Drug-induced pericarditis

Article written and reviewed by Dr Luca Moretti

drug-induced pericarditis is pericardial inflammation attributable to a medication through a clinical causality assessment. It may be isolated or accompanied by myocarditis, a hypersensitivity reaction, a lupus-like syndrome, or systemic toxicity. It must be distinguished from a medication-related effusion without clear inflammation, in which sodium and water retention, altered permeability, or other mechanisms may predominate.

The diagnosis is not simply the appearance of pain or fluid after starting a therapy. The treated disease may itself cause pericarditis: this problem is particularly evident in neoplasms, inflammatory bowel disease, and autoimmune conditions. The timeline, exclusion of alternatives, and course after withdrawal help establish the role of the drug, but rarely provide absolute proof.

Treatment links removal of the suspected exposure with management of the cardiac syndrome and the condition for which the medication was prescribed. Severity depends mainly on myocardial involvement, tamponade, and systemic reaction. Early recognition may allow complete recovery and prevent new exposures, without indiscriminately abandoning entire therapeutic classes when the risk is specific to one agent or remains uncertain.

Implicated drugs and quality of causal evidence

available evidence is heterogeneous and includes spontaneous reports, case series, observational studies, and regulatory information. A recent systematic review confirms the variety of reported agents and the possible severity of complications. However, the number of published cases does not measure incidence among users: it depends on how widely the drug is used, clinician awareness, and reporting propensity. Lists of suspected agents should therefore be interpreted by distinguishing established associations from isolated reports.

aminosalicylates, including mesalazine, may be associated with pericarditis or myopericarditis, often within the first weeks of exposure. The onset of chest pain during this interval requires attention, but inflammatory bowel disease may cause cardiac manifestations independently of treatment. Intestinal disease activity, previous exposures, and response to withdrawal help the assessment. Re-administration may cause rapid recurrence and should not be used as a routine diagnostic test.

drug-induced lupus-like syndrome may include serositis, arthralgia, and constitutional symptoms. Hydralazine and procainamide are classic examples, although their use and risk differ. Long latencies are possible and make it insufficient to limit the history to the most recent prescriptions. Autoantibodies, including an antihistone profile in the appropriate context, support the assessment but do not by themselves prove clinical disease. Hydralazine may also be associated with vasculitic presentations requiring specific renal and pulmonary evaluation.

antineoplastic drugs may cause pericardial injury through different mechanisms. Some cytotoxic agents cause acute cardiac injury; other treatments promote serositis, altered permeability, or fluid retention. Dasatinib is associated mainly with pleural effusions but also with pericardial effusions, and management follows severity, response, and the indications for hematologic therapy. It is not correct to automatically extend the same risk to all tyrosine kinase inhibitors or to interpret every collection as immune-mediated inflammation requiring corticosteroids.

immune checkpoint inhibitors may be associated with pericarditis and effusion, sometimes together with myocarditis. In patients with cancer, however, neoplastic progression, infection, and atypical tumor response are important alternatives. Observational studies show an association with pericardial events without resolving every individual attribution. The presence of malignant cells in fluid does not make a concomitant inflammatory component impossible, whereas negative cytology does not automatically prove immune toxicity.

oral minoxidil carries a regulatory warning concerning effusion and possible tamponade; renal function and fluid retention influence interpretation. This information should not be transferred without distinction to different formulations and modes of use. Clozapine, some biologic therapies, and other medications have been linked to myopericardial events with varying levels of evidence. Anticoagulants may also promote hemopericardium in predisposing circumstances, but medication-facilitated bleeding is not necessarily inflammatory pericarditis caused by the drug.

Pathogenic mechanisms and pathophysiology

hypersensitivity may produce an inflammatory response involving the pericardium and myocardium, sometimes together with skin, liver, or other organs. Eosinophilia, rash, and fever suggest this mechanism but may be absent. The absence of blood eosinophilia does not exclude an immune-mediated cardiac reaction. Severity depends on the organ involved and the speed of progression, not only on the extent of skin manifestations or the amount of fluid identified on echocardiography.

drug-induced autoimmunity may emerge after prolonged exposure through altered immune tolerance. In lupus-like syndromes, serositis is part of a systemic picture whose activity may regress after withdrawal, while some autoantibodies persist longer. This dissociation prevents serologic normalization alone from being used as a treatment goal. Persistence or progression of organ disease should also prompt consideration of a primary autoimmune condition or overlap syndrome.

immune-oncologic toxicity reflects removal of physiologic brakes on the immune response. The process may involve multiple sites and follow a course different from common idiopathic pericarditis. Coexisting myocarditis, myositis, and neuromuscular symptoms radically change the level of urgency. Dysphagia, marked weakness, conduction disturbances, or elevated troponin require active investigation for associated complications, even if pericardial pain appears to be the initial manifestation.

direct injury and endothelial alterations represent mechanisms distinct from hypersensitivity. Some treatments may cause exudation or myocardial injury, whereas others promote edema through sodium retention or changes in permeability. The echocardiographic result may look similar, but the therapeutic choice differs. A diuretic may be useful if fluid overload is present but does not remove active inflammation; conversely, immunosuppression is not a universal response to every effusion arising during drug therapy.

pericardial compression depends on the relationship between the rate of accumulation and the adaptive capacity of the sac. A medication-related effusion may remain stable or progress to tamponade, with reduced filling and cardiac output. Withdrawal of the agent does not guarantee regression quickly enough to avoid drainage. In patients with bleeding, renal injury, or polypharmacy, several factors may accelerate deterioration and should be addressed simultaneously.

myocardial involvement may cause arrhythmias, conduction blocks, ventricular dysfunction, and shock even with little fluid. This possibility explains why severity should not be defined exclusively by the effusion. Fibrous organization and constriction are possible but not inevitable complications; their frequency cannot be reliably estimated from published cases alone. In most clinical reasoning, the priority remains prompt recognition of reversible injury and prevention of reactivation by subsequent exposure.

Clinical presentation and reconstruction of exposure

pericardial pain may be acute, pleuritic, and influenced by position, associated with fever or malaise. Dyspnea and reduced exercise tolerance may instead dominate in effusions without substantial inflammation. Palpitations, syncope, persistent atypical pain, and marked weakness broaden suspicion to myocarditis, ischemia, or arrhythmias. Onset after a new prescription is a clue, but should not prematurely narrow assessment of urgent causes of chest pain.

medication reconciliation includes prescribed medications, over-the-counter products, supplements, compounded preparations, and intermittently taken drugs. For each, the start date, dose, recent increases, interruptions, and restarts are needed. It is useful to distinguish active ingredient from brand name because the same agent may appear in different products. Documentation of oncology cycles and infusions can reveal relationships that do not emerge from a list updated only on the day of the visit.

time window should be compatible with the suspected mechanism. A hypersensitivity reaction may occur early or more rapidly after sensitization; drug-induced lupus may require months or years. An event after withdrawal is not impossible when immune effects persist or biologic exposure is prolonged. There is therefore no single interval beyond which all drug responsibility is excluded, but an atypical latency requires a more careful search for alternatives.

systemic signs help define the phenotype: rash, facial edema, eosinophilia, arthralgia, cytopenias, liver abnormalities, hematuria, or pulmonary symptoms may indicate a multiorgan reaction. Their significance depends on the drug and the underlying disease. Examination should include skin, mucosae, the respiratory system, and signs of congestion. Fever in an immunosuppressed patient should not be attributed to toxicity before infection and neutropenia have been seriously considered.

hemodynamic assessment includes blood pressure, heart rate, perfusion, urine output, and venous pressure. Tachycardia, hypotension, or jugular venous distension suggest a complication but may also result from sepsis, dehydration, or ventricular dysfunction. Pulsus paradoxus may help identify tamponade without being mandatory. The presence of a suspected adverse reaction does not alter the need for stabilization and rapid imaging when signs of hypoperfusion or respiratory deterioration appear.

previous exposures and previously documented reactions deserve particular attention. A previous unexplained episode during the same therapy may increase causal plausibility, but memories should be verified through records when possible. A personal or family history of autoimmune disease may suggest alternatives without excluding a drug contribution. The interview should finally identify the benefit provided by the suspected medication and the consequences of interrupting it, so that a clinically sustainable substitution can be organized.

Diagnosis of the syndrome and causal attribution

pericardial diagnosis precedes attribution to the drug. History, friction rub, ECG, effusion, and signs of inflammation should be integrated according to the clinical picture. Echocardiography and troponin help identify complications and myocardial involvement; C-reactive protein and complete blood count describe the systemic response. An asymptomatic collection without other findings should be defined as an effusion, while keeping separate the question of its relationship to drug therapy.

cardiac magnetic resonance is useful when uncertainty persists about the inflammatory or myocardial component and the patient is stable. It may document edema and enhancement but cannot identify the responsible molecule. CT is selected for suspected neoplasia, thromboembolism, or another thoracic condition. Myocardial or pericardial biopsy is not routinely performed to prove every drug reaction; it is considered when a severe or atypical phenotype raises a question whose answer could concretely change therapy.

The differential diagnosis should include infections, autoimmune diseases, renal failure, hypothyroidism, post-procedural injury, and neoplasia according to context. In a person with intestinal disease, an extraintestinal manifestation must be distinguished from an aminosalicylate effect; in a patient with cancer, neoplastic infiltration, radiotherapy, and opportunistic infections should be considered. Investigation should be targeted: indiscriminate panels may generate irrelevant positive results that make attribution more confusing rather than clearer.

dechallenge, meaning observation after removal of the exposure, is an important element. Temporally coherent improvement strengthens the hypothesis but does not prove it if anti-inflammatory treatment, drainage, or other therapies are started at the same time. Lack of immediate regression does not exclude causality, especially in persistent immune mechanisms or structural injury. Both the interventions performed and the objective course of pain, biomarkers, ventricular function, and fluid should be documented.

rechallenge with recurrence of the event may provide a strong clue when it occurs inadvertently, but should not be provoked to obtain diagnostic certainty. Causality scales may make reasoning transparent without replacing clinical judgment or compensating for missing information. A pharmacovigilance signal indicates an association to be studied, not a known individual probability. The conclusion should express the degree of plausibility and the main remaining alternatives, avoiding unsupported definitive labels.

fluid tests are guided by the indication for drainage and etiologic suspicion. Cytology, cultures, and other analyses may identify competing causes, but absence of tumor cells or microorganisms is not a positive test for toxicity. In suspected drug-induced lupus, autoantibodies and organ assessment complete the picture. Normalization of a single parameter should not outweigh the overall assessment: causality is a longitudinal synthesis, not the result of one isolated test.

Assessment of severity and subsequent investigations

initial risk stratification distinguishes a stable episode without myocardial injury from tamponade, arrhythmias, conduction disturbances, ventricular dysfunction, or a severe systemic reaction. These latter presentations require hospital observation and appropriate monitoring. Even an apparently compensated patient may deteriorate during evolution of an immune toxicity. Surveillance intensity should therefore derive from the phenotype and the agent, as well as from collection size and initial vital signs.

assessment for myocarditis includes troponin trends, serial ECGs, ventricular function, and advanced imaging when indicated. During checkpoint inhibitor treatment, neuromuscular symptoms or electrical abnormalities may require an urgent pathway even with preserved ejection fraction. Normal systolic function does not exclude clinically significant myocardial injury. Suspicion should not be deferred until shock develops because timing of recognition influences the possibility of intervening on the immune response.

multiorgan assessment is necessary when signs of hypersensitivity, drug-induced lupus, or vasculitis appear. Complete blood count with differential, liver function, creatinine, and urinalysis may reveal initially subtle involvement. Immunologic evaluation should follow the phenotype, for example in a patient exposed to hydralazine who has hematuria and pulmonary symptoms. Organ injury may require treatment beyond withdrawal alone, with involvement of the appropriate specialists.

interaction analysis should be performed before adding pericardial therapy. CYP3A4 or P-glycoprotein inhibitors may increase colchicine exposure, especially with renal or hepatic impairment. Anticoagulants, antiplatelet agents, and NSAIDs modify bleeding risk; nephrotoxic drugs may reduce the safety margin. A prescription that is formally correct for pericarditis may become inappropriate in the patient's actual combination, making a complete medication review useful.

The need for the suspected drug should be reassessed with the prescriber. For a replaceable therapy, the balance may favor a definitive alternative; for effective oncologic treatment or an essential psychiatric medication, the decision requires multidisciplinary discussion. This does not mean ignoring toxicity, but defining a safe pathway for the underlying disease. Urgent withdrawal in severe presentations should be accompanied, as soon as possible, by a continuity-of-care plan.

causal documentation should preserve the timeline, dose, phenotype, tests, concomitant therapies, and response. These data are useful both for pharmacovigilance and for future prescriptions. A generic label of allergy to all antineoplastic drugs or all anti-inflammatory drugs is poorly informative and may preclude useful options. It is preferable to describe precisely the suspected agent, event severity, and strength of attribution, updating the judgment if new information emerges during follow-up.

Withdrawal, cardiac therapy, and possible resumption

removal of exposure is central when the causal relationship is plausible, with immediate priority in severe reactions. It should be agreed upon and clinically managed, with alternatives or temporary measures for the treated condition. Withdrawal does not replace treatment of complications: tamponade, shock, or arrhythmias require specific interventions. In mild cases with diagnostic uncertainty, the decision considers causal probability, medication benefit, and the feasibility of close monitoring.

isolated inflammatory pericarditis may require aspirin or NSAIDs and colchicine according to general principles, taking individual contraindications into account. Treatment duration and tapering follow clinical remission and control of inflammation, not merely time since withdrawal. Studies of idiopathic forms support treatment of the syndrome but do not demonstrate identical efficacy for every drug reaction. A retention-related effusion without inflammation instead requires management of the mechanism and fluid balance.

corticosteroids may be needed in important immune-mediated reactions, drug-induced lupus with significant activity, or when initial approaches cannot be used. Dose and speed of intervention depend on organ injury. Checkpoint inhibitor myocarditis requires an urgent specialist pathway with specific immunosuppression and monitoring, different from routine treatment of isolated pericarditis. Before and during therapy, infections, metabolic risks, and impact on the oncologic plan should be considered without delaying lifesaving care.

pericardial drainage is indicated when compression compromises circulation or when other clinical and diagnostic conditions require it. Suspected drug etiology does not justify waiting if the patient deteriorates. If the effusion re-forms, persistence of exposure, alternative mechanisms, and adequacy of drainage should be reassessed. A surgical window and other procedures are selected according to anatomy and course, whereas pericardiectomy is reserved for documented clinically relevant constriction.

resumption of treatment is not an automatic step after recovery. In severe hypersensitivity reactions or those with myocarditis, re-exposure may be particularly risky. In other settings, for example some toxicities from oncologic therapies, it may be discussed after complete reassessment, considering alternatives, expected benefit, previous severity, and monitoring capability. Any dose reduction or substitution should follow agent-specific recommendations; there is no universal protocol valid for every drug.

persistent forms after withdrawal require renewed diagnostic assessment. Self-sustaining inflammation may require prolonged treatment, but neoplastic progression, infection, and primary autoimmune disease should be reconsidered. Interleukin-1 blockade is not a standard treatment for all drug-induced pericarditis: it may be considered only in selected recurrent phenotypes, with evidence limitations and attention to the underlying disease. Prolonging immunosuppression without reassessing the cause exposes patients to avoidable risks.

Prognosis, complications, and prevention of re-exposure

A favorable prognosis is possible when the drug is recognized and removed before structural injury or severe compromise occurs. The speed of recovery varies with mechanism, duration of exposure, and myocardial involvement. Pain and biomarkers may improve before complete disappearance of the effusion, while some immunologic findings persist beyond remission. Assessment of recovery should integrate symptoms, function, and course, avoiding both excessive treatment and premature closure of follow-up.

major complications include tamponade, arrhythmias, conduction blocks, heart failure, and shock when the myocardium is involved. A systemic reaction may add hepatic, renal, hematologic, or respiratory injury. Constriction has been described but should not be presented as a usual outcome based on selected cases. Management of the acute phase and identification of the full spectrum of involved organs are more informative for prognosis than simply placing the agent on a list of possible causes.

subsequent follow-up should confirm regression of fluid, resolution of inflammation, and electrical and ventricular stability. Test frequency and observation duration are adapted to initial severity and the biologic persistence of treatment. In patients with cancer, pericardial monitoring is integrated with oncologic monitoring, keeping open the possibility of an alternative cause if the course diverges from expectations. A new effusion months later should not automatically be attributed to the same episode.

prevention of recurrence requires information accessible to different prescribers. The agent, formulation, type of reaction, and recommendation regarding re-exposure should be documented in the clinical record and communicated to the patient. Substitution with a medication from the same class requires assessment of possible cross-reactivity and alternatives; it is neither always safe nor always prohibited. For aminosalicylates in particular, the choice after a cardiac event requires specialist discussion and careful monitoring.

pharmacovigilance contributes to knowledge of rare reactions and should include clinically relevant suspected cases even without absolute certainty. A useful report describes exposure, latency, objective findings, alternatives considered, and the course after withdrawal. It does not demonstrate a causal link or estimate incidence. Data quality is essential so that subsequent assessments can distinguish a true signal from effects of the treated disease or associations due to polypharmacy.

return to activities depends on remission and on any myocarditis, which requires a more detailed assessment before intense exertion. Symptoms requiring reassessment should be explained, especially chest pain, increasing dyspnea, palpitations, and syncope. The patient should not have to choose alone between stopping an essential therapy and ignoring a possible adverse effect: a shared plan for contact, substitution, and monitoring makes prevention practically applicable.

References
1. Schulz-Menger J et al. 2025 ESC Guidelines for the management of myocarditis and pericarditis. European Heart Journal. 2025;46(40):3952-4041.
2. Klein AL et al. Pericardial Diseases: International Position Statement on New Concepts and Advances in Multimodality Cardiac Imaging. JACC: Cardiovascular Imaging. 2024;17(8):937-988.
3. Adler Y et al. 2015 ESC Guidelines for the diagnosis and management of pericardial diseases: The Task Force for the Diagnosis and Management of Pericardial Diseases of the European Society of Cardiology (ESC). Endorsed by: The European Association for Cardio-Thoracic Surgery (EACTS). European Heart Journal. 2015;36(42):2921-2964.
4. Collini V et al. Drug-induced pericarditis: a systematic review of all published cases. International Journal of Cardiology. 2026;445:134000.
5. Brown G. 5-Aminosalicylic Acid-Associated Myocarditis and Pericarditis: A Narrative Review. Canadian Journal of Hospital Pharmacy. 2016;69(6):466-472.
6. Gong J et al. Pericardial disease in patients treated with immune checkpoint inhibitors. Journal for ImmunoTherapy of Cancer. 2021;9(6):e002771.
7. Lyon AR et al. 2022 ESC Guidelines on cardio-oncology developed in collaboration with the European Hematology Association (EHA), the European Society for Therapeutic Radiology and Oncology (ESTRO) and the International Cardio-Oncology Society (IC-OS). European Heart Journal. 2022;43(41):4229-4361.
8. Mori S et al. Pericardial effusion in oncological patients: current knowledge and principles of management. Cardio-Oncology. 2024;10:8.
9. National Library of Medicine. Minoxidil tablet: prescribing information. DailyMed. n.d. Online regulatory resource. Accessed September 26, 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=ab30461c-f8c8-409d-9e24-d58ed8a34873.
10. European Medicines Agency. Sprycel: European public assessment report and product information. European Medicines Agency. n.d. Institutional online resource. Accessed September 26, 2026. https://www.ema.europa.eu/en/medicines/human/EPAR/sprycel.
11. U.S. National Library of Medicine. COLCHICINE tablet, film coated: prescribing information. DailyMed. n.d. Accessed September 26, 2026. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=50b9c910-49a3-4a57-80f5-dc36c09da4df.
12. Imazio M et al. A randomized trial of colchicine for acute pericarditis. New England Journal of Medicine. 2013;369(16):1522-1528.
13. Spodick DH. Acute cardiac tamponade. New England Journal of Medicine. 2003;349(7):684-690.