Can we “terminate” alloimmune platelet transfusion refractoriness?

Inappropriately low posttransfusion platelet (PLT) increments are largely a challenge in patients with hematologic disease and cancer, in whom 25% to 70% of individuals experience at least one poor increment in the course of treatment.1 Most of these “failures” are attributable to nonimmune disease states and medical interventions associated with accelerated PLT consumption, but the less-common immune causes are more addressable.2 In an era of near-universal leukoreduction, repeated PLT transfusion failure due to alloantibodies (so-called alloimmune refractoriness) has been significantly reduced but remains a problem in 3% to 5% of intensively transfused hypoproliferative thrombocytopenic patients.3, 4 Alloimmune refractoriness can also be a complicating issue for surgical patients in whom prior transfusions or pregnancies have produced broad, high-titer antibodies to Class I human leukocyte antigens (HLAs) or to human PLT antigens (HPAs).5 The majority of antibodies in immunized individuals target HLA rather than HPA, but HPA antibodies can cause transfusion refractoriness on their own or in combination with anti-HLA.6 Currently, the treatment of choice for patients with pathogenic antibodies directed against PLT HLA-A and -B antigens, or HPAs on PLT glycoproteins is selection of apheresis PLTs for transfusion from donors matched for recipient antigens or antigen-negative for identified recipient HLA or HPA antibodies.7 Attempts to produce HLA-depleted “universal” PLTs through acid elution of antigens resulted in inconsistent clinical results, which led to abandonment of this practice in the 1990s.8 PLT cross-matching, while inferior to identically matched units, represents an alternative for patients without broad HLA or high-frequency HPA alloimmunization and does not require large panels of typed donors.9 However, even in hospitals supplied by blood providers with sufficient numbers of typed donors or PLT cross-matching technology, PLT transfusion can be challenging. PLT transfusion refractoriness results in more frequent and delayed transfusion, incurs matching fees, and is associated with longer lengths of stay (LOS) and increased patient morbidity and mortality.1, 10, 11 Longer LOS, bleeding and mortality may result from comorbidities for which transfusion refractoriness is a marker but may also conceivably result from complications of prolonged thrombocytopenia. Akin to attempts in the popular Terminator movie series to nip frustrating resistance in the bud before it even begins (to higher-minded purpose in transfusion medicine), prevention of alloimmunization may avert or at least diminish undesirable adverse events and associated costs and is certainly preferable to treatment of established refractoriness. In the current issue of TRANSFUSION, Waterman and colleagues12 from the BloodWorks Northwest Research Institute report creation of a robust and elegant murine model of PLT-induced alloimmunization to explore both the mechanisms of antibody-mediated PLT clearance and the effect of costimulatory blockade on antibody formation. Previous studies by this group have demonstrated the requirement for murine splenic CD4+ T-cell activation and differentiation into a helper phenotype in the development of humoral alloimmunization to leukoreduced PLTs.13 In this study, the correlation of refractoriness with antibody titer and abrogation of passive transfer of refractoriness in Fcγ receptor–deleted, but not C3-deleted mice provide strong confirmatory evidence of the role of immunoglobulin-mediated PLT destruction in the model. How this translates to humans remains to be seen in light of potential evidence of complement-mediated destruction in refractory patients.14 The authors demonstrated that after four sensitizing transfusions, antibody titer correlated well with 24-hour PLT recovery. By corollary, low- to middling-levels of antibody were not found to predict PLT refractoriness in humans, although the many other reasons for refractoriness in humans beyond allosensitization dilute the antibody-mediated signal measured in human clinical trials.15 The investigators also describe the success of abatacept, a fusion protein incorporating the Fc portion of IgG1 and the extracellular portion of CTLA-4, a binder of the APC costimulatory molecules CD80 and CD86, in blocking both transfusion refractoriness and high-titer HLA alloimmunization. While the utility of costimulatory blockading agents is uncertain outside of organ transplantation in humans, a situation in which their risk benefit may be more balanced, this proof of concept suggests important avenues for therapeutic intervention. The model itself will open the door to using even more sophisticated combinations of immune deficiency and allomismatch to dissect immune mechanisms of PLT refractoriness. However, like all animal models, the translation of findings to the human condition may not be direct. PLT alloimmunization is dependent on a variety of patient and donor or product factors. Despite the relative immunosuppression of hematology-oncology patients, some individuals are able to mount an effective antibody response against PLTs. Children are known to mount more effective immune responses than adults, and previously pregnant females also respond more frequently after transfusion.3, 16 Further, it has been suggested that patients with systemic inflammation are more susceptible to alloimmunizing stimuli.17 Even among immunocompetent individuals, however, most patients never form HLA or HPA antibodies despite repeated alloexposure. Difficulties inherent in finding HLA-identical donors for ongoing cellular transfusions generally preclude prophylactic HLA matching as a means of preventing alloimmunization. The large number of nonresponders also argues against even a relatively targeted implementation of this approach except in the most critical patient subgroups (e.g., infrequently transfused patients who lack PLT glycoprotein IIb/IIIa). Interestingly, there is no clear antibody dose-response to transfused PLTs and their passenger white blood cells (WBCs); even a few alloexposures is sufficient in susceptible individuals to trigger broad anti-HLA reactivity to shared public epitopes on a range of otherwise quite polymorphic HLA antigens.18 However, a narrowing of breadth or disappearance of HLA antibody response is far more common over time (even with repeated allostimulation) than a broadening one.19, 20 Despite this relatively encouraging long-term prognosis, the approximately 65% of patients who experience no short-term improvement in antibody production remain a challenge better managed by avoidance of alloimmunization than treatment. With regard to HPAs, while alloantibodies also appear to wane, short-term management is more difficult for higher-frequency specificities despite the slowly increasing availability of HPA-negative apheresis donors. Most HPA alloimmunization is presently managed using PLT cross-match, which struggles to identify reactivity with antigens not present at high surface density (like HPAs-2, -5 and -15 and, occasionally, glycoprotein IV, which can result in an isoimmune response in deficient recipients). An important observation from the Trial to Reduce Alloimmunization to PLTs (TRAP) was that effective PLT prestorage leukoreduction or UVB irradiation to inactivate passenger WBCs markedly reduced HLA alloimmunity as well as titer and persistence of residual antibody formation, but did not affect the HPA alloimmune response.3, 19 The differential success of these product manipulations suggested different modes of antigen presentation that might be manipulable to down regulate or prevent alloimmune responses. PLTs have the second highest HLA Class I antigen surface density after lymphocytes, but many PLT HLA antigens are denatured and adsorbed from the plasma without accompanying β2-microglobulin and are dissociable from the PLT surface.21 WBCs are thus thought to be a more significant source of HLA Class I peptides. Reduction, but not elimination, of transfused WBCs would be expected to decrease, but not entirely eliminate, the HLA Class I alloimmune response. Residual APCs and lymphocytes may still serve as sources for direct recipient T-cell allorecognition from donor APCs and indirect T-cell allorecognition of recipient APC-processed lymphocyte antigens. Studies of alloimmunization in animal models and patients continue to improve our understanding of HLA and HPA antigen presentation and processing. In murine and monkey studies, it has been shown that older PLT units are not as immunogenic as fresher ones.22, 23 This observation has been attributed to a declining stimulatory effect of lymphocytes and APCs during storage, release of immunosuppressive soluble major histocompatibility complex (MHC) Class I molecules from PLTs, and progressive PLT activation and cytokine secretion, which may also promote immunosuppression.24 In contrast, HPA antigen recognition in humans occurs only via indirect immune stimulation, and leukoreduction that does not reduce HPAs in transfused PLTs would not be expected to affect alloimmunization (nor did it in the TRAP study).3 Despite the success of leukoreduction in reducing transfusion refractoriness (from 13%-14% in the TRAP and Canadian studies to 3%-4%),3, 4 the persistence of some alloimmunization-related refractoriness demonstrates either insufficient leukoreduction or more likely, the duality of antigen presentation. Suggesting the importance of the latter explanation are studies demonstrating that inhibitors of indirect allorecognition markedly reduce murine MHC antibody formation.25 Substances that inhibit inducible NO synthase, tubulin formation, and endosomal acidification have been employed with some success. In the present study in TRANSFUSION, an immunosuppressive agent (abatacept) that modulates T-cell costimulatory signaling was also successful at least when given prophylactically before transfusion. None of these immunomodulatory interventions is likely to be safe or acceptable for wide use in sick hypoproliferative thrombocytopenic patients, but modulation of the immune response remains an intriguing avenue to address residual HLA and HPA alloimmune refractoriness to leukoreduced products. While elucidating mechanisms of PLT alloimmunization, basic science studies to date have not yielded the magic bullet needed to “terminate” the allogeneic immune response to transfused PLTs. More promising in the long term, albeit not a preventative strategy, is PLT pharming of HLA-deficient transfusable PLTs. The feasibility of generating functional HLA Class I–silenced PLTs from CD34+ or induced pluripotent stem cells, using a short hairpin RNA to target β2-microglobulin transcripts has been demonstrated.26, 27 The scalability of such interventions and their cost remain significant barriers to widespread use in the foreseeable future. Also of interest are published studies on the effect of UVB irradiation in the presence of riboflavin. Irreversible nucleic acid damage produced by the Mirasol system (Terumo BCT, Lakewood, CO) is accompanied by the prevention of HLA immune stimulation in vitro and in animal models.28, 29 Unlike the effects of gamma irradiation, which merely reduce WBC viability, treatment with UVB and riboflavin also affects direct antigen presentation and costimulation by donor APCs. Whether there is an effect on indirect presentation remains to be seen as further human studies are published using Mirasol or INTERCEPT-treated PLT products. There is a suggestion that the impairment of direct antigen presentation may have further blunted antibody-mediated PLT increment decrements in MIRACLE trial patients already receiving leukoreduced products.30 Once whole blood pathogen reduction is feasible, the alloimmune contribution of WBCs from leukoreduced red blood cell transfusions will also become clearer since a PLT-only intervention only protects recipients from one source of immunizing WBCs. This area of immunomodulation seems to hold the most promise until a nontoxic alternative is available to more efficiently prevent responses to transfused WBCs and PLTs. Stay tuned to these endeavors, because to paraphrase the words of Arnold Schwarzenegger's Terminator, “They'll be back!” PJN has received an honorarium and research support from Terumo BCT. RRV has no conflicts of interest regarding this editorial. Ralph R. Vassallo, MD, FACP1 e-mail: rvassallo@bloodsystems.org Philip J. Norris, MD2 1Blood Systems, Inc. Scottsdale, AZ 2Blood Systems Research Institute San Francisco, CA

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.