Locally advanced squamous cell carcinoma of the head and neck: A systematic review and Bayesian network meta-analysis of the currently available treatment options

Background: There are still many unresolved questions in the management of locally advanced Head and Neck Cancer (HNC). Many chemotherapeutic drugs and radiotherapy fractionation schemes are available and not all have been evaluated in head-to-head clinical trials. This systematic review and Bayesian network meta-analysis aims to compare the available treatment strategies and chemotherapeutic options for locally advanced HNC. Methods: We performed a search on bibliography databases, trials registries and meetings proceedings for published and unpublished randomized trials from January 1st 2000 to December 1st 2017. Trials had to compare systemic interventions and radiotherapy (RT) approaches for locally advanced, non-metastatic HNC. Trials recruiting patients whose surgery was the first treatment option, sample size less than 20 per arm or that did not use randomization for treatment allocation were excluded from the analysis. Summary estimates on Overall survival (OS), Progression-free survival (PFS) and toxicity outcomes (grade 3–4 mucositis and neutropenia) were extracted from the included studies on a predefined database sheet. Bias was assessed through the Chocrane risk of bias assessment tool. We performed a set of pair-wise meta-analyses using a random effect model. We also performed a random effect network meta-analysis under a Bayesian framework. Findings: From the 57 included trials, including 15,723 patients, was possible to conduct analysis on 26 treatments for OS, 22 treatments for PFS and 10 treatments for toxicity. In terms of OS Concurrent chemoradiotherapy (CCRT) with cisplatin (HR 0.70, 95% CrI [credible interval] 0.62–0.78) and cetuximab on top of CCRT (HR 0.7, 95% CrI 0.5–0.97) are clearly superior to conventional RT alone. Induction chemotherapy (IC) with cisplatin and fluorouracil (HR 0.74, 95% CrI 0.52–0.95), IC with docetaxel, cisplatin, fluorouracil (HR 0.55, 95% CrI 0.54–0.89) and IC with paclitaxel, cisplatin, fluorouracil (HR 0.55, 95% CrI 0.34–0.89) before CCRT are all superior to conventional RT. CCRT with cisplatin is also superior to altered fractionation RT (HR 0.74, 95% CrI 0.64–0.84). Altered fractionation RT is not superior to conventional RT (HR 0.95, 95% CrI 0.85–1.06). Regarding PFS, CCRT with cisplatin (HR 0.72, 95% CrI 0.63–0.83), cisplatin and fluorouracil (HR 0.67, 95% CrI 0.5–0.88), carboplatin (HR 0.63, 95% CrI 0.46–0.87), carboplatin and fluorouracil (HR 0.75, 95% CrI 0.56–1), IC with cisplatin and fluorouracil (HR 0.59, 95% CrI 0.45–0.78), IC with docetaxel, cisplatin and fluorouracil (HR 0.53, 95% CrI 0.41–0.68) and IC with paclitaxel, cisplatin and fluorouracil (HR 0.59, 95% CrI 0.35–0.99) are superior to conventional RT and altered fractionation RT. IC with docetaxel, cisplatin and fluorouracil shows a significant superiority against CCRT with cisplatin (HR 0.73 95% CrI 0.58–0.92). Altered fractionation RT is not superior to conventional RT (HR 0.91, 95% CrI 0.81–1.02). Altered fractionation increases the risk of developing grade 3–4 mucositis compared to conventional RT (OR 3.74 95% 1.64–8.67) Interpretation: CCRT with cisplatin remains the gold standard of treatment. Taxane based IC regimens may have a impact on locally advanced disease. Altered fractionation RT is inferior to CCRT and also does not seem to be meaningfully better than conventionally fractionated RT alone. Its role in locally advanced disease should be reevaluated.