Diagnosis and management of heparin‐induced thrombocytopenia: Third edition

This guideline was compiled according to the BSH process at (https://b-s-h.org.uk/media/16732/bsh-guidance-development-process-dec-5-18.pdf). The Grading of Recommendations Assessment, Development and Evaluation (GRADE) nomenclature was used to evaluate levels of evidence and to assess the strength of recommendations. The GRADE criteria can be found at http://www.gradeworkinggroup.org. A literature search was carried out using the terms given in Appendix until August 2022. The review of the guideline followed the standard BSH guidelines procedure. Following review of the draft guideline by the BSH Haemostasis and Thrombosis Task Force and the BSH Guidelines Committee, it was placed on the members section of the BSH website for comment (sounding board). This guideline updates and widens the scope of the previous British Society for Haematology (BSH) Clinical guidelines for Diagnosis and Management of Heparin-Induced Thrombocytopenia: Second Edition1 to include functional assays in the diagnosis of heparin-induced thrombocytopenia (HIT), when to use direct-acting oral anti-coagulants, and the role of intravenous (IV) immunoglobulins and plasma exchange in the management of HIT and spontaneous HIT. HIT is an immune-mediated, highly pro-thrombotic disorder of platelet activation caused by pathogenic antibodies against a platelet factor 4 (PF4)–heparin complex. It is the most frequent drug-induced immune thrombocytopenia and may lead to life-threatening thrombosis. There are two distinct forms of HIT: type I, also known as heparin-associated thrombocytopenia, which is a non-immunological response to heparin treatment, mediated by a direct interaction between heparin and circulating platelets causing platelet clumping or sequestration, and type II, which is immune mediated. Type I HIT is more frequent than type II, affects around 10%–30% of patients and occurs early, within the first 48–72 h following heparin exposure.2 It generally causes transient, mild thrombocytopenia, and the platelet count returns to normal within 4 days of heparin discontinuation. Type I HIT is benign and is not associated with thrombosis. In contrast, type II HIT is much less frequent, and its incidence ranges from 0.1% to 7% depending on the type of heparin, duration of heparin exposure and patient population. Unfractionated heparin (UFH) is associated with ~10-fold greater risk of HIT than low molecular weight heparin (LMWH).3, 4 HIT typically occurs within 5–14 days of first exposure to heparin and is associated with a significantly increased risk of thrombosis.5 Unlike type I HIT, thrombosis is more frequent in type II HIT and occurs in around 25%–50% of patients.6, 7 When used hereafter in this guideline, 'HIT' refers to type II HIT. The complex formed by the binding of heparin to PF4 acts as an immunogen, leading to immunoglobulin (Ig)G antibody production by B cells. Although IgG is the main driver in the pathogenesis of HIT, there is some evidence to suggest IgM and IgA may also have a pathogenic role.8, 9 The antibodies form a heparin–PF4-IgG molecular complex that binds to platelets via platelet FcγRIIa causing platelet activation and aggregation with the release of more PF4 and microparticles, leading to complement and coagulation activation. Furthermore, activation of monocytes through FcγRIIa leads to the expression of tissue factor and binding of HIT antibodies to PF4/glycosaminoglycan complexes on the surface endothelial cells (ECs) causing their activation, creating a pro-coagulant state.10 Platelet activation, aggregation and the activation of complement, monocytes and ECs all lead to thrombosis with thrombocytopenia. Removal of immune complex-coated platelets by the reticuloendothelial system contributes further to thrombocytopenia.11 Anti-PF4–heparin immune complexes are able to induce NETosis via interaction with FcγRIIa on neutrophils and through neutrophil–platelet interactions. In a microfluidic system and mouse model, it has been shown that HIT immune complexes are able to induce thrombi containing neutrophils, extra-cellular DNA, citrullinated histone H3 and platelets, whereas depletion of neutrophils abolished thrombus formation.12 Thrombosis in HIT may be venous, arterial, microvascular or a combination and can affect virtually any tissue or organ.10, 13, 14 The immunogenicity of PF4–heparin complexes is affected by heparin chain length and the level of sulphation.15 This may explain the higher incidence of HIT following exposure to UFH compared to LMWH and the near absence of risk with fondaparinux.16, 17 The diagnosis of HIT is based on the key aspects of the clinical history combined with confirmation of PF4-heparin antibodies presence by laboratory tests. Final confirmation can come from a demonstration that the antibodies can mediate platelet activation. In these disorders, the antigen site(s) on PF4 that support anti-PF4 antibodies with heparin-independent reactivity are distinct from those with heparin-dependent reactivity seen in classical HIT. Several reviews have addressed the incidence of HIT in different circumstances.24 The incidence of PF4–heparin antibodies is much higher than the HIT syndrome itself.25 Several factors have been reported to affect the incidence of HIT, but these are not consistent between studies, and most of the studies include only a small number of patients. In general, it appears that lower frequencies are seen with LMWH versus UFH, prophylactic versus therapeutic doses, medical versus surgical patients and minor versus major trauma.4, 26 However, to determine which patients merit monitoring, the absolute risk should be considered. In the meta-analysis by Martel, which contained predominantly orthopaedic patients, the incidences were 0.2% with LMWH and 2.6% with UFH.4 The reported incidence of HIT following cardiac surgery was 1.1%, and in patients supported with extracorporeal membrane oxygenation (ECMO), it was 6.6%.27, 28 Although PF4–heparin antibodies developed after cardiac surgery within 30 days in 50% of around 950 cases (as measured in a polyspecific enzyme-linked immunosorbent assay [ELISA]29 using a cut-off optical density [OD] of 0.4 with >50% inhibition by excess heparin), this was not associated with any increase in death or thromboembolism. Therefore, HIT testing should be confined to cases with sufficiently high clinical suspicion that HIT may be present. Audits have confirmed that using lower-specificity assays alone can lead to substantial overdiagnosis of HIT.30 A retrospective analysis of 25 653 medical inpatients found rates of HIT of ≤0.2% in patients on prophylactic LMWH, treatment dose LMWH and prophylactic UFH, but 0.7% on treatment dose UFH.31 In another study of medical inpatients, the incidence of HIT with subcutaneous UFH was 0.8% (CI 0.1–1.6) and 1.3% for those on prophylaxis.32 A study of neurology patients showed similar results.33 A prospective study of medical patients given LMWH for prophylaxis or treatment reported an incidence of 0.8%, but this figure may be an overestimate.34, 35 A Cochrane review concluded the risk of UFH following surgery was 2.2%, compared to 0.5% in patients receiving LMWH.36 The risk of HIT is very low in obstetric patients given LMWH. A systematic review identified 2777 pregnancies in which LMWH was given.37 In the 2603 pregnancies given LMWH as prophylaxis, there were two cases of thrombocytopenia not thought to be related to heparin, and in the 174 given LMWH as treatment, there was one case of thrombocytopenia also not thought to be related to heparin treatment. In a large administrative database including 66 468 antepartum hospitalisations, 66 741 delivery hospitalisations and 16 325 postpartum readmissions where women received pharmacological prophylaxis, 10 women during antepartum, one during delivery and 14 during postpartum had readmissions involving HIT.38 Of these women, none had arterial thrombosis, limb amputation, heart failure or death related to HIT.38 These factors form the basis of the '4Ts' scoring system (Table 1) to assess the pretest probability of HIT.43 A low 4Ts score (≤3) carries a high negative predictive value (0.998; 95% confidence interval, 0.970–1.000) and so heparin use (platelet count permitting) can continue without further testing. An alternative, more detailed HIT Expert Probability (HEP) score was developed and performed better than the 4Ts score in a retrospective study,44 but in a subsequent prospective study, its sensitivity and specificity were similar to the 4Ts score. However, it performed better than the 4Ts score for trainees and in ICU patients, where assessment is complicated by multiple alternative aetiologies.45, 46 Either system is therefore acceptable, but in both cases, the positive predictive value is poor, and a positive score (≥4) should be followed by laboratory testing. The positive predictive value of the 4Ts score in identifying HIT in patients post cardiopulmonary bypass and ECMO was only 0.562 (18/32) and 0.25 (15/60), respectively,27 suggesting that in these patient populations, a low 4Ts score may not be sufficient to exclude HIT.27 A large variability in calculating the 4Ts score by clinicians was noted, and the experience of attending physicians in calculating the score was crucial. The value of the 4Ts score in the diagnosis or exclusion of HIT can be markedly improved by its calculation jointly by the treating physician and an on-call haematologist with experience in the diagnosis and management of HIT.27, 47 There is a good case for platelet monitoring in patients who have a significant risk of developing HIT. The risk/benefit has not been calculated formally, but consensus exists that an incidence <0.1% in a particular patient group does not require monitoring but >1% does. As detailed above, many groups appear to fall between these limits, and recommendations are therefore somewhat subjective. The value of detecting HIT is high. Even if heparin is discontinued, the risk of developing thrombosis within 30 days is 50%.48 In the Nationwide Inpatient Sample (NIS) study of 97 508 discharges coded for HIT, the in-hospital mortality was 10·1% (SE 0.2) compared to 2.1% (0.01) of 149 811 891 discharges for non-HIT (adjusted OR 4.075 [95% CI 3.846–4.317]; p < 0.0001).49 The harm from monitoring arises largely from the false positive rate, making the correct application of the 4Ts score and laboratory testing of paramount importance. False positives will result in an increase in major haemorrhage due to inappropriate therapeutic anti-coagulation in thrombocytopenic patients, sometimes with non-heparin anti-coagulants with a higher bleeding risk. The cost of monitoring is low; initially a full blood count (FBC), which is frequently indicated for clinical care already should be performed. The burden may be high for patients not in hospitals, but this is likely to be unusual. False positives may also increase costs when monitoring of non-heparin anti-coagulants is required. A baseline platelet count should be obtained before heparin initiation. The vast majority of cases occur 4–14 days after starting heparin; therefore, platelet monitoring can be restricted to this time period. When there has been exposure to heparin in the preceding 100 days then monitoring should begin immediately. Given the seriousness of the condition, the rapidity of complications and the availability of a simple test, it seems valuable to monitor on at least alternate days. Historically, functional platelet aggregation (heparin-induced platelet activation [HIPA]) or specific granule content release assays (serotonin-release assay [SRA]) in the presence of heparin, were the only methods available for the demonstration of anti-PF4/heparin complex antibodies. Although these remain the gold standard for clinically relevant antibodies, the methods are complex and limited to specialist centres. More recently, antigen-based assays have become available on automated platforms and stand-alone devices, shortening the processing time to under 1 h. In some cases, there has been a trade-off in sensitivity and specificity for quicker result generation. The use of more than one immunoassay has been reported to increase the sensitivity and specificity of testing where functional assays are not available.50 Rapid screening tests based on qualitative demonstration of anti-PF4–heparin complex antibodies are available on a range of equipment/devices usually within 1 h. They are widely available in routine diagnostic laboratories with a high degree of sensitivity; however, the trade-off is a lack of specificity for those antibodies that cause platelet activation and/or the detection of non-PF4-heparin antibodies.51, 52 Furthermore, there is no quantification of the antibody concentration present, stratification of which has been used in the assessment of the chances of developing clinical HIT.53 Lateral flow-based assays detect antibodies to PF4–heparin complexes that are bound to gold nanoparticles as they move laterally (fluid phase) along a membrane.54 IgG complexes are immobilised onto the membrane to generate a unique visible line. The assay claims a sensitivity of 100% with reduced false positives compared to other techniques. Reports have shown that fresh samples must be used, with technical issues occurring from frozen or lyophilised material.55 The PaGIA uses centrifuge column technology to capture IgG/A/M antibodies for PF4–heparin complexes (bound to red polystyrene beads) as they are spun through a Sephacryl® column. The PIFA relies on the vertical flow of fresh samples in the presence of PF4-coated microspheres through a membrane filter. Low sensitivity and/or specificity have been reported for both assays and, at the time of writing, have been discontinued.56, 57 A latex immunoassay is available on the ACL TOP analyzer platform (Werfen, Warrington, UK), in which the patient's potential PF4-heparin antibodies compete with latex beads coated with HIT-like antibodies for PF4/polyvinyl sulfonate (PVS) complexes. This makes the presence of patient antibodies to PF4-heparin inhibit the 'normal' agglutination expected in an inversely proportional manner. The benefit of this assay is that it is performed on a widely available analyser with sensitivity and specificity superior to manual rapid assays. Reports using stratification of values into weak, moderate and strong positivity in conjunction with chemiluminescent-based assays achieved 98% comparability with the gold standard SRA.50, 58 ELISA methods for the detection of anti-PF4 antibodies vary in the class of immunoglobulin that is detected and in the way that PF4 is presented for antibody binding. Some include a high-dose heparin confirmation step. In general, they have excellent sensitivity, with 0.97 (CI 0.95–0.99) reported in one meta-analysis using the manufacturer's optical density (OD) cut-offs between 0.3 and 0.559 and essentially ruling out HIT. Heparin-exposed patients often make PF4-heparin antibodies of IgG, IgA and IgM class,53 but IgG antibodies are thought to have the predominant capacity for triggering platelet activation.53 The detection of non-pathogenic IgA and IgM classes contributes to the lower specificity of polyspecific methods that detect all three Ig classes described in meta-analyses.53, 59 The specificity of IgG-specific methods was superior to polyspecific ELISAs, with values of 0.87 (CI 0.85–0.88) for IgG specific and 0.82 (CI 0.80–0.84) for polyspecific assays, respectively, in one meta-analysis,59 with similar results described in another meta-analysis.60 Negative predictive values of IgG-specific and polyspecific ELISAs were both 0.99 (CI 0.99–1.00) in a meta-analysis.59 It is preferable to use IgG-specific assays since they offer superior positive predictive value at 0.56 (CI 0.52–0.61) compared to polyspecific methods at 0.32 (CI 0.28–0.35), although they fall well short of those achieved by functional assays. Positive IgG-specific ELISA alone does not confirm the presence of HIT. There is variability between the results obtained by different ELISA kits,59, 61 but in general, the stronger the OD signal in an ELISA, the more likely it is that a functional assay will be positive and therefore that a diagnosis of HIT can be confirmed.53 The probability of strongly positive SRA HIT antibodies being present reached 50% or more when the OD was 1.4 or higher in an IgG-specific ELISA.51 A weak positive OD in the range 0.4–1.0 with either of the two methods indicated a low probability of HIT, as defined by a strongly positive SRA.51 In another study, patients with an OD >1.0 using a commercial kit demonstrated a nearly sixfold increased risk of thrombosis compared to cases with an OD of 0.4–0.99.62 Use of a higher OD threshold improves the specificity of IgG-specific assays to >90%.60 Non-heparin-dependent anti-PF4 antibodies, which can cause VITT and adenovirus-associated VITT-like disorder,23 can be associated with elevated ODs of more than 1.0 in IgG, and polyspecific ELISA methods are being used for HIT diagnosis, with variability between results obtained with different kits.63, 64 Currently used commercial CLIA HIT assays have an analysis time of approximately 35 min (i.e. much shorter than current ELISA methods, which take 2–4 h). This facilitates a rapid turnaround time and the ability to offer 24-h test availability. The sensitivity of the IgG-specific CLIA method is >95%.60, 65 The specificity of IgG-specific CLIA was consistently superior to IgG-specific ELISAs in multiple studies and meta-analyses52, 60, 65, 66 when using the manufacturer's cut-off of 1.0 arbitrary units/mL as a threshold for positivity. A specificity of >94% was reported in most studies.52, 60, 65, 66 Currently, CLIA provides the best combination of sensitivity/specificity and accessibility for HIT diagnosis in the absence of a functional assay.67 Platelet activation assays can be subdivided into assays that measure either platelet aggregation or a specific marker indicative of platelet activation. All these assays require a source of donor platelets that have been proven to be responsive to the presence of patient serum containing anti-PF4 antibodies in addition to heparin. All assays are dependent on the detection of platelet activation in the presence of material from HIT patients and an appropriate concentration of heparin. Confirmation of HIT is supported by inhibition of activation in the presence of an excess heparin concentration. Sources of donor platelets may be whole blood (WB), washed platelets (WPs) or platelet-rich plasma (PRP). Their complexity means they should only be performed in experienced centres. The HIPA relies on the visual inspection of platelet aggregation (over regular intervals up to 45 min) in the presence of high and low doses of heparin in a U-bottomed microtitre plate well while being stirred with a steel ball.68 The assay depends on careful screening/selection of known 'reactive' donors prior to testing. Enhanced sensitivity has been reported by using WP instead of PRP in the assay, considered by some a 'gold standard' reference method.69, 70 HIPA and SRA have a good correlation for positive HIT cases, reaching 84% concordance in a recent retrospective analysis.71 Standard LTA relies on detecting increased light transmission that correlates with the formation of platelet aggregates in the presence of both patient plasma and heparin (0.1–0.5 IU/mL) and which is inhibited by an increased concentration of heparin (100 IU/mL) in the test system. LTA with WP has been reported to be more sensitive than using PRP.72 Overall sensitivity has been reported to be between 85%73 and 69%.74 LTA can also be performed using adenosine triphosphate (ATP) release as a measure of platelet activation. LTA-positive results are defined as >20% aggregation in light transmission or detection of ATP production. HIT antibody detection is confirmed by inhibition of aggregation or ATP release by more than 50% in the presence of 100 IU/mL heparin. Multiple electrode aggregometry, also referred to as heparin-induced multielectrode aggregometry (HIMEA), utilises WB as a source of platelets and has been reported to provide results with a sensitivity ranging from 90%75 to 81%.76 Specificity has been reported to be 95%, which compares well to results obtained using SRA.66, 67 The SRA has been referred to as a 'gold standard' functional assay, mainly due to its high level of specificity and sensitivity.77-79 However, at present, SRA is not available in the United Kingdom. Several alternative methods have been developed to avoid the use of radioisotopes and to detect serotonin release, namely, high-performance liquid chromatography,80 ELISA81 and flow cytometry.82 Key determinants for the choice of assays include time, cost and expertise available to the clinician/service trading off against sensitivity and specificity. A summary of various laboratory assays for the diagnosis of HIT is presented in Table 2. In conjunction with pretest probability scoring, single-assay rapid screening techniques can exclude a number of cases with reasonable certainty with limited time, cost and expertise, making them suitable for peripheral centres. This drops off rapidly if scoring or circumstances are more complex or limited. In this scenario, automated latex-based immune and chemiluminescent assays are preferred and can provide <1 h results 24/7 in a wide range of routine testing environments.61, 66 Combination of these assays with either rapid screening or ELISA-based techniques have been reported to reach near equivalence to the gold standard functional-based assays, although time and cost constraints must be balanced.85 Functional-based assays are currently mostly reserved for expert centres as a follow-up/second-line consideration, often in conjunction with automated assays. Morel-Kopp et al.75 Pishko et al.142 Tardy et al.77 Warkentin et al.79 Fouassier et al.81 These diagnostic criteria mean samples from suspected patients with spontaneous HIT must be sent to specialised laboratories for SRA or another functional assay such as the heparin-induced platelet activation test. This should not delay the initial treatment for these patients. However, strict diagnostic criteria as suggested above will avoid overdiagnosis in patients with unexplained thrombocytopenia and positive PF4-heparin antibodies by ELISA. The pattern of results in laboratory tests for HIT has been summarised in a number of cases of spontaneous HIT.19 More recently, there are several case reports of individuals with monoclonal gammopathy of clinical significance (MGCS)88, 89 and, following a recent adenovirus infection, developing anti-PF4 platelet-activating VITT-like antibodies, causing clinical VITT-like syndrome.23 It is important to have high clinical suspicion and test for the anti-PF4 platelet-activating antibodies using ELISA but not by LIA or CLIA, as the latter two assays typically provide negative results for VITT-like antibodies. At present, there is insufficient evidence to indicate the duration of treatment with a non-heparin anti-coagulant in patients who develop spontaneous HIT. HIT is a pro-thrombotic disorder. Therapeutic anti-coagulation with an alternative non-heparin anti-coagulant is required in a person with HIT when the PTPS is high or the diagnosis is confirmed.1 The choice of future thromboprophylaxis is dependent on the previous history of HIT. Patients with confirmed HIT should be given a diagnosis/alert card at discharge from the hospital (the suggested format of a HIT alert card is provided in Figure S1). In all cases of suspected or proven HIT, any form of heparin should be avoided, including heparin flushes.1, 6, 24 In addition, to curtail the pro-thrombotic effect of PF4/heparin complexes in patients with proven HIT, anti-coagulation with an alternate drug to heparin is necessary.1, 6, 24 The currently available options include both parenteral and oral formulations. The recommended duration of treatment in a patient with confirmed HIT, in the absence of thrombosis, is an alternative anti-coagulant for at least 4 weeks or until the platelet count is greater than 150 × 109/L, whichever is later.1 If HIT is associated with a thrombotic complication, 3 months of therapeutic anti-coagulation is warranted.1 The two direct thrombin inhibitors that can be used in this setting are argatroban and bivalirudin, and both require continuous intravenous infusion.90-92 The initial dose of argatroban may require reduction in critically ill patients with liver dysfunction (drug elimination by hepatobiliary clearance), while dose reduction is required in moderate to severe renal dysfunction for bivalirudin.6, 93, 94 It has been recommended that argatroban monitoring be performed using an activated partial thromboplastin time (APTT) with a therapeutic range of 1.5–3.0 times the patients baseline APTT.95 A number of studies have demonstrated major limitations of APTT for monitoring argatroban90, 96, 97 including the variation of APTT based on the method98 and reagents96 used for assessment of APTT as well as influence by coagulopathies, lupus anti-coagulant and raised factor VIII levels.99 Measurement of argatroban concentration is the preferred method of monitoring, with a target of 0.4–1.5 μg/mL suggested by a Swiss guideline.100 A French guideline recommends a target of 0.25–1.5 μg/mL, especially in patients with a prolonged baseline APTT prior to commencing argatroban, as it is not safe to use APTT to monitor argatroban in such situations.101 A single-centre retrospective study of 133 patients treated with argatroban compared two anti-IIa assays for argatroban: HEMOCLOT™Thrombin Inhibitor assay (HTI) (75 patients, target range 0.4–1.2 μg/mL) and ecarin chromogenic assay (ECA) with monitoring by APTT (68 patients, target range by APTT 50–80 s).99 The study showed a reduction in argatroban dosing requirements by approximately 67% without an increase in thrombosis. There was no difference in the incidence of bleeding between the two groups.99 However, assays to quantify argatroban levels are not readily available in many laboratories in the UK at present. Two other parenteral indirect anti-coagulants can be used in HIT patients. Danaparoid sodium is administered either by intravenous infusion or subcutaneously, while fondaparinux is given subcutaneously.102-105 Both of these drugs are cleared through the kidneys and can be measured if required using calibrated anti-Xa assays. Since HIT is a pro-thrombotic state and vitamin K antagonists (VKAs) can deplete endogenous anti-coagulants, VKAs should be started only after the platelet count has normalised in confirmed HIT.106 At least 5 days of concomitant parenteral therapy are needed during the transition.1 Care should be taken when converting direct thrombin inhibitors (argatroban and bivalirudin) to the VKA because the international normalised ratio (INR) may be affected by the thrombin inhibition. In the case of argatroban, during the transition, the target INR should be set at 4; argatroban is stopped once two INR values are in the desired therapeutic range; and subsequently, a repeat INR should be performed 4–6 h later to obtain an accurate INR. In the last decade, there have been several observational studies, cohort analyses, and case reports of the use of DOACs (oral Xa inhibitors [rivaroxaban, apixaban and edoxaban] and an oral thrombin inhibitor [dabigatran]) for patients with HIT with and without thrombosis.107, 108 An important consideration in many of these cases is that a parenteral anti-coagulant was used first before transition to the DOAC. In a systematic review that included 54 patients with HIT (48% with thrombosis at HIT diagnosis), only one patient had thrombus progression, while three had clinically relevant major bleeding but no HIT-related mortality.109 It is useful to remember that DOACs should not be used when an arterial thrombotic event occurs during HIT.6 The choice of anti-coagulant in patients with HIT, depending on the clinical situation, is summarised in Table 3. Fondaparinux subcutaneously once daily or oral anti-coagulants once platelet count to normalised or returned to baseline The optimal application of these therapies is unclear. IVIG may act by competitively blocking FcγRIIa-receptor-mediated heparin-independent platelet activation.110 PEx likely removes antibodies against PF4/heparin high molecular weight complexes.111 Onuoha et al.112 reported the use of PEx, IVIG and a combination of PEx/IVIG in 113 HIT cases: 26/113 cases used PEx alone and 4/113 cases used IVIG and PEx. A post-treatment platelet count ≥150 × 109/L was achieved in 48% of cases within an average of 6 days. The data suggest 1–2 PEx procedures may be effective in reducing or eliminating PF4–heparin antibodies/immune complexes, with the average volume exchanged being 1.3 (range 1.0–2.0). Another review collated 36 cases of acute HIT treated with IVIG, alongside alternative anti-coagulation110 reported between 2014 and 2019: 21/36 were cases of aHIT, with severe thrombocytopenia (platelet nadir 15) and a high frequency of thrombosis, of which 11/21 were rated as having an 'excellent response' and 6 had a 'good response' to a first course of IVIG, with no thrombotic sequelae. To achieve the best response, it is recommended to use the total dose of 2 g/kg over 48 h based on actual body weight, as the treatment failures are usually seen when patients receive doses lower than the recommended total dose of 2 g/kg. Treatment with IVIG and/or PEx could be considered in patients who have severe HIT syndrome, present with a clear case of aHIT, when there is a lack of access to a non-heparin anti-coagulant or in cases where clinically significant bleeding prevents the use of therapeutic anti-coagulation. Re-exposure to heparin following the history of