Background
Phase III, prospective randomized Intergroup trial (RTOG 9111). 520 patients with Stage III or IV glottic or supraglottic squamous cell carcinoma (T1–T4, N0–N3) enrolled. Three-arm design comparing induction cisplatin/fluorouracil followed by RT (PF→RT) vs concurrent cisplatin-RT vs RT alone for larynx preservation. Primary objective: determine the best non-surgical treatment to preserve the larynx. Median follow-up 10.8 years.
Interventions and follow up
Arm A: Induction PF (cisplatin 100 mg/m² + 5-FU 1000 mg/m²/day × 5d, q3wks × 3 cycles) followed by RT 70 Gy (responders to induction only)
Arm B: Concurrent cisplatin 100 mg/m² days 1, 22, 43 + RT 70 Gy
Arm C: RT alone 70 Gy
Primary endpoint: Laryngectomy-free survival (LFS) — composite of larynx preservation and survival
mFollow up: Median 10.8 year
Arm B: Concurrent cisplatin 100 mg/m² days 1, 22, 43 + RT 70 Gy
Arm C: RT alone 70 Gy
Primary endpoint: Laryngectomy-free survival (LFS) — composite of larynx preservation and survival
mFollow up: Median 10.8 year
Results
LFS (induction PF→RT vs RT alone): HR 0.75 (95% CI 0.59–0.95), P=.02
LFS (concurrent CRT vs RT alone): HR 0.78 (95% CI 0.78–0.98), P=.03
Larynx preservation rate (concurrent CRT vs induction PF→RT): HR 0.58 (95% CI 0.37–0.89), P=.005 — significantly better with concurrent CRT
Larynx preservation (concurrent vs RT alone): P<.001
Overall survival: No significant difference between arms; trend toward worse OS with concurrent vs induction (HR 1.25, P=.08)
Non-cancer-related deaths: Higher with concurrent CRT (30.8%) vs induction (20.8%) vs RT alone (16.9%)
LFS (concurrent CRT vs RT alone): HR 0.78 (95% CI 0.78–0.98), P=.03
Larynx preservation rate (concurrent CRT vs induction PF→RT): HR 0.58 (95% CI 0.37–0.89), P=.005 — significantly better with concurrent CRT
Larynx preservation (concurrent vs RT alone): P<.001
Overall survival: No significant difference between arms; trend toward worse OS with concurrent vs induction (HR 1.25, P=.08)
Non-cancer-related deaths: Higher with concurrent CRT (30.8%) vs induction (20.8%) vs RT alone (16.9%)
Adverse events
Main adverse events: Late effects: no statistically significant differences between arms. But non-disease deaths were higher in the concurrent CRT arm at 10+ years (30.8%), possibly reflecting cumulative treatment toxicity. Dysphagia-related late effects noted in all CRT-containing arms.
Conclusions
Concurrent cisplatin-RT provides the highest rate of larynx preservation compared with induction chemotherapy followed by RT or RT alone. Both chemotherapy-containing regimens improve LFS vs RT alone. However, OS was similar across all three arms and concurrent CRT was associated with more non-cancer deaths at long-term follow-up. Concurrent cisplatin-RT is the preferred organ-preservation strategy for laryngeal cancer.
Key Limitations
Key Limitations: No voice/swallowing functional outcomes reported — a critical limitation since larynx preservation without function is not equivalent to functional preservation. The increase in non-disease deaths with concurrent CRT at 10.8 years raises concerns about late treatment-related mortality. OS was not significantly different across arms, calling into question whether larynx preservation comes at a cost. T4 disease with cartilage invasion had poor outcomes regardless of treatment — these patients are now typically recommended laryngectomy. No comparison with TPF induction, which was not yet standard at time of RTOG 9111 design.
Clinical Context
RTOG 9111 is the landmark trial establishing concurrent cisplatin-RT as the standard organ-preservation approach for resectable advanced laryngeal cancer. It superseded the induction-then-RT approach from the VA Larynx Trial and EORTC 24891. Current practice: concurrent cisplatin (100 mg/m² q3w × 3 doses) + RT 70 Gy for larynx preservation in stage III–IVA laryngeal cancer. T4a with cartilage invasion or large-volume T4b: generally laryngectomy recommended. Functional outcomes and swallowing-related quality of life require ongoing assessment — a preserved larynx that does not function is not therapeutic success.
References