Modeling hemoglobin levels of patients chronically transfused with red blood cells.

Abstract

Good models of chronic red blood cell (RBC) transfusion can improve care by accurately estimating time-dependent hemoglobin (Hb) levels and clinically relevant transfusion parameters. A previously derived mathematical model based on overall Hb mass balance (HMB model) with input parameters of transfused units, transfusion efficiency, RBC lifespan, endogenous Hb, and transfusion interval was applied to three separate clinical studies spanning six transfusion conditions across patient populations with myelodysplastic syndrome (MDS) or thalassemia. The HMB model accurately predicted mean pretransfusion Hb levels for each ensemble cohort of patients with thalassemia or MDS. Dynamic changes in Hb levels were modeled as a function of changes in key input parameters. Improving the 24-h post-transfusion RBC survival from 72% to 86% can be used to either (1) reduce RBC usage by 15%-20% through longer transfusion intervals or (2) increase pretransfusion [Hb] by 8%-11% while maintaining a constant transfusion interval. The endogenous Hb level is introduced in the HMB model which represents the patient's self-contribution to overall Hb levels via the autologous RBC lifecycle and was estimated to be 5.0 g/dL for patients with MDS or thalassemia. Transfusion therapy approaches and complementary therapeutics can target multiple, unique model inputs while monitoring net, overall impact on transfusion efficacy. Applying the HMB model to fit individual patient Hb fluctuations will be explored in the future.