Drug-induced myocarditis

Article written and reviewed by Dr Luca Moretti

Drug-induced myocarditis is inflammatory myocardial injury causally associated with a medication; the definition is more restrictive than cardiotoxicity, which includes contractile dysfunction, ischemia, hypertension, thrombosis and arrhythmias even without an inflammatory infiltrate. Distinguishing the two concepts avoids treating dose-dependent toxicity with immunosuppression or overlooking a potentially fulminant immune reaction.

The association may be mediated by hypersensitivity, specific lymphocyte activation, direct injury with release of neoantigens, or mixed mechanisms. Clozapine and mesalazine are recognizable clinical associations; antimicrobials, anticonvulsants and numerous other molecules cause eosinophilic forms. Immune checkpoint inhibitors have such specific biology and management that they warrant a separate pathway.

Causality is difficult because people taking medications often have diseases capable of causing myocarditis or elevated troponin. A patient treated with mesalazine may have cardiac manifestations of inflammatory bowel disease; someone receiving antibiotics may have sepsis; a person undergoing cancer treatment may develop ischemia, embolism or infection. Temporal sequence alone is necessary but not sufficient.

The true frequency remains uncertain because mild events go unrecognized, whereas spontaneous reports overrepresent severe cases and do not provide reliable denominators; registration trials may exclude at-risk patients or fail to apply uniform criteria. Information therefore derives from integration of cohorts, pharmacovigilance, pathology and pharmacologic plausibility.

Early withdrawal can result in complete recovery, but discontinuing an indispensable therapy causes harm. Clozapine may be the only effective antipsychotic in treatment-resistant psychosis, mesalazine may control colitis, and an antimicrobial may treat a severe infection; the decision must protect the heart while simultaneously ensuring an alternative for the underlying disease.

Etiology, Pathogenesis and Pathophysiology

In the immunoallergic mechanism, a drug or metabolite modifies self-proteins and activates T lymphocytes, cytokines and eosinophils. Injury is often dose-independent and may appear after a period of sensitization; rash and eosinophilia increase plausibility but may be absent and are not required for the definition.

Clozapine causes a phenotype concentrated mainly in the first four weeks, frequently during titration. Hypersensitivity, catecholaminergic excess, oxidative stress and metabolic factors have been proposed; no single mechanism explains every case. Rapid titration, interactions that increase exposure and systemic inflammation may modify risk.

Mesalazine and sulfasalazine may be associated with myocarditis or pericarditis, often during the first weeks; the mechanism is considered hypersensitivity-mediated or immune, but inflammatory bowel disease itself can have cardiac manifestations. Improvement after withdrawal and inadvertent recurrence after re-exposure strengthen causality without making intentional provocation ethical.

Antimicrobials, aromatic anticonvulsants, diuretics, anti-inflammatory drugs and numerous other classes have been described in hypersensitivity myocarditis or DRESS. Possibility is not equivalent to frequency: long lists of case reports should not turn every drug into a common risk; the patient's epidemiology and latency remain discriminating features.

Anthracyclines predominantly produce cumulative cardiomyocyte injury and dysfunction, whereas fluoropyrimidines frequently cause vasospasm or ischemia; catecholamines and stimulants can produce necrosis and Takotsubo syndrome. A secondary infiltrate does not automatically convert these phenotypes into myocarditis; classification should instead follow the clinically dominant mechanism.

Post-vaccination myocarditis is a rare event described with some products, with the best-characterized signal after mRNA vaccines against SARS-CoV-2, especially in adolescent and young adult males after certain doses. Most recognized cases have a favorable course, but severe forms exist; individualized assessment and comparison with infection risk avoid both denial and alarmism.

Drugs may unmask a genetic predisposition or interact with a concomitant infection; a persistent arrhythmic phenotype after resolution requires consideration of inherited cardiomyopathy. Causality may be contributory rather than exclusive: removing the drug is appropriate even when it is not the only factor.

Pharmacokinetics influences recovery time; molecules or antibodies with a long half-life continue to act after withdrawal, reactive metabolites may persist, and renal or hepatic failure may increase exposure. A dechallenge that is not immediately followed by improvement does not exclude causality, whereas improvement may also result from concomitant therapies.

Necrosis and edema may resolve or progress to fibrosis; late risk depends on the amount and location of scar, not solely on elimination of the drug. Conduction block may persist after biomarker normalization if the conduction system has been destroyed.

Injury caused by cocaine, amphetamines and other sympathomimetic substances requires accurate classification; vasospasm, thrombosis, hypertension, dissection, hyperthermia and catecholamines cause necrosis and sometimes infiltrates, but initial treatment follows the hemodynamic and ischemic mechanism. Indiscriminately defining the condition as myocarditis conceals exposure and may lead to unnecessary corticosteroids.

Cellular therapies and some oncologic antibodies may cause cytokine-release syndrome with hypotension and reversible dysfunction. This phenotype is not the same as ICI myocarditis and responds to different interventions; infusion date, fever, hypoxia and systemic markers help distinguish a cytokine storm from autoimmune myocardial infiltration.

Injury may depend on excipients, contaminants or interactions rather than the active ingredient; weight-loss supplements and herbal products may contain undeclared sympathomimetics. Toxicologic analysis and product verification become relevant when the history does not explain the phenotype or when several people experience similar events.

Clinical Manifestations

Chest pain, dyspnea, palpitations, syncope, fever and fatigue are the most common symptoms; the presentation may mimic an acute coronary syndrome, pericarditis or infection. Elevated troponin and ST-T abnormalities support myocardial injury, but the responsible drug is identified only after causality assessment.

Clozapine-associated myocarditis frequently presents with persistent tachycardia, fever, malaise, pain or dyspnea during the first weeks. Tachycardia and a modest rise in C-reactive protein, however, are common during titration and should not trigger an automatic diagnosis; troponin, clinical course and echocardiography increase specificity.

Mesalazine-associated myocarditis may present with pleuritic pain, fever, troponin elevation and effusion, producing a myopericarditis phenotype. Intestinal disease activity, infections and other extraintestinal manifestations enter the comparison; short latency after initiation and rapid resolution after withdrawal are useful but not absolute features.

Rash, eosinophilia, facial edema, hepatitis and nephritis suggest hypersensitivity or DRESS. Cardiac involvement may be delayed and more severe than the cutaneous presentation; new dyspnea during convalescence therefore requires reassessment rather than simple reassurance.

Arrhythmias include ectopy, ventricular tachycardia, atrial fibrillation and atrioventricular block. Preserved function does not protect against electrical risk; syncope, new conduction abnormalities and rising troponin require monitored hospitalization.

Shock may result from biventricular dysfunction, allergic vasoplegia, tamponade or combinations of these mechanisms. Rapid hemodynamic assessment distinguishes the components and guides fluids, vasoactive agents and support; discontinuing the drug is essential but does not produce instantaneous stabilization.

A subclinical presentation may emerge from surveillance protocols, especially with clozapine or oncologic therapies. An isolated biomarker requires confirmation and differential diagnosis; overly sensitive screening without a confirmatory algorithm exposes patients to unnecessary interruption of effective treatments.

In a young person with pain after vaccination, the dose history and temporal window are considered together with a search for recent infection, congenital coronary disease, substance exposure and other causes. The clinical phenotype is treated according to severity; risk communication avoids drawing general conclusions about vaccine safety from a single event.

Pain associated with fever may be interpreted as infection, whereas clozapine-related tachycardia may appear to be an expected effect. The key is discordance: persistence at rest, rising troponin, new ECG abnormalities or deterioration in exercise tolerance; structured protocols reduce both diagnostic delay and discontinuations caused by nonspecific signals.

Chronic cardiotoxicity may emerge months or years after exposure and may not show edema; in this setting cumulative history, strain and function matter more than criteria for acute myocarditis. Assessment of drug-induced disease may therefore sometimes conclude that the patient has a noninflammatory toxic cardiomyopathy, a different but equally important diagnosis.

Investigations and Diagnosis

Medication reconciliation is the first etiologic test; for each molecule, initiation, discontinuation, dose, titration, interactions, previous exposures and adherence are recorded. Over-the-counter drugs, herbal remedies, supplements and recreational substances are included because the patient may not consider them treatments.

Serial ECG and troponin define the dynamics, while natriuretic peptides, complete blood count, eosinophils, C-reactive protein, renal and hepatic function and electrolytes describe phenotype and safety. Because troponin T may rise in myositis and renal failure, values are never interpreted outside the clinical context.

Echocardiography assesses function, strain, effusion and alternative diagnoses; CMR identifies edema and nonischemic injury but does not assign causality. Timing, treatment already started and focality explain imperfect sensitivity.

Endomyocardial biopsy is considered in shock, arrhythmias, conduction block, lack of response or when distinguishing lymphocytes, eosinophils, giant cells and infection changes therapy. The presence of a drug in the history does not make tissue unnecessary in severe cases; multiple samples and cardiovascular pathology improve diagnostic yield.

Scales such as Naranjo or the WHO-UMC method make timing, dechallenge, alternatives and prior knowledge explicit. They were not designed to prove myocarditis on their own and may produce different categories according to data quality; the final assessment describes the degree of causality and uncertainties.

Inadvertent re-exposure with recurrence is strong evidence, but intentional provocation is generally inappropriate. A positive dechallenge is less specific because supportive treatment, corticosteroids and the natural course act simultaneously; causality is not reduced to a single observation.

Infections, acute coronary syndrome, pulmonary embolism, Takotsubo syndrome, genetic cardiomyopathy and the underlying disease are excluded proportionately. In patients with inflammatory bowel or autoimmune disease, systemic activity must be documented; in oncology, cumulative effects of multiple treatments are assessed.

A pharmacovigilance report does not require absolute certainty and is submitted when a reasonable suspicion exists. It includes negative and positive findings, dose, batch when relevant, timing and outcome; aggregation of high-quality reports makes it possible to recognize rare events that trials cannot estimate.

Plasma concentration may support overexposure for molecules with validated therapeutic drug monitoring, but does not prove hypersensitivity; smoking, fever, enzyme inhibitors and metabolic changes alter clozapine levels. A therapeutic level does not exclude myocarditis, whereas a high level prompts a search for interactions and titration issues in addition to withdrawal.

Vaccine attribution requires a compatible window, clinical criteria and comparison with background rates; passive reporting systems generate signals, not causal incidence estimates, and administrative databases require case validation. Age, sex, dose and interval modify risk and prevent transferring one estimate to the entire population.

CMR may distinguish nonischemic patterns, but late gadolinium enhancement does not identify a molecule. Normal coronary angiography does not prove a drug-related origin and a positive dechallenge may coincide with the spontaneous course; certainty arises from convergence of independent evidence.

Treatment and Prognosis

The suspected drug is discontinued when probability and severity justify the risk of stopping it. If it treats an essential disease, a non-cross-reactive and noncardiotoxic alternative is selected; the decision is not delayed until biopsy in shock, but samples are obtained as soon as possible.

Heart failure and shock are treated according to phenotype with diuretics, vasoactive agents and mechanical support when necessary, while arrhythmias and conduction block are treated and monitored; intense exercise is suspended during the active phase and resumption is conditional on clinical and electrical recovery.

Corticosteroids are appropriate in clinically significant immunoallergic, eosinophilic or fulminant forms after infection has been considered. Dose and taper are adapted to response and involved organs; direct noninflammatory cardiotoxicity does not derive the same benefit and requires another strategy.

With clozapine, active protocols during the first four weeks use symptoms, heart rate, C-reactive protein and troponin to increase the likelihood of recognition. Proposed thresholds derive from specific cohorts and must be applied within the local system; plausible myocarditis requires discontinuation and immediate psychiatric collaboration.

Treatment of mesalazine-associated disease consists mainly of withdrawal, with corticosteroids in selected cases; control of intestinal disease is maintained with agreed alternatives. Incorrect attribution to mesalazine can deprive the patient of useful treatment, whereas re-exposure after a convincing event is dangerous.

Post-vaccination myocarditis is treated according to severity and the predominance of pericardial involvement, avoiding automatic regimens. Many cases recover with rest and supportive therapy; forms with dysfunction or arrhythmias follow pathways for complicated myocarditis. Subsequent doses are discussed individually according to risk and updated recommendations.

Prognosis is favorable in many forms recognized early, but shock, necrosis, conduction block and arrhythmias increase mortality and sequelae. Elimination of a drug does not guarantee that immune activity will cease immediately; CMR and monitoring assess residual scar and recurrence risk.

Therapeutic rechallenge is generally avoided after a probable or definite event, especially if severe. It may be discussed exceptionally when benefit is substantial and alternatives are lacking, with informed consent, hospitalization or defined monitoring and discontinuation criteria, but it is never used merely to satisfy diagnostic curiosity.

When the drug has an antidote or concentration-dependent toxicity, decontamination, antidote administration and support may be more important than immunosuppression. In acute exposures, poison-control and toxicology services are consulted; treatment of a sympathomimetic overdose is not modeled on that of DRESS.

Alternative therapy is chosen before discharge; cross-reactivity, the same cardiotoxic mechanism, interactions and the ability to control the underlying disease are considered. An unplanned switch may provoke psychiatric, inflammatory or infectious relapse and prompt the patient to restart the discontinued drug independently.

Complications

Shock and malignant arrhythmias may progress despite withdrawal because the drug or immune response persists. Early referral to a center with circulatory support is appropriate in deteriorating forms; the possibility of recovery justifies temporary support when extracardiac prognosis is favorable.

Residual fibrosis causes cardiomyopathy, ventricular tachycardia and exercise limitation; normalization of ejection fraction does not eliminate risk if late gadolinium enhancement persists. Follow-up is calibrated to the initial event rather than to the absence of symptoms alone.

Atrioventricular block may require pacing and, during the inflammatory phase, reversibility is considered, whereas it may become permanent after septal necrosis. Ventricular risk determines whether a pacemaker alone is sufficient or a device with defibrillation capability is required.

Polypharmacy exposes patients to interactions and substitutions that may be equally risky; a new drug introduced to treat the reaction can confound rash, eosinophilia or QT. Continuous reconciliation and shared prescribing responsibility reduce an iatrogenic cascade.

Stopping clozapine can cause psychotic relapse, stopping an anticonvulsant can cause seizures, and stopping an antimicrobial can allow infection to progress. These complications do not justify continuation of a dangerous culprit but require timely and competent replacement.

An incomplete report may lead to accidental re-exposure or, conversely, avoidance of many unrelated drugs. The discharge letter specifies the molecule, probability, phenotype, severity and alternatives; a medical alert bracelet or personal document may be useful after potentially fatal reactions.

Anxiety and distrust of all treatments are common after an iatrogenic event; a transparent explanation distinguishes molecule-specific risk, class risk and background risk. The patient can then participate in future decisions without abandoning necessary treatments because of generalized fear.

Reproductive risk and pregnancy complicate substitution of antiepileptic, immunologic and psychiatric therapies. A cardiologically safe molecule may be teratogenic, whereas loss of control of maternal disease may harm both mother and fetus; the cardiologist, prescribing specialist and maternal-fetal medicine team construct a solution before a new exposure.

Aggregated pharmacovigilance information is updated over time; an event initially classified as “possible” may become more plausible when cases with the same phenotype emerge, whereas controlled studies may reduce the apparent strength of a signal. Documentation preserves the raw data, not only the label, allowing future reassessment without losing the history.

The duration of immune therapy does not depend on drug half-life alone. When the trigger has initiated an autoimmune circuit, the infiltrate may persist after clearance; in pure toxic injury, prolonged corticosteroids add risk without benefit. Troponin, rhythm, function and, in selected cases, tissue findings guide tapering.

Regulatory authorities may update product information, warnings and risk-minimization plans as a signal becomes established. Clinicians verify current information without converting every warning into an absolute contraindication; risk, dose, population and monitoring are translated into a documented individual decision.

Follow-up of the underlying disease is part of the outcome; recurrent psychosis, reactivated colitis or uncontrolled cancer are clinical consequences of the cardiac decision. Joint indicators make it possible to compare strategies that prevent recurrent myocardial disease while preserving, as far as possible, overall therapeutic effectiveness.