Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You and Your Team.

Learn More →

Proctectomy and Coloanal Anastomosis for Rectal Cancer

Proctectomy and Coloanal Anastomosis for Rectal Cancer Abstract Fueled by a greater understanding of pelvic physiology along with an improved comprehension of rectal cancer spread, we are now able to offer most patients restoration of intestinal continuity following oncologic proctectomy. Coloanal or ultralow colorectal anastomosis can be performed in most patients with midrectal cancers, provided that anal sphincter function is not impaired preoperatively. Functional results may be improved by construction of a colonic pouch with pouch-anal anastomosis. Temporary fecal diversion, usually with a diverting loop ileostomy, may be prudent, especially in patients undergoing neoadjuvant chemoradiation. The overwhelming majority of patients undergoing coloanal anastomosis have carcinoma of the rectum. Until recently, most patients with mid or distal rectal cancers underwent abdominoperineal resection. Technical issues notwithstanding, surgeons were loathe to consider preservation of the distal rectum or anal canal in these patients because of their concern of distal intramural spread of tumor. Careful pathologic studies have shown, however, that distal spread of rectal adenocarcinoma (either in the wall of the rectum or mesorectum) greater than 2 cm is rare, and is associated with advanced lesions and poor prognosis.1-7 On the other hand, direct radial extension of tumor into the mesorectum or pelvic sidewall is of critical importance.8 Cure and the avoidance of local recurrence after treatment of mid and distal rectal cancers depend on the adequacy of lateral pelvic clearance and the removal of all mesorectal tissue in its fascial envelope.9,10 Preservation of anal sphincter function is possible following proctectomy for mid and distal cancers, provided that the distal margin of the tumor is several centimeters above the anorectal muscular ring. Precise definition of coloanal anastomosis is lacking. Removal of the rectum to the level of the levators with anastomosis in, or just above, the anal canal is considered a coloanal anastomosis by some surgeons. Others have a more rigorous definition, requiring anastomosis at the dentate line, and refer to the former as an ultralow colorectal anastomosis. Technique of dissection and anastomosis also varies. The dissection can be performed from the abdomen with distal division of the rectum and anastomosis performed using surgical staplers. Alternatively, some surgeons advocate transanal dissection to define the distal margin for low tumors, with a hand-sewn anastomosis. Regardless of technique, the concept of removal of the rectum and mesorectum with restoration of intestinal continuity in the region of the anal sphincter complex remains constant. Construction of a neorectum using a colonic pouch has received considerable attention recently. Stool frequency and urgency are reduced as compared with straight coloanal anastomosis, probably because of increased reservoir capacity.11-18 The advantage of the pouch-anal anastomosis over a straight coloanal anastomosis seems to be most pronounced in the first 1 to 2 years following construction. The typical colonic pouch is constructed in a "J" fashion using a 6- to 10-cm limb length. Larger pouches may impair emptying, although the functional differences emphasized in some studies of pouch limb length may be negated by changes in pouch volume over time.19,20 Many surgeons perform a temporary diverting stoma to minimize the consequences of anastomotic disruption should it occur. The low level of the anastomosis and the increased length of suture line at risk when a colonic pouch is constructed increase the risk of leak and pelvic sepsis. Anastomotic leak may cause pelvic fibrosis, rendering the neorectum stiff and noncompliant. Patients may thus suffer from tenesmus and fecal incontinence. The most commonly used diverting stoma is the loop ileostomy, because its construction and subsequent takedown are straightforward, and the colonic mesentery is not threatened by such manipulation. Temporary fecal diversion may also be prudent in cases where neoadjuvant chemoradiation is used. Although dramatic reductions in tumor size can be achieved with the use of preoperative chemoradiation, it may predispose to anastomotic problems in patients not undergoing temporary diversion at the time of proctectomy and coloanal anastomosis.21 Patient evaluation Patients with rectal cancer undergo evaluation with digital rectal examination, rigid proctoscopy, and transrectal ultrasound. Invasion of the anorectal muscular ring precludes restoration of intestinal continuity. Surgeons should not be seduced into triumphs of technique over judgment in cases where the tumor is very low in the rectum. It is not advisable to jeopardize oncologic cure because of enthusiasm to perform low anastomosis. As Heister remarked in 1739, speaking of intestinal stomas: " . . . it is surely better to part with one of the conveniences of life than to part with life itself."22 Colonoscopy is performed to search for synchronous neoplasms. Abdominopelvic computed tomography and a chest radiograph are obtained to rule out distal metastases. Transrectal ultrasound and pelvic computed tomography can detect extensive mesorectal lymph node involvement and help guide decisions regarding neoadjuvant treatment. Patients undergoing neoadjuvant external beam radiation with or without chemotherapy are reexamined by the surgeon prior to operation to assess tumor response to therapy. Before considering coloanal anastomosis, sphincter function must be assessed. A detailed continence history and physical examination by an experienced surgeon is probably the most predictive of postoperative function.23 However, anal manometry may help in equivocal cases. Despite the enthusiasm of many patients to restore intestinal continuity at all costs, it should be remembered that an abdominal colostomy is preferable to a perineal colostomy in cases where sphincter function is impaired. Operative preparation Consent for operation should include the possibility of temporary or permanent stoma construction. Stoma sites should be marked preoperatively by the surgeon or enterostomal therapist with the patient in the sitting, standing, and supine positions. A mechanical and antibiotic bowel preparation is used. Intravenous broad-spectrum antibiotics are given perioperatively for 24 hours. Deep venous thrombosis prophylaxis is indicated. Our policy has been to use heparin, 5000 U subcutaneously every 12 hours, and intermittent pneumatic compression stockings. Ureteral catheters should be considered in cases where invasion or impingement on the ureters or bladders is suspected. Although some authors have found that ureteral catheters are not useful in preventing injury, we have found them extremely helpful in facilitating difficult pelvic dissection. The patient is placed in the lithotomy position and a Foley catheter is inserted. Digital rectal examination and proctoscopy with the patient under anesthesia can be helpful in determining tumor level and degree of fixation. We routinely employ rectal irrigation with saline and dilute povidone-iodine prior to operation, and leave a rectal tube in place. Exploration A midline incision is made and the abdomen explored. The presence of unresectable metastases makes coloanal anastomosis ill-advised. If a temporary diverting stoma is constructed, the patient may never be fit enough to have the stoma closed. Even if the patient does not require a diverting stoma, the initial 6 to 18 months following coloanal anastomosis are often marked by frequent loose stools and occasional incontinence. The patient may not survive long enough to benefit from the gradual improvement in function of the neorectum. If the patient is treated with chemotherapy postoperatively, the risk of diarrhea and fecal incontinence are increased further. Low anterior resection with Hartmann closure of the rectal stump is a more prudent option for these patients. Tumor mobility in the pelvis is then assessed. Apparent fixation to pelvic structures may not preclude cure, as inflammatory or fibrotic reaction may be responsible, especially in cases where preoperative radiation or chemoradiation has been used. However, if during the course of pelvic dissection it becomes obvious that tumor invasion of pelvic structures makes complete resection impossible, low anastomosis should be abandoned because of the high risk of persistent tumor growth in the pelvis. Mobilization The left colon is mobilized from its retroperitoneal attachments (Figure 1, A and B). The bloodless plane lateral to the descending colon mesentery is developed first, initially avoiding the pelvic brim, where the ureter and iliac vessels are most prone to injury. The small bowel should be protected during this mobilization because of the risk of penetrating through the thin avascular area at the base of the descending colon mesentery. We routinely accomplish this by carefully placing a large laparotomy pad over the small bowel and retracting it medially. After the descending colon has been mobilized and the ureter identified, the sigmoid is mobilized from its lateral attachments. Several maneuvers are then performed to allow the descending colon to reach into the deep pelvis. We use the descending colon to perform all pelvic anastomoses because the sigmoid colon is narrow and has muscular hypertrophy in most patients. In addition, if preoperative external beam radiation has been used, the sigmoid may have been included in the radiation field and should not be used for the anastomosis. The splenic flexure is completely mobilized. The omentum is separated from the distal half of the transverse colon. An incision is made at the base of the sigmoid and descending colon mesentery on the right from the sacral promentory to the duodenum. It is frequently necessary to divide small filmy attachments of the distal duodenum and proximal jejeunum to the colonic mesentery to completely expose the origin of the inferior mesenteric artery (IMA) and the inferior mesenteric vein (IMV). The IMA is divided at its origin (Figure 1, C). Care should be taken to avoid the lumbar sympathetic plexus during IMA division. The IMV is divided adjacent to the ligament of Treitz (Figure 1, D). The blood supply to the proximal bowel is thus dependent on arcades originating from the middle colic artery. Dissection then proceeds from the IMA origin toward the pelvis. Keeping the hypogastric nerves protected posteriorly, the plane just posterior to the superior hemorrhoidal artery is opened. This plane is used to gain entrance to the presacral space at the sacral promontory, and can be followed in the posterior midline to the anal canal without encountering a blood vessel. Once the splenic flexure, descending colon, and sigmoid have been completely mobilized and the IMA and IMV divided, the proximal colon is divided. Although this step can be deferred until after pelvic dissection, division at this point in the operation allows the surgeon to retract on the rectosigmoid without danger of causing injury to the descending colon. The descending colon mesentery is divided between clamps beginning just proximal to the IMA. The line of mesenteric dissection extends perpendicular to the bowel to avoid jeopardizing proximal blood supply. We routinely divide the colon in an area of soft descending colon, leaving the sigmoid with the specimen. By performing the aforementioned mobilization, the descending colon should reach to the level of the anus. If additional length is needed, the omentum can be completely removed from the transverse colon and congenital attachments between the transverse mesocolon and retroperitoneum divided. It is rarely necessary to divide any of the arcade vessels originating from the middle colic artery, and great care should be exercised to ensure that pulsatile flow remains to the distal bowel. If a straight double-stapled anastomosis is planned, the bowel can be divided between clamps. A purse-string suture is placed and the proximal bowel sized to accommodate the anvil of the end-to-end stapler. If a colonic pouch or hand-sewn anastomosis is planned, the proximal bowel can be transected using a linear cutting stapler. Pelvic dissection then proceeds (Figure 2). The goal is to remove the mesorectum intact within its fascial envelope and to achieve adequate radial and distal margins from the tumor. Sharp dissection under direct vision is required. Blunt tearing maneuvers should be avoided. A strong headlight and lighted retractors are invaluable in this phase of the operation. The peritoneum on either side of the mesorectum is divided. The line of incision should be just medial to the ureter on either side, and extend anteriorly to the seminal vesicles or rectovaginal septum. Posteriorly, the areolar tissue between the mesorectum and the presacral fascia is divided. Sharp dissection with electrocautery or scissors under direct vision can be accomplished with anterior retraction of the rectosigmoid. It is often helpful to use a curved malleable blade to retract the mesorectum as the dissection proceeds to the level of the coccyx. The risk of injury to the basivertebral veins on the anterior surface of the sacrum is minimized if the urge to perform this portion of the dissection bluntly with a hand behind the mesorectum is suppressed. Anterior dissection is facilitated with a deep pelvic retractor placed between the anterior rectal wall and the posterior wall of the bladder. An assistant standing between the patient's legs places anterior retraction of these tissues. In a man, the plane posterior to the seminal vesicles is opened, the fascia of Denonvilliers incised, and the anterior rectal wall separated from the posterior capsule of the prostate. Palpation of the posterior surface of the pubis or palpation of the Foley catheter in the urethra distal to the prostate indicate that the distal aspect of the dissection is adequate. In a woman, the rectovaginal septum is separated to the level of the anal canal. The lateral stalks are then divided. Most of this tissue can be divided using the electrocautery, working from the bloodless plane opened posterior to the mesorectum and proceeding in an anterior direction. In some cases it is prudent to divide the middle hemorrhoidal vessels between clamps and ligate the vessels. In other cases, these vessels can be controlled with cautery alone. Irrespective of the method of division, the lateral stalks should be divided as lateral as possible without jeopardizing the ureters. After completion of this maneuver, the rectum should rise out of the pelvis and the mesorectum should appear as a fatty bilobed structure posterolateral to the rectum encased in a glistening fascial envelope. It is now possible to divide any remaining tissue attachments to the distal rectum and proximal anal canal. The levators should be clearly visible, the mesorectum completely mobilized, and the anal canal exposed. In difficult cases where the exact level of distal dissection is uncertain, the surgeon may place a finger in the anal canal with the other hand palpating through the abdomen. The tumor should be well above the level of distal exposure of the muscular tube of anorectum, optimally 5 cm, although 2 cm has found to be adequate. If the tumor is found to be invading the levators or sphincter complex intraoperatively, abdominoperineal resection should be performed. The pelvic autonomic nerves that innervate the genitalia and bladder are subject to injury during mobilization of the rectosigmoid. The preaortic sympathetic plexus can be drawn up into the resected specimen when the IMA is divided. The hypogastric plexus, located at the aortic bifurcation, and the hypogastric nerves, which are the condensation of those sympathetic fibers, are vulnerable to injury during mobilization of the mesorectum from the sacral promontory and lateral pelvic sidewalls. Sympathetic injury will result in retrograde ejaculation, or a "dry orgasm." The pelvic plexus contains sympathetic fibers from the hypogastric nerves as well as parasympathetic fibers from the pelvic splanchnic nerves. This plexus is superior to the levators, anterior and lateral to the rectum, and between the peritoneum and the endopelvic fascia. Injury to the pelvic plexus can occur at the time of division of the lateral stalks or during the dissection of the rectum from the region of the seminal vesicles and prostate. Injury to both of the pelvic plexuses will result in complete impotence in the male. Distal transection of the anorectum Division of the anal canal or very distal rectum can usually be accomplished through the abdomen using a linear stapler (Figure 3). The site of division should be below the tail of mesorectum so that all mesorectal tissue accompanies the specimen. The rectum is retracted superiorly and posteriorly, with a deep pelvic retractor exposing the anterior surface of the rectum. We frequently use a "perineal pusher," an assistant standing between the patient's legs using folded towels enclosed in a sturdy clamp to manually elevate the perineum in a superior direction. After the stapler has been applied, a long occlusive clamp is placed on the rectum just proximal to the stapler and the bowel divided. It is crucial to have the specimen retrieved by the pathologist and have the radial margins inked prior to opening the specimen. Our pathologists now routinely report radial margin distance on all rectal cancer specimens. Some surgeons are enthusiasts of the combined transanal-transabdominal resection procedure, where the distal margin is defined and the distal dissection is performed from a transanal approach. We have found very few indications for this technique. In most patients, the proximal anal canal can be exposed from the abdomen and a stapling instrument applied below the tumor. If the tumor is so low that division in the proximal anal canal would not provide an adequate distal margin, then perhaps an abdominoperineal procedure would be safer from an oncologic standpoint. There may be a few patients who, because of a narrow pelvis or other aspect of body habitus, might benefit from such a technique. Anastomosis Adequate mobilization of the proximal colon and lack of tension are reconfirmed. Blood supply to the proximal margin can be reassessed as the proximal bowel has now been dependent on unidirectional flow from the middle colic artery for the entire period of pelvic dissection. ("If it looks blue, it is blue.") Poor blood supply to the proximal margin must be corrected prior to anastomosis. We routinely use a double-stapled technique to perform coloanal or ultralow colorectal anastomosis, and with increasing frequency we are constructing a colonic J-pouch (Figure 4). The colon is folded into a J-configuration with an approximately 8-cm limb length. The apex of the pouch is opened and a linear cutting stapler is used to create the pouch. Care must be taken to avoid incorporating the colonic mesentery into the staple line. Usually a single fire of a 75- or 90-mm stapler will create adequate pouch length. If there is a long blind end remaining, this can be corrected with a second fire of the linear cutting stapler. A purse-string suture is placed around the opening at the pouch apex and the anvil of the end-to-end stapler inserted. We usually use a 28- or 29-mm stapler in this situation. Retractors are replaced and hemostasis in the pelvis confirmed. The stapler is placed gently in the anus with adequate lubrication. For low anastomosis, there is often not enough remaining length of the anorectum for the stapler head to slip above the sphincter complex and thus give the feeling of being properly positioned. Great care must be exercised in this situation to avoid forcing the stapler and bursting through the distal staple line. The staple head is aligned and the spike of the stapler is advanced through the distal staple line. Although it is tedious at times, we routinely place a purse-string suture in the distal bowel to make it fit snugly around the spike of the stapler. We feel this prevents extrusion of the distal bowel from the anastomosis during closure of the stapler. Carefully retracting the vagina and any other tissue at risk, the anvil is attached to the spike and the stapler closed and fired. After removal of the stapler, the adequacy of the tissue doughnuts are assessed and the anastomosis inspected gently with a proctoscope. The integrity of the pouch and anastomosis are tested by instilling air via the proctoscope with the pelvis filled with warm saline. Bubbles are an indication of a leak and should be investigated and repaired. If there is a small separation at the anastomotic line low in the pelvis, it may be difficult to expose transabdominally and transanal repair with absorbable suture is indicated. Large leaks may require takedown and reconstruction of the anastomosis, although the short length of distal bowel makes this a difficult procedure. Drainage of the pelvis after proctectomy is controversial. The major potential benefit of postoperative drainage is removal of blood, which is a growth medium for bacteria. We typically place 2 soft, Silastic, closed-suction drains in the pelvis after proctectomy with pouch-anal anastomosis. However, if hemostasis is excellent at the conclusion of the procedure, drains may be unnecessary. We do not routinely make an attempt to close the pelvic peritoneum. If the descending colon mesentery is loose, however, we will reapproximate it to the retroperitoneum. We have had a case of small bowel becoming entrapped posterior to the mesentery of the descending colon, necessitating reoperation. In our opinion, the risk of anastomotic or pouch leak with pelvic sepsis and subsequent fibrosis of the neorectum outweighs the benefits of avoiding temporary fecal diversion in our patient population. In addition, most patients undergoing proctectomy and coloanal anastomosis for rectal cancer will have stage II or stage III lesions. We routinely treat these patients with postoperative chemotherapy. As mentioned above, chemotherapy-induced diarrhea in a patient first adapting to a neorectum will cause fecal incontinence. For these reasons, a temporary loop ileostomy is constructed (Figure 5). After chemotherapy is completed, the pouch and anastomosis are assessed endoscopically and radiographically and ileostomy takedown is performed. We thank John A. Craig, MD, for producing the illustrations. Corresponding author: Thomas E. Read, MD, Section of Colon and Rectal Surgery, 216 S Kingshighway (North Campus), St Louis, MO 63110 (e-mail: readt@msnotes.wustl.edu). References 1. Quer EDahin DMayo C Retrograde intramural spread of carcinoma of the rectum and rectosigmoid. Surg Gynecol Obstet. 1953;9624- 30Google Scholar 2. Grinnell R Distal intramural spread of rectal carcinoma. Surg Gynecol Obstet. 1954;99421- 430Google Scholar 3. Black UWaugh J The intramural extension of carcinoma of the descending colon, sigmoid and rectosigmoid: a pathological study. Surg Gynecol Obstet. 1948;1948457- 464Google Scholar 4. Scott NJackson Pal JTDixon MFQuirke PFinan PJ Total mesorectal excision and local recurrence: a study of tumour spread in the mesorectum distal to rectal cancer. Br J Surg. 1995;821031- 1033Google ScholarCrossref 5. Williams NSDixon MFJohnston D Reappraisal of the 5 centimetre rule of distal excision for carcinoma of the rectum: a study of distal intramural spread and of patients' survival. Br J Surg. 1983;70150- 154Google ScholarCrossref 6. Dukes C The surgical pathology of rectal cancer. Proc R Soc Med. 1943;37131Google Scholar 7. Goligher JDukes CBussey H Local recurences after sphincter-saving excisions for carcinoma of the rectum and rectosigmoid. Br J Surg. 1951;39199- 211Google ScholarCrossref 8. Quirke PDurdey PDixon MFWilliams NS Local recurrence of rectal adenocarcinoma due to inadequate surgical resection: histopathological study of lateral tumour spread and surgical excision. Lancet. 1986;2996- 999Google ScholarCrossref 9. Heald RJRyall RD Recurrence and survival after total mesorectal excision for rectal cancer. Lancet. 1986;11479- 1482Google ScholarCrossref 10. Zaheer SPemberton JHFarouk RDozois RRWolff BGIlstrup D Surgical treatment of adenocarcinoma of the rectum. Ann Surg. 1998;227800- 811Google ScholarCrossref 11. Macmanus J Perforations of the intestine by ingested foreign bodies. Am J Surg. 1941;111393- 402Google ScholarCrossref 12. Mortensen NJRamirez JMTakeuchi NHumphreys MM Colonic J-pouch anal anastomosis after rectal excision for carcinoma: functional outcome. Br J Surg. 1995;82611- 613Google ScholarCrossref 13. Lazorthes FChiotasso PGamagami RAIstvan GChevreau P Late clinical outcome in a randomized prospective comparison of colonic J-pouch and straight coloanal anastomosis. Br J Surg. 1997;841449- 1451Google ScholarCrossref 14. Hallbook OSjodahl R Comparison between the colonic J-pouch anal anastomosis and healthy rectum: clinical and physiological function. Br J Surg. 1997;841437- 1441Google ScholarCrossref 15. Seow-Choen FGoh HS Prospective randomized trial comparing J colonic pouch-anal anastomosis and straight coloanal reconstruction. Br J Surg. 1995;82608- 610Google ScholarCrossref 16. Ho YHTan MSeow CF Prospective randomized controlled study of clinical function and anorectal physiology after low anterior resection: comparison of straight and colonic J pouch anastomoses. Br J Surg. 1996;83978- 980Google ScholarCrossref 17. Ortiz HDe MMArmendariz PRodriguez JChocarro C Coloanal anastomosis: are functional results better with a pouch? Dis Colon Rectum. 1995;38375- 377Google ScholarCrossref 18. Nicholls RJLubowski DZDonaldson DR Comparison of colonic reservoir and straight coloanal reconstruction after rectal excision. Br J Surg. 1988;75318- 320Google ScholarCrossref 19. Lazorthes FGamagami RChiotasso PIstvan GMuhammad S Prospective, randomized study comparing clinical results between small and large colonic J-pouch following coloanal anastomosis. Dis Colon Rectum. 1997;401409- 1413Google ScholarCrossref 20. Hida JYasutomi MFujimoto K et al. Functional outcome after low anterior resection with low anastomosis for rectal cancer using the colonic J-pouch: prospective randomized study for determination of optimum pouch size. Dis Colon Rectum. 1996;39986- 991Google ScholarCrossref 21. Hyams DMMamounas EPPetrelli N et al. A clinical trial to evaluate the worth of preoperative multimodality therapy in patients with operable carcinoma of the rectum: a progress report of National Surgical Breast and Bowel Project Protocol R-03. Dis Colon Rectum. 1997;40131- 139Google ScholarCrossref 22. Heister L A General System of Surgery. London, England W Innys1743; 23. Church JMSaad RSchroeder T et al. Predicting the functional result of anastomoses to the anus: the paradox of preoperative anal resting pressure. Dis Colon Rectum. 1993;36895- 900Google ScholarCrossref http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Archives of Surgery American Medical Association

Proctectomy and Coloanal Anastomosis for Rectal Cancer

Archives of Surgery , Volume 134 (6) – Jun 1, 1999

Loading next page...
 
/lp/american-medical-association/proctectomy-and-coloanal-anastomosis-for-rectal-cancer-BRTvjbHW4O
Publisher
American Medical Association
Copyright
Copyright © 1999 American Medical Association. All Rights Reserved.
ISSN
0004-0010
eISSN
1538-3644
DOI
10.1001/archsurg.134.6.670
Publisher site
See Article on Publisher Site

Abstract

Abstract Fueled by a greater understanding of pelvic physiology along with an improved comprehension of rectal cancer spread, we are now able to offer most patients restoration of intestinal continuity following oncologic proctectomy. Coloanal or ultralow colorectal anastomosis can be performed in most patients with midrectal cancers, provided that anal sphincter function is not impaired preoperatively. Functional results may be improved by construction of a colonic pouch with pouch-anal anastomosis. Temporary fecal diversion, usually with a diverting loop ileostomy, may be prudent, especially in patients undergoing neoadjuvant chemoradiation. The overwhelming majority of patients undergoing coloanal anastomosis have carcinoma of the rectum. Until recently, most patients with mid or distal rectal cancers underwent abdominoperineal resection. Technical issues notwithstanding, surgeons were loathe to consider preservation of the distal rectum or anal canal in these patients because of their concern of distal intramural spread of tumor. Careful pathologic studies have shown, however, that distal spread of rectal adenocarcinoma (either in the wall of the rectum or mesorectum) greater than 2 cm is rare, and is associated with advanced lesions and poor prognosis.1-7 On the other hand, direct radial extension of tumor into the mesorectum or pelvic sidewall is of critical importance.8 Cure and the avoidance of local recurrence after treatment of mid and distal rectal cancers depend on the adequacy of lateral pelvic clearance and the removal of all mesorectal tissue in its fascial envelope.9,10 Preservation of anal sphincter function is possible following proctectomy for mid and distal cancers, provided that the distal margin of the tumor is several centimeters above the anorectal muscular ring. Precise definition of coloanal anastomosis is lacking. Removal of the rectum to the level of the levators with anastomosis in, or just above, the anal canal is considered a coloanal anastomosis by some surgeons. Others have a more rigorous definition, requiring anastomosis at the dentate line, and refer to the former as an ultralow colorectal anastomosis. Technique of dissection and anastomosis also varies. The dissection can be performed from the abdomen with distal division of the rectum and anastomosis performed using surgical staplers. Alternatively, some surgeons advocate transanal dissection to define the distal margin for low tumors, with a hand-sewn anastomosis. Regardless of technique, the concept of removal of the rectum and mesorectum with restoration of intestinal continuity in the region of the anal sphincter complex remains constant. Construction of a neorectum using a colonic pouch has received considerable attention recently. Stool frequency and urgency are reduced as compared with straight coloanal anastomosis, probably because of increased reservoir capacity.11-18 The advantage of the pouch-anal anastomosis over a straight coloanal anastomosis seems to be most pronounced in the first 1 to 2 years following construction. The typical colonic pouch is constructed in a "J" fashion using a 6- to 10-cm limb length. Larger pouches may impair emptying, although the functional differences emphasized in some studies of pouch limb length may be negated by changes in pouch volume over time.19,20 Many surgeons perform a temporary diverting stoma to minimize the consequences of anastomotic disruption should it occur. The low level of the anastomosis and the increased length of suture line at risk when a colonic pouch is constructed increase the risk of leak and pelvic sepsis. Anastomotic leak may cause pelvic fibrosis, rendering the neorectum stiff and noncompliant. Patients may thus suffer from tenesmus and fecal incontinence. The most commonly used diverting stoma is the loop ileostomy, because its construction and subsequent takedown are straightforward, and the colonic mesentery is not threatened by such manipulation. Temporary fecal diversion may also be prudent in cases where neoadjuvant chemoradiation is used. Although dramatic reductions in tumor size can be achieved with the use of preoperative chemoradiation, it may predispose to anastomotic problems in patients not undergoing temporary diversion at the time of proctectomy and coloanal anastomosis.21 Patient evaluation Patients with rectal cancer undergo evaluation with digital rectal examination, rigid proctoscopy, and transrectal ultrasound. Invasion of the anorectal muscular ring precludes restoration of intestinal continuity. Surgeons should not be seduced into triumphs of technique over judgment in cases where the tumor is very low in the rectum. It is not advisable to jeopardize oncologic cure because of enthusiasm to perform low anastomosis. As Heister remarked in 1739, speaking of intestinal stomas: " . . . it is surely better to part with one of the conveniences of life than to part with life itself."22 Colonoscopy is performed to search for synchronous neoplasms. Abdominopelvic computed tomography and a chest radiograph are obtained to rule out distal metastases. Transrectal ultrasound and pelvic computed tomography can detect extensive mesorectal lymph node involvement and help guide decisions regarding neoadjuvant treatment. Patients undergoing neoadjuvant external beam radiation with or without chemotherapy are reexamined by the surgeon prior to operation to assess tumor response to therapy. Before considering coloanal anastomosis, sphincter function must be assessed. A detailed continence history and physical examination by an experienced surgeon is probably the most predictive of postoperative function.23 However, anal manometry may help in equivocal cases. Despite the enthusiasm of many patients to restore intestinal continuity at all costs, it should be remembered that an abdominal colostomy is preferable to a perineal colostomy in cases where sphincter function is impaired. Operative preparation Consent for operation should include the possibility of temporary or permanent stoma construction. Stoma sites should be marked preoperatively by the surgeon or enterostomal therapist with the patient in the sitting, standing, and supine positions. A mechanical and antibiotic bowel preparation is used. Intravenous broad-spectrum antibiotics are given perioperatively for 24 hours. Deep venous thrombosis prophylaxis is indicated. Our policy has been to use heparin, 5000 U subcutaneously every 12 hours, and intermittent pneumatic compression stockings. Ureteral catheters should be considered in cases where invasion or impingement on the ureters or bladders is suspected. Although some authors have found that ureteral catheters are not useful in preventing injury, we have found them extremely helpful in facilitating difficult pelvic dissection. The patient is placed in the lithotomy position and a Foley catheter is inserted. Digital rectal examination and proctoscopy with the patient under anesthesia can be helpful in determining tumor level and degree of fixation. We routinely employ rectal irrigation with saline and dilute povidone-iodine prior to operation, and leave a rectal tube in place. Exploration A midline incision is made and the abdomen explored. The presence of unresectable metastases makes coloanal anastomosis ill-advised. If a temporary diverting stoma is constructed, the patient may never be fit enough to have the stoma closed. Even if the patient does not require a diverting stoma, the initial 6 to 18 months following coloanal anastomosis are often marked by frequent loose stools and occasional incontinence. The patient may not survive long enough to benefit from the gradual improvement in function of the neorectum. If the patient is treated with chemotherapy postoperatively, the risk of diarrhea and fecal incontinence are increased further. Low anterior resection with Hartmann closure of the rectal stump is a more prudent option for these patients. Tumor mobility in the pelvis is then assessed. Apparent fixation to pelvic structures may not preclude cure, as inflammatory or fibrotic reaction may be responsible, especially in cases where preoperative radiation or chemoradiation has been used. However, if during the course of pelvic dissection it becomes obvious that tumor invasion of pelvic structures makes complete resection impossible, low anastomosis should be abandoned because of the high risk of persistent tumor growth in the pelvis. Mobilization The left colon is mobilized from its retroperitoneal attachments (Figure 1, A and B). The bloodless plane lateral to the descending colon mesentery is developed first, initially avoiding the pelvic brim, where the ureter and iliac vessels are most prone to injury. The small bowel should be protected during this mobilization because of the risk of penetrating through the thin avascular area at the base of the descending colon mesentery. We routinely accomplish this by carefully placing a large laparotomy pad over the small bowel and retracting it medially. After the descending colon has been mobilized and the ureter identified, the sigmoid is mobilized from its lateral attachments. Several maneuvers are then performed to allow the descending colon to reach into the deep pelvis. We use the descending colon to perform all pelvic anastomoses because the sigmoid colon is narrow and has muscular hypertrophy in most patients. In addition, if preoperative external beam radiation has been used, the sigmoid may have been included in the radiation field and should not be used for the anastomosis. The splenic flexure is completely mobilized. The omentum is separated from the distal half of the transverse colon. An incision is made at the base of the sigmoid and descending colon mesentery on the right from the sacral promentory to the duodenum. It is frequently necessary to divide small filmy attachments of the distal duodenum and proximal jejeunum to the colonic mesentery to completely expose the origin of the inferior mesenteric artery (IMA) and the inferior mesenteric vein (IMV). The IMA is divided at its origin (Figure 1, C). Care should be taken to avoid the lumbar sympathetic plexus during IMA division. The IMV is divided adjacent to the ligament of Treitz (Figure 1, D). The blood supply to the proximal bowel is thus dependent on arcades originating from the middle colic artery. Dissection then proceeds from the IMA origin toward the pelvis. Keeping the hypogastric nerves protected posteriorly, the plane just posterior to the superior hemorrhoidal artery is opened. This plane is used to gain entrance to the presacral space at the sacral promontory, and can be followed in the posterior midline to the anal canal without encountering a blood vessel. Once the splenic flexure, descending colon, and sigmoid have been completely mobilized and the IMA and IMV divided, the proximal colon is divided. Although this step can be deferred until after pelvic dissection, division at this point in the operation allows the surgeon to retract on the rectosigmoid without danger of causing injury to the descending colon. The descending colon mesentery is divided between clamps beginning just proximal to the IMA. The line of mesenteric dissection extends perpendicular to the bowel to avoid jeopardizing proximal blood supply. We routinely divide the colon in an area of soft descending colon, leaving the sigmoid with the specimen. By performing the aforementioned mobilization, the descending colon should reach to the level of the anus. If additional length is needed, the omentum can be completely removed from the transverse colon and congenital attachments between the transverse mesocolon and retroperitoneum divided. It is rarely necessary to divide any of the arcade vessels originating from the middle colic artery, and great care should be exercised to ensure that pulsatile flow remains to the distal bowel. If a straight double-stapled anastomosis is planned, the bowel can be divided between clamps. A purse-string suture is placed and the proximal bowel sized to accommodate the anvil of the end-to-end stapler. If a colonic pouch or hand-sewn anastomosis is planned, the proximal bowel can be transected using a linear cutting stapler. Pelvic dissection then proceeds (Figure 2). The goal is to remove the mesorectum intact within its fascial envelope and to achieve adequate radial and distal margins from the tumor. Sharp dissection under direct vision is required. Blunt tearing maneuvers should be avoided. A strong headlight and lighted retractors are invaluable in this phase of the operation. The peritoneum on either side of the mesorectum is divided. The line of incision should be just medial to the ureter on either side, and extend anteriorly to the seminal vesicles or rectovaginal septum. Posteriorly, the areolar tissue between the mesorectum and the presacral fascia is divided. Sharp dissection with electrocautery or scissors under direct vision can be accomplished with anterior retraction of the rectosigmoid. It is often helpful to use a curved malleable blade to retract the mesorectum as the dissection proceeds to the level of the coccyx. The risk of injury to the basivertebral veins on the anterior surface of the sacrum is minimized if the urge to perform this portion of the dissection bluntly with a hand behind the mesorectum is suppressed. Anterior dissection is facilitated with a deep pelvic retractor placed between the anterior rectal wall and the posterior wall of the bladder. An assistant standing between the patient's legs places anterior retraction of these tissues. In a man, the plane posterior to the seminal vesicles is opened, the fascia of Denonvilliers incised, and the anterior rectal wall separated from the posterior capsule of the prostate. Palpation of the posterior surface of the pubis or palpation of the Foley catheter in the urethra distal to the prostate indicate that the distal aspect of the dissection is adequate. In a woman, the rectovaginal septum is separated to the level of the anal canal. The lateral stalks are then divided. Most of this tissue can be divided using the electrocautery, working from the bloodless plane opened posterior to the mesorectum and proceeding in an anterior direction. In some cases it is prudent to divide the middle hemorrhoidal vessels between clamps and ligate the vessels. In other cases, these vessels can be controlled with cautery alone. Irrespective of the method of division, the lateral stalks should be divided as lateral as possible without jeopardizing the ureters. After completion of this maneuver, the rectum should rise out of the pelvis and the mesorectum should appear as a fatty bilobed structure posterolateral to the rectum encased in a glistening fascial envelope. It is now possible to divide any remaining tissue attachments to the distal rectum and proximal anal canal. The levators should be clearly visible, the mesorectum completely mobilized, and the anal canal exposed. In difficult cases where the exact level of distal dissection is uncertain, the surgeon may place a finger in the anal canal with the other hand palpating through the abdomen. The tumor should be well above the level of distal exposure of the muscular tube of anorectum, optimally 5 cm, although 2 cm has found to be adequate. If the tumor is found to be invading the levators or sphincter complex intraoperatively, abdominoperineal resection should be performed. The pelvic autonomic nerves that innervate the genitalia and bladder are subject to injury during mobilization of the rectosigmoid. The preaortic sympathetic plexus can be drawn up into the resected specimen when the IMA is divided. The hypogastric plexus, located at the aortic bifurcation, and the hypogastric nerves, which are the condensation of those sympathetic fibers, are vulnerable to injury during mobilization of the mesorectum from the sacral promontory and lateral pelvic sidewalls. Sympathetic injury will result in retrograde ejaculation, or a "dry orgasm." The pelvic plexus contains sympathetic fibers from the hypogastric nerves as well as parasympathetic fibers from the pelvic splanchnic nerves. This plexus is superior to the levators, anterior and lateral to the rectum, and between the peritoneum and the endopelvic fascia. Injury to the pelvic plexus can occur at the time of division of the lateral stalks or during the dissection of the rectum from the region of the seminal vesicles and prostate. Injury to both of the pelvic plexuses will result in complete impotence in the male. Distal transection of the anorectum Division of the anal canal or very distal rectum can usually be accomplished through the abdomen using a linear stapler (Figure 3). The site of division should be below the tail of mesorectum so that all mesorectal tissue accompanies the specimen. The rectum is retracted superiorly and posteriorly, with a deep pelvic retractor exposing the anterior surface of the rectum. We frequently use a "perineal pusher," an assistant standing between the patient's legs using folded towels enclosed in a sturdy clamp to manually elevate the perineum in a superior direction. After the stapler has been applied, a long occlusive clamp is placed on the rectum just proximal to the stapler and the bowel divided. It is crucial to have the specimen retrieved by the pathologist and have the radial margins inked prior to opening the specimen. Our pathologists now routinely report radial margin distance on all rectal cancer specimens. Some surgeons are enthusiasts of the combined transanal-transabdominal resection procedure, where the distal margin is defined and the distal dissection is performed from a transanal approach. We have found very few indications for this technique. In most patients, the proximal anal canal can be exposed from the abdomen and a stapling instrument applied below the tumor. If the tumor is so low that division in the proximal anal canal would not provide an adequate distal margin, then perhaps an abdominoperineal procedure would be safer from an oncologic standpoint. There may be a few patients who, because of a narrow pelvis or other aspect of body habitus, might benefit from such a technique. Anastomosis Adequate mobilization of the proximal colon and lack of tension are reconfirmed. Blood supply to the proximal margin can be reassessed as the proximal bowel has now been dependent on unidirectional flow from the middle colic artery for the entire period of pelvic dissection. ("If it looks blue, it is blue.") Poor blood supply to the proximal margin must be corrected prior to anastomosis. We routinely use a double-stapled technique to perform coloanal or ultralow colorectal anastomosis, and with increasing frequency we are constructing a colonic J-pouch (Figure 4). The colon is folded into a J-configuration with an approximately 8-cm limb length. The apex of the pouch is opened and a linear cutting stapler is used to create the pouch. Care must be taken to avoid incorporating the colonic mesentery into the staple line. Usually a single fire of a 75- or 90-mm stapler will create adequate pouch length. If there is a long blind end remaining, this can be corrected with a second fire of the linear cutting stapler. A purse-string suture is placed around the opening at the pouch apex and the anvil of the end-to-end stapler inserted. We usually use a 28- or 29-mm stapler in this situation. Retractors are replaced and hemostasis in the pelvis confirmed. The stapler is placed gently in the anus with adequate lubrication. For low anastomosis, there is often not enough remaining length of the anorectum for the stapler head to slip above the sphincter complex and thus give the feeling of being properly positioned. Great care must be exercised in this situation to avoid forcing the stapler and bursting through the distal staple line. The staple head is aligned and the spike of the stapler is advanced through the distal staple line. Although it is tedious at times, we routinely place a purse-string suture in the distal bowel to make it fit snugly around the spike of the stapler. We feel this prevents extrusion of the distal bowel from the anastomosis during closure of the stapler. Carefully retracting the vagina and any other tissue at risk, the anvil is attached to the spike and the stapler closed and fired. After removal of the stapler, the adequacy of the tissue doughnuts are assessed and the anastomosis inspected gently with a proctoscope. The integrity of the pouch and anastomosis are tested by instilling air via the proctoscope with the pelvis filled with warm saline. Bubbles are an indication of a leak and should be investigated and repaired. If there is a small separation at the anastomotic line low in the pelvis, it may be difficult to expose transabdominally and transanal repair with absorbable suture is indicated. Large leaks may require takedown and reconstruction of the anastomosis, although the short length of distal bowel makes this a difficult procedure. Drainage of the pelvis after proctectomy is controversial. The major potential benefit of postoperative drainage is removal of blood, which is a growth medium for bacteria. We typically place 2 soft, Silastic, closed-suction drains in the pelvis after proctectomy with pouch-anal anastomosis. However, if hemostasis is excellent at the conclusion of the procedure, drains may be unnecessary. We do not routinely make an attempt to close the pelvic peritoneum. If the descending colon mesentery is loose, however, we will reapproximate it to the retroperitoneum. We have had a case of small bowel becoming entrapped posterior to the mesentery of the descending colon, necessitating reoperation. In our opinion, the risk of anastomotic or pouch leak with pelvic sepsis and subsequent fibrosis of the neorectum outweighs the benefits of avoiding temporary fecal diversion in our patient population. In addition, most patients undergoing proctectomy and coloanal anastomosis for rectal cancer will have stage II or stage III lesions. We routinely treat these patients with postoperative chemotherapy. As mentioned above, chemotherapy-induced diarrhea in a patient first adapting to a neorectum will cause fecal incontinence. For these reasons, a temporary loop ileostomy is constructed (Figure 5). After chemotherapy is completed, the pouch and anastomosis are assessed endoscopically and radiographically and ileostomy takedown is performed. We thank John A. Craig, MD, for producing the illustrations. Corresponding author: Thomas E. Read, MD, Section of Colon and Rectal Surgery, 216 S Kingshighway (North Campus), St Louis, MO 63110 (e-mail: readt@msnotes.wustl.edu). References 1. Quer EDahin DMayo C Retrograde intramural spread of carcinoma of the rectum and rectosigmoid. Surg Gynecol Obstet. 1953;9624- 30Google Scholar 2. Grinnell R Distal intramural spread of rectal carcinoma. Surg Gynecol Obstet. 1954;99421- 430Google Scholar 3. Black UWaugh J The intramural extension of carcinoma of the descending colon, sigmoid and rectosigmoid: a pathological study. Surg Gynecol Obstet. 1948;1948457- 464Google Scholar 4. Scott NJackson Pal JTDixon MFQuirke PFinan PJ Total mesorectal excision and local recurrence: a study of tumour spread in the mesorectum distal to rectal cancer. Br J Surg. 1995;821031- 1033Google ScholarCrossref 5. Williams NSDixon MFJohnston D Reappraisal of the 5 centimetre rule of distal excision for carcinoma of the rectum: a study of distal intramural spread and of patients' survival. Br J Surg. 1983;70150- 154Google ScholarCrossref 6. Dukes C The surgical pathology of rectal cancer. Proc R Soc Med. 1943;37131Google Scholar 7. Goligher JDukes CBussey H Local recurences after sphincter-saving excisions for carcinoma of the rectum and rectosigmoid. Br J Surg. 1951;39199- 211Google ScholarCrossref 8. Quirke PDurdey PDixon MFWilliams NS Local recurrence of rectal adenocarcinoma due to inadequate surgical resection: histopathological study of lateral tumour spread and surgical excision. Lancet. 1986;2996- 999Google ScholarCrossref 9. Heald RJRyall RD Recurrence and survival after total mesorectal excision for rectal cancer. Lancet. 1986;11479- 1482Google ScholarCrossref 10. Zaheer SPemberton JHFarouk RDozois RRWolff BGIlstrup D Surgical treatment of adenocarcinoma of the rectum. Ann Surg. 1998;227800- 811Google ScholarCrossref 11. Macmanus J Perforations of the intestine by ingested foreign bodies. Am J Surg. 1941;111393- 402Google ScholarCrossref 12. Mortensen NJRamirez JMTakeuchi NHumphreys MM Colonic J-pouch anal anastomosis after rectal excision for carcinoma: functional outcome. Br J Surg. 1995;82611- 613Google ScholarCrossref 13. Lazorthes FChiotasso PGamagami RAIstvan GChevreau P Late clinical outcome in a randomized prospective comparison of colonic J-pouch and straight coloanal anastomosis. Br J Surg. 1997;841449- 1451Google ScholarCrossref 14. Hallbook OSjodahl R Comparison between the colonic J-pouch anal anastomosis and healthy rectum: clinical and physiological function. Br J Surg. 1997;841437- 1441Google ScholarCrossref 15. Seow-Choen FGoh HS Prospective randomized trial comparing J colonic pouch-anal anastomosis and straight coloanal reconstruction. Br J Surg. 1995;82608- 610Google ScholarCrossref 16. Ho YHTan MSeow CF Prospective randomized controlled study of clinical function and anorectal physiology after low anterior resection: comparison of straight and colonic J pouch anastomoses. Br J Surg. 1996;83978- 980Google ScholarCrossref 17. Ortiz HDe MMArmendariz PRodriguez JChocarro C Coloanal anastomosis: are functional results better with a pouch? Dis Colon Rectum. 1995;38375- 377Google ScholarCrossref 18. Nicholls RJLubowski DZDonaldson DR Comparison of colonic reservoir and straight coloanal reconstruction after rectal excision. Br J Surg. 1988;75318- 320Google ScholarCrossref 19. Lazorthes FGamagami RChiotasso PIstvan GMuhammad S Prospective, randomized study comparing clinical results between small and large colonic J-pouch following coloanal anastomosis. Dis Colon Rectum. 1997;401409- 1413Google ScholarCrossref 20. Hida JYasutomi MFujimoto K et al. Functional outcome after low anterior resection with low anastomosis for rectal cancer using the colonic J-pouch: prospective randomized study for determination of optimum pouch size. Dis Colon Rectum. 1996;39986- 991Google ScholarCrossref 21. Hyams DMMamounas EPPetrelli N et al. A clinical trial to evaluate the worth of preoperative multimodality therapy in patients with operable carcinoma of the rectum: a progress report of National Surgical Breast and Bowel Project Protocol R-03. Dis Colon Rectum. 1997;40131- 139Google ScholarCrossref 22. Heister L A General System of Surgery. London, England W Innys1743; 23. Church JMSaad RSchroeder T et al. Predicting the functional result of anastomoses to the anus: the paradox of preoperative anal resting pressure. Dis Colon Rectum. 1993;36895- 900Google ScholarCrossref

Journal

Archives of SurgeryAmerican Medical Association

Published: Jun 1, 1999

Keywords: anastomosis, surgical,pelvis,rectal carcinoma,rectum excision,intestines,loop ileostomy,colon,neoadjuvant therapy,radiochemotherapy,anus,cancer,anal sphincter

References