Drug Dose Per Kilogram Lean Body Mass Predicts Hematologic Toxicity From Carboplatin-Doublet Chemotherapy in Advanced Non-Small-Cell Lung Cancer

Variations in lean body mass (LBM) are proposed to contribute to interindividual differences in toxicity from cancer drugs. In advanced non-small-cell lung cancer patients receiving carboplatin-doublets, dose per kilogram of LBM of the nonplatinum drug dosed by body surface area was a significant independent predictor of grade 3/4 hematologic toxicity. Taking body composition into account may improve dose individualization of chemotherapeutic agents. Background: Variations in lean body mass (LBM) have been suggested to explain variations in toxicity from systemic cancer treatment. We investigated if drug doses per kilogram of LBM were associated with severe hematologic toxicity (HT) in patients with stage IIIB/IV non-small-cell lung cancer (NSCLC) enrolled onto randomized trials comparing first-line carboplatin-doublets. Patients and Methods: Patients received carboplatin (AUC [area under the plasma concentration vs. time curve] = 5) plus either pemetrexed 500 mg/m2, gemcitabine 1000 mg/m2, or vinorelbine 60 mg/m(2). LBM was estimated from the cross-sectional muscle area at the third lumbar vertebra on computed tomographic scans. Administered doses on day 1, first cycle, were recalculated as milligram of drug per kilogram of LBM. Primary outcome was Common Terminology Criteria for Adverse Events version 3.0 grade 3/4 HT after cycle 1. Results: Data from 424 patients were analyzed. Mean age was 65 years, 57% were men, and 78% had stage IV disease. Despite dose individualization by body surface area for the nonplatinum drugs, mean (range) doses expressed as mg/kg LBM showed similar to 3-fold range: gemcitabine 38.0 (22.5-61.7) mg/kg LBM, pemetrexed 19.1 (8.1-27.9) mg/kg LBM, and vinorelbine 2.4 (1.4-3.6) mg/kg LBM. For these drugs, dose per kilogram of LBM was associated with HT in adjusted multivariate models (P = .004). Taking mean dose per kilogram LBM for each drug as reference, a 1% increase (odds ratio [OR] = 1.03; 95% confidence interval [CI], 1.01-1.06) or 1% decrease (OR = 0.97; 95% CI, 0.95-0.99) was associated with altered risk of grade 3/4 HT. For doses > 20% above and below mean (14% and 15% of patients, respectively) the risk of grade 3/4 HT was almost doubled (OR = 1.93, 95% CI, 1.21-3.10) and halved (OR = 0.52; 95% CI, 0.32-0.83), respectively. Conclusion: Dose per kilogram of LBM varied considerably and was an independent predictor of HT. Computed tomography-defined LBM may provide a future basis for better dose individualization. (C) 2016 Elsevier Inc. All rights reserved.