Abstract

Superiority of SDP over RDP for PLT transfusions is largely assumed, but unproven. We hypothesized that SDP and RDP used for prophylactic transfusions are clinically equivalent.

Methods: In order to minimize the biases imposed by a heterogeneous population, we studied 33 patients (median age 47 years, range, 22–66), which underwent alloHSCT for AML or high-risk MDS with ABO-identical donors. All received intravenous busulfan (BU)-based myeloablative conditioning regimens: 30 patients received BU 130mg/m2 and fludarabine 40mg/m2 for 4 days, and 4 patients were given BU 0.8mg/kg every 6 hours for 4 days with cyclophosphamide 60mg/kg for 2 days. GvHD prophylaxis consisted of tacrolimus and mini-methotrexate. Recipients of unrelated (UD) or one antigen-mismatch related donor transplants received antithymocyte globulin (ATG). Source of stem cells was peripheral blood (n=23) and bone marrow (n=10). Donors were related (n=28) and UD (n=5). Only 3 patients were identified as alloimunized for RBC antigens before tranplant. Transfusions administered for active bleeding were excluded from this analysis. Prophylactic PLT transfusions were administered for PLT counts below 15 x 10e9/l. The corrected increment (CCI) was calculated according to standard formula, using PLT counts performed in samples collected following transfusions. As all patients received both types of products, each transfusion administered during the first 100 days after alloHSCT was evaluated individually. The RDP and SDP were labeled as ’ineffective’, according to the CCI: cut-offs of 5.5 for the samples collected in the first hour post transfusion, 5.0 for the samples between 1 and 18 hours and 4.5 for the samples collected 18 hours or later.

Results: Median time to PLT engraftment was 13 days (range, 8–51, n=30). Patients received 105 RDP and 41 SDP transfusions. All RDP transfusions were ABO compatible, while 4 out of the 41 SDP (9.75%) were ABO-incompatible (p = 0.006). Median PLT and Hb level before transfusion were similar for both SDP and RDP. Frequency of amphotericin and vancomycin use, and presence of fever on transfusion day was similar on both groups. Median post-transfusion platelet count was 51K/μL (range, 5K-118K) after SDP and 36K/μL (range, 3K–115K) after RDP (p = 0.0004). Median CCI was 14.178 for SDP transfusions versus 7.793 after RDP (p = 0.0001). The difference in post-transfusion PLT counts and CCI favoring the SDP group did not translate into superior clinical efficacy. Median time to another PLT transfusion was 3 days (range, 0–10) after SDP and 2 days (range, 0–10) after RDP (p = 0.3). The median number of new transfusions needed in the following week was similar for both groups (2 transfusions, range 0–7). In the week following any given transfusion, 24% versus 16% of the patients receiving SDP versus RDP did not need other transfusion (p = 0.2); 17% of SDP and 30% of RDP transfusions were labeled ’ineffective’ according to CCI (p = 0.1). There were two episodes (6%) of hemorrhage during the study period (none lethal): gastro-intestinal and hemorrhagic cystitis.

Conclusion: SDP transfusions produced a greater increment in PLT counts after transfusion, but SDP and RDP were equally effective in preventing hemorrhage.

Disclosure: No relevant conflicts of interest to declare.

Author notes

*

Corresponding author