Background
Phase III RCT (INT 0123 / RTOG 94-05). 236 patients with clinical stage T1-4N0-1M0 esophageal cancer (squamous cell carcinoma or adenocarcinoma) selected for non-surgical treatment. All patients received concurrent cisplatin (75 mg/m² day 1) + 5-FU (1000 mg/m²/day × 4 days, q4wks × 4 cycles). Randomized to high-dose radiation (64.8 Gy) vs standard-dose radiation (50.4 Gy). The trial was stopped early after an interim analysis.
Interventions and follow up
Arm A: Concurrent CRT: cisplatin/5-FU × 4 cycles + 64.8 Gy in 36 fractions (1.8 Gy/fx)
Arm B: Concurrent CRT: cisplatin/5-FU × 4 cycles + 50.4 Gy in 28 fractions (1.8 Gy/fx)
Primary endpoint: Overall survival
mFollow up: Median 16.4 months (all patients); 29.5 months (survivors)
Arm B: Concurrent CRT: cisplatin/5-FU × 4 cycles + 50.4 Gy in 28 fractions (1.8 Gy/fx)
Primary endpoint: Overall survival
mFollow up: Median 16.4 months (all patients); 29.5 months (survivors)
Results
Median OS: 13.0 mo (64.8 Gy) vs 18.1 mo (50.4 Gy) — not significantly different
2-year OS: 31% vs 40% — not significantly different
Local/regional failure + persistence: 56% (64.8 Gy) vs 52% (50.4 Gy) — no significant difference
Treatment-related deaths: 11 (64.8 Gy) vs 2 (50.4 Gy) — however, 7 of 11 deaths occurred in patients who had received ≤50.4 Gy
2-year OS: 31% vs 40% — not significantly different
Local/regional failure + persistence: 56% (64.8 Gy) vs 52% (50.4 Gy) — no significant difference
Treatment-related deaths: 11 (64.8 Gy) vs 2 (50.4 Gy) — however, 7 of 11 deaths occurred in patients who had received ≤50.4 Gy
Adverse events
Main adverse events: Higher-dose arm: 11 treatment-related deaths vs 2 in standard arm. However, investigation showed most deaths in the high-dose arm occurred before the dose escalation (during the 50.4 Gy phase), suggesting deaths were not due to the higher dose per se. Grade 3+ toxicities similar between arms.
Conclusions
Escalating concurrent CRT from 50.4 Gy to 64.8 Gy did not improve local/regional control or overall survival in esophageal cancer. There was a higher number of treatment-related deaths in the high-dose arm, though this appears unrelated to the dose escalation itself. The standard radiation dose for concurrent cisplatin/5-FU chemotherapy in esophageal cancer is confirmed to be 50.4 Gy.
Key Limitations
Key Limitations: Trial stopped early after interim analysis — underpowered for the planned endpoint. The apparent increase in treatment-related deaths in the high-dose arm likely reflects baseline patient characteristics rather than dose effect (most deaths occurred before reaching 64.8 Gy). Shorter median follow-up limits assessment of late failures. Does not address whether dose escalation might benefit specific subgroups (e.g., adenocarcinoma, specific tumor lengths). Modern dose-painting and IMRT techniques were not used.
Clinical Context
INT 0123 definitively established 50.4 Gy as the standard radiation dose for concurrent CRT in esophageal cancer. This continues to be the dose used in CROSS (41.4 Gy for neoadjuvant) and definitive settings (50.4 Gy). Dose escalation studies with IMRT (up to 63 Gy via simultaneous integrated boost) remain under investigation but have not yet changed the standard.
References