Aim: Although 40-65% of myelodysplastic syndromes (MDS) patients are thrombocytopenic and require platelet transfusions, there is limited literature on the risk factors predictive of bleeding and the burden of immune mediated platelet refractoriness (PLT-R).

Objectives: To evaluate the prevalence of thrombocytopenia, incidence of bleeding events, platelet transfusion dependency (PLT-TD) and immune-mediated platelet refractoriness (PLT-R) in MDS patients.

Methods: A retrospective analysis of 754 MDS patients enrolled in the South Australian MDS (SA-MDS) registry was performed. Platelet counts <100, <50 and <20 (x109/L) were used to define mild, moderate and severe thrombocytopenia respectively. The severity of bleeding events was classified according to the International Society of Thrombosis and Haemostasis (ISTH) classification. PLT-TD was defined as transfusion of at least one unit of platelets each month for four consecutive months. All other patients were classified as transfusion independent (PLT-TI). Immune mediated PLT-R was defined if a patient had HLA-class I or HPA antibodies, poor platelet increments and required HLA-matched platelets. Medication history with regards to anticoagulation and/or antiplatelet therapy was also collected.

Results: At diagnosis, 332 (45%) patients had thrombocytopenia and 106 (14%) patients had moderate to severe thrombocytopenia.

During the study period, 249 bleeding events were recorded in 162 (21%) patients with a cumulative incidence of 33% (Fig 1A). Of the 249 bleeding events, 85 (34%) were major and 164 (66%) were clinically relevant minor bleeding. Notably, 16, 90 and 5 bleeding events were intracranial, gastrointestinal, intraocular respectively. While 41% (96/235) bleeding events occurred in the setting of moderate to severe thrombocytopenia, 21% and 38% (Fig 1B) of bleeding events occurred at platelet counts of >50-100 and >100x109/L respectively. Twenty-eight percent (69/249) bleeding events were associated with concomitant anticoagulation and/or antiplatelet therapy and importantly, platelets counts were >50x109/L and >100 x109/L in 57 (83%) and 46 (67%) at the time of bleeding events, respectively.

During the disease course, 393 (52%) patients required at least one unit of platelet transfusion. 106 (14%) patients were PLT-TD and had significantly poor overall survival (OS) compared to PLT-TI (26 vs 42 months, p<0.0001).

In total, 30/393 (7%) required HLA-matched platelet transfusions. 20/30 (66%) of PLT-R patients were female receiving disease modifying therapy (DMT). This was substantiated by cox regression analysis, demonstrating that females (HR=5.32, p=0.0006), older age (HR=0.97, p=0.028) and haemoglobin (Hb) at diagnosis (HR=1.03, p=0.009) were independent risk factors for PLT-R. Importantly, 20/76 (25%) female patients receiving platelets and DMT developed immune mediated PLT-R requiring HLA matched platelets.

Conclusions: In our cohort of MDS patients, cumulative incidence of bleeding is 33% and 59% of the bleeding events occurred at platelet counts >50X109/L. For all bleeding events that occurred while on anticoagulation and/or antiplatelet therapy, 83% events occurred with platelet counts >50 x 109/L. Therefore, guidelines for anticoagulation and/or antiplatelet therapy are required for MDS patients.

We also showed that development of PLT-TD is associated with poor OS. Importantly, 1 in 4 female MDS patients receiving platelets and DMT required HLA-matched platelets. Platelet transfusions practices should be optimised for these high-risk groups.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution