CHAPTER 26 Minimally Invasive Low Anterior Resection with Total Mesorectal Excision for Malignancy

The videos associated with this chapter are listed in the Video Contents and can be found on the accompanying DVDs and on Expertconsult.com.

Current resective procedures for rectal cancer can have their origins traced to a radical pelvic extirpation described by Ernest Miles in the early 1900s, an operation now commonly known as the abdominal-perineal resection (APR). In Miles’ original description of this operation, a cylinder of tissue was rapidly excavated from the pelvis (Fig. 26-1A), without preservation of neural structures. Five of his first 12 patients died, though none from hemorrhage. In addition, local recurrence was problematic in Miles’ early experience with this technique. Nevertheless, the Miles procedure did represent an advance in rectal cancer surgery that, before this time, had been operated from a perineal approach.

FIGURE 26-1 Extent of resection for rectal cancer. A, Traditional abdominoperineal resection (Miles procedure); coronal view. B, Contemporary low anterior resection with total mesorectal excision (after Heald).

(Redrawn from Heald RJ: Rectal cancer in the 21st century—radical operations: Anterior resection and abdominoperineal excision. In Mastery of Surgery, ed 5, Philadelphia, 2006, Lippincott Williams & Wilkins, pp 1543–1444.)

With the advances in surgical science and technology in the 20th century, an interest in sphincter-preserving operations for rectal cancer evolved. The introduction of the circular surgical stapler in the late 1970s, which could create an end-to-end anastomosis with a single firing, facilitated the construction of a low rectal anastomosis. In addition, accumulated evidence indicated that the classic 5-cm distal tumor margin for rectal cancer was unnecessary; anastomotic recurrence was not observed to be an issue with a distal margin of 2 cm (or even less). Most patients undergoing operation for rectal cancer used to have an APR; the proportion gradually inverted, and now most patients have a transabdominal resection with anastomosis, without a permanent colostomy.

Although the rate of anastomotic reconstruction for rectal cancer resection has increased, the morbidity from these procedures, in terms of sexual and urinary dysfunction, has remained high. In the 1980s, Heald and colleagues demonstrated that sharp dissection of the mesorectum with preservation of known neural structures in the pelvis and intra-abdominal anastomosis (Fig. 26-1B) could produce excellent survival and recurrence rates without the associated morbidities. The procedure did require more time to perform than the cylindric resection shown in Figure 26-1A. The goal of Heald’s procedure, however, was to remove all of the node-bearing mesorectum, without injuring important autonomic and sensory nerves. Heald’s technique of total mesorectal excision (TME) has now become generally accepted among colorectal surgeons and is emphasized in this chapter.

Anterior resection of the rectum involves the removal of the mid to upper portion of the rectum in continuity with the lower sigmoid using a transabdominal approach (i.e., anterior, as opposed to a perineal resection), and typically reconstructed with a primary anastomosis. Addition of the “low” modifier generally indicates that the extent of resection is to the mid to lower rectum. In this chapter, the definition of low anterior resection (LAR) is the transabdominal removal of a continuous section of large bowel from the lower sigmoid to the mid to lower rectum, with a colorectal anastomosis.

Operative indications

The primary indication for minimally invasive LAR with TME is localized, resectable adenocarcinoma of the rectum. In terms of the commonly used TNM classification system of the American Joint Committee on Cancer (AJCC), rectal tumors amenable to LAR with TME would include T1-3, N0-2 lesions or, generally speaking, stages I to III (Tables 26-1 and 26-2). It should be noted that in 2012, some guidelines (including the National Comprehensive Cancer Network) and experts still consider minimally invasive LAR for rectal cancer to be an experimental procedure, only to be performed within the confines of a clinical trial. As evident from the surgical literature, however, this procedure has been performed commonly (outside of trials) for rectal cancer in the Americas, Europe, and Asia since the mid-1990s. LAR for localized, resectable rectal cancer is intended to be curative; in some circumstances, however, palliative transabdominal resection for metastatic rectal cancer may be performed.

Table 26-1 TNM Staging System for Colorectal Cancer

Table 26-2 Anatomic Stages of the TNM System for Colorectal Cancer

The resectability and the need for neoadjuvant therapy of a rectal cancer is dependent on the distance of the tumor from the anal sphincteric complex (Fig. 26-2) and the depth of tumor invasion (the T classification; see Table 26-1), and whether there are radiologically positive lymph nodes in the mesorectum (the N classification; see Table 26-1). Tumor location, depth of invasion, and nodal status are evaluated with computed tomography (CT), endorectal ultrasound (EUS), and/or magnetic resonance imaging (MRI), as discussed in the Preoperative Evaluation, Testing, and Preparation section.

FIGURE 26-2 Digital determination of resectability (coronal view). The critical distance from the superior extent of the sphincteric complex to the tumor is measured using the examiner’s finger as a ruler.

A T1-2, N0 rectal cancer should undergo resection as primary therapy (Fig. 26-3). If the primary tumor is T1, N0, has a dimension of less than 3 cm, has favorable histopathology, and is within 8 cm of the anal verge (see Fig. 26-2), then local excision also may be acceptable treatment, particularly in the patient with increased operative risk or limited survival. Favorable histopathology has been defined as the absence of (1) perineural and lymphovascular invasion, (2) positive margins, and (3) poor differentiation. Recurrence appears to be higher after transanal compared with transabdominal resection of T1 lesions, and radiologically silent N1 disease may be missed by transanal excision of T1 tumors in up to two thirds of cases. As of 2012, the guidelines for transanal excision are in flux.

FIGURE 26-3 Algorithm for treatment of localized rectal cancer. See Table 26-1 for TNM staging definitions. LE, local excision.

If selected, local excision may be accomplished with a conventional transanal approach or with transanal endoscopic microsurgery (TEM). Tumors located up to 20 cm from the anal verge have been resected with TEM in specialty centers. Of note, although TEM traditionally has been performed with a proprietary instrumental setup, it now appears that the technique can be reproduced using single-port instrumentation borrowed from laparoscopic surgery.

If the tumor is less than 5 cm from the anal verge, then abdominoperineal resection generally is recommended. This usually means that the tumor is within 1 cm of the sphincteric complex (see Fig. 26-2). Sphincter-preserving operations for tumors at or below this level are experimental and will not be covered in this chapter. Tumors that are more than 5 cm from the anal verge generally are amenable to LAR.

If the rectal cancer has been staged at T3-4 or N1-2, then the patient should undergo neoadjuvant chemoradiation (see Fig. 26-3). This treatment typically has involved continuous intravenous 5-fluorouracil along with 45 to 50 Gy (fractionated) of pelvic radiation. This neoadjuvant chemoradiation requires about 6 weeks to complete. The transabdominal resection then should be performed 5 to 10 weeks after completion of the chemoradiation. A T1-2 tumor that is too close to the sphincter mechanism to permit a sphincter-preserving procedure also may undergo neoadjuvant chemoradiation in an attempt to make sphincter preservation possible, but the efficacy of this strategy is not clear.

A number of institutions have freely provided on the Internet their own guidelines for the management of rectal cancer. It would behoove the surgeon who operates on this disease to be familiar with a set of these guidelines.

Alternative Treatments

If the patient is medically unfit or otherwise refuses to undergo a major surgical procedure, then one or more of the following treatment alternatives may be considered: (1) primary chemoradiation with or without stenting (stenting is not effective in the lower rectum); (2) diverting colostomy for an obstructive lesion; (3) transanal excision; (4) TEM; or (5) transanal ablation with a variety of energy sources, including endocavitary radiation (as popularized by Jean Papillon in Lyon, France).

Contraindications

The contraindications to LAR for rectal cancer include (1) tumor invasion into the sphincteric complex; (2) tumor invasion into the pelvic sidewall, bladder (see later), or prostate (i.e., a T4b lesion); (3) tumor less than 2 cm from the sphincteric complex (see Fig. 26-2); (4) fecal incontinence; (5) unresectable M1 disease with minimal or no pelvic symptoms; and (6) medical contraindication to a major surgical procedure. A tumor that invades the sphincteric complex should be treated with an APR. As alluded to previously, there have been reports of LAR for tumors located less than 2 cm from the sphincteric complex, using an intersphincteric dissection (en bloc excision of the internal sphincter, which can be performed robotically), with a coloanal anastomosis performed from the perineum. With preoperative chemoradiation, a high rectal cancer with bladder involvement may be resectable with an LAR and concomitant cystectomy.

A T4b lesion should undergo neoadjuvant chemoradiation in an attempt to reduce the T stage, perhaps making a resection feasible. A patient with rectal cancer who is not continent of stool should undergo an APR (with neoadjuvant therapy as indicated) so that the fecal stream is controlled with an end colostomy. If the patient has unresectable metastatic disease and an asymptomatic or minimally symptomatic primary rectal tumor, then it is reasonable not to undergo a major resection for the primary lesion but to proceed directly with chemotherapy. Simultaneous minimally invasive resection of a rectal adenocarcinoma and hepatic metastases may be feasible, but a description of this is beyond the scope of this chapter. Palliation of a symptomatic rectal cancer in the presence of unresectable metastatic disease should be individualized to each patient depending on functional status and life expectancy, but such treatment might include chemoradiation, proximal diversion, stenting, transanal ablation, or transabdominal resection.

Obesity (body mass index [BMI] >30) is a relative contraindication to a minimally invasive LAR. Although minimally invasive rectal resection has been shown to be feasible in obese patients, it probably would be best not to attempt minimally invasive LAR in a patient with a BMI greater than 30 early on in a surgeon’s experience. In particular, centripetal obesity (i.e., excess adipose distributed in the abdomen), which is common in male patients, can increase the difficulty of dissection deep in the pelvis. The definition of early experience (the number of cases constituting the learning curve) for minimally invasive LAR is controversial and surgeon dependent but has been estimated to be in the range of the first 50 cases.

Preoperative evaluation, testing, and preparation

Ideally, all patients with rectal cancer should be discussed at a multidisciplinary tumor conference in both the preoperative and postoperative phase. Such a conference may include participation of medical oncology, surgical oncology, gastroenterology, radiation oncology, radiology, primary care, nursing, palliative care, and other specialties as needed. Evidence has been accumulating for a number of malignancies that patient outcome is improved if such discussions are performed routinely.

The patient with a rectal mass should undergo complete colonoscopy (including digital rectal examination and anoscopy) with biopsy of the mass. If the mass is located at the rectosigmoid junction, then tattooing the mass with dye (e.g., India ink) at the time of colonoscopy can facilitate the identification of the tumor in the event that the patient goes on to resection. If the biopsy confirms the presence of a rectal adenocarcinoma, then the patient also should undergo CT scanning of the chest, abdomen, and pelvis with oral, intravenous, and (ideally) rectal contrast. In addition, the patient should undergo an endoscopic ultrasound of the rectum to establish the depth of penetration (T stage) of the tumor and whether there is mesorectal adenopathy (N stage). In some centers, an MRI with an endorectal coil has been used to stage rectal tumors with impressive accuracy. A carcinoembryonic antigen (CEA) also is assayed on diagnosis of a rectal cancer. More elaborate tests for metastatic disease (e.g., positron emission tomography [PET]-CT scanning) are not routinely indicated for rectal cancer.

If the patient with rectal cancer is a candidate for anterior resection of the rectum, then the surgeon should perform her or his own proctoscopy with digital rectal examination. Ideally, this should be done on a day before the actual resection. The surgeon should determine tumor diameter, fixation, circumferential location, and the distance of the tumor from the anal verge or sphincteric complex. Traditionally, the relevant distance has been that from the tumor to the anal verge, as measured with a proctoscope or a finger (see Fig. 26-2). As indicated earlier, the minimal distance (tumor to anal verge) for the safe performance of an LAR is about 5 cm. The precise location of the anal verge can vary with sphincter contraction, introduction of an endoscope, and so forth, so the surgeon should be cautious when determining this distance in tumors of the lower rectum.

If the tumor is within reach of the surgeon’s examining finger, then perhaps a more reliable (and relevant) distance to measure would be that between the sphincter mechanism and the tumor (see Fig. 26-2). This distance can be determined by sliding the index finger of the other hand into the anal canal to palpate the ring of the sphincteric complex while the tip of the examining finger is placed into the sulcus between the tumor and the rectal mucosa. Using the finger on the ring of the sphincteric complex, the surgeon touches the point on the examining finger where it crosses the sphincter, withdraws the fingers, and thereby can obtain a direct measure of tumor-sphincter distance. Of note, it is easy to stretch the rectal mucosa and overestimate the tumor-sphincter distance, so the surgeon should not apply vigorous digital pressure. A minimum of 2 cm would be needed between the tumor sulcus proximally and the top of the sphincteric complex distally (i.e., not the anal verge) for the safe performance of an LAR. Of note, the critical distance of the tumor to the anal verge versus the sphincters obviously will not be the same, so the surgeon should use consistent terminology.

If the tumor has a clinical stage of T3 or N1-2, then neoadjuvant therapy should be given, as described previously. Although not required, it may be beneficial to restage the patient with CT or EUS after completion of the neoadjuvant therapy, before the resection. Documentation of disease progression after neoadjuvant therapy may alter the management, which can be discussed at the tumor conference. With regard to response to chemoradiation, an interesting scenario involves the patient who, after undergoing neoadjuvant therapy for rectal cancer as described previously, has a clinical complete response, that is, has no evidence of residual tumor. In 2012, most guidelines dictate that such a patient still should undergo transabdominal resection, followed by additional chemotherapy, even if pathologic examination of the specimen revealed no residual tumor (i.e., a pathologic complete response). Further investigation into this clinical situation might identify a subgroup of patients with rectal cancer who can be treated with primary chemoradiation, analogous to the management of anal canal cancer.

If a temporary ileostomy or permanent colostomy is planned, then an enterostomal therapist should see the patient before the operation. Regarding large bowel preparation, recent meta-analyses have claimed that mechanical bowel preparation is unnecessary and even detrimental to the patient. The landmark data from the 1970s, which demonstrated the efficacy of large bowel preparation in reducing postoperative infection, used a combination of oral antibiotics and mechanical bowel cleansing. Most studies of bowel preparation since then have focused on the effects of mechanical preparation and intravenous antibiotics, but not on oral antibiotics. It is from these studies that recent meta-analyses have based their recommendation that mechanical bowel preparation should be omitted. Because oral antibiotics have been largely ignored in recent studies, the actual status of complete bowel preparation (oral antibiotics plus cathartics) with respect to minimally invasive colon resection in 2012 is controversial. The decision whether to perform a bowel preparation and which type to use is left to surgeon preference. For an LAR, it is reasonable to have the colon and rectum evacuated of solid stool to facilitate passage of endoscopes, staplers, and so forth.

Patient positioning in the operating suite

General endotracheal anesthesia is administered, and an orogastric tube and Foley catheter are inserted. A central venous catheter and arterial line are optional, depending on the patient’s underlying comorbidities. Sequential compression devices are placed on both lower extremities, and the patient is placed in the low lithotomy position with both arms tucked (Fig. 26-4). The angle of thigh flexion should be less than 15 degrees to avoid interference with instrumentation during the splenic flexure mobilization. The buttocks should overhang the edge of the table to facilitate intraoperative sigmoidoscopy and transanal stapler insertion. The legs and feet are padded and strapped to the stirrups to secure the lower half of the patient during extremes of table maneuvering. The upper torso is secured to the operating table using two adjustable soft straps in an X configuration (see Fig. 26-4). The anesthesiologist should secure the head so that it will not roll during extremes of lateral tilt.

FIGURE 26-4 Overhead view of low lithotomy positioning for a minimally invasive low anterior resection. Position of monitor 1 is adjusted according to location of the dissection. No floor equipment should be present caudal to the dotted line (a). (b) Head support. (c) Chest straps. (d) Surgeon position. (e) Surgeon’s monitor.

If the patient has an ample panniculus, then it may be helpful to place an additional soft strap underneath the panniculus at the level of the anterosuperior iliac spines (see Fig. 26-4) to prevent the patient from rolling off the table during lateral tilt. The Foley catheter is taped to the leg, away from the perineum. In men, the scrotum may need to be suspended away from the perineum with a single stay suture (or taped) to the medial thigh. Before the sterile preparation and draping, the operating table is placed through the extremes of rotation and Trendelenburg position to ensure that the patient is immobilized and that surrounding equipment will not hinder table movement. In addition, the surgeon should make sure that a segment of left-sided bed rail is available for clamping of the post for a self-retaining retractor, either for the laparoscopic pelvic dissection or in case of open conversion.

If the patient’s preoperative imaging indicated that the tumor or perineoplastic process was involving a ureter, or if there has been a history of pelvic inflammation, operation, or other event that may complicate the pelvic dissection, then the surgeon can consider placement of ureteral stents. In these situations, a stent can aid in the identification of the ureter and reduce the risk for ureteral injury. If a ureter is transected, then the presence of a stent can aid in identification of the ureteral lumen and subsequent repair. These stents are not without risk for urinary tract complications, so it would be reasonable to employ them selectively, in difficult cases.

After the patient has been positioned, a digital examination is performed to confirm location, size, and mobility of the tumor. A proctoscopy should not be performed at this point; otherwise, the colon will become distended and hamper the laparoscopic dissection. The patient’s abdomen, lower chest, and flanks then are swabbed with an alcohol-based surgical preparation solution (e.g., ChloraPrep); the perineum is scrubbed with a solution of povidone-iodine and detergent (e.g., Betadine Surgical Scrub). The perineum and lower extremities are draped with separate sheets. The abdomen is draped so that the region from the pubis to the xiphoid, out to both flanks, is part of the sterile field. If the patient has a large panniculus with an underlying soft strap across the pelvis, then the prepared field may extend to the inferior extent of the panniculus.

The surgeon stands on the patient’s right, the first assistant stands opposite to the surgeon, the camera operator stands on the right cephalad to the surgeon, and the scrub technician is positioned off the right leg (see Fig. 26-4). Two monitors are used, one on each side of the patient; the surgeon’s monitor (on the patient’s left) may need to be repositioned according to where the dissection occurs (e.g., cephalad for the splenic flexure mobilization, caudad for the mesorectal excision). The entire right side of the patient, from the head to the foot, should be free of obstruction (e.g., intravenous poles, anesthesia machine, energy units) to allow the surgeon unencumbered movement.

Positioning and placement of trocars

At a minimum, minimally invasive LAR requires access to both lower abdominal quadrants. If mobilization of the splenic flexure or fecal diversion is necessary, then three or all four quadrants will need to be accessible. In most patients, initial access may be obtained in the infraumbilical position with open insertion of a blunt 10-mm cannula (Hasson type), and this port subsequently will be used for the camera (Fig. 26-5). Alternatively, peritoneal access may be obtained with a Veress needle or with an optical trocar. If the position of the umbilicus has shifted inferiorly because of obesity, then it may be better to place the camera in the midline midway between the xiphoid process and the symphysis pubis. If there is an increased risk that the procedure may need to be converted to open (e.g., patient’s BMI >35), then it would be prudent to use a vertically oriented incision for the camera port.

FIGURE 26-5 Positioning of ports and possible incisions for a minimally invasive low anterior resection. Options for incision use: 1 (periumbilical) = Hasson cannula, extraction; 3 = extraction, loop colostomy; 4 (Pfannenstiel) = stapler insertion, extraction; 5 = stapler insertion, 100p ileostomy; 6 = extraction, loop colostomy. Red = 5 mm port; blue = 12 mm port.

After the camera port has been inserted and the abdomen has been insufflated to 12 to 15 mm Hg, four additional ports (2 through 5) are placed in the diamond configuration, as shown in Figure 26-5. Although there is no absolute or constant arrangement for trocar placement, the positioning in Figure 26-5 is a general guideline that can be modified based on the patient’s body habitus, previous surgical scars, and so forth. The diamond configuration allows reasonable access to all four quadrants of the abdomen.

Because the shape and anatomic relations of the abdominal wall can be distorted with distention, the actual insertion point for ports 2 through 5 should be determined by the surgeon only after the abdomen has been fully insufflated. If the abdominal wall is not excessively thick, then laparoscope transillumination should be used at each port site to identify and avoid laceration of abdominal wall vessels. The port size at positions 2, 3, and 4 in Figure 26-5 initially can be 5 mm, and then upsized to 12 mm later if necessary. The port at position 5 (superior to the right anterosuperior iliac spine) should be 12 mm to accommodate the endoscopic stapler-cutter, and should be positioned as lateral and inferior as possible to facilitate placement of the stapler across the distal rectum. Furthermore, if the surgeon encounters difficulty during the pelvic dissection, then she or he should not hesitate to place an additional port (typically in the right mid-abdomen) to improve pelvic retraction, instrument triangulation, and so forth.

Operative technique

Exploration

With the patient in a neutral position, the peritoneal surfaces, including the visible portions of the liver, are examined for tumor deposits or other disease. Any suspicious lesion is biopsied and sent for frozen section analysis. As the sensitivity of preoperative imaging has increased, it has become uncommon to discover metastatic disease at laparoscopy that was not suspected preoperatively. If a small solitary metastasis is found (e.g., a superficial liver nodule) and can be resected with minimal difficulty, then it would be reasonable to proceed with an oncologic resection (clearance of the primary tumor with nodes) and a metastasectomy. The patient’s postoperative management can be discussed at a subsequent multidisciplinary tumor conference after all the pathologic data have been acquired. If multiple lesions or otherwise unresectable metastatic disease is found, then the decision to proceed with a palliative resection would depend on the symptoms caused by the primary tumor. For example, if a good-risk patient is suffering from anemia or pain, then a palliative resection (as opposed to a diverting colostomy) would be preferable. If the patient is being treated for tenesmus, then an APR likely will be needed. The acquisition of a good quality CT scan, however, should minimize the chance that a patient will come to operation with undiagnosed metastatic disease.

Proctoscopy

If the surgeon did not perform a proctoscopy on a prior date and the tumor cannot be localized on digital examination at the beginning of the anesthetic, then a rigid proctoscopy should be performed early on to establish the precise location of the rectal tumor. While the assistant compresses the distal sigmoid colon with a grasper, the surgeon can perform a proctoscopy with minimal insufflation to determine the distance between the tumor and the sphincter mechanism. It is important to prevent insufflation of the colon and small bowel during this step; otherwise, the laparoscopic working space will be compromised. If need be, the surgeon can mark the position of the tumor with India ink, or the assistant can place a marking stitch on the serosal surface of the anterior rectum.

Surgical Anatomy

Definitions and descriptions of pelvic anatomy are somewhat arbitrary and have been debated in the surgical and anatomic literature for over 100 years. The following is an abbreviated synthesis of current opinions; this is not intended to be a final statement on the relevant pelvic anatomy. The rectum is about 12 to 15 cm long (see Fig. 26-2), depending on how the measurements are taken, and is arbitrarily divided into upper, middle, and lower thirds. The rectosigmoid junction is near the sacral promontory, typically an angled region of the large bowel where the epiploic appendices of the sigmoid colon become less prominent and where the longitudinal bands of the taenia coli diverge and coalesce, encircling the rectum. The anterolateral portion of the upper to mid rectum is covered with peritoneum, whereas the remainder is extraperitoneal. The peritoneum reflects off the mid rectum onto the bladder or the uterus, forming the rectovesicular pouch (in the male) or rectouterine pouch (in the female; sometimes known as the pouch of Douglas) or, simply, the peritoneal reflection.

The arterial blood supply to the rectum derives mainly from the superior hemorrhoidal (rectal) branches of the inferior mesenteric artery and, to a lesser extent, from the middle and inferior hemorrhoidal arteries, which derive from the internal iliac (hypogastric) arteries (Fig. 26-6). The middle hemorrhoidal vessels were described in older surgical literature to run in a structure called the lateral ligament (discussed under Operative Technique), but this arrangement has been questioned by numerous anatomic studies, and the presence of the middle hemorrhoidal vessels now is considered inconstant. Although the inferior (and middle) hemorrhoidal vessels are the primary supply to the anal canal, these vessels have an anastomotic network that can support a long rectal stump after ligation of the superior hemorrhoidal vessels.