Drug-induced immune thrombocytopenia

INTRODUCTION — Unexplained thrombocytopenia is a common clinical problem, and the possibility of drug-induced thrombocytopenia must be considered, especially in hospitalized patients, in whom new drugs are commonly administered. Drugs can cause thrombocytopenia by several mechanisms including direct bone marrow or other organ toxicity. Many of the cytotoxic drugs used for cancer chemotherapy and other conditions produce dose-dependent reductions in platelet counts due to bone marrow suppression as a known effect of the drug.

This topic review discusses drug-induced immune thrombocytopenia (DITP), in which the mechanism involves antibody-mediated platelet destruction caused by exposure to a drug that leads to isolated thrombocytopenia (without anemia or leukopenia). The "drug" in this context may be a prescribed medication (or its metabolite), over-the-counter medicine, herbal supplement, food, beverage, or other substance.

Other types of drug-induced thrombocytopenia are discussed separately:

●Heparin-induced thrombocytopenia (HIT) – (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia".)

●Drug-induced thrombotic microangiopathy (DITMA) – (See "Drug-induced thrombotic microangiopathy (DITMA)".)

General approaches to evaluating patients with thrombocytopenia in various settings are presented separately:

●Neonates – (See "Neonatal thrombocytopenia: Etiology".)

●Children – (See "Approach to the child with unexplained thrombocytopenia".)

●Adults – (See "Diagnostic approach to thrombocytopenia in adults".)

●Pregnancy – (See "Thrombocytopenia in pregnancy".)

●Pancytopenia – (See "Approach to the adult with pancytopenia" and "Aplastic anemia: Pathogenesis, clinical manifestations, and diagnosis".)

●Thrombotic microangiopathy – (See "Diagnostic approach to suspected TTP, HUS, or other thrombotic microangiopathy (TMA)".)

TERMINOLOGY — We use the following terminology/definitions:

●Pseudothrombocytopenia – Pseudothrombocytopenia is a laboratory artifact in which the patient is not thrombocytopenic; apparent thrombocytopenia is due to platelet clumping in the collection tube due to an in vitro reaction with the anticoagulant. The finding is confirmed by repeating the platelet count using a fresh sample and/or reviewing the peripheral blood smear. Typically, a sample for a repeat platelet count is drawn into a tube containing an anticoagulant other than EDTA (eg, citrate). (See "Diagnostic approach to thrombocytopenia in adults", section on 'Pseudothrombocytopenia'.)

●Thrombocytopenia – Thrombocytopenia is defined as a platelet count below the lower limit of normal (<150,000/microL). Some individuals will have a stable platelet count that may be slightly lower than this value and is normal for that individual; these individuals may not require additional testing or intervention. Conversely, some individuals may have a platelet count above the lower limit of the reference interval that represents a significant decrease from baseline for that individual; in such cases it is prudent to repeat the platelet count in one or two days to determine if there is a declining trend. (See 'Approach' below.)

●Primary immune thrombocytopenia – Primary immune thrombocytopenia (ITP) is an autoimmune phenomenon with no apparent underlying cause. (See "Immune thrombocytopenia (ITP) in children: Clinical features and diagnosis" and "Immune thrombocytopenia (ITP) in adults: Clinical manifestations and diagnosis".)

●Drug-induced immune thrombocytopenia – Drug-induced immune thrombocytopenia (DITP), the focus of this topic review, refers to thrombocytopenia caused by drug-dependent antibody-mediated platelet destruction. DITP is a form of secondary ITP. The "drug" in this context may be a prescribed medication (or its metabolite), over-the-counter medicine, herbal supplement, food, beverage, or other substance.

●Non-immune drug-induced thrombocytopenia – Many drugs used as chemotherapy cause thrombocytopenia by bone marrow suppression. Linezolid is an example. Some drugs have a combined mechanism that includes drug-dependent antibodies and bone marrow suppression. (See 'Mechanisms of DITP' below.)

●Heparin-induced thrombocytopenia – Heparin-induced thrombocytopenia (HIT) is a unique drug reaction in which antibodies against complexes of platelet factor 4 and heparin cause both thrombocytopenia and platelet activation, resulting in venous and/or arterial thrombosis. Patients with suspected HIT require immediate institution of a non-heparin anticoagulant. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia".)

MECHANISMS OF DITP — Thrombocytopenia in DITP is caused by accelerated platelet (and rarely, megakaryocyte) destruction from drug-dependent, platelet-reactive antibodies [1-3]. In almost all cases, drug-dependent antiplatelet antibodies bind non-covalently to specific platelet antigens via their Fab regions. DITP with megakaryocyte destruction from drug-dependent antibodies has been described for eptifibatide [4].

●Quinine has been the most widely studied drug to determine the mechanism of DITP; quinine can also cause drug-induced thrombotic microangiopathy (DITMA). Quinine-dependent antibodies may be derived from a pool of naturally occurring antibodies that are weakly reactive with autologous proteins, such as platelet surface glycoproteins [5-7]. Quinine, and possibly other drugs causing DITP, can become integrated into complementarity-determining regions of these naturally occurring antibodies, creating a hybrid paratope (modified antigen-binding region) that greatly increases the antibody’s binding affinity to platelet surface antigens (figure 1) [6,8]. The phenomenon of naturally occurring antibodies may explain the observation of extremely rapid platelet count drop with some drugs such as GP IIb/IIIa inhibitors (see 'GP IIb/IIIa inhibitors' below). In DITP caused by abciximab, antibodies to murine (mouse) antigens may be responsible.

●Another potential mechanism is that binding of the drug to the platelet surface can cause a conformational change in a surface protein that leads to exposure of a neoepitope, which in turn stimulates the formation of antiplatelet antibodies [9].

●Drug-dependent antibodies may have specificity for major drug metabolites as well as (or instead of) the parent drug. In a study of sulfamethoxazole-induced thrombocytopenia, 10 antibodies that reacted with intact sulfamethoxazole also reacted strongly with the N1-acetyl metabolite [10]. Similar results have been obtained with ibuprofen [11]. Reactions that occur exclusively with a metabolite (but not the parent drug) may explain the common occurrence of negative results with drug-dependent antibody testing [10]. Examples include acetaminophen and naproxen, for which drug-dependent, platelet-reactive antibodies have only been identified for drug metabolites and not for the parent drug [11].

●Drug-dependent antibodies typically do not cross-react with different drugs that have a similar molecular structure (eg, they react to sulfamethoxazole but not to the nearly identical sulfisoxazole, or to quinine but not the optical stereoisomer quinidine) [10,12]. This may support the use of other drugs in the same class in patients with a history of DITP; however, unless the antibody specificity (and lack of cross-reactivity) is known for that particular patient, we generally do not assume other drugs of the same class will be safe. (See 'Role of drug-dependent platelet antibody testing' below and 'Role of drug rechallenge' below.)

●Some drugs can cause acute, immune-mediated thrombocytopenia mediated by drug-dependent antibodies in addition to causing bone marrow suppression. Well-documented examples are oxaliplatin and irinotecan [13,14]. In some cases multiple drug-dependent antibodies are produced [15].

●Other drugs can precipitate immune-mediated thrombocytopenia in which thrombocytopenia begins or continues even after the drug is stopped. Gold, which was previously used to treat certain types of arthritis, had been classically implicated in causing this syndrome. Other examples include alemtuzumab and immune checkpoint inhibitors such as pembrolizumab, nivolumab, and ipilimumab [16]. In one report, 6 of 216 patients developed severe thrombocytopenia 1 to 15 months following the last dose of alemtuzumab [17]. The thrombocytopenia was responsive to traditional treatment for ITP, which resulted in prolonged remissions.

●Vaccines are another example.

•In children, the measles-mumps-rubella (MMR) vaccine has also been associated with thrombocytopenia that has the clinical characteristics of primary ITP [18].

•ITP has also been reported following administration of certain coronavirus disease 2019 (COVID-19) vaccines, though causation has not been definitively established [19].

●In addition to chemotherapy drugs traditionally recognized as cytotoxic to the bone marrow, some other commonly used drugs can cause moderate, but relatively specific suppression of platelet production in the bone marrow. Examples include daptomycin, linezolid, valproic acid, and valacyclovir [20-22].

Platelet antigens that are targeted by drug-dependent antibodies may include a variety of surface glycoproteins, most commonly GP Ib/V/IX or GP IIb/IIIa. Others may also be involved (eg, GP V, platelet-endothelial cell adhesion molecule-1 [PECAM-1]). Some antibodies may bind to multiple regions of a single glycoprotein or to multiple epitopes simultaneously [9,23-26].

COMMONLY IMPLICATED DRUGS

List of drugs — Commonly implicated drugs and their mechanism of thrombocytopenia are listed in the table (table 1) [1,11,27-35]. A database of drugs associated with DITP, updated approximately every two years, is available at www.ouhsc.edu/platelets [36]. The database also includes information updated annually from Versiti Blood Center of Wisconsin on drug-dependent platelet-reactive antibodies.

A 2013 systematic review of DITP that considered both clinical and laboratory criteria for the diagnosis of DITP, including the documentation of drug-dependent platelet antibodies by at least two different laboratories, found the following drugs to be directly implicated in DITP reactions: quinine, quinidine, trimethoprim-sulfamethoxazole, vancomycin, penicillin, rifampin, carbamazepine, ceftriaxone, ibuprofen, mirtazapine, oxaliplatin, suramin, GP IIb/IIIa inhibitors, and heparin [37].

Other drugs have been implicated because they have been identified as causing DITP by three different methods: (i) case reports with definite evidence, (ii) identification of drug-dependent antibodies, (iii) data mining of the US Food and Drug Administration's Adverse Event Reporting System database [11].

Incidence — The overall risk of DITP from any new drug is rare. The absolute incidence depends on the patient population; individuals who are hospitalized are more likely to receive new drugs, including antibiotics, and to have a higher incidence of DITP than healthy outpatients.

●Review articles summarizing epidemiologic data suggest a population incidence of 10 per million population [2,38,39].

●Higher rates of DITP have been reported with specific drugs:

•Quinine – 38 per million

•Trimethoprim-sulfamethoxazole – 26 per million

•GPIIb/IIIa inhibitors – 1 to 2 percent (first exposure); as high as 10 percent (second exposure to abciximab [no longer available in the United States])

●Higher rates have also been reported in patients in the intensive care unit (ICU) [39]. A report of 329 patients in the ICU found thrombocytopenia in 41 percent; of these, 9 percent were attributed to a drug [40]. Other common causes in the ICU include sepsis, liver disease, hypersplenism, and disseminated intravascular coagulation (DIC). (See 'Differential diagnosis' below and "Diagnostic approach to thrombocytopenia in adults", section on 'Acutely ill/intensive care unit'.)

CLINICAL PRESENTATION — The typical patient history includes one or more medications, herbal remedies, foods, beverages, or other substances known to cause DITP. The interval between initiation of a new drug taken daily and DITP is usually less than two weeks. A drug taken daily for several months or longer is rarely associated with DITP. Drugs taken only intermittently may cause DITP even if it has been many years since the previous exposure [2]. The following features define the typical clinical presentation, which can be used to determine the likelihood of DITP.

●Patient age – Drug-induced thrombocytopenia may be less common in children than adults [41]. In a 2013 systematic review of case reports limited to children, 32 substances were identified with a definite or probable causal role in thrombocytopenia [41]. Implicated drugs included antibiotics (trimethoprim-sulfamethoxazole, vancomycin), antiepileptics (carbamazepine, phenobarbital, phenytoin), and spironolactone.

●Timing – In the typical presentation of DITP, the drop in platelet count occurs within two weeks of exposure to the drug. As an example, in a series of 12 patients with thrombocytopenia associated with piperacillin, the median interval was 6.5 days of initial continuous exposure [42]. Another report described 13 patients exposed to abciximab in whom thrombocytopenia developed after an interval of three to six days [43]. If the patient has been sensitized by a previous exposure to the drug, the onset of thrombocytopenia may be more rapid (eg, as early as hours after exposure). Thrombocytopenia due to a GP IIb/IIIa inhibitor may occur within the first 24 hours of the initial exposure and can occur after minutes to hours (see 'GP IIb/IIIa inhibitors' below). There are rare reports of thrombocytopenia caused by vancomycin or piperacillin occurring with the first day of exposure, but previous sensitizing exposure may have occurred in these patients [30,42]. The previous exposure may have been in an unusual form, such as vancomycin in orthopedic cement [44]. It has also been speculated that a previous exposure may have occurred due to the presence of antibiotics in foods, but we are unaware of data to support this [42]. As noted below, most patients with DITP have prompt platelet count recovery upon drug discontinuation and a normal platelet count by one week. (See 'Platelet count recovery/monitoring' below.)

●Severity of thrombocytopenia – The thrombocytopenia in DITP is often severe, with a nadir platelet count <20,000/microL [28,42,45]. An exception is heparin-induced thrombocytopenia (HIT), in which the median nadir platelet count is approximately 60,000/microL and platelet counts <20,000/microL are rare [46]. (See 'Heparins' below and "Clinical presentation and diagnosis of heparin-induced thrombocytopenia", section on 'Thrombocytopenia'.)

●Likely drugs – As described above, DITP has been associated with a number of drugs; the typical history involves a drug with high-quality evidence for causing DITP including laboratory demonstration of drug-dependent immunoglobulin binding to platelets in vitro in two or more independent cases (table 1) [37]. (See 'Commonly implicated drugs' above.)

In most cases, the only associated clinical or laboratory findings in patients with drug-induced thrombocytopenia are those related to thrombocytopenia or bleeding. These may include petechiae, purpura, mucosal bleeding (eg, epistaxis), and/or more serious bleeding. Bleeding risk and prevention is discussed below. (See 'Treatment of bleeding/severe thrombocytopenia' below.)

SPECIAL CASES

Heparins — Heparins, including both unfractionated heparin and low molecular weight (LMW) heparins such as enoxaparin and dalteparin, can cause heparin-induced thrombocytopenia (HIT). HIT is a type of antibody-mediated drug-induced thrombocytopenia, but because of its distinct characteristics (eg, platelet activation and risk of thrombosis), we do not include it in our discussion of DITP. Diagnosis and management of HIT are discussed in detail separately. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia" and "Management of heparin-induced thrombocytopenia".)

GP IIb/IIIa inhibitors — Glycoprotein (GP) IIb/IIIa inhibitors (eg, abciximab, tirofiban, eptifibatide) are used in patients with coronary heart disease or coronary stenting. All patients receiving a GP IIb/IIIa inhibitor should have a baseline platelet count prior to drug initiation, and platelet count monitoring during and after treatment, as discussed separately. (See "Acute non-ST-elevation acute coronary syndromes: Early antiplatelet therapy" and "Antithrombotic therapy for elective percutaneous coronary intervention: General use".)

These drugs can be associated with the rapid onset of DITP (minutes to hours after exposure), often profound, with an incidence on the order of 1 to 4 percent [47-53]. The rapid onset is thought to be explained by the presence of preformed, "naturally occurring" antibodies, especially with abciximab [2,54-57]. Their mechanism of causing thrombocytopenia (eg, opsonization, platelet activation, or both) is unclear [58]. (See 'Mechanisms of DITP' above.)

●Abciximab – In a review of four large randomized trials, the incidence of true thrombocytopenia with abciximab was found to be 3.7 percent (205 of 5476 patients), compared with 1.8 percent (56 of 3079) in those assigned to placebo [53]. In patients receiving placebo, thrombocytopenia would have been due to another drug or non-drug cause. Another pooled analysis that evaluated data from eight randomized trials involving GP IIb/IIIa inhibitors found abciximab-induced thrombocytopenia in 5.2 percent, of which most were considered mild (platelet count <90,000 to 100,000/microL) [51]. The incidence of severe thrombocytopenia (<50,000/microL) was 1 percent with abciximab versus 0.4 percent with placebo.

Antibodies in abciximab-associated DITP recognize murine peptide sequences incorporated into the abciximab Fab fragment that confer specificity for GP IIb/IIIa [59].

●Eptifibatide – In a pooled analysis that evaluated data from eight randomized trials involving GP IIb/IIIa inhibitors, eptifibatide was not associated with an increased incidence of thrombocytopenia [51]. The incidence of thrombocytopenia was not increased with eptifibatide in the PURSUIT trial (eptifibatide versus placebo in 10,948 patients with acute coronary syndrome) [60]. Although eptifibatide-induced thrombocytopenia may be less common than thrombocytopenia induced by abciximab and tirofiban, the occurrence and mechanism have been reported [58]. In addition to causing thrombocytopenia, eptifibatide can also activate platelets resulting in thrombosis [58].

●Tirofiban – The incidence of thrombocytopenia was increased with tirofiban in the PRISM trial (aspirin plus tirofiban versus aspirin plus heparin in 3232 patients with unstable angina); rates were 1.1 percent with tirofiban and 0.4 percent with heparin [61]. However, in a pooled analysis that evaluated data from eight randomized trials involving GP IIb/IIIa inhibitors, tirofiban was not associated with an increased incidence of thrombocytopenia [51].

The GP IIb/IIIa inhibitors are also associated with an increased incidence of pseudothrombocytopenia, a laboratory artifact of no clinical importance [48]. A review of four large randomized trials found the incidence of pseudothrombocytopenia associated with abciximab to be approximately 2 percent (117 of 5476 patients), compared with 0.6 percent (17 of 3079) assigned to placebo [53]. The anticoagulant in the sample collection tube was EDTA in most of the 117 cases, although 14 were documented to occur in the presence of citrate also. The presumed mechanism of EDTA-induced platelet agglutination is thought to involve naturally occurring antiplatelet antibodies that recognize EDTA-induced conformational changes in platelet GP IIb/IIIa.

Our approaches to distinguishing between true thrombocytopenia and pseudothrombocytopenia, and between thrombocytopenia due to a GP IIb/IIIa inhibitor, heparin, or another cause, is presented below. (See 'Diagnosis' below.)

DIAGNOSIS

Approach — Overall, the risk of DITP from any new drug is low [45]. However, among patients with thrombocytopenia, it is important to assess the possibility of DITP as the cause because this diagnosis has major implications for management. DITP should be suspected in a patient who presents with new onset of thrombocytopenia, especially sudden unexpected thrombocytopenia or recurrent episodes of acute thrombocytopenia without an obvious etiology.

DITP is a clinical diagnosis made by excluding other causes of thrombocytopenia and documenting resolution of thrombocytopenia upon drug discontinuation. Recurrent acute episodes of thrombocytopenia after drug re-exposure is strongly suggestive of DITP. Laboratory testing for drug-dependent antiplatelet antibodies is helpful if the result is positive, but false-negative results are common and the turnaround time for the test is generally long, as testing is only done in reference laboratories. (See 'Role of drug-dependent platelet antibody testing' below.)

If the patient has had previous occurrence of thrombocytopenia with previous exposure to the suspected drug, this further supports the diagnosis. Re-exposures should generally be avoided due to the risks of causing recurrent, severe thrombocytopenia and bleeding. (See 'Role of drug rechallenge' below.)

The patient evaluation is focused on identifying clues to other possible diagnoses and determining the likelihood of these other diagnoses versus DITP. For patients receiving multiple drugs, close evaluation of the timing of the drug exposure and the likelihood of DITP for each drug is especially valuable. (See 'Medication review/temporal relationship' below.)

Confirm thrombocytopenia — As with any case of unexpected thrombocytopenia, the possibility of laboratory error or pseudothrombocytopenia (ie, in vitro artifact) should be addressed before conducting an extensive evaluation or stopping a needed medication. This is done by examining the peripheral blood smear for platelet clumping (picture 1) and repeating the platelet count. In vitro platelet clumping can occur due to incomplete mixing of the sample, or to an interaction with the anticoagulant in the sample collection tube. This is especially true for EDTA (the routine anticoagulant employed for blood counts). Appropriate alternative sample tubes include those containing citrate or heparin as the anticoagulant in the tube.

Review of the blood smear is essential for eliminating other possible diagnoses such as an inherited platelet disorder or a thrombotic microangiopathy such as thrombotic thrombocytopenic purpura (TTP). (See 'CBC and blood smear review' below.)

A significant drop in platelet count (eg, by 50 percent or more) is sufficient to trigger review of the blood smear, more intensive platelet count monitoring, or other interventions, even if the platelet count is ≥150,000/microL. The frequency of accelerated monitoring depends on the absolute platelet count, rate of decline, and suspected diagnoses. As examples, such a drop in platelet count in a patient receiving a heparin product or a GP IIb/IIIa inhibitor may be the first indication of heparin-induced thrombocytopenia (HIT) or DITP, respectively. In suspected HIT, heparin is immediately discontinued, a non-heparin anticoagulant initiated, and platelet count monitored daily until it begins to recover. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia".)

Exclude other causes of thrombocytopenia — The likelihood of other possible causes of thrombocytopenia (table 2) depends on the clinical setting (eg, incidental finding in an asymptomatic individual seen as an outpatient, or an acutely ill patient in hospital) [42]. For all patients it is important to assess the complete blood count (CBC) for other abnormalities and review all possible drug exposures, with close attention to the temporal relationship between drug exposure and platelet count decrease.

Clues from the patient history and clinical setting

Incidental finding — An otherwise healthy patient may have thrombocytopenia as an incidental finding on a CBC performed for other reasons, or may develop thrombocytopenia in association with a self-limited illness. Other common causes of isolated thrombocytopenia in asymptomatic patients include the following:

●Viral infection

●Primary immune thrombocytopenia (ITP)

●Previously unrecognized inherited platelet disorder

●Myelodysplastic syndrome (MDS), especially in older patients; however, MDS rarely presents with isolated thrombocytopenia

●Nutrient disorder such as vitamin B12 or folate deficiency, although other cytopenias often are present and folate deficiency is extremely rare in countries that use routine folate supplementation in foods

In some cases, these alternative causes may be distinguished from drug-induced thrombocytopenia by findings on the peripheral blood smear such as abnormal platelet size or granulation (for inherited disorders) or megaloblastic changes (for MDS or nutrient disorders). Other distinguishing factors are discussed below and in separate topic reviews. (See 'Differential diagnosis' below.)

Acutely ill or hospitalized patient — Acutely ill or hospitalized patients may have thrombocytopenia as a result of their underlying illness or its treatment. Other common causes of thrombocytopenia in acutely ill patients include sepsis, acute infection, and/or disseminated intravascular coagulation (DIC). Post-transfusion purpura is often associated with severe bleeding, but it is very rare. Nevertheless, DITP should be considered in any hospitalized patient who develops thrombocytopenia during an acute illness or in the hospital.

Thrombocytopenia in combination with thrombosis is especially concerning because it raises the possibility of HIT (see 'Determine if the patient was exposed to heparin' below) or other serious conditions including antiphospholipid syndrome (APS), paroxysmal nocturnal hemoglobinuria (PNH), or a thrombotic microangiopathy (TMA). These conditions require special interventions and expertise.

Early input from the consulting specialist is advised.

CBC and blood smear review — For all patients, it is important to review the CBC and peripheral blood smear because other abnormalities (cytopenias, morphologic features) may suggest an alternate or additional diagnosis. Examples of important findings that suggest a diagnosis other than drug-induced thrombocytopenia include the following:

●Pancytopenia – Pancytopenia (anemia, leukopenia, and thrombocytopenia) suggests the possibility of aplastic anemia (which could also be drug-induced) or another bone marrow disorder. (See "Aplastic anemia: Pathogenesis, clinical manifestations, and diagnosis", section on 'Drugs' and "Approach to the adult with pancytopenia".)

●Microangiopathic hemolytic anemia – The combination of microangiopathic hemolytic anemia (MAHA; suggested by the presence of schistocytes on the peripheral blood smear) and thrombocytopenia in the setting of drug exposure suggests drug-induced thrombotic microangiopathy (DITMA) or another type of TMA. (See "Drug-induced thrombotic microangiopathy (DITMA)" and "Diagnostic approach to suspected TTP, HUS, or other thrombotic microangiopathy (TMA)".)

●Platelet morphologic abnormalities – Platelet abnormalities such as giant platelets or platelets that lack granules may suggest an inherited platelet disorder. However, not all inherited thrombocytopenias are associated with abnormal platelet morphology. (See "Inherited platelet function disorders (IPFDs)".)

●Other CBC findings – Additional abnormalities on the CBC may suggest an infectious cause (eg, intracellular parasites, white blood cell findings consistent with infection) or a primary bone marrow disorder (eg, megaloblastic changes, immature granulocytes). Evaluation for these conditions is discussed in detail separately. (See "Diagnostic approach to thrombocytopenia in adults".)

Medication review/temporal relationship — Establishing a clear temporal relationship between initiation of a drug and development of thrombocytopenia is essential, although this may be challenging.

A patient with recurrent episodes of acute thrombocytopenia should be suspected of having a drug-induced etiology because the intermittent nature of the thrombocytopenia may correspond to intermittent exposure or ingestion of a food, beverage, or herbal remedy [62]. In addition to the timing of drug ingestion, it is important to take a thorough history and consult lists of commonly implicated drugs. (See 'Commonly implicated drugs' above.)

A detailed history must address all of the following:

●Prescribed drugs

●Over-the-counter medications such as acetaminophen

●Alternative medicines, folk remedies, and herbal preparations [63,64]

●Recent vaccinations

●Medications prescribed by other physicians

●Medications prescribed for family members or other individuals, which the patient may not want to admit to taking

●Certain common beverages associated with thrombocytopenia such as tonic water (which contains quinine) [62,65]

●Foods, such as walnuts and tahini [62,66,67]

Explicit questioning is essential because patients commonly assume that when doctors ask about medications, they are only asking about prescribed medications that are taken regularly [68,69]. As an example, quinine is perhaps the most common cause of drug-induced thrombocytopenia, and it is often overlooked in the patient’s history [68,70]. Quinine tablets are only available by prescription in the United States, but can be obtained over-the-counter in Canada and other countries. Quinine is also an ingredient of beverages (eg, tonic water, bitter lemon) (table 3). Quinine is also used for its fluorescent properties in party drinks such as "jello shots" and "shocktails" [65]. A variety of foods and other substances have been associated with acute, immune-mediated thrombocytopenia [62,67].

Most drugs require prior sensitization, either through daily administration or administration at a previous time. However, some drugs may cause thrombocytopenia at the initial exposure without prior sensitization. This commonly occurs with the glycoprotein IIb/IIIa inhibitors (eg, abciximab, tirofiban, eptifibatide), and has also been seen with other drugs, such as vancomycin and piperacillin [30,42]. In some cases there may have been a previous exposure that was unrecognized [44], as discussed above. (See 'Clinical presentation' above.)

Determine if the patient was exposed to heparin — Heparin-induced thrombocytopenia (HIT) is a unique drug reaction that carries a risk of potentially fatal thrombosis. Heparin discontinuation alone is insufficient to reduce this risk, and the patient must be treated with a non-heparin anticoagulant. While some patients present with thromboses, many do not, and the absence of thrombosis does not exclude HIT.

Heparin exposure is most likely in hospitalized patients but may also occur in outpatients who are receiving injections of low molecular weight (LMW) heparin. Hospitalized patients may have received heparin at the time of an invasive procedure or in heparin flushes, heparin bonded catheters, or some plasma products such as certain prothrombin complex concentrates (PCCs) [71]. In such cases, the medication record may not list heparin as a regular medication. It is thus especially important to review the possibility of heparin administration, and to estimate the likelihood of HIT and the need for more extensive testing and presumptive treatment with a non-heparin anticoagulant while awaiting results of this testing. Estimation of pretest probability of HIT is presented separately. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia", section on 'Evaluation' and "Management of heparin-induced thrombocytopenia", section on 'Stop heparin for a presumptive diagnosis of HIT'.)

Decide which drug(s) to stop — The decision to stop a medication is straightforward for a healthy outpatient who develops new onset thrombocytopenia within one or two weeks of starting a new medication for which evidence for causing thrombocytopenia is already available (table 1). If a drug is discontinued and the platelet count improves, this supports the diagnosis of DITP due to that drug [72]. (See 'Platelet count recovery/monitoring' below.)

However, the decision of which drug(s) to stop, if any, may be more challenging in other settings, especially in hospitalized patients who are often receiving multiple medications.

●Non-medication substance/intermittent exposure – A non-medication substance such as a food or beverage, an over-the-counter herbal product, or a medication taken intermittently may be responsible in a healthy outpatient who has recurrent/intermittent episodes of thrombocytopenia but denies taking any regular medications. In some cases, the patient may disclose taking an herbal product or medication available over-the-counter or prescribed for another individual on further questioning. In others, the patient may be able to identify a food or beverage such as tonic water, which contains quinine (see 'Medication review/temporal relationship' above). These substances can then be discontinued and the platelet count observed. The role of rechallenge to confirm the diagnosis is discussed below.

●Drug not previously implicated in causing thrombocytopenia – In a patient who is prescribed a drug not previously known to cause thrombocytopenia, the importance of continuing the medication must be balanced with the likelihood that the drug versus another diagnosis such as ITP is responsible. If the clinical scenario is compatible with DITP, the drug should be stopped and testing for drug-dependent antiplatelet antibodies should be done (see 'Management' below). A negative test does not exclude the diagnosis, but a positive test supports the diagnosis. Of note, it is important to report newly identified causes of DITP to the appropriate monitoring agency. (See 'Documentation/reporting' below.)

●Multiple medications/hospitalized patient – Several confounding factors may make it difficult to determine whether a drug has caused thrombocytopenia in a patient who is acutely ill or hospitalized, and if so, which drug is most likely to be responsible [42]. Some drugs may be given once, and as a result they no longer appear in the medication charting, and it may be necessary to review other records (eg, anesthesia record) or speak with nursing staff and the patient.

•It may be possible that thrombocytopenia is a result of the underlying illness. This is especially likely in a patient with an acute infection or sepsis [42].

•If multiple drugs have been administered in a short period of time, it may be difficult to determine which one is responsible [42]. Commonly used drugs associated with DITP in hospitalized patients are trimethoprim-sulfamethoxazole, vancomycin, and piperacillin. In these challenging situations, it may be necessary to stop all possible drugs. If one drug is strongly implicated and the others are not, it may be reasonable to discontinue only the strongly implicated drug [72]. Ultimately, the choice between stopping all, some, or one drug is based on clinical judgement that balances the likelihood for each drug and the risks to the patient of drug discontinuation for each drug.

In addition to the drugs listed above (see 'Commonly implicated drugs' above) and in the table (table 1), systematic reviews of evidence for a causal relationship of drugs with thrombocytopenia are periodically updated [27,73-78].

Role of drug-dependent platelet antibody testing — We consider testing for drug-dependent antiplatelet antibodies to be an important component of the evaluation for DITP. This testing is especially helpful for distinguishing among drugs in a patient for whom one of several drugs may be responsible, for identifying (and reporting) a new association that has not previously been reported, and in cases where re-exposure to a drug is likely. A positive test would result in the need for lifelong drug avoidance.

Testing is generally only done in reference laboratories (such as the Blood center of Wisconsin). Flow cytometry-based assays to identify drug-dependent platelet antibodies are most commonly used (figure 2) [1,10,30,79-81]. The cost is relatively low (similar to the cost of HIT antibody testing), and confirmation of the presence of a drug-dependent antiplatelet antibody has important implications for future drug avoidance and for adverse event reporting. (See 'Management' below.)

Although testing for drug-dependent antibodies is helpful when the diagnosis is unclear, this testing is not required for establishing the diagnosis when the suspected drug has a well-established role in causing DITP.

DITP remains a clinical diagnosis (see 'Approach' above). The antibody testing can be performed after the drug has been discontinued as confirmation, because drug-dependent antiplatelet antibodies persist for many years, but the highest yield would be to test during the episode of acute thrombocytopenia. Unlike DITP testing, testing for heparin-induced thrombocytopenia is widely available with a fast turnaround time, and detectable antibodies generally do not persist beyond 100 days [82]. (See "Clinical presentation and diagnosis of heparin-induced thrombocytopenia".)

Importantly, antibody testing, when used, should not delay drug discontinuation in a patient with clinically suspected DITP.

While a positive test for drug-dependent platelet antibodies confirms the diagnosis of DITP, a negative test does not rule it out. The low sensitivity of DITP testing can be explained by the need for specialized test methods including the use of the drug in the platelet washing steps and the possibility that drug metabolites rather than the drug itself can be the trigger (figure 2) [79,80]. Negative results are sometimes seen in patients with a strong clinical likelihood of DITP [10,23,83,84].

Additionally, the drug may cause thrombocytopenia by a mechanism other than DITP such as bone marrow suppression (as with linezolid and daptomycin) or primary ITP (as with alemtuzumab).

As noted above, the clinician must often use clinical judgment to decide whether to discontinue therapy with one or more of the patient's medications. (See 'Decide which drug(s) to stop' above.)

Role of drug rechallenge — We rarely use drug rechallenge intentionally due to the risks of recurrent thrombocytopenia. However, in some cases rechallenge may occur inadvertently, or the drug-induced etiology may only be identified in retrospect. In such cases, recurrence of thrombocytopenia upon re-administration of the drug (or other substance) provides strong evidence that the drug is responsible for thrombocytopenia in that patient. In the majority of patients, we use clinical assessment and/or testing for drug-dependent antibodies rather than rechallenge to establish the diagnosis of DITP. (See 'Role of drug-dependent platelet antibody testing' above.)

Examples of the potentially deleterious effects of rechallenge include the following:

●Severe and symptomatic thrombocytopenia has been described in reports of rechallenges when the implicated agent was a food such as walnuts or herbal preparation [63,67].

●Severe thrombocytopenia following rechallenge with a GP IIb/IIIa inhibitor was illustrated in a prospective review of 550 patients who received a second or third dose of abciximab at least seven days after a first treatment [85]. Thrombocytopenia developed in 4.6 percent, was severe (platelet count <20,000/microL) in 2.4 percent, and occurred after hospital discharge in 0.4 percent. In a second report, the incidence of recurrent severe thrombocytopenia was 4 percent in patients re-exposed to abciximab, but this increased to 12 percent if the interval between exposures was ≤2 weeks [86].

In some cases, thrombocytopenia may develop in association with a type of medication that the patient needs (eg, an antibiotic for bacterial infection). We prefer to substitute an agent of another class if possible, to avoid potential antigenic cross-reactivity with the implicated drug. If an alternative class of medications does not exist, it may be possible to give an agent from the same class without causing thrombocytopenia, but good data to support this practice are lacking for most agents. Among the different GP IIb/IIIa inhibitors, the risk of antibody cross-reactivity is thought to be on the order of 15 percent. (See 'Mechanisms of DITP' above.)

Although we avoid rechallenge if at all possible, there may be rare cases in which drug-dependent antibody testing did not identify an implicated drug (or the patient was told the results were negative but documentation is lacking) and the patient is likely to be re-exposed to a possibly implicated drug (eg, history of thrombocytopenia after acetaminophen). In this setting, drug readministration should be done under supervised conditions with close observation of the platelet count and starting with very low doses of the drug. Recurrence of thrombocytopenia on rechallenge provides strong evidence for avoidance of the drug or substance for life.

DIFFERENTIAL DIAGNOSIS — The differential diagnosis of DITP includes a number of other causes of thrombocytopenia (table 2), which are discussed briefly above (see 'Exclude other causes of thrombocytopenia' above) and in more detail in separate topic reviews. The most commonly seen alternative diagnoses are primary immune thrombocytopenia (ITP) in asymptomatic outpatients and sepsis and/or disseminated intravascular coagulation (DIC) in acutely ill patients.

●Primary ITP – Primary immune thrombocytopenia (ITP) is an autoimmune condition in which antiplatelet antibodies cause thrombocytopenia that may be transient in children and may last for months to years to indefinitely in adults. Like DITP, patients with primary ITP typically present as otherwise healthy individuals who develop new onset thrombocytopenia without an obvious cause. Unlike DITP, primary ITP does not resolve when a drug is discontinued. The distinction between DITP and primary ITP is critical in order to avoid unnecessary treatment directed at ITP in a patient with a drug-induced etiology [70]. (See "Immune thrombocytopenia (ITP) in adults: Clinical manifestations and diagnosis", section on 'Diagnostic evaluation'.)

●Sepsis/DIC – Sepsis and disseminated intravascular coagulation (DIC) cause thrombocytopenia, but the patient is very ill with fever and leukocytosis. DIC may also accompany an underlying malignancy. Like individuals with drug-induced thrombocytopenia, patients with sepsis and/or DIC may be receiving multiple medications, some of which may be on the list of drugs that can cause thrombocytopenia; and they may have bleeding. Unlike patients with drug-induced thrombocytopenia, those with sepsis and/or DIC have other laboratory abnormalities indicative of a systemic infection (leukocytosis or neutropenia) and ongoing coagulation (prolonged PT and PTT, low fibrinogen, elevated D-dimer). (See "Evaluation and management of disseminated intravascular coagulation (DIC) in adults".)

●Post-transfusion purpura – Post-transfusion purpura (PTP) is a rare complication that may occur after transfusion of any platelet-containing blood product (red blood cells, platelets, or granulocytes). Like drug-induced thrombocytopenia, PTP may present as new onset severe thrombocytopenia in a patient who has been hospitalized or acutely ill. Unlike drug-induced thrombocytopenia, PTP requires a history of transfusion in the previous 7 to 10 days, and PTP responds to immunosuppressive therapy (typically, intravenous immune globulin [IVIG]). (See "Immunologic transfusion reactions", section on 'Post-transfusion purpura'.)

More extensive discussions of these and other causes of thrombocytopenia are presented separately. (See "Neonatal thrombocytopenia: Etiology" and "Causes of thrombocytopenia in children" and "Diagnostic approach to thrombocytopenia in adults".)

MANAGEMENT

Drug discontinuation — Drug discontinuation is essential in patients with suspected DITP, regardless of clinical presentation. An alternative medication can almost always be substituted if needed. The decision about which drug(s) to discontinue depends on clinical judgement regarding the degree of suspicion for DITP and the decision regarding which drug(s) are the most likely candidates. (See 'Decide which drug(s) to stop' above.)

Decision to hospitalize — For patients with DITP, the platelet count will recover promptly following withdrawal of the offending agent. The need for hospitalization is stratified according to the severity of findings:

●In a healthy outpatient who has severe thrombocytopenia but no bleeding or only minor purpura, hospitalization may not be required as long as there is no evidence for an alternative diagnosis such as drug-induced thrombotic microangiopathy (DITMA), the patient is reliably able to communicate, and they are able to have daily platelet counts measured.

●If there is any bleeding more than minor purpura (eg, if there is epistaxis, heavy menstrual bleeding, or other bleeding), the patient should be hospitalized for close observation. Platelet transfusions are also appropriate and should be given to control clinically important bleeding. Steroids are often given because the distinction of DITP from ITP is often initially unclear. (See 'Treatment of bleeding/severe thrombocytopenia' below.)

Treatment of bleeding/severe thrombocytopenia

Risk of bleeding and mortality — The likelihood of clinical bleeding depends on a number of factors including comorbidities, other medications, and platelet count. Bleeding risk is greater with DITP than with primary ITP [46]. Bleeding risk generally increases with platelet counts below 40,000 to 50,000/microL, but there is not a strong linear correlation between platelet count and bleeding risk [28]. The risk of severe bleeding, including intracranial hemorrhage, may be higher in DITP than in primary ITP, possibly because the onset of severe thrombocytopenia occurs rapidly.

●In a 1997 series of 309 patients with drug-induced thrombocytopenia, more serious bleeding such as gastrointestinal and intracranial bleeding was seen in 12 and 3 percent, respectively [28].

●In a 1998 review of 247 patients, 23 (9 percent) had serious bleeding and two patients (1 percent) had fatal bleeding, both in association with quinine exposure [27].

Drug-induced thrombocytopenia has been associated with increased mortality in other series, especially in hospitalized patients; however, it may be challenging to distinguish association from causality [87].

Life-threatening bleeding: platelet transfusions — Treatment of bleeding and/or severe thrombocytopenia requires close communication among consulting specialists. For major or life-threatening bleeding, platelet transfusions should be administered without delay. Glucocorticoids and/or intravenous immune globulin (IVIG) may be used because drug-induced thrombocytopenia initially cannot be distinguished from immune thrombocytopenia (ITP).

●Patients with clinically important bleeding and/or platelet count <10,000/microL should receive platelet transfusions [2]. Higher thresholds for platelet transfusion may apply in certain settings (eg, fever, infection, GP IIb/IIIa or other antithrombotic agent). IVIG may also be administered if platelet transfusions are ineffective in stopping bleeding [88]. (See "Platelet transfusion: Indications, ordering, and associated risks", section on 'Actively bleeding patient'.)

●For patients with a presumptive diagnosis of DITP due to a GP IIb/IIIa inhibitor who have severe bleeding or require an emergency invasive procedure, platelet transfusions are often given even if the thrombocytopenia is not severe, because these agents interfere with platelet function. Typically, heparin and aspirin are also discontinued and heparin anticoagulation is reversed with protamine sulfate. One report of four patients with DITP due to abciximab found a rapid platelet count response to platelet transfusions, and no apparent benefit from IVIG [89]. (See "Heparin and LMW heparin: Dosing and adverse effects", section on 'Reversal'.)

Role of IVIG and glucocorticoids

●For severe or life-threatening bleeding, IVIG may also be given, especially if the diagnosis of DITP versus primary ITP has not been resolved. Management of patients with possible primary ITP and severe bleeding is discussed separately.

●When bleeding other than petechiae/purpura is present but not severe, glucocorticoids are often given because initially DITP may be indistinguishable from primary ITP. The role of platelet transfusion in the setting of severe but not life-threatening bleeding is individualized. (See "Initial treatment of immune thrombocytopenia (ITP) in adults", section on 'Choice of glucocorticoid and dosing'.)

Platelet count recovery/monitoring — DITP is reversible upon drug discontinuation, and documentation of platelet count recovery is important in confirming the diagnosis. The platelet count is expected to begin increasing within one to two days of drug discontinuation and return to the patient’s normal range in a week [42]. In retrospective series that have reported on the time to platelet count recovery, typical times to platelet recovery are in the range of seven to eight days [27,28,42].

The frequency of monitoring depends on the severity of thrombocytopenia and patient’s clinical condition (eg, presence of bleeding). For healthy outpatients with a finding of drug-induced thrombocytopenia prompted by petechiae/bruising without other significant bleeding, it may be appropriate to recheck the platelet count daily to be sure that the platelet count is recovering, and then a week later to be sure that the platelet count has recovered to normal. It is appropriate to see the patient for a follow-up evaluation in another month, to be sure that the platelet count has remained normal. It is also important to review the remainder of the complete blood count (CBC), especially in patients with slow recovery, to identify any other abnormality that might develop.

However, some drugs (or drug metabolites) remain in the circulation for longer periods, and drug elimination may be further reduced in individuals with reduced metabolism due to renal or hepatic insufficiency. Platelet count recovery may be slower in these settings.

Additional rare agents have been reported to be associated with prolonged thrombocytopenia that is more similar to primary ITP.

●Vaccine for measles, mumps, and rubella (MMR) have been associated with the development of ITP [18]. Similarly acute and sometimes severe ITP has been described following administration of coronavirus disease 2019 (COVID-19) vaccines [19].

●A similar syndrome has been reported with alemtuzumab, a monoclonal antibody directed against the lymphocyte surface marker CD52 [17].

●Some patients receiving eptifibatide have more prolonged duration of thrombocytopenia, perhaps due to direct megakaryocyte impairment by eptifibatide-induced antibodies [4].

●When gold salts were commonly used for the treatment of rheumatoid arthritis, they could be associated with prolonged thrombocytopenia. Gold-induced thrombocytopenia may resemble primary ITP more than DITP, in that it has a propensity to develop long after therapy has been discontinued, to occur more often in patients with HLA-DR3, to be associated with drug-independent antibodies, and to respond to immunosuppressive agents such as glucocorticoids [90-92].

Once it has become established that the platelet count remains normal, there is no need to perform additional platelet counts unless the patient receives the drug again or develops stigmata of bleeding.

In contrast to platelet count monitoring after DITP is diagnosed, the role of platelet count monitoring to detect new onset DITP with drugs such as GP IIb/IIIa inhibitors or heparin is discussed separately. (See 'GP IIb/IIIa inhibitors' above and "Heparin and LMW heparin: Dosing and adverse effects", section on 'Platelet count monitoring'.)

Documentation/reporting — In patients with DITP it is essential to document the association (with the name of the drug) in the patient's medical record, and to ensure that the patient has this information available, in order to avoid future inadvertent administration of a causative drug. This is especially true when a single agent is strongly implicated by clinical criteria (table 4). Since drug-dependent, platelet-reactive antibodies can persist for many years and cause acute, severe thrombocytopenia with a repeat exposure as long as 10 years after the previous exposure, the patient should be informed that the drug should be avoided indefinitely and should be advised to wear a medical alert bracelet or carry a wallet card indicating this fact.

Reporting of drug-induced thrombocytopenia to the medical community is important for making other clinicians aware of the possibility of thrombocytopenia as an adverse effect of a drug not previously known to cause thrombocytopenia.

●All occurrences of DITP in the United States should be reported on the US Food and Drug Administration's MedWatch website (http://www.fda.gov/Safety/MedWatch/default.htm).

●If a drug has not been previously reported as a cause of thrombocytopenia with definite or probable clinical evidence (table 4) (www.ouhsc.edu/platelets), it is important to document this observation with a published case report [11]. Testing for drug-dependent antibodies should be included in the published case report if not previously documented for that drug.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Immune thrombocytopenia (ITP) and other platelet disorders".)

SUMMARY AND RECOMMENDATIONS

●Definition – Drug-induced immune thrombocytopenia (DITP) refers to thrombocytopenia caused by drug-dependent, antibody-mediated platelet destruction. DITP is a form of secondary ITP. The "drug" may be a prescribed or over-the-counter medication, a metabolite, an herbal supplement, a food or beverage (table 3), or other substance. An apparently "naturally occurring" antibody, a primary ITP-like syndrome, and/or megakaryocyte destruction may be involved. (See 'Terminology' above and 'Mechanisms of DITP' above.)

●List of drugs – Drugs commonly implicated in DITP are listed in the table (table 1). Heparin-induced thrombocytopenia (HIT) can cause mild to moderate thrombocytopenia and potentially fatal arterial and venous thromboses. Glycoprotein (GP) IIb/IIIa inhibitors can cause true DITP and pseudothrombocytopenia. (See 'List of drugs' above and 'Heparins' above and 'GP IIb/IIIa inhibitors' above.)

●Incidence and presentation – The overall risk of DITP from any new drug is rare. Patients typically present within two weeks of exposure with severe thrombocytopenia (platelet count <20,000/microL) and/or bleeding. (See 'Incidence' above and 'Clinical presentation' above.)

●Evaluation and diagnosis – DITP is a clinical diagnosis made by excluding other causes of thrombocytopenia and documenting resolution of thrombocytopenia upon drug discontinuation. Laboratory testing for drug dependent antiplatelet antibodies is often helpful but not required. (See 'Diagnosis' above.)

●Differential diagnosis – The differential diagnosis includes other causes of thrombocytopenia (table 2). In asymptomatic outpatients, primary ITP is the other main cause. In acutely ill patients, sepsis and/or disseminated intravascular coagulation (DIC) are common causes of thrombocytopenia. (See 'Differential diagnosis' above.)

●Management – Drug discontinuation is essential in patients with suspected DITP; if needed, a substitute medication can almost always be provided. The need for hospitalization is stratified according to the severity of findings. (See 'Drug discontinuation' above and 'Decision to hospitalize' above.)

●Treatment of bleeding – For clinically important bleeding, platelet transfusions should be administered without delay. Intravenous immune globulin (IVIG) may also be given if bleeding is severe and primary ITP is a possible diagnosis. When bleeding other than petechiae/purpura is present but not severe, glucocorticoids are often given. (See 'Treatment of bleeding/severe thrombocytopenia' above.)

●Recovery – The platelet count will increase, often beginning one to two days after drug discontinuation, with return to the patient's normal range in a week, unless drug metabolism and elimination are impaired by impaired kidney or liver function. All drug-induced thrombocytopenia should be communicated to the patient and documented in the medical record. Publication of a case report is encouraged for drugs not previously known to cause DITP. In the United States a report should be filed on the US Food and Drug Administration's MedWatch website. (See 'Platelet count recovery/monitoring' above and 'Documentation/reporting' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges James N George, MD, who contributed to earlier versions of this topic review.