Background
Phase II/III RCT (PRODIGE5/ACCORD17). 267 patients with esophageal cancer (85% squamous cell carcinoma, 15% adenocarcinoma) treated with definitive concurrent CRT (50 Gy in 25 fractions). Randomized to FOLFOX (leucovorin + oxaliplatin + 5-FU, q2wks × 3 cycles during RT + 3 additional cycles post-RT) vs cisplatin/5-FU (cisplatin 75 mg/m² day 1 + 5-FU 1000 mg/m²/day × 4 days, q4wks × 4 cycles during and after RT). Evaluated whether FOLFOX is non-inferior to the standard cisplatin/5-FU regimen as concurrent CRT.
Interventions and follow up
Arm A: FOLFOX (leucovorin 200 mg/m² day 1, oxaliplatin 85 mg/m² day 1, 5-FU 400 mg/m² bolus day 1 then 1600 mg/m²/46h) × 3 cycles concurrent with 50 Gy, then 3 cycles adjuvant
Arm B: Cisplatin 75 mg/m² day 1 + 5-FU 1000 mg/m²/day × 4 days, q4wks × 2 cycles concurrent with 50 Gy, then 2 additional cycle
Primary endpoint: Progression-free survival (PFS) — non-inferiority
mFollow up: Median 25.3 month
Arm B: Cisplatin 75 mg/m² day 1 + 5-FU 1000 mg/m²/day × 4 days, q4wks × 2 cycles concurrent with 50 Gy, then 2 additional cycle
Primary endpoint: Progression-free survival (PFS) — non-inferiority
mFollow up: Median 25.3 month
Results
Median PFS: 9.7 mo (FOLFOX) vs 9.4 mo (cisplatin/5-FU), HR 1.01 (95% CI 0.75–1.34) — non-inferiority not formally met but clinically similar
Median OS: 17.5 mo (FOLFOX) vs 17.6 mo (cisplatin/5-FU) — not significantly different
2-year OS: 36.6% vs 38.6%
Complete response rate: 18.8% vs 15.6%
Median OS: 17.5 mo (FOLFOX) vs 17.6 mo (cisplatin/5-FU) — not significantly different
2-year OS: 36.6% vs 38.6%
Complete response rate: 18.8% vs 15.6%
Adverse events
Main adverse events: FOLFOX arm: significantly lower rates of grade 3–4 neutropenia (41.5% vs 62.5%), thrombopenia, and mucositis compared to cisplatin/5-FU. FOLFOX arm: higher rates of peripheral neuropathy (grade 2–3). No significant difference in grade ≥3 renal toxicity.
Conclusions
FOLFOX achieved similar PFS and OS to cisplatin/5-FU as concurrent CRT for esophageal cancer, with a more favorable hematologic toxicity profile. Formal non-inferiority was not established due to the confidence interval exceeding the pre-specified margin, but the PFS outcomes were essentially identical. FOLFOX represents an acceptable, better-tolerated alternative to cisplatin/5-FU for definitive CRT.
Key Limitations
Key Limitations: Non-inferiority was not formally demonstrated per pre-specified statistical criteria (CI exceeded bound). Majority SCC histology (85%) — results may not fully apply to adenocarcinoma. The FOLFOX regimen schedule (biweekly) differs from typical concurrent CRT regimens and requires careful coordination with RT days. Long-term follow-up beyond 2 years was limited. No surgical consolidation arm — purely definitive CRT comparison.
Clinical Context
PRODIGE5 positioned FOLFOX + 50 Gy as a clinically acceptable alternative to cisplatin/5-FU for definitive esophageal CRT, particularly in patients with renal impairment (cisplatin-ineligible) or hematologic fragility. NCCN and ESMO guidelines now include FOLFOX as an option for definitive concurrent CRT. The carboplatin/paclitaxel regimen (from CROSS) is also increasingly used in neoadjuvant settings. Selecting between regimens is largely driven by toxicity profile, tumor histology, and institutional practice.
References