Clinical meta-analyses of targeted therapies in adenocarcinoma

Although the interpretation of the data reported in meta-analyses may hide several issues, it is undoubtable that this methodological approach may significantly contribute to implement the results of clinical trials, and represent a useful and practical tool for the evidence-based medicine process. Indeed, level-one recommendations should consider well-conducted meta-analyses as well as large and adequately powered randomized trials as the main contributors for the definition of guidelines for clinical practice. In addition, the role of meta-analyses for issues whereas conflicting data (and/or unpowered results) are provided, is well established. In the field of lung cancer, meta-analyses already participated to change the current standard, and are now facing the challenging issues of predictive biomarkers of prognosis and/or efficacy of targeted agents. With this aim, the meta-analytic approach helped in the recent years to implement the quantification of the magnitude of the benefit of targeted agents, and added new insights by interpreting the data coming from clinical trials by integrating them with biomarkers. The treatment-interaction analyses according to putative predictive factors of efficacy may clarify unknown issues and generate new hypotheses for future perspectives. The current review attempts to put in the context of the clinical data of targeted agents for lung cancer all the pros and cons of the meta-analytic process published to date, and critically analyze all the potential perspectives which this methodology may add for both current practice and forthcoming research. © 2013 Springer-Verlag France.