Responses of smoking and nonsmoking cancer patients to drug treatment: A protocol for systematic review and meta-analysis.

Introduction: Smoking is well-known to increase cancer risk, particularly risk of lung cancer, and negatively affects efficacy of cancer treatment. However, recent evidence suggests that among cancer patients, paradoxically, smokers respond to treatment better than non-smokers. We propose to conduct a focused review and meta-analysis to compare response to drug treatment between smoking and non-smoking cancer patients. Methods and design: We will collect data from large clinical trials of therapies for cancer patients which have included smokers and non-smokers. We will search PubMed, PMC/ MEDLINE, SCOPUS, Embase, and the registries for clinical trials and four major clinical journals up to June 30, 2019. Search terms will be "Drug name" phase-3" or "Drug name" phase-III." Data collection will be focused on randomized clinical trials of cancer drugs that enrolled at least 100 participants and reporting treatment results from smoking and nonsmoking patients. Initial selection criteria will be clinical trial studies of drug treatment of 100 or more cancer patients, and reporting hazard ratios (HR) for smokers and non-smokers. Two persons will be searching such publications independently, or data will be provided, double checked, or confirmed by authors. Multiple sub-group analyses will be conducted by at least two persons to avoid bias or experimental errors. Discussion: The results will clarify whether smoking and response to treatment of cancer are linked not. Our results may possibly identify drug/s that work better among cancer patients who are smokers.