Pathological spectrum of immune checkpoint myocarditis: The fast track ICI cohort study

Background Endomyocardial biopsy (EMB) is the gold standard for myocarditis diagnosis. However, the Dallas pathological criteria(1) prevail over multimodal cardiovascular imaging and focus on viral myocarditis. Immune checkpoint inhibitors myocarditis(2) are easier to diagnose in severe cases, but non-severe, asymptomatic forms have emerged. The value of EMB has not been prospectively assessed(3). The lack of predictors for disease severity highlights the need to identify low-risk populations. Purpose To assess the correlation between clinical features and histology characteristics in patients with suspected ICI-myocarditis. Methods This single-centre prospective sub-study of a larger cohort included adults for suspected cardiovascular ICI-related IRAEs who underwent EMB performed when deemed necessary by the MDT to rule in or out ICI-myocarditis. EMBs were reviewed by few experienced pathologists using current guidelines, including CD3/CD68 immunohistochemistry.The primary endpoint was the correlation between EMB findings and clinical features. Results Among 175 consecutive patients referred, we performed a total of 48 biopsies. Of these, 4 were normal, 4 showed definite myocarditis (i.e., >7 T cells/mm² and myocardial loss), 37 showed significant myocardial inflammation (>7 T cells/mm²) but no cell loss, and 1 showed cardiac metastasis.The median number of lymphocytes in the definite myocarditis group was 114/mm² versus 16/mm² in the borderline pattern (Dallas criteria). The clinical characteristics of patients with definite myocarditis and the histological/clinical/radiological parameters of EMB population are depicted in Figure 1. Patients with a definite myocarditis pattern were more likely to: Be thymoma patients [2 (50%) versus 1 (2.7%); p=0.001], have non-lung/non-breast cancers [3 (75%) versus 3 (16.2%); p=0.028], stage IV cancer [4 (100%) versus 10 (27%); p=0.003], higher median CPK levels [741 U/L (IQR 122-1648) versus 61 (IQR 40-100); p=0.035], have received ICIs without other cancer drugs [2 (50%) versus 3 (8.1%); p=0.015] and have undergone fewer cycles of ICIs [2 (IQR 1.25-2) versus 7 (IQR 3-9); p=0.002] and were more frequently hospitalized [4 (100%) versus 8 (21.6%); p=0.001] . Breast cancer patients [28 (78.4%) versus 0 (0%); p=0.001] treated with a combination of ICIs and other chemotherapies [32 (86.5%) versus 1 (25%)p=0.003] and anthracyclines [22 (64.7%) versus 0 (0%); p=0.029] were more likely to present with a borderline pattern(Figure 2). N° of T-cells were significantly correlated with the levels of troponin normalized to the upper normal limit (R=0.3, P=0.028) (Figure 2). Conclusions ICI-myocarditis shows a complex biopsy spectrum and is not a binary phenomenon. T-cell infiltration without necrosis or symptoms remains of unclear significance. We found that some clinical features correlate with histological Dallas criteria. Further research is needed to establish a gold standard for ICI-myocarditis.