Jover R et al. (2006) Mismatch repair status in the prediction of benefit from adjuvant fluorouracil chemotherapy in colorectal cancer. Gut 55: 848–855

Mismatch-repair (MMR) deficiency is present in approximately 15% of colorectal cancers. Although 5-fluorouracil has been shown to be effective against stage III colorectal cancer, in vitro and clinical retrospective studies suggest that MMR-deficient tumors might be resistant to this agent. Jover et al. carried out a prospective study to investigate the relationship between MMR deficiency and adjuvant 5-fluorouracil therapy in 754 patients with colorectal cancer, 66 of whom had MMR-deficient disease. Median follow-up was 728.5 days.

In patients with stage II or III colorectal cancer, adjuvant 5-fluorouracil was associated with better overall and disease-free survival in patients with MMR-competent tumors, but had no effect on these outcomes in patients with MMR-deficient tumors. MMR status had no effect on probability of survival in patients who received 5-fluorouracil. In patients who did not receive 5-fluorouracil, MMR-deficient disease was associated with a better probability of survival than MMR-competent disease, although this result did not reach statistical significance.

Given the lack of efficacy of 5-fluorouracil in MMR-deficient colorectal cancer, the authors recommend that patients with stage II disease should be treated with surgery alone. Trials of other chemotherapeutic agents, such as irinotecan, should be undertaken for patients with stage III disease. According to the authors, the molecular characteristics of tumors need to be taken into account in the design of adjuvant chemotherapy regimens for colorectal cancer, and the use of immunohistochemistry or microsatellite instability analysis should be routine in the design of such regimens.