Meta-Analysis of Induction Chemotherapy as a Selection Marker for Chemoradiation in the Head and Neck

Objective: Many trials incorporate induction chemotherapy (IC) in selecting for organ preservation in head and neck squamous cell carcinomas (HNSCC). However, few studies examine IC response in predicting for chemoradiation therapy (CRT) response. This meta-analysis aims to determine the predictive accuracy of IC for subsequent response to CRT and overall survival (OS). Data sources: Medline, EMBASE, Cochrane register. Methods: A systematic search identified studies from database inception to October 2016 that used IC prior to CRT as definitive treatment for advanced HNSCC. The sensitivities and specificities of IC response predicting for complete CRT response were calculated, and the results were pooled in a summary receiver operating curve. One-, 2- and 5-year OS data were extracted. Results: Seven studies (n = 423 patients) were analyzed for response and six (n = 439) for OS. Pooled median sensitivity and specificity of IC response predicting CRT response were 0.95 (95% confidence interval [CI]: 0.72-0.98) and 0.43 (95% CI: 0.00-0.61), respectively. Patients were more likely to respond to CRT given previous response to IC (positive likelihood ratio = 1.6; 95% CI: 1.21-2.11) and less likely to respond to CRT if they failed to respond to IC (negative likelihood ratio = 0.16; 95% CI: 0.07-0.38). At 2 years, good response to IC was a statistically significant prognostic marker with a risk ratio of 1.35 (95% CI: 1.12-1.64). Conclusion: Our data suggests that patients with poor IC response will have poorer response to CRT and should be directed to other modalities. In contrast, good IC response does not guarantee a favorable outcome to CRT; however, because these patients are likely to have better prognoses, they should be offered salvage therapies of curative intent despite treatment failure.