Minimizing Therapy and Maximizing Outcomes in Rectal Cancer

  • Joel E. Tepper and
  • Bert H. O'Neil
JCO Dec 10, 2011:4604-4606; published online on November 7, 2011

It has been accepted for decades that surgical therapy is essential for cure of rectal cancer. However, as with other tumor types, studies have demonstrated that a small percentage of patients can be cured with primary radiation therapy (RT).These early studies treated patients who refused surgical resection (most often because of nonacceptance of colostomy) or who could not undergo resection secondary to comorbid medical conditions. Surgical resection carries significant acute morbidity as well as substantial long-term alteration of bowel function (especially for low-lying tumors), so minimizing surgery has the potential to benefit patients significantly.

Perhaps the largest such series, from the Princess Margaret Hospital (Toronto, Ontario, Canada), demonstrated a 5-year actuarial survival of 27% (48% for mobile tumors). Not surprisingly, the Princess Margaret Hospital series showed better local control and survival rates for patients with less locally advanced disease. Interestingly, patients in this series who had clinical complete remissions were noted to achieve those remissions slowly (often at 4 to 8 months), suggesting that our current paradigm of waiting only 6 weeks after radiation therapy to proceed to surgery may underestimate the true benefits of RT. Several other studies have also demonstrated that longer periods between RT and surgery lead to higher rates of pathologic complete response (pCR), an observation that lends support to the concept of an approach with delayed or omitted surgery.

A seminal article by Habr-Gama  was the first to seriously raise the possibility of avoiding surgical resection altogether in a substantial number of patients through use of chemoradiotherapy. In that study, 71 patients (28.6% of initial cohort of 265 patients) with a protocol-defined complete clinical response at 8 weeks, defined by computed tomography imaging and endoscopic evaluation, were observed without surgery, and outcomes were compared with those of patients with pCR at surgery. With a mean follow-up of 57 months, no cancer-related deaths occurred in the observed group, no patient experienced pelvic recurrence, and the overall recurrence rate was only 7%. Furthermore, 8.3% of patients who underwent surgery had pCR, despite lack of apparent clinical CR.

In the article that accompanies this editorial, Maas report on 21 patients (11% of 192 treated initially) with localized rectal cancer and a clinical CR to preoperative RT and concurrent chemotherapy who declined surgery and who were then managed in a consistent and well-defined manner. To be eligible for the study, patients had to have convincing evidence of a clinical CR to the neoadjuvant therapy as evaluated by clinical examination, pelvic magnetic resonance imaging, endoscopy, and biopsy. They were then observed closely so that local recurrences could be detected early and surgical salvage employed when necessary.

Although the follow-up is still short, the results have been impressive, with only one local failure to date (and one patient who experienced failure seems to have had successful surgical salvage). These patients did as well as patients who had neoadjuvant therapy and who were found to have a pCR after surgery. Importantly, this result was achieved with less toxicity and better short-term bowel function in the observed patients compared with those who underwent surgical resection. Thus, the early results from this study support and enhance the results reported by Habr-Gama  and raise the possibility that this approach could be generalized to a larger portion of the patient population.

It is notable that over the last decade, there has been a substantial improvement in chemotherapy for colon and rectal cancers, both in adjuvant and metastatic settings. In fact, a small group of patients with metastatic disease experiences CRs to combination chemotherapy, and a few of those patients may be cured with chemotherapy alone. One would expect that a course of chemotherapy such as FOLFOX (fluorouracil, leucovorin, and oxaliplatin) could potentially increase the pCR rate associated with a nonsurgical treatment regimen and thereby allow a substantially higher percentage of patients to be treated without surgery. In fact, the approach of neoadjuvant RT plus extended chemotherapy before surgery has now been examined in several trials and seems to be both feasible and interesting in terms of pCR rate and outcome, with the caution that studies thus far have been small and have had limited follow-up duration. This also raises the more general possibility that chemotherapy could be used to replace one of the two local modalities now used in the treatment of rectal cancer, with the hope that this will reduce the long-term adverse impact of treatment with all three modalities.

Building on the success of chemotherapy in colorectal cancer, a potentially complementary approach to primary RT involves the use of chemotherapy and surgery to avoid RT. Pelvic radiation is hardly benign, occasionally resulting in long-term bowel and bladder issues and an increased risk of fracture of the pelvic bones Indeed, a study presented recently by Schrag used preoperative fluorouracil, leucovorin, and oxaliplatin plus bevacizumab with no RT in carefully selected patients, with a surprising 23% pCR rate achieved with the chemotherapy regimen and excellent local control (again with relatively short follow-up in selected group of patients). The new US Alliance cooperative group (previously Cancer and Leukemia Group B, American College of Surgeons Oncology Group, and North Central Cancer Treatment Group) plans to study this approach further (without use of bevacizumab) in a multicenter randomized trial. However, this study will include only patients with tumors located higher in the pelvis (> 5 cm from anal verge) and without the highest-risk features for local recurrence (including no tumor within 3 mm of expected mesorectal resection margin).

Thus, one can now imagine a new possible treatment algorithm that needs to be studied, evaluated, and modified over the next 5 or more years, suggested by recent studies, including the current report by Maas et al,7 which might end up looking something like the following:

Very low-risk (T1-T2, N0), high tumors.: Surgical resection alone.

Moderate-risk (T3 or N+), high tumors.: Neoadjuvant chemotherapy → surgery: RT used only for patients with pathology demonstrating a high risk of local failure at time of resection.

Low-lying, low- to moderate-risk tumors.: Chemoradiotherapy therapy → chemotherapy → observation: surgery used if no clinical CR.

High-risk tumors.: Neoadjuvant chemoradiotherapy therapy → surgery → adjuvant chemotherapy or chemoradiotherapy → chemotherapy → surgery.

It is also likely that development of an algorithm of this sort could benefit from the increasing power of genomic evaluations that might in the future allow for better prediction of which tumors would most likely respond best to RT and/or chemotherapy as well as provide improved categorization of low and high risk. At the present time, the treatment of rectal cancers is split into two categories: patients receive either surgery only or neoadjuvant RT with concurrent chemotherapy followed by adjuvant chemotherapy. Because all of our treatment modalities have improved, we now have the opportunity to tailor further the primary therapy to one that best suits the individual patient and to truly maximize the benefit to the patient while minimizing the extent of the therapeutic interventions.