Platelet Immunology

6 August 2005
16:00 to 19:30
Harbourside Meeting Room 4
Sydney Convention and Exhibition Centre

Chairman:  T Warkentin, Canada
Co-Chairs: JB Bussel, USA; BH Chong, Australia; D. Cines, USA; A. Greinacher, Germany;
V. Kiefel, Germany; H. Kroll, Germany; M.F. Murphy, UK; G.P. Visentin, USA

Committee Co-Chairs (in attendance): J. Bussel, B. Chong, A. Greinacher, H. Kroll
Committee Co-Chairs (not in attendance): D. Cines, V. Kiefel, M. Murphy, G. Visentin

The program was divided into three parts: (I) Alloimmune Thrombocytopenia, (II) Autoimmune Thrombocytopenia, and (III) Drug-Induced Thrombocytopenia.

ALLOIMMUNE THROMBOCYTOPENIA (Chairs: C. Kaplan, H. Kroll)

W. Ouwehand (Cambridge, UK): Collaborative study to establish the first international standard for quantitation of anti-HPA-1a. W. Ouwehand, speaking on behalf of Paul Metcalfe, summarized the development and validation of a reagent, NIBSC code 03/152, for the quantitation of anti-HPA-1a alloantibodies in patient serum. The process by which the ISBT approved this reagent was reviewed, which included evaluation using this reagent of three sera of varying anti-HPA-1a titers by 39 laboratories in 24 countries. These included 8 laboratories represented at the Platelet Immunology SSC (Aster/Curtis; Greinacher/Eichler; Kaplan; Kelton/Smith/Warkentin; Kiefel; Kroll/Santoso; Murphy; Ouwehand), all of whom had one or more members that voted in favor of the proposal (and none opposed). The final vote was 11 in favor (none opposed) for recommending that the ISTH approve the reagent (NIBSC code 03/152) as an anti-HPA-1a (anti-Pl A1) standard. This matter was to be considered at the meeting of the Standardization committee later in the Sydney ISTH meeting.

G. Bertrand (Paris, France): Quantitation of anti-HPA-1a alloantibodies using the MAIPA procedure. A method to quantitate anti-HPA-1a alloantibodies using the monoclonal antibody immobilization of platelet antigens (MAIPA) method was summarized, which constructed a standard curve using diluted high-titer serum (from 1:1 to 1:128), and employed computerized fitting of the data for analysis. It was concluded that this method was useful for quantitating antibody titer, thus permitting studies of correlations between antibody titer and fetal/neonatal thrombocytopenia.

A. Greinacher (Greifswald, Germany) on behalf of V. Kiefel (Rostock, Germany): Treatment of NAIT: update on experience with random donor platelets for treatment of NAIT. Anecdotal evidence was provided suggesting that most newborns with severe NAIT may benefit from transfusion of random donor platelets. (Current recommendations are that these infants be transfused with platelets negative for HPA-1a/5b alloantigens, and/or intravenous gammaglobulin and/or maternal platelets.) Sixteen cases were reviewed from 7 centres (6 German, 1 Canadian) in whom platelets increased rapidly to >40 in 15 of the infants (the platelet count rose to >80 in 11 of 16 infants). No side effects were observed. It was concluded that treatment of NAIT with random platelets is a potential option for managing this situation, which has the advantage of avoiding delays from some of the other options. Discussion included the comment that there are limited data in the literature indicating that this treatment can be beneficial. It was noted that in the U.K. 8 blood centre locations have HPA-1a/5b-negative platelets available, allowing for platelets to be available usually within 2 hours.

H. Kroll [Presenter]/S. Santoso (Giessen, Germany): Use of novel cell lines for diagnostics of NAIT/Proposal for the distribution of reference material. It would be useful to have target cells available for alloantibody testing that bear uncommon/rare platelet alloantigens. EBV- transformed cell lines have been prepared for HPA-1-5,6,7,8,9, and 15. Two laboratories ( Giessen , Germany ; Potters Bar , UK ) act as repository laboratories to ensure storage, viability, etc. of these cells. Laboratories that define new platelet alloantigens should ensure immortalization of B cells, or make available of material for expert laboratories to perform this. There is a need for development of stable cell lines. PROPOSAL: to establish a registry of alloimmunization to alloantigens other than HPA-1-5/15, i.e., low frequency alloantigens (coordinator: H. Kroll). This would provide information on antibody frequency and heterogeneity, situations of non-immunization despite alloantigen mismatch, variability in clinical severity, which methods for detection are appropriate, and clinical information on items such as outcomes of previous pregnancies, duration of survival of transfused platelets, etc. In discussions, an issue arose as to whether patient informed consent was required even when non-identifying information would be placed into a registry.

H. Kroll [Presenter]/S. Santoso (Giessen, Germany): Heterogeneity of alloantibodies in alloimmune thrombocytopenia. F actors contributing to variable severity of NAIT were reviewed. Questions were raised about whether anti-HPA-1a antibodies impair fibrinogen-gpIIb/IIIa interactions. A notable finding was that in fibrinogen-mediated cell adherence experiments most CHO cells (HPA-1a/1b) did not have impaired adhesion in presence of anti-HPA-1a antibodies, but some did, especially using PTP sera (blood obtained at time of severe thrombocytopenia). In the case of PTP, these features may be associated with disease pathogenesis. Another issue is heterogeneity of anti-HPA-3a alloantibodies, with variable binding (despite evident disease) that appears to be assay-dependent (in some cases, only whole blood assays show binding). Issues such as platelet handling/storage and lysis appear to be relevant. Reactivity is optimized if HPA-3a/3a platelets are used. Some conclusions were that most anti-HPA-3a alloantibodies were detectable by flow cytometry (though there are problems with HLA interference). All anti-HPA-3a alloantibodies were detectable on day 0 by MAIPA, with decreased reactivity on longer storage. Thus, the MAIPA is a suitable assay when fresh, homozygous platelets are used. PROPOSAL: to establish a registry of alloimmunization involving anti-HPA-3a. The aim of this project would be to develop recommendations for optimal testing for anti-HPA-3a alloantibodies.

C. Kaplan (Paris, France): Heterogeneity of anti-HPA-9bw alloantibodies. NAIT due to anti-HPA-9bw was summarized, which accounts for about 2% of NAIT, usually quite severe, and often with poor platelet count increments following random donor platelet transfusion. It was noted that no one monoclonal antibody has been identified that permits detection of all anti-HPA-9bw alloantibodies.

AUTOIMMUNE THROMBOCYTOPENIA (Chairs: J. Bussel, B. Chong)

J. Bussel (New York, USA): Definition of autoimmune thrombocytopenic purpura: preliminary report of a working party. A working party is being organized to define better ITP, including an algorithmic approach for its diagnosis that includes certain aspects of laboratory testing. Some issues with existing definitions were reviewed; e.g., it was noted that the ASH guidelines on ITP indicate that “no other causes of thrombocytopenia” is a criterion, but details on how exactly to ascertain the absence of other causes of thrombocytopenia are not listed. The presentation focused on two interrelated issues, beginning with what investigations might ‘rule out’ ITP. Various infections as causes of apparent immune thrombocytopenia (or ITP-mimicking illnesses) were listed (HIV, HCV, Helicobacter pylori, CMV, dengue, malaria), as were clinical and laboratory features suggesting hereditary thrombocytopenia (family history, small/large platelets, hearing abnormalities, renal failure, cataracts, mental retardation, leukemia/lymphoma, thalassemia trait or myelodysplasia in X-linked inheritance, absent radii/other orthopedic problems). The possibility of an ITP-mimicking drug-induced AITP syndrome was raised. Early myelodysplasia can resemble ITP. Also, how does one deal with related disorders that historically have conferred the term “secondary ITP”, such as certain infections (HIV), humoral immunodeficiency (antibody, complement), Hashimoto’s thyroiditis, Evan’s syndrome, neoplasia (CLL, Hodgkins lymphoma, NHL, SLE, APS, etc.). Then, an algorithmic approach was discussed in regards to how one might ‘rule in’ ITP. Three distinct approaches could include (a) positive test for platelet-reactive autoantibodies; (b) platelet count response to IVIgG or anti-RhD; or (c) evolution of illness over continuing follow-up (say, a 6-month period). In summary, the following approach was suggested as a first step for discussion: (a) development of diagnostic “check list” to improve diagnosis by listing those features of history, examination, and laboratory testing that improve specificity of diagnosis; (b) study of these features could help to identify those laboratory tests that should be performed as a routine (?HCV ?immunoglobulin levels ?anti-phospholipid antibodies ?H pylori); (c) evaluate various laboratory tests for anti-platelet antibodies, TPO, etc., that might be helpful diagnostically.

T. Kuehne (Switzerland) on behalf of P Imbach and D Provan: PARC Registries and EHA Working Party update. An update was given on existing registries of pediatric and adult chronic ITP. The first registry provided information on the natural history of ITP. Subsequent registries deal with such issues as disease heterogeneity, biological and genetic markers, and so forth. This presentation was made as a matter of information for committee members.

DRUG-INDUCED THROMBOCYTOPENIA: D-ITP (Chairs: B. Chong, H. Kroll)

A. Greinacher (Greifswald, Germany) (on behalf also of R Aster, B Chong, B Curtis, H Kroll, S Santoso, T Warkentin): Proposal for definition of what is required by the SSC to accept putative drug-induced immune thrombocytopenia (D-ITP), including proposal for a registry of those drugs agreed to cause D-ITP, e.g., web-based database.

Both scientific and drug safety reasons were listed for aiming to improve the diagnosis of D-ITP. Careful analysis of patients with HIT and GPIIb/IIIa thrombocytopenia have shown that they are different from “classic” D-ITP. As the majority of drugs currently listed in textbooks as inducing D-ITP has been selected by clinical criteria only, the SSC favors a rigorous approach in which both clinical and laboratory criteria need to be fulfilled to establish a drug as causing D-ITP. This may allow better definition of common pathogenetic mechanisms or development of improved techniques for detecting the relevant antibodies, thus providing a positive impact on improving diagnosis. PROPOSAL: The Working Party proposes to generate a set of clinico-pathologic criteria for establishing a drug as definitively causing D-ITP. Additionally, the Working party proposes to generate a web site on which those drugs (and the relevant criteria met) are listed. Committee members discussed some of the potential inclusion criteria, such as the number of well-documented cases needed to establish a drug as causing D-ITP (?1 ?3, etc.), sharing of serum among a minimum number of laboratories (?2 ?4), the minimum platelet count nadir (e.g., <20, <100), and other considerations. These will be developed by the Working Party. The audience voted in favor of this proposal. A question arose as to whether the ISTH has special requirements for establishing such a website.

DRUG-INDUCED THROMBOCYTOPENIA: HIT (Chairs: Y. Gruel, T. Warkentin)

A. Greinacher (Greifswald, Germany): Frequency of anti-PF4/heparin antibodies of the IgG, IgA and IgM class: the Greifswald experience.Sera from 755 patients with clinically-suspected HIT were studied over a 9-month period. Only about 15% of all referred sera tested positive in any of the assays (HIPA, EIA). Only about 40% of these positive sera had platelet-activating antibodies (positive HIPA). About 70% of all positive sera tested had anti-PF4/heparin antibodies of IgG class. Of those testing positive, only 3% were only positive in the HIPA, indicating that a negative EIA is good at excluding HIT.Analysis of cases suggest that in patients who only have IgA and IgM class antibodies, other clinical reasons for thrombocytopenia and/or thrombosis were generally evident. The data suggest that there may be considerable over-diagnosis of HIT.

T. Warkentin ( Hamilton , ON ): Relevance of anti-PF4/heparin antibodies of the IgM and IgA class. The Hamilton prospective trials . Data were presented evaluating the systematic serologic investigation (by serotonin release assay [SRA], EIA-IgG, EIA-IgM, EIA-IgA, EIA-GTI) of patients who participated in large prospective trials of heparin therapy for orthopedic surgery. The data show that the SRA, EIA-IgG, and EIA-GTI all have high sensitivity for detecting clinically-significant antibodies. However, specificity was SRA > EIA-IgG > EIA-GTI. It was further shown that median levels of IgA and IgM class antibodies in patients with clinical HIT were indistinguishable (in terms of their overall degree of reactivity) from patients who formed a non-HIT immune response (as judged by a positive reaction in the EIA-GTI). The data suggest that the operating characteristics of the EIA for diagnosis of HIT likely would be improved by detecting only antibodies of IgG class.

A. Leyte ( Amsterdam , Netherlands ), Y. Gruel ( Tours , France ), T. Warkentin ( Hamilton , ON ): Scoring system for clinical HIT. Experience using a clinical scoring system for HIT (the 4 T’s) was presented. The first speaker ( Leyte ) presented experience in an intensive care unit (ICU) population. Using the 4 T’s plus criteria presented by B Chong at the Paris meeting, the authors concluded that only 0 to 7% of thrombocytopenic patients in their study had probable or definite ICU. Moreover, the discrepancy between clinical criteria and results of EIA testing was large. It was suggested that the ICU population is a special one that will require special attention. The second speaker (Gruel) showed data from a prospective multi-centre evaluation of the 4 T’s in which several laboratory assays (SRA, EIA, and rapid particle gel assay [Diamed]) were assessed. The scoring system showed high (100%) negative predictive value, based upon the SRA. However, 9/42 low score patients tested positive in the EIA (but with a negative SRA). This indicates that the SRA provides much greater diagnostic specificity than does the EIA. All three assays were positive in the 5 patients that had a high clinical score for HIT. The third speaker (Warkentin) showed that the 4 T’s scoring system had high negative predictive value for excluding HIT. The positive predictive value depended on the setting in which it was performed, with high positive predictive value of a high clinical score seen in Hamilton (performed by a hematologist) whereas it was lower in Germany (used by multiple physicians when requesting laboratory testing for HIT antibodies).