CSS sets predict risk category, RFS in colorectal cancer

Because of the overall favorable survival with surgery alone for most patients with stage II colon cancer, there is ongoing controversy about the potential role for adjuvant therapy and risk stratification for these patients. Current practice is to test the cancers of patients with stage II disease for either loss of DNA–mismatch proteins or the consequent genomic manifestation, microsatellite instability, which is present in up to 15% of patients and conveys a superior prognosis. Some studies indicate a potential survival detriment with therapy with fluorinated pyrimidine monochemotherapy in these patients vs. observation alone. Stage II patients at higher risk for recurrence have been identified by clinical or pathologic characteristics and, across studies, T4 status has consistently predicted a higher risk for recurrence. Treatment guidelines indicate that either observation or treatment with postoperative chemotherapy with a fluorinated pyrimidine with or without oxaliplatin can be considered. However, adjuvant therapy trials have not shown a clear-cut survival benefit for adjuvant treatment of even high-risk stage II disease. Several gene signature profiles have been identified that can risk stratify patients with stage II disease into high, intermediate and low risk for recurrence. Although potentially prognostic, these signatures have not been predictive of a benefit for adjuvant chemotherapy in high-risk patients.
Gao and colleagues now report in JAMA Oncology what they term a “combinatory cancer hallmark-based gene signature” to predict recurrence and potential chemotherapy benefit in stage II colorectal cancer. They report potentially greater accuracy in risk stratification with the combined gene signature sets. Although 416 patients received adjuvant chemotherapy (the nature of which is never fully defined), only a small patient number of genomic signature-identified high-risk patients were actually treated and appeared to achieve a survival benefit from treatment compared with no treatment. The intermediate-risk patients treated with chemotherapy actually had a worse survival outcome compared with no treatment, but this is not sufficiently elaborated upon. Whether this is a function of presence of DNA–mismatch repair or microsatellite instability-high status is not clarified; the observation of a detrimental effect of fluoropyrimidine monochemotherapy in these patients has again already been reported. Although these data and observations are encouraging and hypothesis generating, the small patient numbers and retrospective nature of the analyses, as well as the potential impact of inclusion of oxaliplatin as part of adjuvant therapy and the feasibility of applying the proposed gene signature, indicate that further study and validation are needed.