André · Journal of clinical oncology : official journal of the American Society of Clinical Oncology 2015 · randomized controlled trial · n=2246

Adjuvant Fluorouracil, Leucovorin, and Oxaliplatin in Stage II to III Colon Cancer: Updated 10-Year Survival and Outcomes According to BRAF Mutation and Mismatch Repair Status of the MOSAIC Study.

Cited 682 times in the scientific literature.

Level 2 - randomized trial

Randomized controlled trial with 10-year follow-up

PubMed 26527776 · doi:10.1200/JCO.2015.63.4238 · record verified 2026-08-30

What was done

Ten-year survival update of the MOSAIC multicenter randomized controlled trial evaluating adjuvant bolus/infusional fluorouracil plus leucovorin (LV5FU2) versus LV5FU2 plus oxaliplatin (FOLFOX4) in 2,246 patients with resected stage II to III colon cancer. Mismatch repair (MMR) status and BRAF mutation status were also retrospectively analyzed in available tumor specimens from 1,008 patients to evaluate disease-free survival (DFS) and overall survival (OS).

What was found

After a median follow-up of 9.5 years, 10-year OS rates for LV5FU2 versus FOLFOX4 were 67.1% versus 71.7% in the overall population (HR 0.85; P = .043), 79.5% versus 78.4% for stage II disease (HR 1.00; P = .980), and 59.0% versus 67.1% for stage III disease (HR 0.80; P = .016). Among 1,008 evaluated samples, 95 (9.4%) had deficient MMR (dMMR) and 94 (10.4%) had BRAF mutations. BRAF status was not prognostic for OS (P = .965), while dMMR was an independent prognostic factor (HR 2.02; 95% CI, 1.15 to 3.55; P = .014). In stage II to III dMMR tumors, FOLFOX4 hazard ratios were 0.48 (95% CI, 0.20 to 1.12) for DFS and 0.41 (95% CI, 0.16 to 1.07) for OS; in BRAF-mutated tumors, HRs were 0.50 (95% CI, 0.25 to 1.00) for DFS and 0.66 (95% CI, 0.31 to 1.42) for OS.

Why it matters

These results confirm a durable 10-year overall survival advantage of oxaliplatin-based adjuvant chemotherapy in stage III colon cancer, while demonstrating no long-term survival benefit for the unselected stage II population.

Limits

Tumor tissue for biomarker assessment was available in less than half the trial cohort (1,008 of 2,246), introducing potential selection bias. Biomarker subgroups had small sample sizes and low event counts, resulting in wide confidence intervals that crossed 1.0 for DFS and OS hazard ratios.

Cited by