To treat, or not to treat—the decision to use adjuvant chemotherapy plagues medical oncologists and patients harnessed with the diagnosis of stage II colon cancer. A look to the literature does not simplify the decision, as significant controversy exists regarding the magnitude of benefit associated with 6 months of adjuvant chemotherapy. Dr. Kopetz and colleagues provide a well-organized review of the current literature examining the benefit of adjuvant chemotherapy in stage II disease, and discuss potential prognostic markers that may help determine who would most likely benefit from treatment.
Although individual trials have not had enough patients with stage II disease to adequately power an analysis of efficacy, investigators have used meta-analyses in attempts to define the benefit of adjuvant chemotherapy. They have failed to show a significant benefit in overall survival and typically nonsignificant improvements in disease-free survival.[1,2] Based on these data, the American Society of Clinical Oncology (ASCO) recommended that the routine use of adjuvant chemotherapy for stage II colon cancer was not recommended, as direct evidence from previous fluorouracil(Drug information on fluorouracil) (5-FU)-based trials did not demonstrate a survival benefit for chemotherapy.[3] The National Comprehensive Cancer Network (NCCN) guidelines mirror this sentiment, stating that the "benefit of adjuvant chemotherapy does not improve survival by more than 5%."[4]
Adjuvant 5-FU–Based Therapy: Recent Data
The QUick And Simple And Reliable (QUASAR) trial is the largest (and only) study to show a statistically significant benefit for adjuvant therapy in stage II colorectal cancer. Updated results with a median follow-up of 5.5 years were published recently and demonstrated a significant disease-free and overall survival benefit for adjuvant 5-FU therapy. In this study of 3,239 patients, 91% with stage II disease and 71% with colon cancer, the hazard ratio was 0.82 (P = .008) for overall survival and 0.78 (P = .001) for recurrence.
Further, the proportional risk reduction with the use of chemotherapy was the same regardless of whether patients received weekly chemotherapy or chemotherapy every 4 weeks, and did not differ between those with colon or rectal cancer. The benefit of chemotherapy in preventing recurrence was lost after 2 years, was not seen in patients over 70 years of age, and came at the cost of a reduction of quality of life only during the 6 months of treatment, with one possible toxicity-related death. The 5-year overall survival benefit for all patients was 3.6%, with a benefit of 5.4% in high-risk patients and 3.6% in low-risk patients.[5]
As noted by Dr. Kopetz, current adjuvant therapy for stage III colon cancer is FOLFOX, a combination of infusional 5-FU and oxaliplatin(Drug information on oxaliplatin) (Eloxatin), based on the results of the MOSAIC study, which demonstrated an improvement in disease-free and overall survival over infusional 5-FU alone.[6] With a median follow-up of 6 years, an improvement of 3.8% in disease-free survival (P = .258) was demonstrated for all stage II patients, while in the high-risk group, disease-free survival was increased by 7.2% with a hazard ratio of 0.74. However, no improvement in overall survival was demonstrated.[7] Based on the improvement seen in this study as well as that from the QUASAR study, it is possible that the benefit of FOLFOX therapy over surgery alone may be approximately 7%, but this figure is based on notoriously unreliable cross-trial comparisons.[8]
The Potential Cost of Adjuvant Therapy
Dr. Kopetz accurately points out that the risk-benefit ratio for adjuvant therapy must favor the benefit and that the reduction in risk associated with treatment must be adequate. The NCCN guidelines recommend that a discussion with patients contemplating adjuvant therapy for stage II disease should incorporate the prognosis, patient preferences, comorbidities, presence of poor-risk features, and a thorough conversation about the risks of therapy and the anticipated absolute and relative benefits of treatment[4].
In QUASAR, quality-of-life measurements were worse in patients receiving chemotherapy, but only for the duration of therapy.[5] In patients receiving 5-FU and oxaliplatin in either the MOSAIC study or the recently reported National Surgical Adjuvant Breast and Bowel Project (NSABP) C-07 adjuvant study, neurotoxicity was a major issue for 8% to 12% of patients during treatment.[6,9] Further, 11% of patients treated with FOLFOX in the MOSAIC study continued to experience grade 1 neuropathy at 4 years after treatment, while 10% of patients treated with bolus weekly 5-FU and oxaliplatin (FLOX) had unresolved neuropathy at 27 months after treatment.[7,10] In fact, 20% of patients receiving FLOX had an increase in neurotoxicity symptoms at 18 months compared to baseline, and 22% withdrew from the study due to neurotoxicity. The average amount of time to resolution of neurotoxicity was 9 months, and the effect on quality of life has yet to be determined.[10]
Additional toxicities to be considered when discussing adjuvant chemotherapy in stage II disease include the risk of neutropenia (41% with FOLFOX vs 4.7% with 5-FU) and diarrhea (10.8% with FOLFOX vs 6.7% with 5-FU)[7]. In C-07, severe gastrointestinal toxicity with bowel wall injury was seen more frequently in the FLOX arm (5.5% vs 3.0%), requiring antidiarrheals, antibiotics, intravenous fluids, and hospitalization.[9] Grade 3/4 diarrhea was also more frequent in the weekly regimens.[7,9] Thus, a thorough discussion of toxicities and their likelihood is important in framing the risk-benefit ratio of adjuvant 5-FU/oxaliplatin–based chemotherapy.
Predicting Recurrence
Dr. Kopetz and colleagues provide a comprehensive discussion of the classic histologic features thought to predict a poor prognosis, while recognizing that there is no evidence to demonstrate that these prognostic factors in any way predict response to adjuvant therapy. However, the current NCCN guidelines highlight poor features such as high histologic grade, lymphovascular invasion, bowel obstruction or perforation, less than 12 lymph nodes examined, close or positive margins, and T4 disease, and provide a category 2A recommendation for the optional use of adjuvant 5-FU/oxaliplatin, capecitabine(Drug information on capecitabine) (Xeloda), or 5-FU in addition to observation for these patients.[4]
The emergence of gene-expression profiling to help predict which tumors are at highest likelihood of recurrence may provide better guidance than traditional histopathologic features in determining who should receive chemotherapy. In stage II disease, gene-expression profiles with varying numbers of genes have been shown to accurately predict recurrence. Wang et al reported the initial results of a 23-gene signature with a 78% predictive accuracy for recurrence. Barrier et al validated this signature and reported an accuracy of 80% using a 30-gene profile.[11,12] Johnston et al reported that a 48-gene signature could predict early relapse with 100% accuracy in 32 patients.[13] While these early studies of small numbers of patients show promising accuracy, gene-expression profiling will require validation in large cohorts of patients before routine integration of these techniques into clinical decision-making can be recommended.
While adjuvant therapy remains a controversy in stage II colorectal cancer, the inclusion of these patients in dedicated clinical trials is imperative. With the current Intergroup trial, E5202, researchers are attempting to address the benefit of adjuvant therapy in the context of potential prognostic markers, which should be a high priority in this patient population. Only through dedicated clinical investigation of this patient population will we be able to truly determine the role of adjuvant therapy in stage II disease.
—Crystal S. Denlinger, MD
—Paul F. Engstrom, MD
Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
