Approaches to management of HIT in complex scenarios, including cardiac surgery

Visual Abstract

Abstract

Although heparin-induced thrombocytopenia (HIT) presents management challenges for any population, it adds complexity to the management of certain patient populations, including those undergoing cardiac surgery and those with refractory HIT and/or acute bleeding. For each of these scenarios, we review alternative management strategies when standard therapies—heparin cessation and the initiation of a nonheparin anticoagulant—are either insufficient or not practicable. In patients with HIT undergoing cardiac surgery, we review the clinical experience for heparin reexposure using therapeutic plasma exchange (TPE) or antiplatelet therapy. In patients with refractory HIT despite adequate nonheparin anticoagulation, we address the use of intravenous immune globulin, TPE, and rituximab. Finally, in patients with active bleeding, we discuss bleeding management and the risks associated with platelet transfusion. Although they may facilitate a patient-centered approach, most of these strategies are supported by limited evidence.

Learning Objectives

• Evaluate alternative strategies for managing HIT in patients undergoing cardiac surgery

• Compare therapeutic options for patients who, despite adequate nonheparin anticoagulation, develop refractory HIT

• Identify strategies for managing patients with HIT and active bleeding

CLINICAL CASE

A 47-year-old woman underwent mitral and aortic valve replacement surgery. During surgery, she received unfractionated heparin (UFH). On postoperative day 2, warfarin was initiated. Her platelet counts, which were 274 000/uL on admission, decreased to 54 000/uL by postoperative day 5. Ultrasound showed new bilateral upper-extremity thromboses. The suspected diagnosis was heparin-induced thrombocytopenia (HIT). The 4Ts probability score was 6.¹ Anti–heparin platelet factor 4 (PF4-H) screen (immunoglobulin A [IgA], IgG, IgM), using the HemosIL HIT- Ab(PF4-H) latex particle enhanced immuno-turbidimetric assay was 5.594  U/mL (reference range <1.000). Pending a confirmatory serotonin release assay (SRA), warfarin was reversed and she was initiated on bivalirudin.

Introduction

Challenges of HIT Management in Complex Scenarios

Although HIT presents management challenges for any population, in certain scenarios, its management can be especially complex. Complex management challenges occur in patients with HIT who 1) are planned to undergo cardiac surgery, 2) have progressive thrombosis or thrombocytopenia despite adequate therapy, and 3) develop active bleeding while on therapeutic anticoagulation. These scenarios are relatively rare and effective management can be hindered by limited data. Therefore, we have summarized potential management approaches for the practicing hematologist to consider.

CLINICAL CASE (continued)

After the SRA returned positive, the patient became progressively unstable. Therefore, the surgeon elected to return to the operating room for redo mitral valve replacement surgery. However, instead of intraoperative anticoagulation with bivalirudin, the surgeon requested UFH.

Cardiac surgery

The cardiopulmonary bypass circuit and the need for anticoagulation

Cardiac surgery is marked by tissue factor release and inflammation.² Contributing to this inflammatory milieu is the cardiopulmonary bypass (CPB) circuit. When the CPB circuit comes in contact with blood, it activates the contact pathway.³ The subsequent coagulation activation leads to thrombin generation and thrombus formation. Hence, CPB requires therapeutic anticoagulation.

Since its first use in 1953, the preferred CPB anticoagulant is UFH. Today, UFH remains the preferred CPB anticoagulant due to its short half-life, thrombosis prevention effectiveness, and well-established protocols for protamine reversal.⁴ In patients diagnosed with HIT, however, UFH increases thrombosis risk.⁵ Therefore, UFH reexposure during cardiac surgery could be considered unsafe.

Timing of UFH use during cardiac surgery

The safety of UFH use during cardiac surgery depends on the time elapsed since the HIT diagnosis. In patients with remote HIT, the incidence of severe bleeding or recurrent HIT is approximately 4%.⁶ In a study of 10 patients with remote HIT undergoing cardiac surgery, short-term UFH did not increase HIT antibodies during a 10-day follow-up period.⁷ In another study of 17 patients with remote HIT undergoing cardiac surgery,⁸ intraoperative UFH resulted in enzyme immunoassay seroconversion in 11 (65%), SRA seroconversion in 8 (47%), and recurrent HIT with thrombosis in 1 (6%). Therefore, despite potential immune-related concerns, short-term UFH may be considered beneficial if confined to the surgical period alone.

In patients with acute or subacute HIT (positive functional and/or immunoassay), UFH reexposure confers a high risk of thrombosis (44%) and mortality (33%).⁹ For these patients, UFH re-exposure is supported by limited data. Primarily limited to case reports and series, strategies reported in the literature include 1) intraoperative use of a nonheparin anticoagulant,^10,11 2) intraoperative UFH following therapeutic plasma exchange (TPE),^12,13 and 3) intraoperative UFH following administration of a potent antiplatelet agent.^14-16 These strategies are explored in subsequent sections.

Intraoperative use of a nonheparin anticoagulant

Nonheparin anticoagulants that have been studied for use during cardiac surgery include danaparoid, recombinant hirudin, and bivalirudin (Table 1).^10,11,17,18 Although argatroban is widely used in the nonsurgical context and has limited use in cardiac surgery,^19-21 bivalirudin is more commonly used for HIT-associated cardiac surgery. In an online survey of cardiovascular anesthesiologists and extracorporeal technologists, 74% of respondents reported the use of bivalirudin in their institutional cardiac surgery protocol.²²

Table 1.

Nonheparin anticoagulants used during cardiac surgery in HIT

Author/Year (N) study type	Population (n)	Surgery (n)	Nonheparin anticoagulant and mean dose	Outcomes
Fernandes 200017 (6) Retrospective CS	HIT Ab + (3) HIT Ab + SRA + (1) HIT Ab − (2)	OHT (1) MVR (1) CABG (4)	Danaparoid bolus +/−7 u/kg/h infusion to target anti-Xa of 1.5-2.0 units/mL	Fibrin strands in operative field and/or CPB circuit in 50% Postop stroke in 1 patient Postop blood loss 330-3250  mL
Koster 200018 (57) Retrospective CS	HIT Ab + (6) HIPAA + (46) Remote HIT (5)	CABG (27) Valve surgery (16) CABG + valve surgery (8) Aneurysm repair (4) Other (2)	Hirudin 0.25  mg/kg intravenous bolus followed by a continuous infusion of 0.5  mg/min	24-hour blood loss 50-2200  mL CPB clotting in 2 patients 3 unrelated deaths
Koster 200710 (49) Prospective single-arm study	HIT Ab + (35) HIT Ab & functional assay + (7) Clinical HIT with Tpenia (1) Remote HIT (6)	On-pump CABG	Bivalirudin 1  mg/kg + 2.5  mg/kg/h	Procedural success  =  94%a Intraop blood loss: 575 ± 524 mL (50-1900) 24-h blood loss  =  936 ± 525 mL (600-2745)
Dyke 200711 (51)b Prospective single-arm study	HIT Ab + (35) HIT Ab + and Tpenia (6) Tpenia only (5) HIT + thrombosis (5) HIT not confirmed (1)	Off-pump CABG	Bivalirudin 0.75  mg/kg + 1.75  mg/kg/h	Procedural success  =  92%a Intraop blood loss: 404 ± 420  mL (50-2140) 24-h blood loss  =  998 ± 595 mL (100-3200)

^a Procedural success was defined as the absence of death, myocardial infarction, stroke, or repeat revascularization by postoperative day 7 or hospital discharge, whichever occurred first.

^b Although only 51 patients are in the study, the total number of HIT diagnoses adds up to 52.

Abs, antibodies; CABG, coronary artery bypass grafting; CS, case series; CPB, cardiopulmonary bypass; OHT, orthotopic heart transplant; Tpenia, thrombocytopenia.

Additionally, in 2 prospective single-arm studies of patients with HIT, intraoperative bivalirudin was assessed for its effect on procedural success—the absence of death, myocardial infarction, repeat operation for coronary revascularization, and stroke within the first 7 days of surgery or by hospital discharge.^10,11 In 49 patients, procedural success at 7 days, 30 days, and 12 weeks was achieved in 94%, 86%, and 82%, respectively.¹⁰ In 51 patients, procedural success at 7 days, 30 days, and 12 weeks was achieved in 92%, 88%, and 88%, respectively.¹¹ Together, these studies demonstrate clinical experience with bivalirudin use in patients with HIT.

Although there is clinical experience with bivalirudin in cardiac surgery,²² its intraoperative use may be associated with increased transfusion. In a randomized trial comparing intraoperative bivalirudin to UFH, a higher percentage of patients in the bivalirudin group received transfusion of packed red blood cells (57.1% vs 51.9%), platelets (19.4% vs 9.6%), and plasma (27.6% vs 23.1%).²³ Furthermore, a higher percentage of patients on bivalirudin received 2 or more packed red blood units (51.0% vs 44.2%). Additionally, compared to historical patients treated with UFH, bivalirudin is associated with increased clot formation within the surgical field.²⁴ Therefore, in patients with HIT undergoing cardiac surgery, strategies have been pursued to facilitate UFH reexposure.

Intraoperative heparin following therapeutic plasma exchange

Where UFH reexposure is desired intraoperatively, circulating anti PF4-H antibodies can be removed with TPE.²⁵ During TPE (sometimes referred to as plasmapheresis), plasma and its associated pathologic antibodies are removed and replaced with a replacement solution (typically albumin, albumin/plasma, or plasma).²⁶ By removing HIT-associated immune complexes,^25,27 TPE may facilitate UFH reexposure during cardiac surgery.

TPE was evaluated in a retrospective study investigating a single perioperative TPE in 24 patients with positive PF4-H before anticipated heparin CPB use (Table 2).¹³ Although there were no HIT-related deaths, postoperative thrombosis was reported in 3 patients. Of these, 2 instances were possibly HIT-related. In another retrospective study of 11 patients undergoing a single TPE procedure prior to CPB, TPE reduced anti PF4-H titers by 50% to 84% (Table 2).¹² Although no deaths were attributed to HIT, 1 patient (9%) developed an ischemic foot. Though limited to case series and reports, these studies may suggest the utility of TPE.

Table 2.

Studies reporting TPE prior to cardiac surgery in HIT

Author/Year (N) study type	Population (n)	Surgery (n)	TPE prescription	Outcomes
Welsby 201012 (11) Retrospective CS	HIT + within 2 wk of surgery (9) HIT + within 2 mo of surgery (2) HIPA − (11) Median preop OD 0.8e (0.7–2.2)	AVR (1) CABG (1) OHT (6) VAD (3)	1 TPE in ORbc 1.3 PV Plasma	OD ↓50% → 84% 1 patient with left foot loss due to IABP. 3 deaths (3  mos, 6  mos, and 1 year post op)
Ramu 201847 (4) Retrospective CS	HIT and SRA + (2); HIT + SRA NR (2)	OHT (1) VAD (2) AVR, MVR (1)	1 intraop TPE, 4-7 preop TPE	No perioperative thrombosis
Moreno Duarte13 (24) Retrospective CS	HIT Ab + (10)a HIT Ab − (2) HIT Ab NR (12) Median PLT  =  160  ×  103/mm3 Preop AC: bivalirudin (88%), argatroban (4%), warfarin (4%)	OHT (6) VAD (13) Other (5)	1 TPE in ORb 1.0 PV plasma	In 11 patients, median preop OD 1.99 → 0.34 after TPE 3 deaths (hemorrhagic stroke, respiratory failure, sepsis) 3 thrombotic events within 7 days of surgery (stroke, DVT, and NR)d

^a All patients had a history of HIT. Pre-TPE HIT antibody test results were reported in 12 patients.

^b TPE was done in the OR after the induction of general anesthesia, immediately prior to or after heparinization during cardiopulmonary bypass.

^c A subset of patients in this study received postoperative TPE. However, this number was not reported.

^d Two thrombotic events happened before platelet recovery and the third event (NR) after platelet recovery.

^e In this study, a positive anti-PF4-H OD was >0.4.

Ab, antibody; AC, anticoagulant; AVR, aortic valve replacement; CABG, coronary artery bypass grafting; CS, case series; HIPA, heparin-induced platelet aggregation; MVR, mitral valve replacement; NR, not reported; OD, optical density; OHT, orthotopic heart transplant; PLT, platelet count; PV, plasma volume; VAD, ventricular assist device.

However, there are no established protocols for TPE use in cardiac surgery. In a systematic review evaluating 30 patients with HIT that received TPE prior to cardiac surgery,²⁸ the average time to TPE was 7 days. The average number of procedures was 4 (range, 1-11). Most patients received TPE at least 1 day prior to or on the day of cardiac surgery. The primary replacement fluid was plasma, and the average plasma volume exchanged was 1.3. In 4 cases, intravenous immune globulin (IVIG) was given in combination with, prior to, or following TPE.²⁸ In patients with HIT undergoing cardiac surgery, there have been no published reports of IVIG used as monotherapy.

It is important to note that TPE use for cardiac surgery is limited by institutional preference and availability. In a 2019 survey evaluating TPE use in HIT, 29 out of 78 institutions (37%) reported TPE use to facilitate UFH reexposure during cardiac surgery.²⁷ Therefore, TPE may be preferentially used at some institutions.

Intraoperative heparin with a potent antiplatelet agent

Antiplatelet agents can effectively inhibit heparin-induced platelet activation and aggregation.²⁹ For this reason, potent antiplatelet agents have been used to facilitate UFH re-exposure in patients with HIT undergoing cardiac surgery. Antiplatelet agents that have been studied in this role include iloprost, cangrelor, abciximab, and tirofiban (Table 3).^14-16,29

Table 3.

Antiplatelet agents for unfractionated heparin reexposure in HIT

Author/Year (N) study type	Population (n)	Surgery (n)	Antiplatelet agent and mean dose	Outcomes
Koster 200129 (47) Retrospective CS	HIPA + (35) Remote HIT (12)	Aortic surgery, transplant, reoperation, combined proceduresa	bTirofiban 10  µg/kg bolus + 0.15  µg/kg infusion	24-h mean postop blood loss 410 ± 180  mL
Palatianos 201514 (110) Retrospective case control	HIT Ab + (64) HIT Ab + & Tpenia (46)	CABG (74) valve replacement (49) AAA (12)	cIloprost 7.63  ng/kg/min (range, 3–24)	No significant differences from 118 HIT negative controls
Lee 201816 (6) Retrospective CS	HIT Ab & SRA + (2) HIT Ab + SRA NR (2) HIT Ab + SRA − (2)	VAD (6)	dAbciximab 0.25  mg/kg bolus + 0.125 ug/kg	Median postop ICU stay  =  9 days (range 5-76); hospital stay  =  22 days (range 18-132)
Gernhofer 20194 (10) Retrospective CS	HIT Ab & SRA + (3) HIT Ab + SRA − (4) HIT Ab + no SRA (1) HIT Ab – (1) Remote HIT – no testing (1)	AVR (1) Pulmonary TE (6) OHT (1) VAD (2)	eCangrelor 30  µg/kg bolus + 2–4  µg/kg/min	No thrombotic events. One death in a patient with advanced intracardiac malignancy.

^a The types of surgery were not specified.

^b Tirofiban was administered 10 minutes prior to heparin.

^c Iloprost target dose was initially determined based on the dose that diminished the heparin-induced platelet aggregation assay to less than 5%.

^d Abciximab was followed by heparin, 3 minutes later, prior to cardiopulmonary bypass.

^e Cangrelor infusion rate was adjusted to the P2Y12 reaction unit (PRU) values throughout CPB. After cangrelor achieved satisfactory platelet inhibition, heparin was administered 10 minutes later.

AAA, aortic abdominal aneurysm repair; Ab, antibody; AVR, aortic valve replacement; CABG, coronary artery bypass grafting; CS, case series; HIPA, heparin-induced platelet aggregation; OHT, orthotopic heart transplant; TE, thromboendarterectomy; Tpenia, thrombocytopenia; VAD, ventricular assist device.

Iloprost to facilitate intraoperative UFH during cardiac surgery was reviewed in a retrospective series of 110 cases (42% acute HIT) and 118 controls (Table 3).¹⁴ Prior to UFH infusion, iloprost was titrated to a heparin-induced platelet aggregation of less than 5%. Compared to the control group, there were similar rates of postoperative thrombosis (5.1% vs 5.4%) and 30-day postoperative mortality (8.2% vs 8.5%).

Cangrelor to facilitate intraoperative UFH during cardiac surgery was evaluated in a case series of 10 patients (30% acute HIT; Table 3).¹⁵ Prior to UFH administration, cangrelor was titrated to platelet inhibition using the VerifyNow (Werfen, Barcelona, Spain) P2Y12 reaction unit point of care assay. Although postoperative thrombosis was not reported, death occurred in 1 patient with intracardiac cancer whose life support was withdrawn.

Abciximab to facilitate intraoperative UFH infusion during cardiac surgery was evaluated in 6 patients (67% acute HIT; Table 3).¹⁶ Abciximab was administered 3 minutes prior to UFH infusion. Although no postoperative thrombotic events occurred, 1 patient (17%) developed major postoperative bleeding.

Tirofiban to facilitate intraoperative UFH infusion during cardiac surgery was evaluated in 47 patients (74% acute HIT; see Table 3).²⁹ Tirofiban was administered 10 minutes prior to UFH infusion. Although there were no postoperative thrombotic complications, 2 patients died from myocardial failure.

Taken together, these studies show clinical experience using potent antiplatelet agents to achieve platelet inhibition prior to and concurrent with UFH re-exposure.

Summary of strategies used for acute or subacute HIT in cardiac surgery

For urgent/emergent cardiac surgery in patients with acute or subacute HIT, the preferred strategy may be the use of a nonheparin anticoagulant.²² Other treatment options include TPE (+/− IVIG) or platelet inhibition with a potent antiplatelet agent prior to UFH reexposure.

CLINICAL CASE (continued)

Following a single perioperative TPE, the patient received intraoperative UFH. Postoperatively, she was restarted on bivalirudin. Despite bivalirudin at consistent therapeutic levels, the patient developed persistent thrombocytopenia. Also, on ultrasound, there was concern for progressive left-upper-extremity venous thrombosis.

Refractory HIT

Persistent thrombocytopenia or recurrent thrombosis may occur despite therapeutic nonheparin anticoagulation

Despite appropriate therapy, persistent thrombocytopenia or new thromboembolic events can occur. In a retrospective multicenter study of 442 patients with positive PF4-H (80% acute HIT),³⁰ despite appropriate nonheparin anticoagulation, the primary composite outcome of death, limb amputation/gangrene, or new thrombosis occurred in 36% (113/316). In patients who developed new thromboses (74% venous and 19% arterial) or limb amputation/gangrene, 64% of cases occurred while on a nonheparin anticoagulant. These events may identify a group of patients with primary refractory HIT.²⁸

Prior to a diagnosis of refractory HIT, however, the first practical step is to verify the anticoagulation quality (Is the patient consistently therapeutic?) and efficacy (Are there any pharmacokinetic or pharmacodynamic concerns?). Quality/efficacy concerns should prompt consideration of an alternative nonheparin anticoagulant. Subsequently, if thrombocytopenia persists and/or new or progressive thrombosis occurs, refractory HIT should be considered.

Refractory HIT may indicate the presence of heparin-independent HIT antibodies

Refractory HIT is characterized by prolonged thrombocytopenia (>1 week) or progressive thrombosis despite heparin cessation and the initiation of therapeutic nonheparin anticoagulation.⁵ Its pathogenesis may involve heparin-independent platelet activation by anti-PF4 antibodies.^31,32 These heparin-independent antibodies are also thought to underlie autoimmune HIT. Although rare, if autoimmune HIT is suspected in this context, it may require alternative management. Published experience has included the use of IVIG, TPE, and other immunomodulatory therapies.

IVIG for refractory HIT

The most commonly reported therapy for refractory HIT treatment is IVIG. Because platelet Fc receptors bind IgG with high avidity,³³ IVIG is thought to improve platelet counts by inhibiting FcγRIIa-mediated platelet activation through competition with plasma IgG.³⁴ Clinical response of refractory HIT to IVIG was demonstrated in a small case series of 3 patients.³² In these patients, refractory HIT was marked by prolonged thrombocytopenia and recurrent thromboembolism. Following the administration of IVIG (range, 2.0-2.5  g/kg), platelet counts improved over 3 days. Nevertheless, despite IVIG-induced in vitro inhibition of platelet activation, anti PF4-H enzyme-linked immunosorbent assay levels remained unchanged. Also, platelet activation, as demonstrated by the PF4-dependent P-selectin expression assay and SRA (in 1 patient), remained positive.

Similar to the results of the above case series, platelet recovery following IVIG has been demonstrated in several case reports. In a systematic review summarizing case report data for IVIG in refractory HIT,²⁸ the average IVIG dose was 2.2  g/kg administered over an average of 3 days. In these patients, the average time from HIT diagnosis to IVIG infusion was 11 days. As highlighted above, the data have significant limitations. Nevertheless, given the increased risks of prolonged thrombocytopenia and/or recurrent thrombosis, IVIG may be considered in patients who develop progressive thrombosis despite adequate nonheparin anticoagulation and/or persistent thrombocytopenia (>1 week) despite heparin cessation.

TPE for refractory HIT

When compared to IVIG, TPE in patients with refractory HIT is reported less frequently.²⁸ However, a number of case reports suggest a temporal relationship between TPE use and improved platelet counts.^35,36 In 1 case report of severe refractory HIT marked by persistent thrombocytopenia and distal pedal gangrene, platelet counts improved from 38 to 178   ×   10⁹/L after 4 days of TPE.³⁷

Although there was a heterogenous application of TPE in this population, a systematic review showed that the average prescription was 4 TPE procedures.²⁸ The replacement solution varied: albumin alone, albumin/plasma, or plasma alone. The volume exchanged was 1 to 1.5 plasma volumes. Although there is limited published experience of TPE in patients with refractory HIT, it may have possible utility in patients with refractory HIT unresponsive to IVIG.

Other therapies for refractory HIT

In patients with refractory HIT, the use of rituximab as a salvage immunosuppressant has been reported following IVIG and TPE.^38,39 Additionally, there are ongoing early-phase preclinical studies of a novel therapeutic—Fc-Modified KKO.⁴⁰ While KKO is a mouse monoclonal IgG2b_κ anti PF4-H/polyanion monoclonal antibody that mimics pathogenic HIT antibodies, Fc-modified KKO is a deglycosylated version that acts by blocking FcγRIIa-mediated binding. In so doing, it prevents platelet activation and decreases complement activation. In mouse models, Fc-modified KKO reversed thrombocytopenia and decreased thrombus size. These effects need further evaluation in future clinical studies.

Summary of refractory HIT management strategies

From the above data, limited evidence exists to support any specific therapies in patients with refractory HIT. Nevertheless, based on the available clinical experience and considering its ease of administration and availability, practicing hematologists could consider IVIG. Where IVIG elicits no response, TPE, where available, could be considered as a possible rescue therapy. Other possible therapies to consider include rituximab. The utility of promising novel therapies, still in early preclinical studies, will become evident over time.

CLINICAL CASE (continued)

While on bivalirudin, the patient developed an acute lower gastrointestinal bleed. Due to active bleeding and the need for ongoing transfusion, bivalirudin was held.

Patients with active bleeding

Bleeding has been reported in patients with HIT

Although the primary clinical impact of HIT is a prothrombotic state, bleeding in patients with HIT may be associated with the use of therapeutic anticoagulation.⁴¹ In a retrospective large database study of 6332 HIT hospitalizations, bleeding events, as identified by international classification of diseases (ICD) codes, occurred in 358 cases (5.7%).⁴² Bleeding events occurred most commonly in the gastrointestinal (2.7%) and genitourinary tracts (2.2%) and less commonly in the central nervous system (1%). Due to the reliance on ICD codes alone, the bleeding rates identified in this study may be lower than those in other studies.

When ICD bleeding definitions are combined with transfusion and laboratory data, the incidence of major bleeding among patients with HIT can range from 36% to 48%.^30,43 In a retrospective study of 310 patients, among 44 patients (14%) with confirmed HIT, the incidence of major bleeding was 39% (17/44).⁴³ Fatal bleeding occurred in 9% (4/44). Although bleeding incidence in this study was highest in patients who received anticoagulant therapy despite being HIT negative (44%; n  =  51/116), these data suggest that the incidence of major bleeding in patients with HIT is higher than previously reported. In an earlier study of 442 patients, the incidence of major bleeding among patients with HIT and thrombosis (n  =  71) was 48%. In patients with isolated HIT without thrombosis (n  =  284), it was 36%.³⁰ Therefore, in patients with HIT treated with a nonheparin anticoagulant, the incidence of major bleeding is not trivial.

Possible factors associated with bleeding in HIT

Of the available studies in the literature evaluating bleeding complications in patients with HIT, none has evaluated bleeding-risk factors in patients with HIT alone. Most of the available studies have evaluated patients with suspected HIT in aggregate. In the earlier cited study of 310 patients with suspected HIT (14% acute HIT),⁴³ factors associated with an increased adjusted hazard of major bleeding were intensive care unit admission (HR, 2.24; 95% CI, 1.44-3.47), a platelet count of less than 25  ×  10⁹/L (HR, 2.13, 1.10-4.12), and renal dysfunction (HR, 1.56, 1.06-2.27). Bleeding risk was not significantly associated with the duration of anticoagulant therapy. In the 422-patient retrospective study evaluating patients with HIT (80%) and without HIT (20%),³⁰ nonheparin anticoagulation was also found to have no statistically significant effect on major bleeding (HR, 0.83; CI, 0.57-1.21; P  =  .33). Therefore, although these studies did not specifically evaluate patients with HIT alone, they provide insights into the possible factors associated with bleeding risk.

Management of bleeding in patients with HIT

Because clinically significant bleeding is a strong contraindication to therapeutic anticoagulation, in patients with HIT who develop active bleeding, therapeutic anticoagulation is routinely stopped. Following anticoagulant cessation, strategies are pursued to decrease bleeding risk. In patients who have active bleeding with severe thrombocytopenia (platelets <50  ×  10⁹/L), strategies reported in the literature include platelet transfusion.^44,45

Platelet transfusions, however, may be associated with increased thrombosis risk. In the earlier described large database study of 6332 HIT cases, platelet transfusions in 7.1% of patients were associated with a higher adjusted odds of arterial thrombosis (odds ratio [OR], 3.4; 95% CI, 1.2-9.5) and mortality (OR, 5.2; 95% CI, 2.6-10.5) but not venous thrombosis (OR, 0.8; 95% CI, 0.4-1.7).⁴² Therefore, due to increased thrombosis risk, unless there is a life-threatening bleed, platelet transfusions should generally be avoided.

In patients for whom platelet transfusion is being considered, potential thrombotic complications must be weighed against the risks of uncontrolled bleeding. In patients with severe thrombocytopenia, platelet transfusions may result in bleeding cessation. In a retrospective study of 37 patients with positive anti-PF4-H enzyme-linked immunosorbent assay, 65% (n  =  24) were transfused for bleeding prophylaxis and 35% (n  =  13) for bleeding control. Of those transfused for bleeding control, bleeding stopped within 24 hours of platelet transfusion in 62%.⁴⁴ In 6 nonresponders, 2 died from bleeding complications and 1 from withdrawal of life support. Thrombotic complications were not identified. In a smaller retrospective case series of 13 patients with SRA-positive HIT, 4 patients received platelet transfusion (3 for active bleeding).⁴⁵ Bleeding cessation within 24 hours occurred in 2 out of 3 (67%) one patient with coagulopathy experienced persistent bleeding. None of these patients developed a thrombotic complication.

Summary of bleeding-management strategies

In the absence of studies that adequately weigh the adverse effects of bleeding risk vs platelet transfusion, risk assessment may need to be individualized. Nevertheless, given the high mortality risk of uncontrolled major bleeding,⁴⁶ platelet transfusion may play a role in managing life-threatening bleeding where there is severe thrombocytopenia.

CLINICAL CASE (continued)

The patient underwent successful transcatheter arterial embolization with bleeding cessation. Subsequently, bivalirudin was cautiously restarted. Over the next 2 weeks, platelet counts returned to baseline. The patient was transitioned to apixaban to complete 3 months of therapeutic anticoagulation.

Conclusion

Managing HIT in complex clinical scenarios can be challenging because they are rare and have a limited evidence base. Therefore, management requires a multidisciplinary approach tailored to the patient. In patients with acute or subacute HIT undergoing urgent or emergent cardiovascular surgery, the preferred strategy may be the perioperative use of a parenteral nonheparin anticoagulant. In select patients, clinical experience may support a role for heparin reexposure using TPE or potent antiplatelet agents. In patients with refractory HIT, practical steps should be taken to evaluate the nonheparin anticoagulant's quality or efficacy. Where there is no clinical improvement, clinical experience suggests a possible role for IVIG, TPE, or rituximab. Where bleeding complications occur, strategies should be employed to avoid platelet transfusion. However, in patients with severe thrombocytopenia, platelet transfusion may have a possible role in managing uncontrolled life-threatening bleeding. Although the outlined strategies lack a strong evidence base, future studies may help clarify the strategies that deliver the best outcomes.

Conflict-of-interest disclosure

Oluwatayo Adeoye: no competing financial interests to declare.

Guoliang Zheng: no competing financial interests to declare.

Oluwatoyosi A. Onwuemene: has received honoraria from Sanofi. This relationship is unrelated to the content presented in this manuscript.

Off-label drug use

Oluwatayo Adeoye: The article discussed off-label use of intravenous immune globulin, rituximab, iloprost, cangrelor, and abciximab.

Guoliang Zheng: The article discussed off-label use of intravenous immune globulin, rituximab, iloprost, cangrelor, and abciximab.

Oluwatoyosi A. Onwuemene: The article discussed off-label use of intravenous immune globulin, rituximab, iloprost, cangrelor, and abciximab.