Chapter 41B. Perspective on Rectal Cancer

Dr Goldberg and Dr Bleday have admirably summarized the current literature regarding the diagnosis, evaluation, and treatment of rectal cancer. In broad strokes, I agree with what they say and wish primarily to highlight a few important issues.

In the description of the anatomy, the authors emphasize their preference for the anorectal ring as the anatomic landmark when evaluating the level of the tumor. Yet in other sections of the chapter, the anal verge and the dentate line are mentioned as the distal landmark rather than the anorectal ring. This is consistent with the confusion that exists in colorectal and general surgery and confounds and confuses recommendations made for approaching rectal tumors. I personally prefer the dentate line because it is a clear tissue transition not altered by patient body habitus. Consider, for example, a lesion at 5 cm from the anal verge. Heavyset patients may have a longer distance from the anal verge to the dentate line (4 cm), leaving the lesion quite close to the dentate line (1 cm above). In contrast, the distance from anal verge to the dentate may be very short in thin patients (1 cm), and the lesion may actually reside relatively high in the rectum (4 cm above the dentate line). This variability holds true for the other landmark mentioned, the anorectal ring. The muscular funnel that comprises the anal sphincters may be long in young muscular patients and shorter in others. These variations lead to unclear recommendations as to how to approach lesions at various heights. The literature would benefit from a standardization of landmarks so that authors and clinicians attempting to follow the recommendations in articles could compare outcomes across studies.

I agree with the authors' recommendation for a CT scan of the chest, abdomen, and pelvis in the preoperative evaluation of patients with rectal cancer and would add that a preoperative PET scan adds value when used selectively to assess abnormalities identified on CT scans. This approach is preferred to the routine use of PET scans as it is more cost-effective. It is better than a follow-up CT in 3 months after surgery in that it allows for earlier identification and treatment of metastatic and may obviate the need for surgical intervention.

The 7th edition of the American Joint Committee on Cancer (AJCC) TNM (tumor-node-metastasis) staging system published in 2010 developed new classifications of stages II and III tumors following the recommendations of the Hindgut Taskforce.1,2 As noted by the authors, the tumors are evaluated about depth of tumor invasion (T), nodal involvement (N), and distant metastases (M). Stage 0 tumors are T0 or Tis, N0 and M0. Stage I is T1 or T2, N0, M0. Stage II is T3 or T4, N0, M0. Stage III is Any T stage, N1 or N2, and M0. Stage IV is Any T stage, Any N stage, and M1. Stages II and III can be subdivided and these subdivisions were modified in the 7th edition based on the Surveillance, Epidemiology, and End Results (SEER) and the National Cancer Data Base (NCDB) data. Traditionally stage II was divided into IIA and IIB, T3, N0, M0, and T4, N0, M0, respectively. Stage III was separated into three stages: IIIA (T1 or T2, N1, M1); IIIB (T3 or T4, N1, M0); and IIIC (Any T, N2, M0). T4bN0 is associated with poorer survival and is now classified as IIC (previously IIB). In contrast, the following three tumors appear to have somewhat better survivals upgrading their classifications to IIIB, T1 or T2, N2a, T1 or T2, N2b and T3, N2a. Finally, T4b, N1a and T4b, NIb are IIIC (previously IIIB). These reclassifications are important for prognostication and treatment planning.

Perioperative management is an arena in which data appear to drive practice less rigorously than in other realms in surgery despite the push to practice evidence-based, outcomes-driven medicine. A case in point is the issue of bowel preparation. The authors note despite the lack of evidence supporting oral mechanical bowel preparation and evidence from Cochrane reviews and a meta-analysis from Pineda et al that oral mechanical bowel preparation may be harmful, it is still their preference to mechanically bowel-prep their patients prior to surgery.3,4 They are not alone in this practice. Nearly 99% of colorectal surgeons surveyed in 2003 did the same.5 Unfortunately, this survey predates the larger discussion and simply reinforces the need to educate the practicing physicians. We abandoned routine oral mechanical bowel preparation at our institution over 3 years ago because there are no data to support its routine use. We have continued oral mechanical bowel preparation when intraoperative colonoscopy is planned for tumor localization or clearing of the proximal bowel. Oral mechanical bowel preparation is associated with increased anastomotic leaks and wound complications in many series. This may result from dehydration, increased intraoperative fluid requirements (secondary to the dehydration), decreased core temperature (secondary to rehydration), and overresuscitation leading to edematous bowel. We have not noted any negative impact on bowel handling on either open or laparoscopic cases as suggested by the authors. Many other centers have abandoned routine oral mechanical bowel preparation based on the literature for all cases and simply perform one or two preoperative enemas to clear the distal bowel of feces to allow passage of a stapler. Others have eliminated bowel preparation for all right-sided lesions. The issue is far from settled as highlighted by a debate at the SSAT in 2009 when the presentation to eliminate oral mechanical bowel preparation won the debate unanimously and yet not one individual out of 250 attending did agree to change their practice based on the data. A large multicenter North American trial is still desired.

Wound infection rates are impacted by factors other than oral mechanical bowel preparation. The Surgical Care Improvement Project (SCIP) has outlined recommendations that have been adopted by the Centers of Medicare and Medicaid Services (CMS) and the Centers for Disease Control (CDC) as performance measures. These recommendations emphasize appropriate timing of antibiotic delivery, within 60 minutes of incision, discontinuation of antibiotics within 24 hours (although there are no data to support more than one preoperative dose with appropriate intraoperative redosing as needed),6 appropriate hair removal, maintenance of normothermia, and “early” removal of urinary catheters.7–11 In teaching hospitals the discontinuation of antibiotics proved to be a challenge. Therefore we moved to a single dose of ertapenem when appropriate as it provides a 24-hour coverage obviating the need for any postoperative “prophylaxis.” The timing of urinary catheter removal has not been well studied in the colorectal population, and colorectal pelvic cases are explicitly exempted from the urinary catheter removal performance measure.

The authors provide an excellent discussion of the controversies surrounding local excision of rectal cancers as would be expected, given Dr Bleday's leading role in defining patient populations appropriately treated in this fashion. Articles by You and others highlight the potential risks and benefits of transanal excision.12–16 It appears T1 lesions within 5 cm of the dentate line with favorable histology and clear margins seem appropriately treated in this fashion. Whether T2 lesions may be treated in this fashion if they meet the aforementioned criteria and are treated with either preoperative combined modality therapy (chemotherapy plus radiation therapy) or postoperative therapy remains unclear. The potential for decreased operative morbidity and mortality with local excision is clear. Recently the presumed functional benefit and improved quality of life (QOL) associated with transanal excision were questioned when local excision was compared to low anterior resection (LAR).

It is clear that appropriate patient selection and adherence to preoperative selection criteria are critical, and ultimately this is an individual decision made after lengthy informed discussions with the patient and the patient's family. Unfortunately the current selection criteria are still lacking an ability to predict who will fail locally or with distant disease, and we await biologic markers. However, even with improved markers, questions will remain regarding treatment of primary rectal tumors by local excision in attempts to preserve gastrointestinal continuity. For instance, in a patient with markers that predict a high likelihood of early metastasis, does the method of removing the primary rectal tumor impact development of systemic disease and overall survival? Are local excision and early chemotherapy to treat microscopic disease preferable to a large operation because of the potential for immunosuppression associated with a more invasive procedure?

As with any other operative approach, surgical technique (and most likely surgeon volume and possibly hospital volume) plays a role. This is clear in total mesorectal excision (TME), and it seems reasonable to assume that it would be true for local excision approaches, whether TEM or transanal excision. This issue is at the heart of the dilemma in that surgeons often feel that their personal experience is not equal to that published in large series where technique and selection may be harder to define.

The authors give a nice description of transcoccygeal surgery. I have not had the opportunity to use this approach feeling the morbidity of a transsacral colocutaneous fistula outweighs the potential benefits. Selected patients with significant comorbidities and increasing operative risk might be candidates for this approach. My practice has been to proceed with an LAR or abdominoperineal resection (APR) instead of the transcoccygeal approach.

Quality of life (QOL) is a much-understudied subject in the treatment of rectal cancer. We are just starting to accumulate data to help us answer how best to treat these patients with a focus on long-term QOL issues. QOL impacts our decisions to perform lateral lymph node dissections, provide pelvic radiation therapy, create a colonic J-pouch, and pursue palliative procedures. The experience in North America with lateral pelvic node dissection, as highlighted by the authors, suggests that the complications associated with the procedure outweigh potential benefits. However, a true comparison of TME plus lateral pelvic node dissection versus TME plus pelvic irradiation in the treatment of patients with rectal cancer with a focus on local failure, overall survival, and QOL is lacking.

Pelvic radiation for the treatment of rectal cancer is standard of care for T3 or T4 lesions and Any T with nodal disease. However, there is clearly room to define the functional impact of radiation therapy on the function of the residual rectum. Many patients suffer frequent bowel movements after resection of the rectum with clustering of their movements (LAR syndrome). Radiation therapy negatively impacts the reservoir function of the rectum, leaving some to question the routine use of radiation after a well-performed TME.

Colonic J-pouches favorably impact the frequency of bowel movements in the first year. The pouch does not appear to be associated with significant long-term benefit and many patients have difficulty evacuating the pouches, leaving some to question the advisability of creating a pouch for short-term benefit. One note of caution, the authors state that they select either sigmoid or descending colon to create the pouch. In Western cultures the surgeon needs to be sure that the sigmoid is healthy and not involved with diverticular disease that would limit distensibility of the pouch.

Palliation of the primary rectal lesion in a patient with established distant disease is a challenging problem that is best approached with a multispecialty team, often a tumor board. We have chosen to be aggressive in our treatment of metastatic disease in the well-selected patient, believing the metastatic disease presents the biggest challenge to overall survival. In a patient with an asymptomatic primary, we have offered chemotherapy to treat the metastatic disease looking for a tumor response. In those patients where a response is seen, we may proceed with chemoradiation therapy of the primary if indicated by imaging studies. If the metastatic disease does not progress during this time interval, we may then proceed with resection of the primary with simultaneous resection of the metastatic disease, especially if we are able to achieve these goals laparoscopically with the assistance of an experienced laparoscopic liver surgeon. We may also proceed with resection of the primary alone or metastatic disease in staged procedures. We are ever vigilant regarding the primary tumor and local invasion of surrounding structures, believing the pain associated with local invasion a significant issue that we control poorly. If the primary appears to be encroaching on the sidewalls, we offer resection for palliation.

We have been aggressive, as have the authors, with locally advanced primary or recurrent disease believing posterior exenteration, pelvic exenteration, and exenteration including sacrectomy to be the best methods to control the tumor and associated symptoms in those with locally advanced disease.

Finally, whenever possible, we have preferred stenting of the primary rectal tumor in the patient with advanced metastatic disease to fecal diversion as it avoids a stoma and the morbidity of surgery with associated time loss due to hospitalization and recovery.

The description of TME is excellent, and the importance of this technique cannot be understated. Pathologic assessment of the specimen and evaluation of the adequacy of mesorectal excision has been shown to predict 5-year survival.17

I do not perform oral mechanical bowel preparation routinely, reserving that for cases in which I anticipate intraoperative colonoscopy might be necessary to localize a lesion. I provide one dose of ertapenem and no postoperative prophylactic antibiotics because no data support its use. Unless contraindicated, 5000 units of subcutaneous heparin and sequential compression stockings are used routinely. All vessels are taken with an electrosurgical device unless there is difficulty with hemorrhage. No drains are placed routinely as no data support this practice and increased complications are seen with routine usage.18,19

Postoperative care is another arena in which major advances and challenges to the traditional dogma await young investigators. One example is “early refeeding.” Experience in laparoscopic surgery started to challenge our beliefs; 80–90% of patients tolerated liquids within 24 hours of laparoscopic procedures.20,21 This practice was expanded to the open patient population. A meta-analysis found a statistically significant decrease in infection rates and lengths of stay and increased rates of vomiting.22 A Cochrane review suggested that early refeeding is safe and may reduce post-operative complications.23 We have practiced a patient-controlled diet for 18 years offering on postoperative day 1 a postsurgical diet (tea, coffee, chicken broth, toast, crackers, juice, English muffin, bagel). This is followed by a regular diet of choice (diabetic, etc.) emphasizing chicken, fish, rice, bread, pasta, and potatoes. The patient is encouraged to eat what sounds good and small meals (six per day) and avoid rich foods and foods that have “always made [them] sick.”