Portal de Periódicos da CAPES

Clinical sample investigation (CSI) hematology: pinpointing the precise onset of heparin‐induced thrombocytopenia (HIT)

2007; Elsevier BV; Volume: 5; Issue: 3 Linguagem: Inglês

10.1111/j.1538-7836.2007.02360.x

1538-7933

Theodore E. Warkentin, Jo‐Ann I. Sheppard,

Venous Thromboembolism Diagnosis and Management

Prospective studies show that immune heparin‐induced thrombocytopenia (HIT) typically begins 5–10 days after cardiac surgery, even though almost all patients have previously been exposed to unfractionated heparin (UFH) [1Pouplard C. May M.A. Regina S. Marchand M. Fusciardi J. Gruel Y. Changes in platelet count after cardiac surgery can effectively predict the development of pathogenic heparin‐dependent antibodies.Br J Haematol. 2005; 128: 837-41Crossref PubMed Scopus (140) Google Scholar]. The reason is that preoperative UFH exposure (e.g. at heart catheterization or for treatment of myocardial ischemia) is relatively unlikely to produce high levels of heparin‐dependent platelet‐activating antibodies, whereas formation of such pathologic 'HIT antibodies', beginning 5 days after cardiac surgery, is common [2Warkentin T.E. Greinacher A. Heparin‐induced thrombocytopenia: recognition, treatment, and prevention. The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy.Chest. 2004; 126: 311S-37SAbstract Full Text Full Text PDF PubMed Scopus (756) Google Scholar, 3Warkentin T.E. Sheppard J.I. No significant improvement in diagnostic specificity of an anti‐PF4/polyanion immunoassay with use of high heparin confirmatory procedure.J Thromb Haemost. 2006; 4: 281-2Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar]. However, we recently encountered a patient case that proved to be an exception to this general rule, as the platelet count fell dramatically and progressively during the early post‐cardiac surgery period (Fig. 1). On postoperative day 8, the hematology service was consulted when a strong positive platelet serotonin release assay (SRA) result for HIT antibodies was reported by the laboratory (using serum obtained on postoperative day 2): 100% serotonin release in the presence of therapeutic concentrations of heparin (normal, <10% release); 3% release if Fc receptor‐blocking monoclonal antibody was added prior to addition of heparin; 43% release in the absence of heparin; and 0% release in the presence of 100 IU mL−1 heparin. In addition, an in‐house enzyme immunoassay (EIA) that detects IgG antibodies against PF4–heparin complexes (anti‐PF4/H‐EIA) was also strongly positive [2.64 optical density (OD]) units; normal <0.45 units]. Upon review of the serial platelet counts, it was difficult to ascertain precisely when the HIT began, as the largest proportion of the platelet count fall coincided with the use of cardiopulmonary bypass (CPB) during cardiac surgery that was performed for acute flail mitral valve disease (with associated use of blood products: 2 units of red cell concentrates, 1 unit of fresh frozen plasma, and 1 unit of 5% albumin). Furthermore, the thrombocytopenia steadily worsened during the first 5 postoperative days, without any clear abrupt drop that might have pinpointed the onset of HIT. However, we were able to use a 'clinical sample investigation' (CSI) hematology approach to determine the precise timing of anti‐PF4/heparin seroconversion, as follows. First, we identified and retrieved all of the patient's as‐yet‐undiscarded blood specimens (collected into three different anticoagulants) that were still available within the four different sections of the laboratory: (i) hematology [ethylenediaminetetraacetic acid (EDTA)]; (ii) transfusion medicine (EDTA); (iii) coagulation (sodium citrate); and (iv) clinical chemistry (lithium heparin). Then, we performed the anti‐PF4/H‐EIA using all of these available samples. Although the plasma specimens collected into EDTA and sodium citrate confirmed strong positive test results for anti‐PF4/heparin antibodies of IgG class, they were unable to provide the precise date of seroconversion, due to lack of available samples during the crucial time period. However, 10 consecutive daily lithium heparin‐anticoagulated blood samples, collected beginning 3 days prior to cardiac surgery, were still available from the chemistry laboratory. When these samples were tested 'as is' in the EIA, all of the results were negative, consistent with false‐negative results resulting from collection into the lithium heparin anticoagulant. As there are 51 units of lithium heparin in a vacutainer yielding a 3‐mL blood draw (or about 2 mL of plasma), this would result in a heparin concentration of about 0.5 U mL−1 using plasma subsequently diluted 1 : 50 in the EIA; this amount of heparin interferes with binding of HIT antibodies to the PF4–heparin complexes bound to the solid phase (in a parallel experiment, we showed that heparin at concentrations of 0.20 U mL−1 or greater in the EIA causes inhibition of reactivity in the EIA). We therefore used a positively charged anion exchanger, Ecteola cellulose (Sigma Aldrich, Inc., St Louis, MO, USA), to 'deheparinize' the lithium heparin plasma samples. According to the manufacturer's product insert (for procedure no. H9895/H 9896), and as reported in the literature [4Visentin G.P. Ford S.E. Scott P.J. Aster R.H. Antibodies from patients with heparin‐induced thrombocytopenia/thrombosis are specific for platelet factor 4 complexed with heparin or bound to endothelial cells.J Clin Invest. 1994; 93: 81-8Crossref PubMed Scopus (544) Google Scholar], 70 mg of Ecteola cellulose is capable of binding up to 20 U of heparin per mL of plasma. In parallel experiments, we corroborated data in the literature [4Visentin G.P. Ford S.E. Scott P.J. Aster R.H. Antibodies from patients with heparin‐induced thrombocytopenia/thrombosis are specific for platelet factor 4 complexed with heparin or bound to endothelial cells.J Clin Invest. 1994; 93: 81-8Crossref PubMed Scopus (544) Google Scholar] that this amount of Ecteola cellulose successfully removed up to 100 U mL−1 heparin added to known HIT serum (after a 10 min incubation and centrifugation for 10 min at 1500 × g), as assessed by recovery of reactivity in the anti‐PF4/H‐EIA. As shown in Fig. 1, after removal of heparin, the lithium heparin‐anticoagulated plasma samples confirmed a strong positive test result in the anti‐PF4/H‐EIA of similar magnitude to that observed with the EDTA‐anticoagulated samples. Moreover, assessment of the serial plasma samples showed that anti‐PF4/heparin antibody seroconversion occurred on the day prior to surgery (0.57 OD units; normal, 0.45 units), and was strongly positive (1.74 units) on blood obtained on the day of surgery. In addition, the platelet SRA was also positive (40% serotonin release) for the (deheparinized) sample obtained on the morning of cardiac surgery, as well as for the sample obtained on postoperative day 1 (80% release). Interestingly, the platelet count had already fallen by 20% (from 250 to 200 × 109 L−1) on the morning of surgery, suggesting that this patient underwent cardiac surgery coincident with the early phase of clinically evident HIT. Despite minimal exposure to postoperative heparin (no postoperative heparin 'flushes' was given, and only one 5000‐U subcutaneous injection of UFH was administered on postoperative day 3), the platelet count continued to fall progressively and dramatically, reaching its nadir (28 × 109 L−1) on postoperative day 5. This course is consistent both clinically and serologically with 'delayed‐onset HIT', as the platelet count fall continued after cessation of heparin, and because the patient's serum (postoperative day 2 sample) produced considerable platelet activation (43% serotonin release; normal, <10%), even in the absence of added heparin [5Warkentin T.E. Kelton J.G. Delayed‐onset heparin‐induced thrombocytopenia and thrombosis.Ann Intern Med. 2001; 135: 502-6Crossref PubMed Scopus (375) Google Scholar]. In this patient case, it is clear that the perioperative HIT episode was triggered by the preoperative use of therapeutic‐dose UFH that was begun 6 days prior to cardiac surgery. Interestingly, despite CPB anticoagulation with UFH in a patient in the early phase of clinical HIT, no overt thrombosis developed. Routine duplex ultrasonography of all four limbs showed only a very small thrombus at the right internal jugular catheter site (performed on postoperative days 8 and 13). When the diagnosis of HIT was made on postoperative day 8, the prophylactic‐dose danaparoid was increased to a therapeutic‐dose regimen, as per consensus conference recommendations [2Warkentin T.E. Greinacher A. Heparin‐induced thrombocytopenia: recognition, treatment, and prevention. The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy.Chest. 2004; 126: 311S-37SAbstract Full Text Full Text PDF PubMed Scopus (756) Google Scholar], with achievement of therapeutic levels of danaparoid thereafter. The importance of achieving therapeutic anticoagulation was shown by serial tests for disseminated intravascular coagulation (DIC): a positive test result for fibrin monomers was observed on postoperative day 8 (despite 72 h of prophylactic‐dose danaparoid), which became negative when therapeutic‐dose danaparoid levels were attained. The patient remained otherwise thrombosis‐free at 6‐week follow‐up. In summary, this case illustrates the potential value of a CSI hematology approach, in which the retrieval and testing of as‐yet‐undiscarded blood samples – including, even, heparin‐anticoagulanted plasma following Ecteola cellulose 'deheparinization' procedure – can be used to pinpoint the onset of anti‐PF4/heparin IgG seroconversion, and thus bring diagnostic clarity to a potentially confusing and complex clinical situation. The authors state that they have no conflict of interest.