;4:601-606.

407. Buls JG, Goldberg SM. Modern management of hemorrhoids. Surg Clin North Am 1978;58:469-479. [PubMed: 675462]

408. Shafik A. Recurrent anal fistula: is it a true recurrence or “de novo” fistula? Coloproctology 1995;17:249-253.

409. Marks CG, Ritchie JK. Anal fistulas at St Mark’s Hospital. Br J Surg 1977;64:84-87. [PubMed: 890252]

410. Shafik A. Complete rectal prolapse: a technique for repair. Coloproctology 1987;9:345-352.

Copyright ©2006 The McGraw-Hill Companies. All rights reserved.
Privacy Notice. Any use is subject to the Terms of Use and Notice. Additional Credits and Copyright Information.

Leave a Reply

Time limit is exhausted. Please reload the CAPTCHA.


apply_now Pepperstone Group Limited