Phase III randomized trial comparing weekly versus three-weekly (W3W) cisplatin in patients receiving chemoradiation for locally advanced head and neck cancer.

Abstract

6007 Background: Chemoradiation (CRT) with cisplatin 100mg/m2 given 3-weekly is the standard of care in locally advanced head and neck squamous cell cancer (HNSCC). Substituting low dose weekly cisplatin has pharmacological rationale due to lower toxicity and enhanced radiosensitization but has never been compared to 3-weekly cisplatin. Methods: Phase III non-inferiority trial (CTRI/2012/10/003062) in patients with Stage III or IV (non-metastatic) HNSCC planned for radical CRT. Pts were stratified for T-category, N-category and intent of therapy (adjuvant versus definitive) and centrally randomized 1:1 to cisplatin 30mg/m2 weekly or 100mg/m2 3-weekly concurrently with radiotherapy. Primary endpoint was locoregional control (LRC); secondary endpoints included toxicity, compliance, response, progression-free survival and overall survival. The upper boundary of non-inferiority margin was set at 15%, assuming LRC of 60% with power 80% and alpha 0.05. Results: 300 pts were randomized,150 to each arm; median age 44 years (range: 25-67), males 89%; 71% were smokeless tobacco users. 61% were T4, 71% > N2. 93% pts received CRT as adjuvant therapy for high-risk disease; indication was perinodal extension in 84%, close/ positive margins 8%. Median total treatment time was 86 days (IQR: 79-95). Median RT dose was 60 Gy (IQR 60-60 Gy) using shrinking field technique. In the weekly arm, 133 pts (88.7%) received > 6 cycles; 14 (9%) required dose reduction, median cumulative cisplatin dose 210 mg/m2 (IQR 180-210). In the 3-weekly arm, 143 pts (95%) received > 2 cycles, median cumulative cisplatin dose 300 mg/m2 (IQR 200-300); 12 (8%) required dose reduction. At a median follow up of 20 months (range: 1-49), locoregional relapses (LRR) occurred in 42.2% pts in the weekly arm and 29.6% pts in the 3-weekly arm, leading to an absolute difference in LRR of 12.7% (95%CI:1.89-23.41), p=0.035 by Gray’s test, HR-1.58 (95%CI:1.02-2.46). Acute > grade 3 toxicity occurred in 85.3% pts in 3-weekly arm and 70.7% pts in weekly arm, p=0.002; 30.7% pts in the 3-weekly arm and 14% patients in the weekly arm required hospitalization for management of toxicity, p=0.001. Conclusions: 3-weekly cisplatin leads to 42% relative reduction in locoregional recurrence; it is superior to weekly cisplatin and should be the preferred regimen in CRT for HNSCC. Clinical trial information: CTRI/2012/10/003062.