OD-03. Red blood cell transfusion dosing for chronic anemia in children of different ages receiving anticancer therapy
Polina S. Kuga, Evgeny A. Bakin, Maxim A. Kucher, Boris A. Baryshev, Ludmila S. Zubarovskaya, Alexander D. Kulagin
RM Gorbacheva Research Institute, Pavlov University, St. Petersburg, Russia
Dr. Ludmila S. Zubarovskaya, phone: +7 (921) 973-14-31, e-mail: email@example.com
Red blood cell (RBC) transfusion is an important supportive therapy modality for children with different malignancies during diagnostic workup, chemotherapy or hematopoietic stem cell transplantation (HSCT). There is no professional community consensus towards RBC dosing and single transfusion effectiveness assessment in pediatric patients with chronic anemia. Therefore, the situations associated with high risk of post-transfusion reactions and volemic disorders are still possible. Our purpose was to construct a predictive model for hemoglobin (Hb) level increase after single RBC infusion suitable for children of different ages with chronic anemia due to cancer or anticancer treatment received.
Materials and methods
The data on 753 RBC transfusions performed for 175 patients admitted to RM Gorbacheva Research Institute in December 2018 to September 2020 for diagnostic workup, chemotherapy or HSCT were included in the analysis. With a median age of 4 (3 months-11years) years these patients weighted from 5 to 30 kg. The male to female ratio was 1.65:1 (109 boys and 66 girls). All patients were transfused leukoreduced irradiated RBCs due to moderate to severe chronic anemia without signs of hemorrhagic syndrome. The data was processed using regression model based on the LASSO method. Age, sex, diagnosis, therapy modality, presence of hepato/splenomegaly, transfusion medium shelf life, BMI, BSA, temperature, initial haemoglobin level, and volume of transfusion related to body weight were chosen as the initial set of independent variables. The hemoglobin increase value was a dependent variable. The maximal value providing an error that exceeded the minimum value by no more than one standard deviation was set as a regulating value. Thus, the final model included age, presence of MDS or JMML, presence of hepatosplenomegaly, excess body weight, and febrile fever presence. The root square error of prediction based on cross-validation results was 5.8 g/L. The calculation was carried out using the glmnet package of the R programming language, version 3.7.
There were no post-transfusion reactions and complications registered in our patients. In infants the excess (more than 31 g/l) increase in the Hb level was registered in 23.5% cases (16 transfusions). In 10 (62.5%) of these children there was a subsequent GFR decrease by up to 22% compared to pre-transfusion value. In children older than 1 year the excess increase in Hb level was evident in 11.8% cases (81 transfusions). In this group the GFR decrease incidence was 49.4% (40 transfusions) with maximal decrease of up to 30.9% compared to baseline value. This data suggest that the safety window for Hb level increase after the single RBC transfusion is 10 to 30 g/ll. Currently, a prognostic model for hemoglobin increase after RBC transfusion in children is being constructed. By the time of this publication we included 18 patients in the prospective part of the study (31 transfusions), and the online dose calculator interface was created. Using the preliminary calculated data, it is possible to predict the hemoglobin level dynamics after transfusion, and to increase the safety of treatment.
The preliminary data demonstrates the feasibility of optimal RBC volume calculation. Conversion of program interface created into the freely available smartphone and PC application may help to predict the post-transfusion Hb increment making the whole supportive therapy for children with cancer more effective and safe.
Anticancer therapy, erythrocyte transfusion, children, predictive model.