Continuous Platelet Transfusion As a Potential Desensitization Regimen in HLA Class I Alloimmune-Mediated Platelet Refractoriness

Background: The development of HLA antibodies is responsible for 4-8% of platelet transfusion refractoriness (PTR). The presence of HLA antibodies is a clinical complication that is generally managed by the selection of products that are negative for the antigens cognate to the patient's antibodies. Often, this selection is facilitated by targeted recruitment of donors with known HLA types. However, for very broadly HLA-alloimmunized patients, compatible products may be exceedingly scarce or completely unavailable, precluding the ability to consistently provide products the patient will likely increment to and benefit from. Here we describe a case of a patient with severe aplastic anemia (platelet count chronically <1 K/uL) and platelet transfusion refractoriness secondary to broad HLA-alloimmunization (98% PRA). HLA compatible products were unavailable due to the patient's rare HLA type and breadth of antibodies, and attempts at procuring compatible products via multiple national agency searches were unsuccessful. Thus, the patient could only be supplied with HLA incompatible products, which yielded insufficient post-transfusion platelet corrected-count increments (CCI). Her course was eventually complicated by multifocal intracranial hemorrhages, necessitating ICU admission with aggressive platelet transfusion support, immune-targeting therapy consisting of IVIg, and aminocaproic acid and recombinant activated factor VIIa.

Methods: A continuous platelet infusion, or "platelet drip," was employed in order to provide platelet transfusion support in the face of her severe thrombocytopenia with profound refractoriness. This consisted of consecutive transfusions of 4-6 equivalent unit apheresis platelet products ("doses") transfused over 4 hours with a 15 minute post-transfusion CBC obtained from Q8 to Q12 hours to assess platelet increments. This ultimately resulted in 24-30 equivalent unit platelet product infusions/day.

Results: Two days into the continuous platelet infusion, the patient began incrementing appropriately to occasional platelet products (CCI>>5) and, consequently, the infusions were able to sustain a platelet count >10K/uL, with a range of 20-70 K/uL, over the subsequent week. The improved counts, along with the patient's improved clinical condition and stable imaging findings showing resolving intracranial hemorrhages, allowed the transfusion support to be goal-directed targeting a prophylactic platelet threshhold of 10K/uL. Continued monitoring of the patient's CCIs enabled the identification of permissive antigens (i.e., allogeneic platelets with HLA antigens predicted to be incompatible with the patient's serum by solid-phase testing but were found to result in satisfactory increments); this knowledge allowed progressive widening of her permissive donor pool and enabled continued prophylactic support as needed. After three weeks the patient was able to be discharged from the ICU.

Subsequent HLA class I antibody screens using flow-cytometry based single antigen bead (SAB) technology revealed progressive decrease in the patient's mean fluorescence intensity for most of her antibody specificities, with the majority of her class I A antibodies falling to undetectable or near-undetectable levels (while previously >5000 MFI). Simple regression analysis (RStudio 1.0.153, R version 3.4.1) was used to test to how level of exposure to individual HLA platelet antigens (as measured by the number of doses with a given antigen) affected subsequent MFI levels. The results of the regression indicate that pre MFI and exposure explain 70.6% of the variance (R²=0.706, F(2,78)=93.7, p=<2.2e-16). It was found that exposure is a significant predictor of decreased MFI (b=-107.20, p=0.01422). High initial MFI was strongly associated with high subsequent MFI (b=0.75, p= <2.2e-16).

Conclusions: This data raises the possibility that large bolus platelet infusions could not only be of immediate therapeutic benefit in patients with HLA-alloimmune-mediated platelet refractoriness, but may also accelerate the evanescence of the responsible antibodies. Further research is warranted to explore this possibility, and to ascertain the potential magnitude of this effect in comparison to competing factors such as natural evanescence and/or B-cell immunosuppressive or immunomodulatory therapies.