Are We Overtreating Some Patients With Rectal Cancer?

In this issue of ONCOLOGY, Dr. Rothenberger and colleagues have collated clinicopathologic data with the theme of local recurrence and selective use of adjuvant therapy. They conclude that the data suggest we continue to overtreat some patients with rectal cancer. As a generalization, I completely agree with the authors. However, it is quite difficult to take outcomes data from large numbers of patients and selectively apply the end results to the prospective management of an individual patient with rectal cancer in the absence of highly accurate preoperative staging. The largest group that has been overtreated comprises those with upper rectal cancer or rectosigmoid cancer, ie, those with the lower edge of the cancer at or above the peritoneal reflection. Except for very bulky cancers, these tumors are quite low-risk for local failure. From an adjuvant perspective, we should utilize the "colon cancer" paradigm and not the rectal cancer algorithms. It is frequently unclear to the radiation therapist or medical oncologist as to the exact location of the tumor. It is very important that this large subset of patients with "rectal cancer" be clarified via tumor board or individual referral. Node-positive patients should receive single-modality adjuvant chemotherapy, not chemoradiation. Optimal management of patients with cancer in rectal polyps has always been difficult. For both colon and rectum, following polypectomy, the exact site of the tumor is frequently problematic because of the rapid healing at the polypectomy site. All such sites should be marked with India ink as soon as a pathology report suggests any possibility of invasive cancer, even if this means repeating the colonoscopy. This approach will allow a leisurely review of the pathology slides, second opinion, and so forth, before an appropriate treatment decision is made. 'Invasive' Cancer Without Metastatic Disease
The authors suggest that muscularis mucosa is equivalent to the basement membrane. Colon and rectum histology is quite different from many other epithelial sites. "Invasion" is usually associated with penetration in the lamina propria. In the large bowel, however, lymphatics are present only beneath the muscularis mucosa. This is clarified by the sixth edition of the American Joint Committee on Cancer's TNM staging manual, in which in situ cancer involves not only intraepithelial cancer (clearly noninvasive) but also cancers that invade the lamina propia without extension throughout the muscularis mucosa. Hence, patients may have "invasive cancer" but be at no risk for metastatic disease. These issues are well described by the authors and cannot be overemphasized. Lymph Node Metastases
The authors also discuss the risk of lymph node metastases based on level of invasion in the pedunculated or sessile polyp. One of the problems in collating the clinicopathologic results of such patients is that the databases used for analyses are those compiled from patients who undergo radical resection. Many of these patients have very superficial cancers, but they are likely ulcerated. This is quite different from the soft, superficial, pedunculated, or sessile cancer with invasion. Based on my interpretation of the data, I believe that the risk of lymph node metastasis is much less than the 10% quoted for these T1 cancers. The same issue is addressed by the authors in their section on the management of T1 and T2 rectal cancers. I agree completely that many patients with stage I rectal cancer are overtreated, particularly with the common use of adjuvant chemoradiation therapy. The authors point out that the main reason for this is an inability to reliably predict stage I cancer prior to treatment. Multicenter studies from the Netherlands and Sweden suggest that even with optimal pelvic clearance using total mesorectal excision and sharp dissection, stage III patients with negative radial margins still have local failure rates in the 15%-to-20% range. As pointed out, the Dutch study suggests that this local failure rate is reduced to < 5% with preoperative radiation therapy. Compelling data show that the use of postoperative chemoradiation therapy is far less effective than the preoperative sequence in the adjuvant therapy of rectal cancer. In addition, there is likely a greater impact on late bowel dysfunction associated with the postoperative sequence. Stage I Cancers
From this perspective, in the absence of highly reliable staging, patients who are at "some" risk for positive lymph nodes should receive preoperative adjuvant therapy. The major challenge is a more accurate assessment of the clinical stage I patient. Because patients with transmural tumors are at high risk for lymph node metastases, this group should be included in the neoadjuvant treatment algorithm. The fully mobile clinical stage I cancers represent the major problem, as discussed in the article by Dr. Rothenberger and coauthors. Even in such cases, if the tumor is extremely low-lying, I would recommend neoadjuvant therapy. Although the risk of local failure from a positive radial margin or lymph node spread is probably small, the concern about direct tumor cell implantation during the resection or reconstruction is a real theoretical concern and I believe justifies adjuvant therapy. The challenge is the patient with midrectal cancer, clinical stage I, in which an optimal surgical procedure has been associated with a high degree of cure. Many of these patients can be treated without neoadjuvant therapy and then selective use of postoperative therapy. Use of the colonic J-pouch for reconstruction will likely reduce the adverse effects on late bowel function associated with postoperative therapy. We all await molecular or genomic assays on the primary tumor that are highly predictive of regional spread, to facilitate these individual patient decisions. Improved ultrasound and magnetic resonance imaging may also complement molecular staging initiatives.