When Cancer Immunotherapy Turns on the Heart

Immune checkpoint inhibitors have reshaped modern oncology by releasing the brakes on T cells and letting the immune system attack tumors that once defied treatment. Drugs targeting PD-1, PD-L1, and CTLA-4, including pembrolizumab, nivolumab, ipilimumab, atezolizumab, and newer approvals such as toripalimab for advanced nasopharyngeal carcinoma, dostarlimab for mismatch repair–deficient endometrial cancer, and tislelizumab for esophageal and gastroesophageal junction cancers, have earned their place in the oncology arsenal. Yet the same immune activation that destroys cancer cells can spill over into healthy tissue. A comprehensive narrative review published in Supportive Care in Cancer synthesizes what is known about the cardiovascular consequences of this therapy, a spectrum that includes myocarditis, pericarditis, arrhythmias, heart failure, vasculitis, and thromboembolism. The central message is a paradox of rarity and severity: these cardiac toxicities occur in only a small fraction of patients, but when they strike, particularly as myocarditis, they can be rapidly fatal.

The numbers tell a nuanced story. Immune-related adverse events overall are common, with one cited study reporting that roughly 90 percent of patients on anti-CTLA-4 therapy and 70 percent of those on anti-PD-1 or anti-PD-L1 agents experience some immune toxicity. Cardiovascular events, by contrast, remain uncommon, with most clinical trial populations showing rates between 0.1 and 1.14 percent. A retrospective analysis of 424 patients found that 62, or 14.6 percent, developed new cardiovascular disorders after starting checkpoint blockade, though this broad category included conditions well beyond inflammation of the heart muscle. In a safety meta-analysis, dyslipidemia was the most frequent cardiovascular adverse event at 19.3 per 1000 patients, while myocarditis was the least common at 3.2 per 1000. Cerebral arterial ischemia, heart failure, and myocardial infarction each occurred two to three times more often than myocarditis.

Severity, not frequency, is what makes ICI-associated myocarditis the most feared complication. Early studies reported case-fatality rates approaching 50 percent, but the review is careful to note that these figures come largely from initial case series and pharmacovigilance databases and likely overestimate contemporary risk as recognition and treatment have improved. One meta-analysis of 4751 patients identified cardiovascular immune-related events in 1.3 percent, with myocarditis accounting for just over half of those cases, and 15 patients, or 24.6 percent of the affected group, died from cardiovascular causes. Regimen matters: the same analysis reported cardiovascular event rates of 3.1 percent with monotherapy, 5.8 percent with dual checkpoint blockade, and 3.7 percent when ICIs were combined with chemotherapy. Combination PD-1 and CTLA-4 therapy raises myocarditis incidence roughly two- to threefold compared with PD-1 inhibition alone, and female patients and those over 75 appear at elevated risk, as do patients with pre-existing cardiovascular disease, autoimmune conditions, diabetes, or obesity.

The biology underlying this toxicity is a story of immune tolerance collapsing in the heart. Blocking PD-1, PD-L1, or CTLA-4 unleashes T-cell activation that can turn against cardiomyocytes. CD8-positive cytotoxic T cells attack heart muscle cells directly through perforin- and granzyme-mediated killing, producing apoptosis and necrosis, while CD4-positive T cells amplify the assault by secreting inflammatory cytokines including TNF-alpha, interferon-gamma, and IL-6. The shared antigen hypothesis adds a further layer: T cells expanded against tumor antigens may cross-react with similar proteins expressed by cardiac tissue, and clonal expansion of T cells with high affinity for cardiac epitopes has been observed in ICI-associated myocarditis. Autoantibodies against cardiac proteins, complement activation, and loss of B-cell tolerance may compound the injury. Autopsy studies reveal a telling pattern, with CD8-positive T cells predominating in myocarditis after PD-1 inhibitors and CD4-positive cells more prominent after CTLA-4 blockade.

Timing is one of the most clinically important features. Myocarditis typically appears early, with a median onset of about 30 days, most often after the first or second infusion, and a broader reported range of 2 to 54 weeks. The presentation is notoriously variable, spanning asymptomatic troponin elevations, acute coronary syndrome–like chest pain with normal coronary arteries, new-onset heart failure, and fulminant disease with cardiogenic shock and malignant arrhythmias. High-grade atrioventricular block and ventricular tachycardia or fibrillation are particularly ominous because they can precede overt heart failure and cause sudden cardiac death. Any unexplained conduction abnormality in a patient on checkpoint blockade should trigger immediate cardiac evaluation. A diagnostic clue worth remembering: when skeletal myositis or myasthenia gravis appears after ICI initiation, clinicians should suspect concurrent myocarditis. Notably, nearly half of affected patients retain normal systolic function despite clear evidence of myocardial injury.

Diagnosis demands a multimodal approach because no single test is definitive. Cardiac troponin remains the cornerstone biomarker, with troponin I preferred for its cardiac specificity when skeletal muscle inflammation coexists, although one study found troponin T positive in 93 percent of confirmed cases versus 64 percent for troponin I, suggesting higher sensitivity for troponin T in some settings. Electrocardiography is rapid and informative but imperfect: in one case series, 89 percent of patients had abnormal ECGs at diagnosis despite mostly normal baselines, and QRS prolongation has been linked to worse outcomes including complete heart block and shock. A normal ECG, however, does not exclude myocarditis. Echocardiography is accessible but limited, with roughly 51 percent of confirmed myocarditis cases showing normal left ventricular ejection fraction at diagnosis, including 38 percent who later developed major adverse cardiac events.

Cardiac magnetic resonance imaging has emerged as the gold-standard noninvasive test, detecting myocardial edema, inflammation, and fibrosis through T2-weighted imaging and late gadolinium enhancement. Timing matters here too: in one study, late gadolinium enhancement appeared in only about 21 percent of patients imaged within the first three days of symptoms but in 72 percent of those imaged on day four or later. Advanced techniques such as T1 and T2 mapping and extracellular volume quantification have pushed reported sensitivities above 90 percent, and in a multicenter cohort of 136 cases, 95 percent met at least one CMR criterion when mapping was included, compared with roughly 53 percent meeting classic edema criteria alone. Endomyocardial biopsy remains the histopathologic gold standard in principle but is underused in practice because it is invasive, prone to sampling error in patchy disease, and often bypassed when clinical suspicion is high and treatment delays can be fatal. It is reserved for uncertain or atypical cases, suspected giant cell or eosinophilic myocarditis, or nonresponse to standard therapy.

Beyond the heart muscle itself, the review catalogs a wider cardiovascular footprint. ICI-associated pericarditis can progress to cardiac tamponade, with a systematic review finding tamponade physiology in 41 percent of reported pericardial disease cases and invasive drainage required in 68 percent; cytology revealed malignant cells in about 25 percent of cases initially attributed to immunotherapy, underscoring the need to exclude other causes. Arrhythmias range from atrial fibrillation to lethal ventricular events, and a disproportionality analysis of the FDA Adverse Event Reporting System showed significantly increased arrhythmia reports with anti-PD-1 and anti-PD-L1 agents. Rare vasculitic syndromes have also been described, including aortitis, which FDG PET/CT identified as pathologic uptake in the aorta or major branches in approximately 63 percent of evaluated patients in one series. Venous thromboembolism, whose risk appears modestly elevated in larger cohort analyses, particularly when ICIs are combined with chemotherapy, rounds out the spectrum.

Management hinges on speed. High-dose corticosteroids are the cornerstone, with methylprednisolone at 1 to 2 mg/kg recommended when myocarditis is suspected and pulse-dose 1 g/day reserved for inadequate response, though some experts advocate upfront pulse dosing in patients with markedly elevated troponins. Guideline frameworks diverge in important ways: ASCO and NCCN grade severity on a CTCAE-based four-tier scale and reserve high-dose steroids for grade 2 or higher and grade 3 or higher disease respectively, while the 2022 European Society of Cardiology guidelines instead classify myocarditis as fulminant versus non-fulminant and recommend high-dose steroids for confirmed cases regardless of grade, alongside withholding ICIs in all suspected cases during work-up. Steroid-refractory disease may prompt escalation to abatacept, alemtuzumab, tacrolimus, tocilizumab, mycophenolate mofetil, intravenous immunoglobulin, or plasmapheresis, though a case-control study linked infliximab to higher cardiovascular death, making it contraindicated in heart failure. Mechanical circulatory support with ECMO can bridge patients to recovery.

The review’s authors are candid about the limits of the evidence base. As a narrative rather than systematic review, it carries selection bias, and the underlying literature is dominated by retrospective cohorts, case series, and pharmacovigilance analyses, with few prospective studies and no adequately powered randomized trials of management strategies. Reported incidence and fatality estimates vary widely with case definitions and surveillance intensity, and emerging biomarkers such as IL-17A rest largely on preclinical data requiring validation. Those caveats frame the field’s central challenge: as immunotherapy expands across cancer types and treatment lines, harmonizing divergent grading systems, validating predictive biomarkers, and building prospective multicenter studies will be essential to turning a rare but lethal complication into one that clinicians can reliably anticipate, detect, and defuse.

Subject of Research: Immune checkpoint inhibitor–associated cardiotoxicity: incidence, diagnosis, and management

Article Title: Immune checkpoint inhibitor–associated cardiotoxicity: incidence, diagnosis, and management

Article References: Haddad, S. F., Philip, L., Padrón, L., Zoghbi, M., Jacobson, K., Burk, K., Chaftari, P., Markides, D. M., Dendy, S., Krishnamani, P., Mendoza Hernández, A., & Cruz Carreras, M. T. (2026). Immune checkpoint inhibitor–associated cardiotoxicity: incidence, diagnosis, and management. Supportive Care in Cancer, 34(11), Article 1153. https://doi.org/10.1007/s00520-026-11362-x

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11362-x

Keywords: Immune, checkpoint, inhibitor, associated, cardiotoxicity, incidence, diagnosis, management, scientific research