This issue of Clinical Chemistry and Laboratory Medicine includes two reports related to antibody-platelet interactions involving heparin and the concept of heparin-induced thrombocytopenia (HIT) associated with thrombosis (‘HITT’). Vianello and colleagues compared the performance of three different immunoassays in the diagnosis of heparin-induced thrombocytopenia in 114 consecutive patients with clinically suspected HIT , whilst Panzer and colleagues report serological features of anti-protamine and anti-protamine-heparin antibodies in six patients initially suspected clinically to have HIT but lacking laboratory evidence of platelet activating anti-PF4-heparin antibodies .
Central to pathogenesis of ‘classical’ HIT are anti-PF4-heparin antibodies, which are capable of inducing platelet activation and subsequent thrombocytopenia and prothrombotic manifestations through cross-linking of FcγRIIa . Diagnosing HIT remains a significant clinical and laboratory challenge. It can be argued, however, that the pendulum is swinging from a lack of awareness of HIT to an over-diagnosis of this disorder, based on non-pathological falls in platelet count and/or the results of poorly specific laboratory tests. Whereas under-diagnosis of HIT potentially exposes patients to major organ damage and possible life-threatening thrombosis, the false diagnosis of HIT needlessly exposes patients to increased risk of complications, such as bleeding from the use of alternative non-heparin anticoagulants. Currently recommended diagnostic paradigms combine clinical scoring systems, such as the 4Ts or the HIT Expert Probability (HEP) Score with laboratory tests, which are either designed to detect anti-PF4 heparin antibodies in vitro (immunoassays) or confirm if the antibodies in patient serum/plasma are capable of inducing platelet activation in the presence of heparin (‘functional assays’) [3, 4]. Whilst both the clinical prediction rules and immunoassays have a high negative predictive value (NPV) and can be used to effectively rule out HIT in patients with a low clinical probability score and/or no detected anti-PF4 heparin antibodies, substantial problems arise with alternate combinations. For example, in a recent meta-analysis of trials using the 4Ts prediction rule, the positive predictive value (PPV) of a high score (>6) was only 64% . Therefore, even in patients with a high clinical probability of HIT, it is strongly recommended to use laboratory testing to confirm the presence of (pathogenic) PF4-heparin antibodies. However, laboratory assays for identification of HIT have their own strengths and weaknesses [3, 6].
Functional assays, such as the serotonin release assay (SRA) and the heparin-induced platelet activation assay (HIPA), are considered the gold standard for detection of pathogenic, platelet-activating heparin-dependent antibodies. Whilst these assays are characterized by high specificity and PPV, their technical complexity plus need for fresh and reactive platelets limits their use to tertiary referral centers or large commercial laboratories and additionally precludes their routine diagnostic use. Therefore, because of their wide availability and high sensitivity, immunoassays have become the mainstay in the laboratory diagnosis of HIT. In general, the immunoassays for HIT have high sensitivity, but poorer specificity for HITT than functional assays.
Vianello and colleagues  comparatively evaluated data from three immunoassays against the HIPA, which they used to identify 29 (25.4%) HIT positive patients, all of whom also had a 4Ts score ≥4. The single best performing assay was the HemosIL® AcuStar HIT-IgG (PF4-H) (Instrumentation Laboratory), which is based on the chemiluminescence detection of anti-PF4-heparin antibodies bound to PF4-polyvinyl sulfonate. At the recommended cut-off of 1 U/mL, this assay had PPV and NPV of 80% and 99%, respectively. Another ELISA-based assay, the Lifecodes PF4 IgG, showed a very high sensitivity (100%) at the recommended cut-off, but this was at the expense of specificity (45%). When the authors used an optimized cut-off (≥0.92 OD) based on the receiver-operating characteristic (ROC) analysis, the specificity increased to 88%. The lateral-flow immunoassay for the detection of HIT antibodies (Stago LFI- STic Expert® HIT) is designed to evaluate one patient sample within 15 min without the need for special equipment . This assay classified correctly 103 out of 114 patients, which equates to PPV and NPV of 75% and 97%, respectively. The authors also investigated a sequential combinatorial approach in order to increase specificity. As none of the sequential testing performed better than HemosILR AcuStar HIT-IgG alone as a single test at a cut-off of 1.13 U/mL, Vianello and colleagues  concluded that a single test approach based on the HemosIL® AcuStar HIT with optimized thresholds represented the best strategy to reduce over-diagnosis of HIT whilst preserving test sensitivity.
There are, however, some potentially specific characteristics of the particular patient cohort studied, which may have influenced the study interpretation and conclusions. First, the rate of ‘true HIT’ (defined as intermediate/high 4Ts score and positive HIPA) in their cohort of consecutively referred patients was much higher (at 25.9%) than typically reported in the literature (10%–15%) . This high rate might suggest a more stringent pre-selection criteria applied by physicians at the authors’ institution. However, the data summarized in Table 1 appear at odds with this scenario; given that the median 4Ts score for the HIT negative subpopulation was 3 with the interquartile range between 1 and 6. Thus, it appears that in contrast to most recommendations , a significant proportion of (at least 42) patients with a low 4Ts score (≤3) were referred for serological investigation. Consistent with previous reports of relatively lower incidence of HIT in medical patients [8, 9], patients exposed to heparin for medical reasons were the most represented demographic in the HIT negative group (51%). The highest rate of HIT diagnosis was observed among orthopedic patients (50%, albeit a small group overall), closely followed by cardiovascular surgical patients (43%), which also represented the largest subgroup in the HIT positive subpopulation. Considering some of these peculiarities, it is difficult to predict whether the findings of the study by Vianello and colleagues  will be widely generalizable, and thus confirmatory studies are required.
Importantly, two HIT-negative patients reported by Vianello and colleagues  still had a clinical course complicated by thrombotic events. It is feasible, then, in the light of the second article featured here , that these patients may have had anti-protamine-heparin antibodies, which can also be platelet-activating and have been recently reported in patients with clinical manifestations similar to HIT . Protamine, a cationic DNA-binding protein derived from salmon sperm, is administered as protamine sulfate to reverse the anticoagulant effect of heparin following cardiopulmonary bypass surgery. Protamine is also used as a stabilizer in long-acting isophane insulin [11–13]. Immunogenicity of protamine has been recognized for decades [14–17] in several patient groups including vasectomized men, diabetic patients on isophane insulin and patients requiring reversal of heparin with protamine sulfate. The most widely reported and feared side effects of intravenous protamine sulfate include an immediate anaphylaxis, or delayed onset non-cardiogenic pulmonary edema or anasarca . First reports regarding anti-protamine-heparin antibodies in so called ‘pseudo-HIT’ surfaced about 5 years ago  followed by confirmation of high immunogenicity of the protamine-heparin complex in a mouse model . Panzer and colleagues  employed four commercially available immunoassays to extensively assess for anti-protamine (protamine-heparin) antibodies in six (5 cardiopulmonary bypass surgery and 1 diabetic) patients with clinically suspected, but serologically excluded classical HIT. Two of those patients succumbed to a thrombotic event. The authors evaluated two poly-specific assays (Stago Asserachrom HPIA and Diamed PaGIA), and two IgG-specific assays (Milenia Biotec LFI-HIT and Hyphen Biomed Zymutest HIA IgG). Patient samples were also tested by functional assays (heparin-platelet aggregation test and HIPA) to confirm if the antibodies were functionally capable of inducing platelet activation. Zymutest HIA uses platelet lysate as a source of PF4 (and possibly other proteins capable of binding to heparin), which is immobilized onto protamine-coated ELISA plate. The authors also used a variant of Zymutest HIA in which heparin was surface-immobilized by streptavidin-biotin system to test the patient plasma with either platelet lysate or different concentrations of protamine sulfate. Furthermore, the authors tested for antibody binding to plates coated directly with protamine.
Whilst functional assays failed to detect platelet-activating heparin dependent antibodies in any of the patients’ samples, all patients tested strongly positive by Zymutest HIA (but additionally were not inhibited by excess heparin) and borderline positive by the Asserachrom HPIA. The LFI-HIT was negative with all tested samples whilst only one patient tested positive by PaGIA. Interestingly, the Zymutest HIA reactivity was not dependent on the presence PF4 or other heparin binding proteins found in platelet lysate suggesting that the IgG antibodies were not directed at PF4-heparin but most likely toward the protamine-heparin complex. This interpretation was further corroborated by strong binding to biotinylated heparin in the presence of protamine sulfate. However, strong binding was also observed to plates coated with protamine only. Whether this reflected a polyclonal mixture of antibodies directed to protamine-heparin complexes plus protamine, or antibodies capable of identifying both is not entirely clear, but Panzer and colleagues  concluded that the antibodies detected in the plasma samples from these six patients were ‘anti-protamine antibodies with cross-reactivity against protamine-heparin complexes’. It also appears that Zymutest HIA is not only capable of detecting IgG antibodies against PF4-heparin but also against protamine-heparin complex or protamine alone.
Of interest, one of these six patients (patient #2) was the index case that reportedly inspired Bakchoul et al.  to define the frequency of anti-protamine-heparin antibodies before and after cardiopulmonary bypass in a cohort of 591 patients. The authors concluded in that study that only anti-protamine-heparin IgG antibodies (prevalence of 4.1% by Day 10 post-surgery) were pathogenic as evidenced by their ability to induce platelet activation by cross-linking of FcγRIIa only in the presence of both heparin and protamine in vitro as well as decreased survival of human platelets using a mouse model. The presence of these transient antibodies (most disappeared by 120 days post-surgery) was found to be an independent risk factor for early thromboembolic complications in this patient cohort. Of interest, 63% of patients in this cohort who tested positive for anti-protamine-heparin IgG, also tested positive for anti-PF4-heparin IgG. Although this high coincidence may be a reflection of the common underlying process leading to immunogenicity of several antigens, onset of thrombocytopenia and thrombosis appears to be a major clinical difference between protamine-induced thrombocytopenia (early onset) and heparin-induced thrombocytopenia (1–2 weeks after surgery) .
In conclusion, HIT is a difficult condition to diagnose, both clinically and by laboratory testing. Clinical diagnosis is aided by scoring systems, such as 4Ts and HEP , but these remain imperfect, and some patients with high scores will show absence of pathogenic anti-PF4-heparin antibodies . However, laboratory testing for HIT is also imperfect, and there is some non-concordance of assays such that some detect more cases than others . Moreover, a proportion of patients found to be negative in these assays may indeed have other pathogenic features, and some will be identified to have anti-protamine-heparin antibodies . The story of heparin-induced thrombocytopenia associated with thrombosis is therefore a story still in the making.
Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Financial support: None declared.
Employment or leadership: None declared.
Honorarium: None declared.
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
1. Vianello F, Sambado L, Scarparo P, Lombardi A, Bernardi D, Plebani M, et al. Comparison of three different immunoassays in the diagnosis of heparin-induced thrombocytopenia. Clin Chem Lab Med 2015;53:257–63.10.1515/cclm-2014-0809Search in Google Scholar PubMed
2. Panzer S, Schiferer A, Steinlechner B, Drouet L, Amiral J. Serological features of antibodies to protamine inducing thrombocytopenia and thrombosis. Clin Chem Lab Med 2015; 53:249–55.10.1515/cclm-2014-0664Search in Google Scholar PubMed
3. Cuker A. Clinical and laboratory diagnosis of heparin-induced thrombocytopenia: an integrated approach. Semin Thromb Hemost 2014;40:106–14.Search in Google Scholar
5. Cuker A, Gimotty PA, Crowther MA, Warkentin TE. Predictive value of the 4Ts scoring system for heparin-induced thrombocytopenia: a systematic review and meta-analysis. Blood 2012;120:4160–7.10.1182/blood-2012-07-443051Search in Google Scholar PubMed PubMed Central
6. Smock KJ, Ledford-Kraemer MR, Meijer P, Hsu P, Van Cott EM. Proficiency testing results for heparin-induced thrombocytopenia in North America. Semin Thromb Hemost 2014;40:254–60.10.1055/s-0034-1365842Search in Google Scholar PubMed
8. Creekmore FM, Oderda GM, Pendleton RC, Brixner DI. Incidence and economic implications of heparin-induced thrombocytopenia in medical patients receiving prophylaxis for venous thromboembolism. Pharmacotherapy 2006;26:1438–45.10.1592/phco.26.10.1438Search in Google Scholar PubMed
9. Girolami B, Prandoni P, Stefani PM, Tanduo C, Sabbion P, Eichler P, et al. The incidence of heparin-induced thrombocytopenia in hospitalized medical patients treated with subcutaneous unfractionated heparin: a prospective cohort study. Blood 2003;101:2955–9.10.1182/blood-2002-07-2201Search in Google Scholar PubMed
10. Bakchoul T, Zöllner H, Amiral J, Panzer S, Selleng S, Kohlmann T, et al. Anti-protamine-heparin antibodies: incidence, clinical relevance, and pathogenesis. Blood 2013;121:2821–7.10.1182/blood-2012-10-460691Search in Google Scholar PubMed
12. Ellerhorst JA, Comstock JP, Nell LJ. Protamine antibody production in diabetic subjects treated with NPH insulin. Am J Med Sci 1990;299:298–301.10.1097/00000441-199005000-00002Search in Google Scholar PubMed
13. Federlin K, Becker F, Hammes P, Marhoffer W, Discher T. Diabetes mellitus and immunology. Klinisches Labor 1991;37:206–10.Search in Google Scholar
14. Samuel T, Kolk AH, Rumke P, Van Lis JM. Autoimmunity to sperm antigens in vasectomized men. Clin Exp Immunol 1975;21:65–74.Search in Google Scholar
16. Weiss ME, Nyhan D, Peng Z, Horrow JC, Lowenstein E, Hirshman C, et al. Association of protamine IgE and IgG antibodies with life-threatening reactions to intravenous protamine. N Engl J Med 1989;320:886–92.10.1056/NEJM198904063201402Search in Google Scholar PubMed
17. Adourian UA, Hirshman CA, Adkinson Jr NF, Weiss ME. Immunoreactivity of protamine preparations used to reverse heparin anticoagulation. Anesthesiology 1990;73:328–31.10.1097/00000542-199008000-00021Search in Google Scholar PubMed
18. Amiral J, Peyrafitte M, Vissac A. Association of protamine sulfate antibodies with ‘Pseudo-HIT’ in heparin treated patients. Hamostaseologie 2009;29:PP4.-9:A65 (Abstract).Search in Google Scholar
19. Chudasama SL, Espinasse B, Hwang F, Qi R, Joglekar M, Afonina G, et al. Heparin modifies the immunogenicity of positively charged proteins. Blood 2010;116:6046–53.10.1182/blood-2010-06-292938Search in Google Scholar PubMed PubMed Central
©2015 by De Gruyter