Abstract
Transfusion of red blood cells (RBC) is a common intervention to treat and prevent complications in sickle cell disease (SCD). However, frequent transfusions may lead to erythrocyte alloimmunization, thereby complicating donor matching procedures and posing patients at risk for hemolytic transfusion reactions. Little information is available about the risk of alloimmunization of sickle cell patients living in European countries. In the Netherlands extensive matching procedures to prevent alloimmunization were introduced a decade ago, but the effect on alloimmunization has not been evaluated yet.
The primary aim of this study is to evaluate the cumulative incidence of first alloantibody formation in a Dutch cohort of transfused SCD patients, and to compare this with a general Dutch RBC-transfused population. In addition, the effect of extended RBC matching protocols on the incidence of alloantibody formation in SCD and potential clinical determinants of alloimmunization will be assessed.
We conducted a retrospective cohort study and collected data on SCD patients (genotypes HbSS, HbSC, HbSβ0 and HbSβ+ thalassemia), diagnosed in three Dutch Sickle Cell Treatment Centers that received non-extended matched (ABO, RhD) RBC transfusions between 1984-2004 and extended matched (at least ABO, Rhesus phenotype, Kell) RBC transfusions between 2004-2011.
In addition, we compared this population with a general population of 3 042 patients that received non-extended matched (ABO, RhD) RBC transfusions between 2005-2009 in the Leiden University Medical Center (Zalpuri et al. 2012). Cohorts were not matched for ethnicity. Alloimmunization risk was calculated as Kaplan-Meier incidence with cumulative number of transfusions as time variable. The association with the clinical determinants gender, SCD-phenotype and ethnicity was analyzed with Cox-regression analysis.
A total of 291 SCD patients received 7 957 RBC units. Alloantibody formation occurred in 52 (17.9%) patients. The cumulative incidence of alloimmunization was 9% after 5 RBC units, 15% after 10, 24% after 20 and 34% after 40 RBC units. Multivariate analysis, correcting for the cumulative number of transfusions, demonstrated a significantly increased risk of alloantibody formation in our SCD cohort when compared to a general population of transfused patients (HR 7.5 (95% CI: 5.06-11.14), where the cumulative incidence of alloimmunization was 1.1% after 5, 2.4% after 10, 3.4% after 20 and 6.5% after 40 RBC units. No association could be demonstrated between alloantibody formation and clinical determinants such as gender, SCD-phenotype or ethnicity. However, a significant reduction in alloimmunization was observed in SCD patients that received their first transfusion from the year 2004 onwards, after preventive matching for Rhesus phenotype and Kell was introduced for SCD patients (HR 0.48 (95% CI: 0.24-0.97)).
The overall rate of first RBC alloantibody formation in our cohort was 17.9% and the risk of alloimmunization increased substantially with an increasing number of RBC transfusions. A unique comparison with a general cohort of Dutch transfused patients demonstrates a significantly higher risk of alloantibody formation in SCD, acknowledging earlier findings. This may partially be explained by differences in RBC antigens between patients of African descent and the predominantly Caucasian donors. Besides the number of RBC units, no other clinical risk factors for allo-immunization in SCD could be identified. The effectiveness of extended RBC matching protocols in the prevention of alloimmunization for chronically transfused patients in the participating centers was confirmed.
Kaplan-Meier survival curve for RBC alloimmunization incidence according to the number of transfused units: A comparison between SCD patients that received their first transfusion ever before or after 2004, and a general transfused population (Zalpuri et al. 2012).
Kaplan-Meier survival curve for RBC alloimmunization incidence according to the number of transfused units: A comparison between SCD patients that received their first transfusion ever before or after 2004, and a general transfused population (Zalpuri et al. 2012).
No relevant conflicts of interest to declare.
