Background
Phase 3 RCT. 396 patients with stage IIIA(N2) NSCLC. Enrolled 1994–2001 at SWOG/RTOG/ECOG/NCCTG institutions. Eligible if radiographically involved N2 nodes, no malignant pleural effusion, and ≤10% weight loss.
Interventions and follow up
Arm A: Cisplatin 50 mg/m² days 1, 8, 29, 36 + etoposide 50 mg/m² days 1–5, 29–33 concurrent with RT 45 Gy → thoracotomy (if stable/responding) → cisplatin/etoposide × 2 more cycle
Arm B: Same induction cisplatin/etoposide + RT 61 Gy (definitive, no surgery) → cisplatin/etoposide × 2 more cycle
Primary endpoint: OS
mFollow up: Median 22.5 month
Arm B: Same induction cisplatin/etoposide + RT 61 Gy (definitive, no surgery) → cisplatin/etoposide × 2 more cycle
Primary endpoint: OS
mFollow up: Median 22.5 month
Results
OS: 23.6 vs 22.2 months, HR 0.87 (95% CI 0.70–1.10), P=.24 — not significant
PFS: 12.8 vs 10.5 months, HR 0.77 (95% CI 0.62–0.96), P=.017
5-yr OS: 27% (surgery) vs 20% (definitive CRT)
Lobectomy subgroup OS: HR 0.63 (favoring surgery)
Pneumonectomy subgroup OS: HR 1.26 (favoring definitive CRT)
PFS: 12.8 vs 10.5 months, HR 0.77 (95% CI 0.62–0.96), P=.017
5-yr OS: 27% (surgery) vs 20% (definitive CRT)
Lobectomy subgroup OS: HR 0.63 (favoring surgery)
Pneumonectomy subgroup OS: HR 1.26 (favoring definitive CRT)
Adverse events
Main adverse events: Treatment-related mortality: 8% surgery arm vs 2% definitive CRT arm. Pneumonectomy mortality: 26% (7/27 patients requiring pneumonectomy died). Grade ≥3 esophagitis: comparable (~22% both arms, slightly higher in definitive RT arm with 61 Gy). Pulmonary morbidity higher in surgery arm.
Conclusions
Adding surgery to concurrent chemoRT did not improve OS vs definitive CRT alone in stage IIIA(N2) NSCLC (HR 0.87, p=0.24). Modest PFS improvement was offset by high surgical mortality, particularly among pneumonectomy patients. The trial effectively discouraged routine trimodality therapy for N2 disease, though lobectomy patients appeared to benefit.
Key Limitations
Key Limitations: Unexpectedly high pneumonectomy mortality (26%) substantially drove the null OS result — a rate that would not be acceptable at high-volume thoracic centers today. The pneumonectomy:lobectomy ratio was unfavorable. Modern surgical techniques may achieve lower perioperative mortality. The definitive RT dose of 61 Gy is suboptimal by current standards (RTOG 0617 established 60 Gy as standard). Stage IIIA(N2) is heterogeneous (single vs multi-station N2); separate analyses were not performed. Enrolled 1994–2001 — pre-PET, pre-EBUS staging era, with likely nodal staging inaccuracies.
Clinical Context
Intergroup 0139 effectively established definitive concurrent CRT as standard for unresectable stage IIIA(N2) NSCLC, a position reinforced by PACIFIC (durvalumab consolidation). However, highly selected patients with minimal N2 disease where lobectomy (not pneumonectomy) is feasible may still benefit from trimodality therapy at experienced centers. NCCN guidelines support surgical consideration for carefully selected single-station N2 NSCLC after multidisciplinary evaluation.
References
References: Albain KS et al, Lancet 2009 (Intergroup 0139)