Diagnosis and Management of Platelet Alloimmunization

Platelet refractoriness remains a clinical challenge associated with an increased risk of bleeding, prolonged hospital stays, and decreased survival. Poor one-hour post-transfusion increments typically represent immune platelet refractoriness, whereas an adequate one-hour increment followed by a suboptimal 18- to 24-hour count suggests peripheral consumption or sequestration. In the absence of a direct prospective comparison of the three strategies commonly used to support patients who have platelet alloimmunization – human leukocyte antigen (HLA) matching, HLA antibody avoidance, and platelet cross-matching – two recent reviews have concluded there is no definitive evidence that any of these strategies improves hemorrhage control or mortality.^1,8  The provision of HLA-matched platelets can be very efficient in centralized transfusion systems as it has been in the United Kingdom.² 

Apheresis platelets stored in additive solution, which contain 2/3 less donor plasma, have become available in the United States, with a longer track record in Europe. They are associated with lower platelet increments at one hour, but are less likely to cause allergic transfusion reactions or hemolysis, compared with standard platelets suspended in 100 percent donor plasma.³  Platelets that have undergone psoralen- UVA treatment to inactivate bacteria (and other infectious agents) are also known to produce lower post-transfusion increments. These considerations need to be taken into account when evaluating patients who seem to be refractory to platelet transfusion.

A recent study in an animal model suggests that immune-mediated clearance of MHC mismatched platelets can occur in the absence of anti-platelet alloantibodies, in mice that are deficient in B cells. Allo-reactive CD8+ T cells appear to mediate antibodyindependent platelet clearance.^4,5  If this finding is corroborated in humans, immune platelet refractoriness may actually account for some refractory cases now presumed to be non-immune. A putative T-cell—mediated mechanism would also favor HLA matching over cross-matching as a management strategy.

What is your approach to the diagnosis and management of platelet alloimmunization?

Platelet refractoriness occurs in 5 to 15 percent of patients who receive chronic platelet transfusions.¹  The patient’s size and the number of platelets transfused should be factored into the assessment of refractoriness. For example, one measure, the corrected count increment (CCI), is computed as follows: CCI = (platelet increment after transfusion/μl) x (body surface area in m2) ÷ (platelet dose x 1011). For the purposes of this calculation, assume that each single-donor apheresis unit contains 3 x 10¹¹ platelets or that each whole-blood-derived platelet concentrate contains 5.5 x 10¹⁰ platelets. Using this formula, if the platelet count increased by 20,000/μL in a patient who had a body surface area of 2.0 m² and who received one apheresis unit of platelets, the CCI is 20,000/μL x 2 ÷ by 3 = 13,333/μL.

Refractoriness is defined as a CCI value below 2,500 platelets/μL at 18 to 24 hours post-transfusion or a value below 4,500 platelets/μL at 10 to 60 minutes post-transfusion²  on two to three consecutive platelet transfusion episodes. A less rigorous approach is to assume that, for an average size, non-refractory adult, the platelet count increment will be at least 10,000/μl to 20,000/μL one hour after the transfusion of either one unit of apheresis platelets or an equivalent dose of pooled platelets (5-6 combined random-donor platelet concentrates are equivalent to 1 apheresis unit), recognizing that the number of platelets per component unit may vary widely.

Non-immune causes of platelet refractoriness predominate.

Non-immune factors are present in the majority (72-88%) of transfusion-refractory patients, and immune causes are present in a minority (25-39%).² Non-immune factors, most of which are prevalent in a high proportion of hematology-oncology patients requiring prolonged platelet transfusion support, include splenomegaly, fever, infection, DIC, bleeding, and drugs such as vancomycin and amphotericin, are associated with platelet refractoriness.^2,3  The mechanism for refractoriness associated with various drugs is partially mediated by drug-dependent platelet antibodies, although specific testing for such is not widely available and unlikely to be of practical help.

Use of fresh ABO-matched platelets can improve transfusion response.

Platelets express blood group A and B antigens, but they are often transfused without ABO matching. Major ABO incompatibility (such as group A platelets transfused into a group O recipient) can decrease post-transfusion increments. A trial of fresh ABO-matched platelets can be a valuable temporizing measure while investigation of the basis of platelet transfusion refractoriness is ongoing.

HLA alloimmunization is prevented by the transfusion of leuko-reduced red blood cells and platelets.

Antibodies to HLA antigens account for the overwhelming majority of cases of immune platelet refractoriness, with antibodies to platelet-specific antigens being much less common. HLA alloimmunization may occur in response to prior pregnancies or to transfusions, although only a subset of alloimmunized individuals demonstrates immune platelet refractoriness. The TRAP study⁴  showed that filtration-removal of leukocytes and ultraviolet B irradiation to inactivate leukocytes were equally effective in preventing the development of platelet transfusion refractoriness, which occurred in 16 percent of control patients, compared with 7 to 10 percent of patients who received leuko-reduced or irradiated platelets. On the other hand, HLA antibodies developed in 45 percent of control patients compared with 17 to 21 percent of intervention patients. Outcomes were equivalent for filtered apheresis platelets and for filtered pooled platelets. In Canada, universal prestorage leuko-reduction of platelets has lessened the incidence of alloimmune platelet refractoriness from 14 to 4 percent.⁵  Almost all apheresis platelet units and more than 80 percent of packed red blood cell units in the United States are leuko-reduced by filtration either at the time of collection or immediately prior to storage.

HLA typing and antibody testing are complementary approaches.

Platelets express only HLA Class I antigens. For patients who are candidates for allogeneic stem cell transplantation, HLA typing results may already be known. Most HLA laboratories have adopted high-throughput molecular methods for genotyping HLA Class I and II antigens. A sequence-specific oligonucleotide probe method requires only small amounts of DNA and therefore can be performed on samples from neutropenic patients. Low-resolution HLA-A and B typing (Class I antigens) is adequate for the management of platelet alloimmunization, while high-resolution typing, including sequencing, is reserved for HLA Class II typing to select stem cell donors.

HLA antibody detection can be performed using a variety of methods.³ Multiplex flow cytometric bead assays are more sensitive than traditional ELISA.⁶  In the former assay, patient serum or plasma is incubated with color-coded microbeads that are coated with HLA antigens. Flurochrome-labeled anti-human IgG is added, and a flow analyzer is used to determine the color code of the reactive beads with a computer algorithm determining the specific antigens to which the antibody is reactive. The panel-reactive antibody (PRA) represents the percent of HLA targets to which the patient has made antibodies. PRA can be determined by using the traditional lymphocytotoxic assay, by ELISA, or by fluorescence-based microbead method. Serial assays are useful in assessing candidates for organ transplantation, but less so for management of platelet-refractory patients because a numerical PRA result (the percentage of the population to which the patient has HLA antibodies) does not provide actionable information to guide platelet selection.

Strategies to select platelets for refractory patients include HLA-matching, avoidance of known HLA antibody specificities, and platelet cross-matching.

For the purpose of platelet donor selection, a grading system (with designated categories A, B1, B2 C, D) is employed. A perfect four-antigen match (2 loci each at HLA-A and HLA-B) is grade A. In a B1 match, all of the donor’s HLA antigens are present in the recipient, but the donor lacks one of the recipient’s HLA antigens; in a B2-match, all of the donor’s HLA antigens are present in the recipient, but the donor lacks two of the recipient’s HLA antigens; in a C-match, the donor has one HLA antigen that is not present in the recipient; in a D-match, the donor has more than one HLA antigen that is not present in the recipient. High-grade matched donors (grade A, B1, or B2) are specifically recruited to donate platelets for a particular patient, but transfusion with grade C or D “matched” units is unlikely to produce a clinically meaningful incremental increase in the platelet count. This grading approach to matching also allows for categorization of antigens into cross-reactive groups (CREG). For example, HLA-A1 and A36 are within the same CREG, so if a patient has the A36 antigen and no available donor platelet is A36 positive, then an A1 donor platelet – typically more prevalent in the Caucasian population – can be used in a grade B match. HLA Matchmaker is an epitope-based computer algorithm used in some centers to identify permissible donor platelets that are more likely to yield adequate platelet increment increases without being HLA matched.⁷  Despite the resources invested in the management of patients who are refractory to platelet transfusion, a recent review of the literature identified no studies that were adequately powered to detect an effect of transfusion of HLA-matched platelets on mortality or hemorrhage.⁸  Prospective studies utilizing current technology and examining clinical outcomes are needed to evaluate the effectiveness of HLA-matched platelet transfusion.⁸ 

For management of the transfusion-refractory patient, available data argue that selection of donors with HLA antigens against which the recipient does not have antibodies is a better strategy than HLA matching. In one observational study involving 29 refractory patients and a database of more than 7,000 HLA-typed donors, a mean of 39 donors were HLA grade A or B matched, but a mean of 1,426 donors were identified as permissible by antibody exclusion.⁹  Post-transfusion platelet count increments were comparable for the two strategies. HLA antibody testing should be repeated at periodic intervals because antibody specificities may evolve over time.

Platelet cross-matching tests the patient’s serum against samples of available donor apheresis platelets using a solid-phase adherence assay or an ELISA. A recent study found a mean CCI of 7,000 at one hour in more than 400 cross-matched platelet doses transfused to 71 refractory patients.¹⁰  Platelet cross-matching can be done within a few hours compared with several days for HLA testing, but it does involve frequent repeat testing because the shelf-life of platelets is five days. Automated platforms are invaluable in making this approach efficient and practical.¹  A recent shortage of the commercial kits used for platelet cross-matching in the United States is expected to be resolved by early 2014.

A comparison of some of the advantages and disadvantages of the three strategies for dealing with refractoriness to platelet transfusion is contained in the table. Choice of method depends on local resources, and communication between the hematologist and the transfusion service is critical to ensure that donor selection is appropriate and that valuable resources are not wasted or used inappropriately.

Anti-fibrinolytic agents can be a useful adjunct for mucosal bleeding.

Other approaches to ameliorating the consequences of alloimmune platelet refractoriness include infusion of IVIgG, citric acid treatment of platelets to remove Class I HLA epitopes, and infusion of recombinant activated FVII in actively bleeding patients. Despite anecdotal reports of success, none of these approaches has been validated for clinical use. Use of family members as platelet donors for patients who are potential allogeneic SCT candidates is controversial, based on a theoretical concern for inducing alloimmunization that may jeopardize engraftment. Anti-fibrinolytic agents such as epsilon-aminocaproic acid and tranexamic acid, however, can be useful in platelet-refractory patients with oral mucosal bleeding.