Heparin-induced thrombocytopenia

Heparin-induced thrombocytopenia (HIT) is an immune-mediated syndrome of thrombocytopenia and thrombosis as a result of heparin exposure. HIT is the most common type of drug-induced, immune-mediated thrombocytopenia and is associated with high morbidity and mortality if undiagnosed
^[1]
.

The wide use of heparin in hospitalised patients demonstrates that HIT can develop in patients receiving the therapy for a number of indications, such as atrial fibrillation, venous thromboembolism (VTE), invasive procedures, surgery, via dialysis and even thromboprophylaxis. HIT is more common in surgical patients than in medical patients. Other groups that have shown to be at higher risk are women and older patients
^[2]
.

Studies have documented a variable incidence of HIT, which ranges from <0.1% to 5% depending on the patient population and type of heparin received. In incidence, HIT is ten times higher with unfractionated heparin (UFH) compared with low molecular weight heparin (LMWH)
^[3]
. Up to 50% of patients who develop HIT will present with a thrombus, with venous thrombus being more common than arterial thrombosis, occurring at a 4:1 ratio^[4]
. The thrombotic
sequelae can be limb- and life-threatening, therefore, it is vital not to misdiagnose.

Pathophysiology, pathogenesis and risk factors

The pathophysiology of HIT is similar to a bacterial host defence mechanism where immunoglobulin G (IgG) antibodies bind to the heparin-platelet factor 4 (PF4) complexes. PF4 is a cytokine, a positively charged molecule of unknown function, released by alpha granules of platelets, that binds to the negatively charged heparin in the plasma. This complex undergoes a conformational change, exposing a neo -antigen to which the IgG antibodies bind. The formation of the anti-PF4–heparin complex binds to FcγIIa receptors on circulating platelets, which are removed by splenic macrophages, resulting in platelet consumption and ultimately causing thrombocytopenia. The anti-PF4–heparin complex also leads to platelet activation and aggregation, as well as production of platelet-derived microparticles, promoting excessive thrombin generation, resulting in a hypercoagulable state and leading to thrombosis. Warkentin and Kelton

described the duration of seropositivity as ranging from 50 to 85 days after stopping heparin, but this depended on the type of assay performed
^[5]
.

Not all patients who develop the anti-PF4–heparin antibodies will clinically present with HIT, demonstrating that some of these complexes are not pathogenic. The high incidence of HIT associated with surgical patients is related to the concentration of circulating PF4; these molecules are released at the point of trauma to the endothelial cells
^[6]
, increasing the chances of antigenic complexes forming. In addition, the wide spread use of UFH
intraoperatively increases the chances of HIT developing.

The occurrence of HIT is dependent on the duration and formulation of heparin administered because molecular weight, chain length and degree of sulphation affects the binding to PF4. UFH is derived from animal sources (porcine-origin is only available in the UK) and has a molecular weight ranging from 3,000 to 30,000 Daltons. It is a negatively charged sulphated glycosaminoglycan and has a high binding affinity to PF4. Conversely, LWMHs are depolymerised products of UFH, with smaller molecular weights ranging from 2,000 to 10,000 Daltons and have a lesser binding affinity to PF4. These structural differences explain why HIT is more likely with UFH (1–5%) than with LMWH (0.1–1%)
^[2]
.

Diagnosis

The diagnosis of HIT is based on a combination of clinical manifestations and laboratory tests. Thrombocytopenia, a general term for a low blood platelet count, in hospitalised patients can be the result of several factors, such as sepsis, viral infection, drugs other than heparin, intravascular devices (e.g. cardiopulmonary bypass, intra-aortic balloon pumps, extracorporeal membrane oxygenation [ECMO]), fluid resuscitation (with dilution), disseminated intravascular clotting (DIC) and other autoimmune conditions; therefore, HIT may remain undiagnosed
^[7]
. Conversely, to prevent the over-diagnosis of this rare syndrome, clinical features can be reviewed using the pre-test 4T scoring system (see Table 1) as a differential diagnosis tool to determine the likelihood of HIT. A low 4T score of ≤3 can exclude HIT, therefore, no further investigations would be required
^[8]
. In 2012, a meta-analysis showed that a negative predictive value of a low probability 4T score is 99.8%
^[9]
. Treatment of HIT, as described later, is indicated at a 4T score of >3, while performing laboratory tests for further investigations.

Other HIT probability scores have been developed that factor in specific patient groups, such as patients who have had cardiopulmonary bypass (CPB), which has a good predictive value of 97%
^[10]
. This particular score encompasses the platelet count pattern, time from CPB and CPB duration.

Clinical features

Thrombocytopenia

Although the definition of thrombocytopenia is a platelet count less than 150×10

⁹/L, a fall in platelets by 50% from baseline can be a marker of HIT. The onset of thrombocytopenia is usually between five and ten days of receiving heparin. A faster onset, within minutes to hours, can occur in patients who have had a recent exposure (typically within the past 100 days) to heparin and, therefore, developed anti-PF4–heparin antibodies
^[2],[11]
.

The British Society of Haematology recommend that all patients have a baseline platelet count before heparin administration because this can assist in the diagnosis of HIT if fit were to develop^[2]
. For patients who are considered at a greater risk of developing HIT, such as surgical patients receiving UFH, regular monitoring of platelets is recommended every two to three days.

Thromboembolic events

Despite thrombocytopenia, bleeding is an uncommon complication of HIT. The occurrence of thrombosis is associated with a mortality rate of 30% in HIT
^[1]
. These events most commonly present as deep vein thrombosis or pulmonary embolism and, less so with arterial thrombus, presenting as a heart attack, stroke or peripheral arterial thrombosis leading to limb necrosis that requires amputation. Post-operative patients are more likely to develop a VTE, whereas arterial thrombosis is more prevalent in patients with a history of cardiovascular disease
^[12]

Heparin-induced skin lesions

Skin lesions are rare and can occur at the site of injection, which can become necrotic. This inflammatory response is usually accompanied by fever^[2]
.

Laboratory testing

Following the initial screening stages (i.e. the 4T test and clinical features), laboratory tests should be conducted to test for the presence of antibodies. For suspected HIT, the two main tests are immunoassays and functional tests.

The immunoassay, specifically enzyme-linked immunosorbent assay (ELISA), is used to detect the IgG -only or IgG/A/M antibodies that bind to the PF4–heparin complexes. If the ELISA test is positive, it is repeated using high-dose heparin (100units/ml) as a confirmatory step. These tests are highly sensitive but not specific because non-pathogenic antibodies are also detected, therefore, the chances of false positives are likely with the potential of over-diagnosis, which highlights the importance of the initial screening stage. The result of this test is measured in optical density; a reading above one would indicate a high risk of thrombosis
^[2]
.

For patients who score between four and five on the 4Ts score, which is classified as intermediate, a negative particle gel immunoassay can be conducted. This can be used as rapid test to rule out diagnosis, however, it is less sensitive. A positive result would indicate further specific tests are required and an alternative anticoagulant should be initiated^[2]
.

The heparin-induced platelet activation (HIPA) test and serotonin release assay (SRA) are examples of these functional tests. The HIPA test detects the heparin-dependent platelet aggregation. This is conducted by introducing heparin low dose (0.1–0.5units/ml) and high dose (100units/ml) to the patient’s plasma. This is then quality controlled with a known HIT patient’s plasma
^[11]
. The SRA is a gold standard test, but it involves radiation techniques. Functional tests are the preferred methods and are only conducted in specialist laboratories. Washed platelets achieve greater sensitivity^[2]
.

Treatment

The initial management of suspected HIT is to stop all heparin, including heparin locks and flushes. As a result, the platelet count starts to rise within two to three days and typically reaches baseline within four to ten days, whereas it can take up to two to three months for the antibodies to clear. Prophylactic platelet transfusion is not recommended as it may exacerbate the patient’s prothrombotic state, but it can be considered in patients with concurrent bleeding complications in view of proven efficacy
^[2],[11]
.

An alternative non-heparin anticoagulant at a treatment dose should be offered even if the patient has no clinical presentation of a thrombosis. The available alternative agents should be selected based on the patient’s clinical status. These are argatroban, bivalirudin, danaparoid and fondaparinux, which are all administered parenterally. Table 2 shows the pharmacological differences and dosing of all agents.

For surgeries that require intraoperative heparin, it is recommended that the procedure be delayed until the antibodies have disappeared.

Vitamin K antagonists (VKAs), such as warfarin, should not be initiated until the platelet count has recovered because these agents are also prothrombotic since they inhibit the body’s natural anticoagulants; proteins C and S. If the patient is on a coumarin at the point of HIT diagnosis, it should be reversed. Once the patient is able to start taking warfarin, there should be an overlap period with the alternative non-heparin anticoagulant. Warfarin should be continued for one or three months
^[2]
for HIT and thrombosis associated with HIT, respectively.

Argatroban is a direct thrombin inhibitor and reversibly binds to thrombin. It is licensed for the treatment of HIT and is administered as a continuous infusion, requiring dose adjustments according to activated partial thromboplastin time ratio (APTTr) with a starting infusion rate of 2µg/kg/min, with a maximum recommended dose of 10µg/kg/min^[13]
. Argatroban reaches steady state within 1 to 3 hours, therefore, the APTTr should be taken 2 hours after initiating the infusion to guide the dose adjustments, with subsequent APTTr to be taken less frequently when in range. Lower doses starting at 0.5µg/kg/min in hepatic impairment or critically ill patients are recommended. In view of argatroban being metabolised via the hepatic system, a dose reduction is required alongside more frequent monitoring of APTTr. Research in the critical care setting has demonstrated that patients are more sensitive to argatroban and require low doses to avoid excessive anticoagulation^[14]
.

However, argatroban use increases prothrombin time, causing a false rise in international normalised ratio (INR), therefore, it can be challenging when switching a patient to longer-term management with a VKA. Consequently, concurrent argatroban and a coumarin should be administered for at least five days, with two consecutive days of INR≥4. An upper target INR for this period is not given, but to prevent over-anticoagulation the recommendation would be to stop argatroban with an INR of >5. In this situation, the argatroban should be stopped for four hours and then INR repeated
^[2]
.

The benefits of argatroban are that it is rapid acting, which significantly reduces the risk of new thrombosis and, in turn, reduces mortality rates^[15]
. This agent would be preferable in patients with end-stage renal disease who are on haemodialysis because no dose adjustments are needed^[9]
. Argatroban has been demonstrated to be a successful anticoagulant in patients with multiple organ dysfunction syndrome and in patients undergoing percutaneous coronary intervention (PCI).

Bivalirudin is an oligopeptide analogue of hirudin, which inhibits thrombin. It is licensed for use in patients undergoing PCI, but its use in HIT is off-label. It is administered as a continuous infusion requiring dose adjustments according to APTTr, with a starting infusion rate of 0.15–2mg/kg/min.

The American College of Chest Physicians (ACCP)
^[11]
recommend the use of bivalirudin for patients diagnosed with HIT who require urgent cardiac surgery or PCI. Cohort studies have demonstrated that the rate of death, Q wave myocardial infarction, repeat operation of cardiac revascularisation or stroke were similar for patients treated with bivalirudin with or without HIT.

The advantage of bivalirudin is that it has the shortest half-life (25 mins) of all the agents, and coagulation reaches baseline within 1 to 2 hours of stopping the infusion. It is a unique anticoagulant because its clearance is predominantly by proteolysis. However, the half-life extends up to four hours in patients on dialysis, therefore, they require a lower initial dose and frequent monitoring
^[16]
.

Danaparoid is a heparinoid, containing a mixture of heparin sulphate, dermatan sulphate and chondroitin sulphate (which inhibits factor Xa, as well as factor II at a lesser degree). This is administered intravenously at a therapeutic dose, as outlined in table 2. Within the European Union, prophylactic doses are also being used, however, these are not recommended in the UK after reports of treatment failures. Monitoring anti-Xa assay and dose adjustments are required for patients with severe renal impairment or body weight above 90kg^[11]
.

In vivo


studies have documented rare cross-reactivity of HIT with danaparoid. Furthermore, this agent has a long half-life, so rapid reversal after discontinuation will not occur. This agent is usually selected for patients with normal renal function. Prothrombin time and INR are not effected by danaparoid, so the transition to a VKA is not as complicated as with agents that influence these monitoring parameters. The majority of evidence for the treatment in pregnant patients who develop HIT is for danaparoid, therefore, it is suggested as a first-line therapy by the ACCP in this patient group
^[11]
.

Fondaparinux sodium is a synthetic pentasaccharide that inhibits activated factor X. Dosage is based on weight and administered subcutaneously once a day, which is ideal for ease of use outside the critical care setting. It does not require routine monitoring, however, it is contraindicated in patients with a creatinine clearance of 20ml/min

^[17]
.

Although this agent has a similar structure to heparin, fondaparinux has weak affinity for PF4 and the incidence of HIT with this agent is minimal. It is successfully used as a treatment for HIT, however, its use for this indication is off-label, as well as controversial in view of its immungenicity

^[18]
.
Fondaparinux can also be considered in pregnant HIT patients if danaparoid is unavailable.

Challenges with treatment options

These treatments are more expensive than heparin, associated with a relatively increased risk of bleeding and no antidote is currently available for any of them. However, this is offset by the short half-life of selected anticoagulants. The challenge remains when transitioning to an oral anticoagulant for long-term management because of the potential of thrombosis if a VKA is prematurely initiated and/or the parenteral anticoagulant is prematurely discontinued.

Possible future therapies

Observational data show that the treatment of acute HIT can be simplified with possible oral options being investigated for efficacy in this context, such as direct oral anticoagulants (DOACs)
^[19]
. Suggestions have been made that high-dose intravenous immunoglobulins (IVIG) may have an adjunctive effect alongside the alternative anticoagulant, however, data are limited.