Background
Phase III, randomized controlled trial (RTOG 9501 / Intergroup). 459 patients with previously untreated squamous cell carcinoma of the head and neck who had undergone complete resection with high-risk pathologic features (two or more positive lymph nodes, extracapsular extension, or microscopically involved resection margins). Randomized after surgery.
Interventions and follow up
Arm A: Concurrent cisplatin 100 mg/m² IV on days 1, 22, and 43 + radiotherapy 60–66 Gy in 30–33 fractions over 6–6.6 week
Arm B: Radiotherapy alone, identical dose and schedule
Primary endpoint: Local-regional control
mFollow up: Median 45.9 months (primary analysis); 9.4 years (2012 long-term update)
Arm B: Radiotherapy alone, identical dose and schedule
Primary endpoint: Local-regional control
mFollow up: Median 45.9 months (primary analysis); 9.4 years (2012 long-term update)
Results
Local-regional control: HR 0.61 (95% CI 0.41–0.91), P=.01; 2-year rate 82% vs 72%
Disease-free survival: HR 0.78 (95% CI 0.61–0.99), P=.04
Overall survival: HR 0.84 (95% CI 0.65–1.09), P=.19 — not significant (primary analysis)
2012 long-term update (DFS): HR 0.78, P=.05; OS HR 0.84, P=.12 — no significant OS benefit at 9.4 years
Disease-free survival: HR 0.78 (95% CI 0.61–0.99), P=.04
Overall survival: HR 0.84 (95% CI 0.65–1.09), P=.19 — not significant (primary analysis)
2012 long-term update (DFS): HR 0.78, P=.05; OS HR 0.84, P=.12 — no significant OS benefit at 9.4 years
Adverse events
Main adverse events: Grade ≥3 AEs: 77% (CRT) vs 34% (RT alone), P<.001. Four treatment-related deaths in CRT arm. Mucositis grade ≥3 substantially higher in CRT arm. Late toxicities: comparable between groups at long-term follow-up.
Conclusions
Postoperative concurrent cisplatin + radiotherapy significantly improved locoregional control and DFS compared with radiotherapy alone in high-risk resected head and neck squamous cell carcinoma. No significant overall survival benefit was demonstrated. Together with EORTC 22931, this established adjuvant CRT as standard of care for patients with extracapsular extension and/or positive margins.
Key Limitations
Key Limitations: OS benefit not demonstrated despite improved locoregional control — raises questions about distant metastasis and competing causes of death. High-risk eligibility (2+ positive nodes, ECE, or positive margins) included heterogeneous populations; the combined EORTC/RTOG pooled analysis later clarified that only ECE and positive margins consistently predicted benefit. The 2012 long-term follow-up showed DFS and LRC benefit persisted but OS remained non-significant. Toxicity burden (77% grade ≥3) is substantial and must be weighed against individual patient fitness.
Clinical Context
Combined with EORTC 22931 (Bernier NEJM 2004), RTOG 9501 established concurrent cisplatin-RT as the standard adjuvant treatment for resected H&N cancer with high-risk features. The Bernier/Cooper pooled analysis (Head Neck 2005) clarified that benefit is most consistent for patients with extracapsular extension or positive margins — not merely multiple positive nodes. These two trials remain the cornerstone of post-operative H&N management guidelines (NCCN, ESMO).