Chapter 41A. Perspective on Rectal and Anal Cancer

Introduction

For many years the treatment of rectal cancer has involved removal of the rectum and the mesorectal envelope through a laparotomy, an operation commonly known as total mesorectal excision (TME).1 For surgeons performing this procedure, the main surgical consideration was whether to preserve the sphincter and restore continuity of the bowel or remove the entire rectum and anal canal leaving the patient with a permanent colostomy. In recent years, an improved understanding of the biology of rectal cancer and the causes of local recurrence,2,3 coupled with advances in imaging,4 surgical techniques,5,6 and the use of radiation and systemic chemotherapy7 have expanded the available surgical options. Selection between the different surgical therapies is based predominately on the stage and location of the tumor. Other factors such as patient age, overall health, functional status, and personal wishes and expectations also need to be taken into consideration when deciding on an appropriate surgical approach.

Patient Evaluation

Treatment decisions in patients with rectal cancer can be influenced by the presence of synchronous tumors, by the locoregional extension of the disease, and by the presence of distant metastasis. Therefore, every patient should undergo a complete evaluation before outlining a treatment plan.

A complete colonoscopy is important to exclude synchronous polyps and cancers, but locoregional staging is essential to guide the initial therapy. A digital rectal examination (DRE) provides useful information because the mobility of the tumor in relation to the rectal wall is an indication of the depth of tumor invasion. The DRE is particularly useful in assessing the relationship of the tumor to the levator muscle and the external anal sphincter, and deciding between the different treatment options. A proctoscopic examination is the best method to assess the distance of the tumor from the anal verge, the only anatomical landmark that can be seen simultaneously with the distance marks of the rigid scope.

In addition to a thorough clinical examination, every rectal cancer patient should undergo adequate local and regional staging with the help of the best available imaging technology.4 Endorectal ultrasound (ERUS) is a useful technique for staging early rectal cancer as it provides detailed images of the different layers of the rectal wall and demonstrates the disruption of those layers by the tumor. Magnetic resonance imaging (MRI) is most useful for staging locally advanced rectal cancer because it provides a broader view of the pelvis and the best images of the fascia propria of the rectum. The new-generation computed tomography (CT) scanners also provide high-resolution cross-sectional images of the rectum, the mesorectum, and surrounding pelvic structures, and can be used for the locoregional staging of rectal cancers when high-quality MRI is not available. A chest x-ray and a CT scan of the abdomen and pelvis are also commonly included in any patient assessment to diagnose metastatic disease. Occasionally, other tests such as a triple-phase CT of the liver or a positron emission tomography-CT (PET-CT) may be necessary to confirm the diagnosis of liver or pulmonary metastasis.

Choosing the Surgical Approach

At the completion of the evaluation, the surgeon must decide whether the patient requires a TME or can be treated with a local form of therapy, such as local excision (LE). To make the right choice, the surgeon should take into consideration both the location and characteristics of the tumor and the overall status of the patient.

Local Excision

Patients with early-stage rectal cancer, that is, tumors localized to the bowel wall that have not penetrated beyond the muscularis propria and do not involve the mesorectal lymph nodes, can potentially be treated with LE, thus avoiding some of the mortality, morbidity, and functional consequences of removing the entire rectum.8 Local excision of low rectal cancers can be performed by a conventional transanal excision (TAE) or by transanal endoscopic microsurgery (TEM), an operation that uses a special operating proctoscope, endoscopic imaging, and long surgical instruments similar to those used for laparoscopy.9 The oncologic outcomes are similar with both LE techniques for tumors of the same stage and location.10 However, the advantage of TEM is that it permits the use of LE in tumors located in the mid and upper rectum, which would otherwise be out of reach using TAE. But this advantage is relative as LE should only be considered for patients with early-stage distal rectal cancers in whom a TME would otherwise require a coloanal anastomosis or an abdominoperineal resection (APR) of the rectum.

The initial decision to perform LE should be based on clinical staging and imaging studies; only patients with small mobile tumors, localized to the rectal wall and without mesorectal nodes suspicious for lymph node metastasis according to optimal imaging, should be considered for LE.8 However, the decision to accept LE as the only treatment should be based on the pathologic examination of the LE specimen. For fit and healthy patients treated with curative intent, LE as the only form of therapy should only be offered for distal T1N0 rectal cancers with favorable histologic features (well or moderately differentiated, without lymphovascular invasion, mucinous component, or signet ring cells) and negative resection margins. Patients with positive margins or tumors with unfavorable histologic features should be offered a TME. Patients found to have T2 tumors after an LE should be offered a TME because the 5-year survival rate after LE as the only form of therapy for T2 tumors is lower compared to TME.11 Two prospective phase II trials have suggested that postoperative radiation and chemotherapy decrease the risk of local recurrence after LE for T2 rectal cancer provided the surgical margins are negative and the mesorectum is free of nodes in preoperative imaging staging.12,13

Patients with clinically staged T2N0 tumors, that is, those with a complete break of the submucosa but no penetration into the perirectal fat and without evidence of mesorectal nodes in ERUS imaging, deserve special consideration. While still potentially resectable for cure with LE, these tumors carry a significant risk of occult nodal metastasis, and, if confirmed to be T2 tumors on histopathologic examination, they are associated with a high rate of local recurrence when they are treated with LE alone. Therefore, chemoradiation (CRT) before LE has been explored as an option for patients with distal uT2uN0 rectal cancer interested in an organ preservation approach.8 The ACOSOG Z6041 (American College of Surgeons Oncology Group Z6041) trial investigated the efficacy of CRT before LE in this subset of rectal cancer patients, but the long-term oncologic outcomes for these patients are not available yet, and therefore CRT before LE for ultrasound-staged T2N0 tumors should still be considered an experimental treatment.14

Total Mesorectal Excision

The majority of rectal cancer patients with tumors that have penetrated the muscularis propria or metastasized to the mesorectal lymph nodes require the removal of the rectum and the mesorectal envelope. This operation should be performed by sharp dissection within the areolar space between the fascia propria of the rectum and the presacral fascia. A blunt dissection increases the risks of tearing the mesorectum potentially leaving nests of cancer cells behind or causing bleeding from inadvertent tearing of the presacral veins. The risks of injuring the hypogastric nerves or the branches of the pelvic plexuses are also considerably reduced if they are visualized and protected during a sharp mesorectal excision along well-defined anatomical planes. Rectal perforation with the associated risks of pelvic infection, tumor cell spillage, and compromised sphincter preservation are also less likely when the dissection is performed outside the fascia propria of the rectum. The importance of performing a sharp mesorectal excision has been highlighted by a number of pathologic audits that have linked the completeness of the mesorectal excision to the risk of local and distant tumor recurrence.3

The need to remove the entire mesorectum for tumors located in the upper rectum has been a matter of controversy for years. There is now conclusive evidence that rectal cancers rarely spread distally in the mesorectum beyond 5 cm, measured from the lower end of the tumor.15 Therefore, for tumors located in the upper rectum, the mesorectum can be safely transected transversely, without coning, approximately 5 cm distal to the lower margin of the tumor. This operation is often called tumor-specific mesorectal excision. For tumors located in the mid or lower rectum, a 5-cm mesorectal clearance requires a complete TME.16

The mesorectum tapers distally as the rectum approaches the levator hiatus and ends slightly above the level of the anorectal ring. Distal to that point, the muscularis propria of the rectum is in contact with the levator muscle. For most mid and distal rectal cancers, the rectum is transected below the end of the mesorectum and the intestinal continuity is reestablished by a double-stapling technique. Whenever possible, a colonic J-pouch or a side-to-end anastomosis should be performed to provide some reservoir capacity and reduce, at least temporarily, the urgency and frequency associated with the sphincter-saving procedure. For patients with tumors located close to the anorectal ring, the surgeon must decide whether removal of the tumor with a negative margin is compatible with sphincter preservation or requires an APR of the rectum. For many years the decision between these two surgical options was based primarily on the possibility of obtaining a negative distal resection margin in the bowel wall. While a 2-cm margin of normal rectal wall distal to the tumor is desirable, a margin as short as 1 cm does not seem to increase the risk of recurrence, particularly in patients treated with neoadjuvant CRT.17 In recent years surgeons have learned that in most patients with very distal rectal cancers, the need for an APR of the rectum is due to the inability to obtain a negative circumferential resection margin rather than a negative distal margin. In general, an APR of the rectum becomes necessary when a distal rectal cancer has penetrated beyond the muscularis propria and infiltrates the levator muscle or the external anal sphincter. In these patients, the dissection in the intersphincteric plane in an attempt to preserve the sphincter will result in a positive circumferential resection margin with the consequent risk of local tumor recurrence. Tumor fixation on DRE is a good indicator of tumor infiltration of the levator muscle or the external anal sphincter, but high-resolution MRI and CT scans that provide axial, saggital, and coronal views can demonstrate the relationship of the tumor to the levator muscle and the external anal sphincter with a high degree of accuracy.

Historically, an APR of the rectum has been associated with higher local recurrence rates compared to the sphincter-saving procedures.18 This difference has been attributed in part to the higher rate of positive circumferential resection margins associated with the APR in tumors that infiltrate the levator muscle or the external anal sphincter.19 When the pelvic dissection is carried out distally to the level of the levator hiatus or beyond and the upper portion of the levator muscle is not removed, the risk of leaving tumor behind is very high. Therefore, when performing an APR for rectal cancer that infiltrates the levators or the external sphincter, the mesorectal dissection should stop at the upper level of the levators. During the perineal dissection, the levators should be divided at the apex of the ischiorectal fossa where they insert in the white line of the obturator fascia. This operation, known as a cylindrical APR, has been shown to decrease the risk of local recurrence compared to conventional APR.20 Some surgeons prefer to perform the perineal portion of the APR with the patient in the prone-jackknife position because it provides better visualization of all anatomical structures, improved ergonomics, and better use of assistants.

Some patients with low rectal cancers that do not infiltrate the levator muscle or the external sphincter but are too close to the pelvic floor to perform a double-stapling technique are still potential candidates for a sphincter-saving procedure with a hand-sewn coloanal anastomosis. In this procedure the dissection of the portion of the rectum distal to the tumor is performed through the anus. This transanal approach provides simultaneous visualization of the distal end of the tumor and the anatomical landmarks in the anal canal, in particular the dentate line and the anal verge. A circular incision is made in the rectal wall at least 1 cm below the level of the tumor. This incision is carried though the mucosa, submucosa, and muscularis propria/internal sphincter until the intersphincteric space is reached. The transanal dissection is carried out in the intersphincteric space separating the distal rectum from the levator muscle posteriorly and laterally and from the urethra and the prostate or vagina anteriorly, provided that the intersphincteric space is free of tumor. This transanal dissection of the distal portion of the rectum can be performed before or after the transabdominal mesorectal dissection. If the mesorectal dissection has been performed from the pelvis first, the transanal dissection is continued until the rectum is totally mobilized. When the transanal approach is the first step of the operation, the lumen of the distal rectum is closed with interrupted sutures and the patient is repositioned to perform the abdominal and pelvic aspects of the operation. Either way, once the specimen is removed, the distal end of the colon is anastomosed to the anal canal with interrupted absorbable stitches. Patients with distal rectal cancer treated with a TME and a sphincter-saving procedure should receive a loop ileostomy because of the high risk of anastomotic leak.

Minimally Invasive Tme

In most centers a TME is performed though a midline or low transverse laparotomy. However, many surgeons use minimally invasive techniques for the treatment of rectal cancer because of the potential gains of a faster recovery and improved short-term quality-of-life outcomes.

The length of the incision in rectal cancer surgery is dictated by the need to mobilize the left colon and take down the splenic flexure rather than by the mesorectal dissection. Therefore, dividing the inferior mesenteric artery and vein, taking down the splenic flexure, and mobilizing the left colon laparoscopically or with the help of a hand-assisted device, help reduce the length of the incision even if the mesorectal dissection is performed open through a low midline or low transverse incision. A laparoscopic or robotic mesorectal dissection reduces the size of the abdominal incision even further and expedites recovery without compromising the quality of the operation or the oncologic outcomes compared to open mesorectal dissection.21,22

For a laparoscopic TME, the patient is placed in a modified lithotomy position with the legs in stirrups and the hips fully extended. The patient needs to be well secured to the operating table to avoid sliding when the table is placed in steep Trendelenburg's position and/or lateral rotation. Once the pneumoperitoneum is created by either a Veress needle in the left upper quadrant or by placing the Hasson trocar in the periumbilical area, additional trocars are placed in the right upper, right lower, and left lower quadrants. Once the peritoneum is inspected and the presence of peritoneal carcinomatosis excluded, the operation begins by identifying the inferior mesenteric artery and its branches at the root of the left colon mesentery. A space is developed underneath the superior rectal vessels, and, once the left ureter is identified, the vessels are divided with a stapling device, between vascular clips, or with a bipolar energy device. Care has to be taken to avoid injuring the hypogastric plexus that lies close to the aorta. Next, the inferior mesenteric vein is dissected and divided close to the ligament of Treitz. The mesentery of the sigmoid and descending colon is lifted from the retroperitoneal structures by blunt dissection, from the inferior border of the pancreas to the pelvic inlet. The lateral attachments of the colon to the parietal peritoneum along the line of Told are divided from the pelvic inlet to the splenic flexure. Finally, the splenic flexure is completely mobilized after separating the omentum from the left side of the transverse colon. Some surgeons routinely take down the splenic flexure and mobilize the entire left colon in every patient undergoing a sphincter-saving TME for rectal cancer to ensure a tension-free anastomosis using the end of the descending colon. The blood supply of the left colon is never an issue provided the left branch of the middle colic vessels and the marginal vessels are preserved.

Once the colon is completely mobilized, the mesorectal dissection starts by applying traction anteriorly from the stump of the superior rectal vessels to open the areolar space behind the fascia propria of the rectum at the level of the promontory. The hypogastric nerves, clearly identified at this level in their course toward the pelvic sidewalls, should be carefully separated from the fascia propria of the rectum and brushed posteriorly and laterally. The peritoneum is opened on both sides of the rectum all the way to the cul-de-sac, and the rectum is lifted from the concavity of the sacrum by sharply dividing the areolar attachments of the fascia propria of the rectum to the presacral fascia. The lateral stalks are divided next using electrocautery, although armonic scalpel or bipolar coagulation can also be used. The anterior dissection to separate the rectum from the urogenital organs is performed last. The anterior dissection can be performed in different planes, depending on the location of the tumor. For anterior tumors the dissection should be carried in front of Denonvilliers' fascia to avoid dissecting into the tumor. For other tumors, the dissection can be safely performed behind Denonvilliers' fascia.

The laparoscopic TME is a technically demanding procedure because of the two-dimensional visualization, the use of long and rigid instruments, the difficulty handling the rectum and providing traction and countertraction during the dissection, and the unnatural position for the surgeon. Consequently conversion rates for laparoscopic TME in prospective randomized trials have been high.21 The robotic da Vinci (Intuitive Surgical, Inc., Sunnyvale, CA) platform eliminates some of these difficulties by providing tridimensional visualization, articulating instruments that resemble the human wrist, improved scale of motion, enhanced surgeon control of camera and instruments, and improved ergonomics. A number of retrospective case series have reported that a robotic TME is safe and provides similar outcomes compared to open or laparoscopic TME.23 However, the changes in patient position needed to take advantage of the gravity required to keep the small bowel away from the area of dissection represents a handicap to the use of the robot because it requires more than one docking of the instruments to the patient. Therefore, most surgeons perform a hybrid procedure with laparoscopic control of the vessels, mobilization of the left colon, and takedown of the splenic flexure and robotic mesorectal dissection. Robotic techniques are still evolving, and with new instrumentation and improved trocar placement it may be possible to perform the entire operation without the need to reposition the patient.

Use of Neoadjuvant Crt

Another important consideration for patients with rectal cancer is the use of neoadjuvant CRT to improve local tumor control. Information accumulated over several decades has proven that for patients with locally advanced rectal cancer the use of pelvic radiation with or without chemotherapy decreases the risk of local recurrence. For many years, pelvic radiation was given after surgery for patients with tumors found at surgery to penetrate the perirectal fact or involve the regional lymph nodes.24 With recent advances in imaging techniques that can accurately stage tumors before surgery, patients diagnosed with locally advanced rectal cancer now receive pelvic radiation, usually associated with sensitizing chemotherapy, prior to surgery. Indeed there is now conclusive evidence suggesting that the use of neoadjuvant CRT is more effective for local tumor control and safer than adjuvant CRT.25

Current guidelines in the United States recommend that all patients with clinical stage II or III rectal cancers should be treated with 5 weeks of hyperfractionated radiation and sensitizing chemotherapy, followed by TME 6–8 weeks later, and postoperative adjuvant chemotherapy.26 In Europe, where most rectal cancer patients are staged with phased-array MRI, patients are stratified into three different risk groups according to the penetration of the tumor into the mesorectum and its relationship to the fascia propria of the rectum.27 Patients with early rectal cancer are treated with TME alone. Patients with T3 or node-positive rectal cancer that is not close to the fascia propria of the rectum are treated by short course radiation (5 cGy/d for 5 consecutive days) followed by TME 1 week later. Patients with locally advanced tumors that are close to or reach the fascia propria of the rectum are treated with hyperfractionated radiation and sensitizing chemotherapy followed by TME, or extended surgery as needed to achieve an R0 resection, 6 weeks after completion of the CRT.

Patients with Metastatic Disease

Almost one-third of rectal cancer patients present with distant metastasis at the time of diagnosis. The treatment options in these patients are multiple and need to be individualized according to the local-regional stage and the symptoms of the primary tumor, the extent and potential resectability of the metastatic disease and the comorbid conditions, and performance status of the patient. Treatment decisions in these patients require a multidisciplinary approach with input from several medical and surgical specialists.

Patients with distant metastases that are resectable or borderline resectable at the time of diagnose often require multimodality therapy, including systemic chemotherapy, neoadjuvant CRT of the primary tumor, and surgery for the primary and the metastatic disease. The sequence and timing of these interventions need to be individualized according to the tumor and the patient, but also according to the response of the tumor to the different treatments. Asymptomatic patients should be treated initially with systemic chemotherapy with reevaluation after 2 months to assess tumor response and plan the surgery for both the primary tumor and the distant metastasis. Depending on the tumor response, some patients benefit from additional cycles of chemotherapy. Patients with locally advanced primary rectal cancers that approach the circumferential resection margin and have a high risk for local recurrence often need short-course radiation (5 Gy/d for 5 days) or CRT before surgery.

Fit and otherwise healthy patients may be candidates for synchronous resections of the primary tumor and the distant metastasis, particularly if the rectal resection can be performed laparoscopically. However synchronous resections are associated with high morbidity and some patients are better treated by staged resections. If the decision is to perform staged resections, the primary tumor should be treated first because the inability to achieve a curative resection for the rectal primary, or the finding of unexpected peritoneal disease at the time of the first procedure, may impact the treatment plan for the distant metastasis.A number of rectal cancer patients with resectable metastasis present with obstructive symptoms, rectal bleeding, or rectal pain at the time of diagnosis. While patients with severe symptoms may require an intervention such as tumor resection, diverting stoma, stenting, or even electrocoagulation, many of them experience a significant symptomatic improvement after a few weeks of CRT. Once the CRT is completed, patients can be treated with systemic chemotherapy.

In the past, asymptomatic patients with unresectable metastasis were offered surgery to treat or prevent debilitating complications related to the primary tumor such as obstruction, bleeding, or perforation. But multiple studies have proven that the proportion of these patients requiring an operation to treat complications is relatively small. Therefore asymptomatic patients with unresectable metastasis should also be treated initially with systemic chemotherapy. The treatment of the primary tumor will depend on the response to the chemotherapy and the development of symptoms. Patients with unresectable distant metastasis and symptoms either at diagnosis or after systemic chemotherapy may require local palliative interventions. These palliative procedures may include resection, diverting stoma, stenting, tumor ablation by laser or electrocautery, or even CRT, depending on the type and severity of the symptoms, the size of the tumor, the performance status, the comorbid conditions, and life expectancy of the patient.

In the last three decades of the 20th century, anal cancer changed from being a disease primarily treated by surgery to a disease treated with chemotherapy and radiation. These changes resulted in improved patient survival and quality of life. However, they have been paralleled by a dramatic increase in the incidence of anal cancer due primarily to the AIDS epidemics.

The surgeon still plays a pivotal role in the management of patients with anal cancer at different levels. The surgeon can play a role in the primary prevention of anal cancer by diagnosing and treating precancerous conditions, diagnosing and directing the initial therapy, following the patient to assess the tumor response to the multimodality therapy, and diagnosing and treating the recurrences. Therefore surgeons should be familiar with all aspects of the disease, from the etiology and pathogenesis to the treatment of advanced disease. All of these aspects are well covered in Chap. 42.

The term cancer of the anus includes a number of different neoplasms with different etiology, pathogenesis, treatment, and prognosis that have in common only their anatomical locations. Therefore a precise understanding of the embryology, anatomy, and histopathology of the anal region is essential. The anal canal extends from the anorectal ring, represented by the impromptu of the puborectalis in the distal rectal wall, to the anal verge, which corresponds to the palpable groove created by the distal end of the internal anal sphincter. Outside the anal verge is the anal margin, which corresponds to a circular area of skin extending 5 cm from the anal verge around the anus. Tumors with their center located in the anal canal are considered cancers of the anal canal, but in some cases the classification of an adenocarcinoma as rectal or anal, or a squamous cell carcinoma as originating in the anal canal or in the anal margin, may be difficult. However, this distinction may be relatively irrelevant, because an adenocarcinoma invading the distal rectum and anal canal should be treated as a distal rectal cancer with neoadjuvant CRT and APR of the rectum and anal canal, while a squamous cell carcinoma of the anal margin extending to the anal canal should be treated primarily with CRT.

As emphasized in Chap. 42, the anal canal is a complex embryologic region where different types of epithelia overlap over a significant distance. For example, squamous epithelium could be found several centimeters within the rectum. Therefore it is not unusual to find a squamous cell carcinoma located entirely within the distal rectum. In addition, there are cases of squamous cell carcinoma located exclusively in the mesorectal lymph nodes secondarily invading the rectum without conclusive evidence of a primary tumor in the anal canal per se. These considerations are important, because tumors in the anorectal region should be treated according to their histology. In other words, a squamous cell carcinoma of the distal rectum should be treated as a squamous cell carcinoma of the anal canal.

One of the most controversial issues in the management of patients with anal cancer is primary prevention by diagnosing and treating precancerous conditions. As described in Chap. 42, a number of institutions are now performing anal Pap smear in high-risk patients groups. Patients with positive Pap smear are often referred to surgeons for high-resolution anoscopy. These procedures are only performed at a handful of institutions and the data are still limited, but the diagnosis and ablation of areas of high-grade squamous intraepithelial lesions (H-SIL) seems to reduce the risk of invasive squamous cell carcinoma of the anus in high-risk patients. The true effectiveness of this intervention, measured as the number of cancers prevented and the complications caused, is still to be proven. It is likely that the recently introduced prophylactic and therapeutic vaccine will prevent anal cancer with less discomfort and potential complications than the surgical or ablative therapies.

Shortly after the introduction of multimodality therapy as definitive therapy, patients were typically followed for 6–8 weeks after the completion of the CRT. A biopsy of the tumor bed was performed, and an APR of the rectum was recommended if the biopsy was positive. However, these premature biopsies led to many unnecessary permanent colostomies. As a result patients are now followed clinically with DRE and anoscopy, and biopsies are reserved for patients with persistent or recurrent palpable, visible, or symptomatic lesions. In most institutions follow-up examinations are performed at 3–4 months for the first 3 years and every 6 months for 2 additional years. The benefit of using ultrasound in the follow-up has not proven to be effective.

Not every patient with squamous cell carcinoma of the anus responds to multimodal therapy, and 20–30% of tumors will never disappear completely or will relapse after an apparent initial complete response. Surgery is the best treatment for patients who have failed multimodality therapy, but only 30% of those who have a surgical resection survive 5 years.28–34 A radical resection with negative resection margins is the main prognostic factor and the only hope for cure in these patients, and therefore every effort should be made to diagnose recurrences early when they are still resectable and to plan the operation adequately to remove the entire tumor en block with negative margins. An APR in patients with anal cancer, particularly in those with squamous cell carcinoma of the anus who have failed multimodality treatment, usually requires a wider skin excision and complete removal of the ischiorectal fat and the levator muscle. These patients are therefore at a greater risk of infection and dehiscence of the perineal wound with the consequent morbidity. Therefore many of these patients benefit from a reconstruction of the perineum using either a rectus muscle flap or unilateral or bilateral gluteal flaps. In some patients salvage surgery requires extensive exenterative procedures with more complex reconstructions involving multiple specialists.30–34

The author thanks Nicola Solomon, PhD, for assistance in writing and editing the manuscript.