Purpose: The combination of docetaxel and gemcitabine was tested in several studies in patients with lung, breast, and pancreatic cancers and other tumor entities. Some studies reported cases of severe or even fatal pulmonary toxicity that led to early termination of some trials. We created a meta-analysis model of published studies to identify explanatory factors for docetaxel-gemcitabine-dependent pulmonary toxicity.
Methods: We searched MEDLINE/Pubmed, EMBASE, and Cochrane Clinical Trials database for prospective full-text studies that used a schedule of docetaxel and gemcitabine to treat a malignant disease. We performed a meta-analysis for proportions using the arcsine transformation and a meta-regression using a generalized linear mixed model based on a binomial distribution and a logit link.
Results: We included 103 trials with 113 treatment arms comprising 5,065 patients (major entities included non-small cell lung cancer (n = 2,550), breast cancer (n = 1,119), pancreatic cancer (n = 466), and urothelial cancer (n = 161)). For the incidence of severe lung toxicity (common toxicity criteria [CTC] grades 3-5), we found a combined estimate of 2.70% (95% CI 2.26, 3.14). The estimate for the proportion of fatal cases was 0.35% (95% CI 0.21, 0.58). We found that the sequence of the chemotherapy schedule had no influence on the incidence of severe pulmonary adverse events (F-test F = 0.65, df = 3,113, P = 0.58) nor did the study phase, treatment line or ethnicity of the participants. We found that patients with breast cancer, compared to lung cancer patients, developed severe lung toxicity less frequently (OR = 0.18, 95% CI (0.09, 0.36)).
Conclusion: We could not demonstrate that a particular chemotherapy sequence of docetaxel-gemcitabine is associated with excess pulmonary toxicity. Patients with lung cancer are at a higher risk for severe pulmonary side effects with docetaxel-gemcitabine than are patients with breast cancer.