Platelet Immunology

June 30, 2006
Oslo Kongressenter , Norway

Chair: T. Warkentin (Canada)
Committee Co-Chairs (in attendance): A. Greinacher, Y. Gruel, V. Kiefel, H. Kroll
Committee Co-Chairs (not in attendance): 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: H. Kroll, T. Warkentin)

P. Metcalfe: Proposal for adoption of an anti-HPA-3a reference reagent by the ISTH. Anti-HPA-3a is the third or fourth most common cause of neonatal alloimmune thrombocytopenia (NAIT), but the pathogenic anti-HPA-3a alloantibodies can be difficult to detect. Therefore, there was interest in the development of a reference anti-HPA-3a reagent. A reagent (03-190) was prepared from plasma obtained from a mother with two babies born with severe NAIT due to anti-HPA-3a. Forty-nine laboratories from 23 centers participated in the evaluation of this material, including internationally-recognized laboratories (Kiefel, Kroll/Santoso, Murphy, Kaplan, Ouwehand, Aster/Curtis). Studies were aimed at determining the extent to which the participating laboratories could identify the antibodies, including minimum titer of detectability. Almost all of the labs (except for two) could detect the antibodies. Dilutions of anti-HPA-3a up to 1:64 were detected. It was determined that a minimum dilution of 1:8 should be detected by a laboratory (if not, the assays used are not sensitive enough). Prior to the meeting, the draft document (prepared by P. Metcalfe) describing the performance of this reagent among the laboratories was circulated to committee members to review.

Vote by committee members on proposal to adopt anti-HPA-3a reference reagent (NIBSC 03-190) by the ISTH: FOR (Chong, Imbach, Kaplan, Kiefel, Kroll, Greinacher, Gruel, Smith [in absentia], Warkentin); AGAINST (none against).

C. Kaplan: Predictive value of anti-HPA-1a alloantibody concentrations for severity of fetal alloimmune thrombocytopenia. About 1 in 800 live births is affected by NAIT. Severity of thrombocytopenia increases with subsequent pregnancies. The MAIPA was used to test for a correlation between the antibody concentration in the MAIPA against severity of thrombocytopenia. In the method reported, a correction factor for hemodilution is performed (using albumin levels). Significant correlation was seen between anti-HPA-1a concentrations of >250 AU/mL and fetal thrombocytpenia, irrespective of the gestational age. The sensitivity and negative predictive values were improved when the testing was performed before 28 weeks of gestation. The correction factor for hemodilution is important to perform. Of interest, significant decline in antibody levels occurs toward the end of pregnancy (presumably reflecting increase in transplacental passage of the antibodies).

C. Kaplan :New mutation altering HPA-1a genotyping. Laboratory diagnosis relies on testing for alloantibodies (usually by MAIPA) and assessing parental antigen incompatibility. In a case presented, no maternal antibody was detected by MAIPA, but there was apparent HPA-1b incompatibility shown by genetic testing (mother: HPA-1b/b, father HPA-1a/a). However, by platelet typing by MAIPA, the mother was found to be heterozygous. Platelet genotyping by PCR-RFLP also showed the mother to be heterozygous. A silent mutation was found within the antisense primer of the PCR-SSP that let to false typing assignation (Bertrand et al. Transfusion 2006; in press). This illustrates the importance of using complementary methodologies to avoid false diagnosis of incompatibility.

H. Kroll/S. Santoso: Registry of alloimmunization against low-frequency HPAs and atypical anti-HPA-3a. A new alloantigen (Swi a) has been recently described (Kroll 2006), representing GPIa 3347 (T to C) substitution. It would be useful to have a registry on these (and others to be discovered) rare alloantigens. The speaker indicated that he is working with committee members to develop such a registry (ongoing work). Recently, a questionnaire was submitted regarding interest in such a registry regarding rare alloantibodies, as well as with atypical anti-HPA-3a alloantibodies. A questionnaire regarding these issues was sent to 52 labs; to date, 6 labs responded.

Technical issues of anti-3a alloantibody detection were discussed, including time of test platelet storage (which leads to loss of reactivity with some anti-sera). The speaker indicated that he will work with committee members to establish a registry of alloimmunization to these rare (low frequency) alloantigens.

AUTOIMMUNE THROMBOCYTOPENIA (Chairs: B. Chong, V. Kiefel)

P. Imbach: Update: Activities of the Intercontinental Childhood ITP Study Group (ICIS). An overview of ITP pathogenesis was presented, highlighting the involvement of many aspects of the immune system. An historical perspective of IVIgG therapy was also summarized, and the dramatic growth of this therapy was emphasized (currently, 57 tons/yr sold world-wide). ICIS Registries 1 and II are closed. Currently, the Splenectomy Registry and the PARC-ITP Registry are open. More than 2000 patients were entered into Registry I; a slight male predominance was seen, which was greatest in children under 1 year of age. The natural history of ITP was such that 11% of patients still had a platelet count <20 at 1 year follow-up. Standardized approaches to management (e.g., systematic approaches to treatment of three stages of bleeding severity) were discussed.

F. Rodeghiero; on behalf of the Working Group on Thrombocytopenias of the European Hematology Association. Heterogeneity of terminology and clinical definitions in adult ITP: A critical appraisal from literature analysis. The speaker demonstated the magnitude of terminology heterogeneity and definitions for ITP in the literature, and ways to achieve consensus. A total of 1247 papers were retrieved (2000-2005). Definitions of ITP (platelet count thresholds, time-related parameters, grade of severity) were reviewed. First-line therapies were reviewed. Definitions were assessed for various factors, including definition of ITP, platelet cutoff for initial treatment, platelet levels to define response, time to assess response, bleeding score, platelet level to define chronic phase, time point to assess for meeting the criteria to diagnose “chronic” ITP, criteria for splenectomy, definition of response to splenectomy, etc. Numerous differences in criteria were found, with no literature consensus. For example, for starting treatment, platelet count thresholds varied widely (<20, 3/21 publications; <30, 13/21; <50, 3/21; other criteria, 2/21). Similarly, complete and partial response criteria varied widely in the literature, as did timing of assessment for first-line therapy, and duration of response (e.g., short term defined as 3-10 days, and long term defined as 3-12 months). Re: definition of chronic ITP, platelet thresholds ranged from <50 to <100 to <150 to < “normal range”, and duration of thrombocytopenia ranged from >6 weeks, to >3 months, to >6 months, to >2 years. All of the items reviewed showed considerable heterogeneity in definition. Thus, there are considerable limitations in the interpretation of the ITP literature. The speaker reported on plans to develop consensus on these issues by a Working Group on Thrombocytopenias of the European Hematology Association, with various planned upcoming meetings (Sep 2006 ICIS Meeting; Nov 2006 Vicenza meeting Proposal and Discussion on Topics and Methodology; Dec 2006 ASH Congress; June 2007 EHA Meeting Presentation of Final Terminology Proposal. The participants were listed as: Rodeghiero [Chair], Provan [Co-Chair], Godeau, Fenaux, Imbach, Bussel.

B. Chong [on behalf of the Working Party]: Definition of ITP (integrated clinical and laboratory definition: an update of current activities). Dr. Chong gave the presentation on behalf of J. Bussel, who was unable to attend. The starting point for a definition of ITP is that of thrombocytopenia caused by platelet-reactive autoantibodies, and that is fundamentally based upon clinical exclusion of other causes of thrombocytopenia. Discussion was held regarding which specific conditions need to be excluded? (HIV, HCV, splenomegaly, congenital thrombocytopenia, etc.). Should response to treatment be a criterion? (but then, what if treatment is not required). The speaker advised generating a list of disorders to be ruled out, along with a guide to what tests ought to be performed. Parameters for performing bone marrow aspirate were discussed, e.g., age >60 years, no response to treatment, prior to planned splenectomy or certain immunosuppressive therapies. Studies of “phase III” assays for platelet glycoprotein-associated assays show sensitivity in the 50-66% range, and specificity in the 80-90% range. Important unresolved issues include absence of a “gold standard” laboratory assay for diagnosis of ITP, lack of agreement on which specific non-ITP causes of thrombocytopenia need to be excluded, diagnostic value of phase III assays, differences between adults and children. The speaker presented the framework of a potential diagnostic algorithm that incorporated important steps as formalized exclusion of non-ITP causes of thrombocytopenia, performance of a phase III assay, and response to immunosuppressive therapy.

V. Kiefel; Impact of abciximab therapy on autoantibody detection in thrombocytopenic patients. The speaker showed that about 70% of patients have elevated GP-PAIgG (by MAIPA, SZ21) if they were tested soon after receiving abciximab. However, there was no good correlation between any drop in platelet counts and the level of GP-PAIgG post-abciximab. In theory, such antibodies could react against either abciximab itself, or against a new antigen (ligand-induced binding site [LIBS]). In all cases investigated, preincubation with human Fab showed nearly complete absorption. The speaker speculated that natural anti-abciximab/GPIIbIIIa IgG reacts with the papainic cleavage site of the drug.

DRUG-DEPENDENT THROMBOCYTOPENIA: D-ITP (Chairs: V. Kiefel)

A. Greinacher (on behalf also of R Aster, B Curtis, B Chong, H Kroll, S Santoso, T Warkentin): Registry of drugs causing immune thrombocytopenic purpura (D-ITP syndrome by serum exchange.

Numerous drugs are suggested as causing immune thrombocytopenia, but the “true” list of causative drugs may be much smaller. It was proposed that a systematic approach to identify drugs that cause D-ITP—in which confirmatory diagnostic testing in at least two labs (with serum exchange)—is crucial for “proving” that a drug causes immune thrombocytopenia (high specificity is assumed). This concept was examined by an exchange of several putative drug-induced thrombocytopenia sera (vancomycin, quinine, trimethoprim-sulfamethoxazole) between the McMaster and Greifswald laboratories. Concordant results were seen with each of the sera tested (though techniques varied, i.e., flow cytometry in McMaster, MAIPA whole platelet ELISA in Greifswald ). Issues addressed include: serum shipping, ethics (anonymity of patients whose serum is tested is required, including avoiding even use of the date-of-birth, given the rare nature of these reactions; however, this raises issues on how to be sure distinct sera are being studied); use of generic versus brand names; how to handle drugs; how to report results, including on a registry data base (positive vs negative only? quantitative data including all control data?); appropriate controls to use, especially when a known positive serum may not necessarily be available (normal serum control; platelet-reactive alloantibodies when drug-dependent serum is not available?); what sort of clinical information should be available in the registry?

B. Chong: Drug-induced immune thrombocytopenia .

Dr Chong presented some of his viewpoints regarding D-ITP, including the concept that current assays are both specific and sensitive (provided that the correct drug and drug metabolites are studied), and if patient serum is used after cessation of drug. Either drug-dependent and patient serum-dependent PAIgG or GP-specific assays can be used, although the latter assays are more sensitive. Use of indirect (serum-based) assays is preferred, compared with using patient platelets when thrombocytopenia is present. For quinine-induced thrombocytopenia investigated in his laboratory, 8/13 sera tested positive with flow cytometry, whereas all 13 reacted against GPIb-IX, and 3/13 reacted against GPIIb/IIIa, in the GP-specific assay.

Regarding ELISAs for HIT, the cutoff influences the sensitivity and specificity of the assays. For the functional tests, international and Australian surveys previously reported that expert labs can perform functional assays for HIT antibodies well, but not necessarily other (non-expert) labs.

HEPARIN-INDUCED THROMBOCYTOPENIA (HIT) (Chair: V. Kiefel)

T. Warkentin and A. Greinacher: Gender and HIT . Data from three studies were presented showing a female predominance in HIT. The common odds ratio is ~1.5-2.0. Also, analysis of data from individual patient studies shows that there are three separate risk factors for HIT: (a) type of heparin (UFH > LMWH), (b) type of patient (surgery > medicine); and (c) gender (female > males). The implication is that in some settings (e.g., females receiving postoperative UFH thromboprophylaxis) the impact of using LMWH (to prevent HIT) is most effective.

T. Warkentin, A. Greinacher, Y. Gruel, R. H. Aster, B. Chong: Conceptualizing the sensitivity-specificity tradeoffs of laboratory testing for D-ITP: influence of antibody classes. The concept was presented that current PF4/heparin ELISAs and washed platelet activation assays are inherently very sensitive for detection of HIT antibodies, but lack specificity because non-pathogenic antibodies are frequently generated in heparin-treated patients, and because not all patients even with “strong” HIT antibodies develop thrombocytopenia. ELISAs inherently have lower specificity than platelet activation assays, since the former are more likely to detect non-pathogenic antibodies. This situation is the reverse of laboratory tests for D-ITP, in which sensitivity is not 100% (because some cases are caused by metabolites that may not necessarily be included in the diagnostic test system) but specificity is believed to be high. The implications of these concepts regarding evaluation and interpretation of diagnostic tests was presented. An advanced manuscript discussing these concepts has been prepared.

A. Greinacher: Discrepancies in PF4/polyanion-immunoassays and platelet activation assays in HIT: influence of antibody classes. Important issues in test discrepancy were pointed out. For example, sometimes commercial immunoassays have used degraded PF4 in which the antigenic complexes were lost (this problem is not necessarily shown by using a “positive” control, as these are not HIT sera, but usually monoclonal anti-PF4 antibodies). A small number (~0.4%) of samples referred to the Greifswald laboratory are positive in the washed platelet activation assay (HIPA test), but negative for IgG antibodies in the PF4/heparin ELISA. Results of seven discrepant sera (HIPA positive but PF4/heparin ELISA) were further investigated—after IgG depeletion, the HIPA test became nonreactive. This suggests that an IgG against another (non-PF4/heparin) antigen was likely present. The data further suggested that ELISA OD titers are useful for predicting risk of HIT. There is also information that IgA and IgM antibodies do not interfere with the platelet activation assay. The conclusion is that the “combination of a washed functional (platelet activation) assay and an antigen assay is still the gold standard.”

T. Warkentin, A. Greinacher, Y. Gruel, B. Chong: Proposal for HIT scoring system. The historical context of scoring systems for HIT was presented. Published evaluations of the ‘4 T’s’ clinical scoring system were shown. It is evident that this scoring system has a high negative predictive value, although the positive predictive value varies among different clinical settings. The working party on HIT is preparing a manuscript dealing with this scoring system.