Biopsy-proven myocarditis in the intensive care unit: diagnostic and prognostic insights from a monocentric cohort

Background Myocarditis may have fulminant presentation, requiring cardiac intensive cardiac unit (ICU) admission for invasive diagnosis and treatment, including mechanical circulatory support (MCS). ICU management, length of stay and therapeutic strategies have not been specifically studied in endomyocardial biopsy (EMB)-proven myocarditis cohorts. Purpose The aims of our study are: (1)To investigate details of ICU treatments and hospital-stay in EMB-proven myocarditis patients requiring ICU admission, and (2)to evaluate long-term prognostic factors. Methods Our retrospective longitudinal study included EMB-proven myocarditis patients admitted to our third-level referral Hospital ICU, in 1993-2025. All EMBs underwent guideline-based analysis, including polymerase chain reaction (PCR) search for viral genomes. Predictors of death and heart transplant (HTx) were identified by univariable and multivariable regression. Results Among 428 patients with EMB-proven myocarditis, 65 (15%) required ICU admission. Admission SCAI shock stages were A (56%), B (11%), C (20%), D (10%), and E (3%). Lymphocytic myocarditis was the predominant histotype (90%) but rare histotypes were also found (eosinophilic 5%, giant cell 3%, polymorphic 2%). Notably, viral genomes were found on EMB in 58% of cases. Virus-negative cases were treated with immunosuppression (27/65). While 28 patients were directly admitted to our center, 37 came from a spoke center. About half cases received inotropes (for a median duration of 8 days) and 22% necessitated MCS (for a median duration of 15.5 days). Patients from spoke centers more often required MCS (35% vs 3.6% at the hub) and had longer non-hub stays (median 20.5 days, IQR 11.5–43.5). At 66 months follow-up, the composite endpoint of death/HTx occurred in 15% of patients (12% death, 3% HTx). Event rates clustered in patients with higher SCAI stage, need for vasopressors/MCS, and/or lower LVEF. Long-term survival tended to be lower in patients with viral PCR positivity (p=0.06) on EMB and was similar, when Cardiac Magnetic Resonance (CMR) was performed in the acute phase, if CMR was performed before or after EMB (p=0.76). An inverse relationship was found between left ventricular ejection fraction (LVEF) and SCAI shock class (p=0.046). On univariable analysis, lower LVEF (HR 1.06 per 1% LVEF increase), inotrope (HR 4.44) and MCS use (HR 9.10) predicted death/HTx; in multivariable models, both LVEF (HR 1.05 per 1% increase) and MCS use (HR 5.03) were confirmed as death/HTx independent predictors. Conclusions In EMB-proven myocarditis patients requiring ICU admission, outcomes are driven by hemodynamic severity and need for intensive cardiac support. EMB remains indispensable for etiological diagnosis, revealing a high prevalence of virus positive cases and a trend towards lower survival in these cases. Our data support early SCAI staging and invasive assessment of myocarditis patients admitted in the ICU.