Benefits and risks of using erythropoiesis-stimulating agents (ESAs) in lung cancer patients: study-level and patient-level meta-analyses.

In anemic patients receiving myelosuppressive chemotherapy, erythropoiesis-stimulating agents (ESAs) raise hemoglobin levels and reduce transfusion requirements, but ESA-related safety concerns exist. To evaluate ESA benefits and risks in lung cancer, we conducted meta-analyses of data from controlled ESA trials conducted in lung cancer patients. Study-level analyses included controlled ESA trials reporting lung cancer mortality, identified from the 2006 Cochrane ESA report and from a systematic search for studies published through December 2010. Patient-level analyses included data from lung cancer patients receiving chemotherapy in Amgen studies evaluating darbepoetin alfa (DA) vs placebo. Study-level and patient-level analyses examined deaths, progression, and transfusion incidence. Patient-level analyses also examined adverse events (AEs) and fatigue. In a study-level meta-analysis of nine ESA studies of 2342 patients receiving chemotherapy, the ESA odds ratio (OR) was 0.87 (95% confidence interval [CI] 0.69-1.09) for mortality; the overall random-effects risk difference (95% CI) for mortality was -0.02 (-0.06, 0.02). The ESA OR (95% CI) for disease progression in five chemotherapy studies reporting progression was 0.84 (0.65-1.09). The ESA odds ratio (95% CI) was 0.34 (0.28-0.41) for transfusion incidence. In a patient-level meta-analysis of four studies evaluating 1009 patients through follow-up, the median survival time was 41 weeks with DA and 38 weeks with placebo. During the combined study and follow-up periods, 80% of placebo-group patients and 74% of DA patients died (mortality hazard ratio [HR] 0.90 [95% CI, 0.78-1.03] for DA); results were similar for small cell lung cancer and non-small cell lung cancer. Overall, 87% of placebo patients and 84% of DA patients progressed or died. Fewer DA patients had transfusions (week 5 through end-of-study, DA 19%, placebo 43%). AEs included thrombotic/embolic events (DA 10.5%, placebo 7.2%), cerebrovascular disorders (DA 3.7%, placebo 4.2%), pulmonary edema (DA 0.4%, placebo 1.0%) and pulmonary embolism (DA 1.8%, placebo 0.6%). These meta-analyses suggest that ESAs reduce transfusions without increasing mortality or disease progression in lung cancer patients undergoing chemotherapy.