A vision for a quicker definitive diagnosis of HIT

In this issue of Blood, Kanack and colleagues¹ provide proof-of-concept for a new approach to definitive diagnosis of heparin-induced thrombocytopenia (HIT). Using cryopreserved platelets in an enzyme-linked immunosorbent assay (ELISA)-based platelet activation assay with serum from patients suspected of HIT, they could distinguish HIT-positive from HIT-negative samples.

There is a major unmet need in HIT diagnostics for a rapid test that can be performed in a standard laboratory setting. Kanack and colleagues present a clear vision for such a test: an off-the-shelf reagent, cryopreserved platelets, used in a readily performed assay, ELISA, generating accurate, timely, and locally available definitive HIT testing. HIT is diagnosed by both clinical and laboratory findings. Clinically, affected individuals develop thrombocytopenia, and often thrombosis, while receiving heparin. Even when treated appropriately, many patients have not only new or progressive thrombosis, but also major bleeding while on nonheparin anticoagulants, especially in the setting of compromised vital organ function. HIT is particularly problematic among patients with cardiovascular disease, notably those undergoing interventional procedures or support, as well as the medically ill in intensive care units. Unfortunately, thrombocytopenia is common in those same patients, so the burden of suspected HIT is high. Clinical scoring systems, such as the 4 Ts, help to evaluate the likelihood of HIT based on the initial clinical suspicion.² The current standard of care in suspected HIT is to stop all forms of heparin, send diagnostic testing, initiate nonheparin anticoagulation with a direct thrombin inhibitor, and monitor platelet counts, signs of thrombosis, and signs of bleeding.³ 

HIT diagnostic testing has 2 phases. First is an initial immunoassay for the presence of patient antibodies directed against platelet factor 4 (PF4) complexed with a polyanion. (The medicinal glycosaminoglycan heparin is a polyanion.) These PF4 immunoassays are useful when they show a negative result, essentially ruling out HIT. However, a positive result is not specific enough to be diagnostic for HIT, necessitating the second phase, a definitive platelet activation assay. As pointed out by the authors, current definitive HIT platelet activation assays are complex and occur in specialized referral laboratories. As a result, there is a delay in the definitive diagnosis of HIT, exposing patients suspected of HIT but ultimately proven otherwise to expensive anticoagulants and excessive bleeding risk.

Why is current definitive HIT testing so complex? One reason is that from 1 to 4 healthy human donors are required to provide fresh (ie, that day) platelets for the test. Furthermore, there is marked interindividual variation in healthy human platelet responses to HIT immune complexes; therefore, so-called known highly responsive donors, or pedigreed donors, are usually used. A second reason is that the assays of the fresh platelets are not routine in general hospital clinical laboratories, including the radioactive serotonin release assay (SRA), heparin-induced washed platelet aggregation, or P-selectin flow cytometry (one version of which developed by the authors using exogenous PF4 called the platelet excitation assay [PEA]).

It is in this context that Kanack and colleagues provide clever alternatives. Although the cryopreservation of platelets in dimethyl sulfoxide- or trehalose-containing buffers has a long record of providing activatable platelets for transfusion under challenging circumstances,⁴ this methodology is appreciated most by transfusion medicine specialists and less so by practicing hematologists or hemostasis clinical laboratory colleagues. In this work, trehalose-based cryopreservation is shown by the authors to provide platelets that remain able to be activated after thawing, even after shipping on dry ice to another testing center. Current HIT immune complex-mediated platelet activation tests use aggregation, release of dense granules (serotonin), or release of alpha granules (surface exposure of P-selectin from the alpha granule membrane). Kanack and colleagues studied the release of the alpha granule protein thrombospondin-1 (TSP-1), as assayed by a commercially available ELISA. The authors contend, with a strong rationale, that such an ELISA is more readily performed in general hospital laboratories.

How close is this work to achieving this vision? The authors used a limited number of known patient samples, already identified as positive or negative in the conventional assays, that is, SRA and PEA. Clinical information on the patients providing the samples was very limited. They show that, for these patient serum samples, the SRA and PEA are concordant. After presenting initial trial experiments, the authors optimized a method for cryopreservation, thawing and TSP-1 ELISA. In particular, whether their TSP release assay (TRA) is conducted by provision of exogenous PF4 (PF4-TRA) or exogenous unfractionated heparin (hep-TRA), they find that the ratio of sample TSP-1 to that induced by a concurrent normal control serum is diagnostic. If the ratio exceeds 2, it is HIT. If the ratio is close to 1, it is not HIT. Importantly, the ratio should be high with therapeutic heparin (on order 0.5 U/ml) and low at high heparin (on order 100 U/ml).

What is needed next for the vision of the authors to become reality? First, the interindividual variability in healthy donor platelet responses must be resolved. The authors suggest the use of pooled platelets or pedigreed donors in the future. Whichever is chosen needs validation. Second, the control serum should be better described; for example, how many donors provide it, and is it heat inactivated or frozen? Third, this ELISA requires that the laboratory coat the plates overnight before the assay the next day. How will that influence laboratory practice and throughput? Fourth, recombinant human PF4 at a high concentration in PF4-TRA can be expensive; for HIT testing, hep-TRA seems more promising. Finally, HIT platelet activation assays, even in the best of hands, continue to have standardization concerns.⁵ True validated standardization will require that coded patient samples be distributed among independent reference laboratories, which then use identical protocols and reagents to make the call of the assay results. Breaking the code, scoring the results, and linking to the clinical course will begin to make it clear if these new approaches can be used to improve care of patients and to help providers.

Conflict-of-interest disclosure: S.E.M. has an intellectual property interest and is a member of the Scientific Advisory Board of Veralox Therapeutics.