Immune checkpoint inhibitor myocarditis

Article written and reviewed by Dr Luca Moretti

Immune checkpoint inhibitor myocarditis is an inflammatory toxicity associated with antibodies directed against CTLA-4, PD-1 or PD-L1. These therapies have transformed the prognosis of many cancers by releasing T lymphocytes from inhibitory signals, but the same activation may break tolerance to the heart. The event is uncommon relative to the growing use of ICIs, yet it retains a high risk of arrhythmias, conduction block and death.

The phenotype does not coincide with classic viral myocarditis. Ventricular function may remain normal while conduction worsens, CMR may initially be negative and troponin elevation may accompany extensive myositis. Diagnosis therefore rests on a combination of exposure, biomarkers, ECG, imaging, exclusion of alternatives and, in selected cases, biopsy.

Overlap with myositis and myasthenia, referred to as triple-M syndrome, is a crucial clinical feature. Ptosis, diplopia, dysphagia, neck weakness and reduced respiratory muscle strength may precede ventilatory failure that is not explained by the echocardiogram; the heart and respiratory muscles must be assessed simultaneously from the first suspicion.

Onset often occurs during the first weeks or after the first doses, especially with combinations of checkpoint inhibitors, but late events have been described. Probability does not fall to zero after prolonged treatment; every new troponin rise, arrhythmia or symptom requires proportionate assessment without automatically attributing the event to immunotherapy.

Therapeutic decisions involve cardio-oncology, oncology, intensive care, neurology, immunology and electrophysiology; withholding the ICI and suppressing immunity may reduce tumor control, whereas delaying corticosteroids in severe disease can be fatal. Cardiac risk and the oncologic objective are managed within the same plan rather than in separate sequence.

Etiology, Pathogenesis and Pathophysiology

CTLA-4 limits initial T-cell activation, whereas PD-1 and PD-L1 attenuate the response in peripheral tissues and the tumor microenvironment. Therapeutic antibodies remove these brakes and promote stronger antitumor activity; loss of tolerance, however, may expand clones capable of recognizing cardiac antigens.

Pathologic studies show infiltrates of T lymphocytes, often CD8+, and macrophages in the myocardium and conduction system. Shared clones among tumor, heart and skeletal muscle support the hypothesis of common antigens; necrosis and distribution vary, explaining why small foci may cause complete heart block without global dysfunction.

Combination anti-CTLA-4 and anti-PD-1 therapy increases immune activation and risk compared with monotherapy in several cohorts, without making the event predictable in an individual patient. Pre-existing autoimmune or cardiovascular disease, other cardiotoxic treatments and tumor characteristics may contribute; no baseline biomarker identifies with certainty who will develop myocarditis.

Concomitant myositis causes elevated CK and may increase troponin T through skeletal-muscle involvement. Troponin I often provides greater cardiac specificity in this setting, but does not replace clinical assessment; myasthenia may be seronegative and respiratory weakness can emerge rapidly, requiring serial functional measurements.

Inflammation may involve the pericardium, coronary arteries and vessels in addition to the myocardium; ICIs are also associated with noninflammatory dysfunction, Takotsubo syndrome and acceleration of atherosclerotic events. Elevated troponin during immunotherapy therefore does not define a single toxicity, and ischemia remains one of the principal differential diagnoses.

Prolonged target occupancy and immunologic memory explain why discontinuation does not immediately stop the process. Corticosteroids reduce inflammatory transcription and cellular trafficking, whereas abatacept restores an inhibitory signal through CTLA-4; strategies directed against JAK or other pathways seek to control refractory responses without indiscriminately eliminating all immunity.

The electrical substrate evolves with edema, necrosis and scar. During the active phase it may change within hours, causing progression from PR prolongation to complete heart block or ventricular tachycardia. After resolution, septal and ventricular fibrosis may maintain risk and the need for a device.

Severity is not defined by ejection fraction alone; troponin, conduction, arrhythmias, hemodynamics, respiratory function and multiorgan overlap contribute to classification. This approach prevents labeling as “mild” a form with preserved function but rapidly progressive block or myasthenia.

The tumor may contribute through shared antigens, inflammation, infiltration and concomitant treatments; thymoma and cancers associated with neuromuscular autoimmunity warrant particularly careful surveillance, but no histologic type makes the event inevitable. Thoracic radiotherapy, anthracyclines and targeted therapies add substrates that may confound causality and reduce reserve.

The microbiota, T-cell repertoire and immunogenetic variants are being studied as predictors, but they do not yet have a clinical application capable of reliably selecting patients. Screening should not deprive a person of cancer therapy on the basis of experimental markers; prospective research requires uniform definitions and samples obtained before corticosteroids.

Necrosis of the conduction system may involve far less myocardial mass than is needed to reduce ejection fraction. This anatomic principle explains the discordance between echocardiography and risk; surveillance therefore gives equal weight to ECG, troponin and neuromuscular symptoms rather than relying on a hierarchy centered only on function.

Clinical Manifestations

Dyspnea, chest pain, palpitations, fatigue and syncope are common but nonspecific. Some patients are identified through an asymptomatic rise in troponin; others present with cardiac arrest, shock or complete heart block. The speed of progression requires an urgent pathway for every signal accompanied by ECG abnormalities or symptoms.

Systolic function is preserved in a substantial proportion of cases and should not be reassuring by itself. Abnormal strain may precede a fall in ejection fraction; pericardial effusion, regional abnormalities and right ventricular dysfunction add information, but no echocardiographic finding is pathognomonic.

Arrhythmias include ventricular tachycardia, ventricular fibrillation, atrial fibrillation and bradyarrhythmias; a new bundle-branch block or PR prolongation may precede advanced block. Telemetry and serial ECGs are used because a normal tracing at triage does not describe the entire trajectory.

Ptosis, diplopia, dysarthria, dysphagia, proximal or neck weakness, and dyspnea with an echocardiogram that is not severely impaired suggest concomitant myositis or myasthenia. Oxygen saturation and respiratory rate may initially remain acceptable despite reduced vital capacity; neurologic assessment is not postponed until hypercapnia appears.

Other immune-related toxicities include hepatitis, colitis, pneumonitis, nephritis, endocrinopathies and rash; hypotension may result from adrenal insufficiency or sepsis as well as cardiogenic shock. Multiorgan assessment clarifies support needs and prevents every abnormality from being attributed to the heart.

Fever and inflammatory markers are not mandatory; troponin may remain elevated for days or weeks and its decline does not always parallel symptoms. A new rise during taper requires distinction among recurrence, myositis, ischemia and renal dysfunction.

The nonsevere form includes stable patients without arrhythmias, conduction block, significant dysfunction or dangerous overlap, but the classification is dynamic. Initial monitoring must be able to detect a rapid shift in category; therapy is adapted to the trajectory rather than only to the snapshot at presentation.

In patients with advanced cancer, dyspnea and fatigue are common and may result from anemia, pulmonary embolism, progression, infection or deconditioning. A systematic approach avoids both overdiagnosis and delay; a palliative goal does not exclude treatment of reversible toxicity, but it modifies the proportionality of support.

Pericarditis may present in isolation or together with myocarditis and sometimes with a significant effusion. Anti-inflammatory drugs and colchicine may control an uncomplicated pericardial phenotype but do not replace immunosuppression when troponin, conduction or function indicate myocarditis; tamponade and cardiogenic shock require immediate echocardiographic distinction.

ICI pneumonitis, pulmonary embolism, metastases and infection share dyspnea and hypoxemia; chest CT, blood-gas analysis and perfusion assessment complement echocardiography. Reduced vital capacity with limited pulmonary imaging abnormalities suggests respiratory muscle weakness and justifies closer monitoring.

Investigations and Diagnosis

Before immunotherapy, cardiovascular history, ECG, troponin and risk assessment establish a reference; echocardiography and other tests are added in appropriate profiles. During the first doses, surveillance strategies vary among guidelines and centers; a baseline value prevents a chronic elevation from being interpreted as new.

When acute disease is suspected, ECG, troponin I or T, natriuretic peptides, CK, complete blood count, electrolytes, renal and hepatic function and inflammatory markers are repeated. Elevated troponin is central to clinical criteria, but it requires an explanation; CK and transaminases help identify myositis, while TSH and cortisol assess endocrinopathies when indicated.

Echocardiography with strain describes function and supports monitoring; CMR applies the modified Lake Louise criteria, looks for edema, hyperemia and late gadolinium enhancement and assesses alternative diagnoses. In ICI myocarditis, sensitivity may be lower than in conventional forms, especially when performed early; a negative CMR does not close a high-probability case.

Coronary angiography or coronary CT is performed when ischemia is plausible; pulmonary embolism, sepsis, tumor progression, Takotsubo syndrome and toxicity from other drugs are investigated. The possibility of multiple mechanisms means that the algorithm should not stop at the first abnormality.

Endomyocardial biopsy confirms infiltrate and necrosis and is particularly valuable when imaging is negative or discordant, the presentation is severe or second-line therapy depends on certainty. Focality causes false negatives and requires multiple samples; procedural risk is balanced against that of unnecessary massive immunosuppression.

Cardio-oncology diagnostic criteria combine troponin with CMR, histology and minor criteria such as clinical syndrome, arrhythmias, dysfunction and concomitant toxicities. They allow definite or clinical categories, but do not replace judgment; the level of certainty is updated as results become available.

Neuromuscular assessment includes neurologic examination, CK, acetylcholine-receptor antibodies and other selected tests, electromyography and serial respiratory measurements. Negative antibodies do not exclude the syndrome. Dysphagia requires swallowing assessment and aspiration protection.

Therapeutic response is not used as the sole diagnostic test; troponin may fall after corticosteroids in other conditions as well, and function may improve because of support. Conversely, an incomplete response may reflect delay, scar or an alternative diagnosis, not simply an insufficient dose.

Comparison with baseline values is essential in patients with chronic troponin elevation due to renal failure, amyloidosis, cardiomyopathy or metastases; a dynamic rise and new ECG abnormalities carry more weight than isolated exceeding of the reference limit. If troponin T is influenced by myositis, troponin I and the overall picture help, but there is no mathematical conversion between the two tests.

Late CMR may show scar even after edema and troponin have decreased, whereas an early negative scan may precede later findings. Repeating the examination is reasonable if the answer changes the diagnosis or resumption of cancer therapy; ritual repetition without therapeutic consequences adds delay and contrast exposure without benefit.

Biopsy material is also sent for infectious analyses and alternative diagnoses because the oncology patient is immunologically complex. Tumor infiltration, amyloidosis, rejection in transplant recipients and pre-existing myocarditis may be identified; the finding must be correlated with ICI exposure rather than merely described as lymphocytic.

Treatment and Prognosis

The ICI is withheld immediately when there is clinically relevant suspicion and the patient is monitored. In severe forms or when probability is high, high-dose intravenous methylprednisolone is started early without waiting for all results; retrospective studies associate earlier administration with fewer cardiovascular events.

Guidelines and consensus documents differ in details of dosing for nonsevere forms, while converging on the need for high-dose pulses in fulminant, arrhythmic or conduction-block presentations. Response is followed through symptoms, troponin, ECG, hemodynamics and concomitant organ involvement; tapering is gradual and slowed if biomarkers or conduction worsen.

Shock, ventricular tachycardia and conduction block receive intensive care, cardioversion, pacing and circulatory support when indicated. Respiratory function is supported early in neuromuscular overlap; waiting for obvious fatigue can be dangerous because ventilatory failure may deteriorate abruptly.

In refractory cases, depending on center and phenotype, mycophenolate, immunoglobulins, plasma exchange, antithymocyte globulin, abatacept, ruxolitinib and other strategies are used. Evidence comes largely from case series and nonrandomized studies; infliximab requires particular caution and is generally avoided in moderate or severe heart failure.

Abatacept is a CTLA-4 fusion protein that reduces T-cell costimulation by binding CD80/CD86. Strategies using doses adjusted to receptor occupancy, sometimes combined with ruxolitinib, have shown promising results in severe cohorts but remain to be confirmed; prospective trials are defining dose, timing and population.

Every intensive immunosuppressive regimen requires infection prophylaxis and investigation; pneumocystosis, viral reactivations, bacteremia and fungal infections may mimic recurrence and worsen outcome. Immunologic benefit is balanced against possible weakening of antitumor activity.

Historical prognosis has been severe, but recognition of less fulminant forms and faster treatment are changing estimates. Conduction block, ventricular arrhythmias, shock, persistently elevated troponin and neuromuscular overlap worsen outcome; normal ejection fraction alone does not identify a favorable prognosis.

Resumption of immunotherapy after severe myocarditis is generally avoided; after a nonsevere form with limited oncologic alternatives, it may be discussed exceptionally after resolution, review of diagnostic certainty and informed consent. Drug, class, combination, monitoring and discontinuation thresholds are defined before dosing.

Before second-line treatment, failure and severity are defined; troponin that does not decrease, progressive conduction abnormalities, arrhythmias or increasing support indicate uncontrolled activity, but persistently low levels may reflect scar. Excessive escalation exposes patients to infection without correcting irreversible injury; multimodal reassessment precedes new combinations when stability permits.

The effect on antitumor immunity varies among agents, doses and duration; corticosteroids required for severe toxicity are not withheld for oncologic concerns, but immunosuppression is reduced to the minimum effective exposure after control. Tumor imaging and subsequent-line decisions continue during cardiac follow-up so that freedom from toxicity is not achieved at the cost of all tumor control.

Complications

Complete heart block may appear rapidly and require temporary pacing; recovery of conduction is possible, but necrosis and scarring may make it permanent. Selection of a permanent device also considers ventricular arrhythmias and oncologic prognosis.

Cardiogenic shock may require ECMO or other support as a bridge to immunologic recovery; candidacy and goals are shared with oncology, taking into account tumor response and reversibility. Cancer does not automatically exclude temporary support, just as oncologic innovation does not make every intervention proportionate.

Triple-M syndrome can cause respiratory failure, aspiration and myasthenic crisis. Some cardiovascular and antiarrhythmic drugs worsen neuromuscular transmission and require caution; cardiology and neurology coordinate ventilation and immunotherapy while avoiding conflicting recommendations.

Opportunistic infections increase with pulse therapy, combinations and ventilation. New fever, a pulmonary infiltrate or renewed troponin elevation require samples and imaging before everything is attributed to immune toxicity; antimicrobial therapy and immunomodulation may need to proceed together.

Residual scar retains a risk of ventricular tachycardia and reduced reserve; Holter monitoring, echocardiography, CMR and functional assessment guide follow-up. Troponin normalization is important but does not by itself represent complete electrical recovery.

Permanent discontinuation of the ICI may reduce oncologic options and have substantial psychological impact. The patient receives a transparent explanation of the relationship between tumor control and cardiac risk; alternative strategies are discussed without presenting cardiology as an isolated obstacle to treatment.

Recurrences during taper or after rechallenge are possible; a written plan establishes tests, warning symptoms and rapid access. Self-administration of corticosteroids or independent resumption of the ICI is dangerous because it may mask signals and delay multidisciplinary management.

Endocrinopathies may persist permanently and require replacement, whereas myocarditis may resolve. Hypothyroidism, hyperthyroidism and adrenal insufficiency modify heart rate, blood pressure and energy and may be mistaken for recurrence; cardio-oncology follow-up maintains links with endocrinology and primary care.

Prospective registries and multidisciplinary case review improve quality because the event is rare and historical definitions are heterogeneous. Corticosteroid dose, timing, biopsy, overlap and outcomes are recorded in a standardized manner; this infrastructure helps distinguish real improvements from changes in the diagnostic threshold.

Opportunistic prophylaxis is adapted to corticosteroid dose and duration and to additional immunomodulators. Screening for hepatitis, tuberculosis and latent infections precedes second-line therapy when time permits, whereas in shock, sampling and prophylaxis do not delay treatment. Antimicrobial management is updated with lymphocyte recovery and tapering.

Experimental biomarkers, immunophenotyping and CD86 occupancy may make immunomodulation more precise, but they are not yet available everywhere or validated for every regimen. Their use in an expert center is distinguished from the standard of care; innovation is accompanied by prospective recording and clear communication of uncertainty.

The pathway ends with an explicit oncologic plan: alternative therapy, possible observation, exceptional criteria for rechallenge and surveillance. The patient should not be left between two specialists who defer the decision to each other; a multidisciplinary conference documents the balance and identifies who will coordinate follow-up.