Higher Rate of Red Blood Cell (RBC) Alloimmunization in RBC-Transfused Myelodysplastic Syndrome Compared to Acute Myeloid Leukemia

Background: Up to 90% of patients with myelodysplastic syndrome (MDS) and acute myeloid leukemia require red blood cell (RBC) transfusions and 11–14% of RBC-transfused MDS patients develop RBC alloantibodies (Singhal et al., Haematologica 2017; Chhetri et al., Haematologica 2019), most commonly against Rhesus (Rh) and Kell (K).

Method: Myeloid neoplasm (MN; n=1961) patients enrolled in the South Australian Myeloid Neoplasm registry with complete data to a censor date of January 1, 2023, were included for further analysis. We first characterized the RBC transfusion burden and alloimmunization rate in RBC-transfused patients. Then, we assessed the risk factors associated with RBC alloimmunization and its impact on transfusion requirements.

Results: Our study of 1961 MN patients included MDS (n=1077, 54.9%), AML (n=644; 32.8%), MDS and myeloproliferative neoplasm (MPN) overlap syndrome (n=222, 11.3%) and related MN/bone marrow failure syndrome (n=18; 0.9%). Although the majority of MN patients (n=1660, 84.7%) were de novo, 15.3% (n=301) had therapy-related MN (t-MN) following prior exposure to cytotoxic therapies for independent primary cancers. The median age at MN diagnosis was 69.2 years (interquartile range 58.6, 78.4) and 63.2% (n=1240) of patients were male.

The majority of patients (n=1452, 74.0%) had died by the censor date. The overall survival (OS) of AML was significantly shorter than MDS and MDS-MPN overlap cases (10.8 vs. 36.3 vs. 30.9 months; P <0.0001). The median OS of t-MDS (14.2 vs. 42.0; P < 0.0001) and t-AML (6.9 vs. 11.7 months; P <0.0001) was significantly worse than their de novo counterparts.

After MN diagnosis, 1732 (88.3%) patients received at least one RBC unit and 11% (n=191) of these patients developed 264 alloantibodies. The most common alloantibodies were E (n=69, 45.1%), K (n=37, 24.2%), C (n=25, 16.3%), D (n=23, 15.0%), and Jka (n=22, 14.4%). The rate of alloimmunization was significantly higher in RBC-transfused MDS than AML (12.3% vs. 3.6%; P < 0.001), despite similar rates of antibody screens between the two groups (42.9% vs. 46.2%; P = 0.062), which could explain the higher RBC transfusion requirement in MDS compared to AML patients (40.75 vs. 30.40 units; P < 0.001). Higher alloimmunization in MDS compared to AML is probably related to the highly myelosuppressive treatment used in AML. Furthermore, the RBC alloimmunization rate was significantly higher in t-MN compared to de novo (12.5% vs. 8.1%; P =0.023), despite receiving a comparable number of RBCs (37.0 vs. 35.8 units; P = 0.7). Although the RBC alloimmunization rate was higher in females compared to males, it was not statistically significant (10.5% vs. 7.8%; P =0.06).

Next, we assessed the clinical consequences of RBC alloimmunization. RBC autoantibodies were detected in 16 MN patients and alloimmunized patients were significantly more likely to develop autoantibodies (7.2% vs. 0.3%; P < 0.001). The RBC transfusion burden in alloimmunized patients was significantly higher compared to non-alloimmunized patients (72.1±77.2 vs. 32.9±50.0 units; P<0.001). In RBC-alloimmunized patients, the RBC transfusion burden significantly increased after alloantibody formation compared to before the alloimmunization (22.5±42.0 vs. 50.1±57.2 units; P <0.0001).

Conclusion: To our knowledge, this is the largest cohort characterizing RBC alloimmunization in MN (n~2000). The key findings of the study include the high burden of RBC alloimmunization in RBC-transfused MDS compared to AML. Antibodies against Rh and Kell were most prevalent and RBC-alloimmunization is associated with increased autoantibody formation. Importantly, RBC alloimmunization is associated with increased transfusion burden. This study provides compelling evidence of extended phenotyping to provide Rh and Kell phenotype-matched RBC transfusions to MDS patients.