Value of the pretransplant evaluation in predicting toxic day-100 mortality among blood stem-cell and bone marrow transplant recipients.

Abstract

PURPOSE To determine the value of pretransplant studies in predicting day 100 nonrelapse toxic mortality following high-dose therapy. PATIENTS AND METHODS A retrospective review of 383 consecutive hematopoietic stem-cell transplants was performed with attention to toxic mortality and pretransplant factors. Univariate log-rank analysis was used to yield the most significant cut-off values for individual factors. Multivariate analysis using Cox proportional hazards regression determined factors independently predictive of early toxic death. RESULTS Nonrelapse toxic mortality before day 100 occurred in 23 of 383 (6.0%) transplant recipients. Factors associated with an increased risk of toxic death by univariate analysis included forced expiratory volume in 1 second (FEV1) less than 78% of predicted (P = .0002), allogeneic versus autologous transplant (P = .0003), diffusion capacity of carbon monoxide less than 52% of predicted (P = .002), serum creatinine concentration greater than 1.1 mg/dL (P = .003), Eastern Cooperative Oncology Group performance status greater than 0 (P = .006), preparative regimen containing total-body irradiation versus chemotherapy alone (P = .006), marrow versus blood stem cell (P = .01), serum ALT greater than 50 IU/L (P = .02), diagnosis of hematologic disorder versus solid tumor (P = .06), serum bilirubin level greater than 1.1 mg/dL (P = .08), left ventricular ejection fraction (P = .09), and growth factor use (P = .09). In the multivariate model, transplant type (relative risk, 4.2), FEV1 (relative risk, 4.5), performance status (relative risk, 3.7), serum creatinine (relative risk, 3.8), and serum bilirubin (relative risk, 3.7) were found to be independent predictors of early toxic mortality. CONCLUSION The pretransplant evaluation is a useful tool to identify patients at risk for early toxic mortality following high-dose therapy.