Background
Pooled retrospective comparative analysis of two independent Phase III RCTs — EORTC 22931 (Bernier, NEJM 2004) and RTOG 9501 (Cooper, NEJM 2004) — both comparing adjuvant concurrent cisplatin-based chemoradiotherapy vs radiotherapy alone after surgery for high-risk head and neck squamous cell carcinoma. Combined N=750 patients. Analysis performed to identify which risk features most predicted benefit from adding chemotherapy to post-operative RT.
Interventions and follow up
Arm A: Surgery + concurrent cisplatin + radiotherapy
Arm B: Surgery + radiotherapy alone
Primary endpoint: Identify risk factors predicting benefit from concurrent chemoradiotherapy in post-operative setting
mFollow up: Per individual trial
Arm B: Surgery + radiotherapy alone
Primary endpoint: Identify risk factors predicting benefit from concurrent chemoradiotherapy in post-operative setting
mFollow up: Per individual trial
Results
Key findings: Extracapsular extension (ECE) and/or microscopically involved surgical margins were the only risk factors for which the impact of CRT was significant in both trials. Trend in favor of CRT also seen for stage III–IV, perineural infiltration, vascular embolisms, and/or clinically enlarged level IV–V nodes from oral cavity/oropharynx tumors
2+ positive lymph nodes (without ECE): Did NOT show consistent benefit from addition of chemotherapy in this analysis
2+ positive lymph nodes (without ECE): Did NOT show consistent benefit from addition of chemotherapy in this analysis
Adverse events
Main adverse events: As per original individual trials (see EORTC 22931 and RTOG 9501 entries).
Conclusions
The pooled analysis clarified that microscopically involved resection margins and extracapsular spread from neck nodes are the strongest indications for post-operative concurrent chemoradiotherapy. Patients with ≥2 positive nodes without ECE may not derive benefit from adding chemotherapy to adjuvant RT.
Key Limitations
Key Limitations: Retrospective pooled analysis with different eligibility criteria across the two trials — EORTC included more advanced-stage patients and used slightly different RT doses; RTOG required ≥2 positive nodes as an entry criterion. The conclusion regarding ≥2 positive nodes is derived from subgroup analysis and should be interpreted cautiously. This analysis was not prospectively designed and cannot fully account for selection bias between institutional populations. Both trials used cisplatin 100 mg/m² — whether other regimens show equivalent benefit is unknown.
Clinical Context
This analysis defined the "high-risk" criteria most relevant for adjuvant CRT: positive margins and ECE. NCCN and ESMO now list these two features as the primary indications for post-operative concurrent chemoradiotherapy. Patients with only multiple positive nodes (without ECE) may be observed or treated with RT alone, based on this analysis. The paper is frequently cited to justify de-escalation of chemotherapy in post-operative H&N management.