LAUSR

e14579 Background: We retrospectively analysed the cases of immune checkpoint inhibitor(ICI) -associated myocarditis managed in PUMCH. We aim to investigate the clinical characteristics of myocarditis and the risk factors affecting the prognosis. Methods: Medical records of all checkpoint inhibitor-related myocarditis were reviewed retrospectively. Descriptive statistics were used for the report of clinical outcomes. Univariate and multivariate Logistic analyses were performed to establish a prognostic model. Results: The median age of 19 patients was 64 years old (range 39̃80 years). The median time of onset from the iniatial of ICIs was 21 days (ranging from 2 to 375 days). Seven cases were grade 4 (36.8%), 2 cases were grade 3 (10.5%), 8 cases were grade 2 (42.1%), and 2 cases were grade 1 (10.5%). Clinical symptoms included dyspnea (13/19, 68.4%), fatigue (11/19, 57.9%), myalgia/myasthenia (9/19, 47.4%), sweating (6/19, 31.6%) and palpitation (6/19, 31.6%) et al. 47.4% (9/19) of the patients showed cardiac insufficiency/hemodynamic instability at first, and 52.6% (10/19) of the patients showed symptoms related to myositis first. cTnI and CK increased in all patients. The level of cTnI was significantly higher in grade 3̃4 than that in grade 1-2. NTproBNP was elevated in all grade 4 cases and one grade 3 case. The LVEF decreased only in 4 patients with grade 4 myocarditis (4/19, 22.2%). The electrocardiogram showed ventricular arrhythmia in 4 cases and complete atrioventricular block (3°AVB) in 3 cases. All patients were treated with corticosteroids (GCS). Grade 4 (6 cases), grade 3 (1 case) and grade 2 (2 cases) were given GCS pulse dose. The median total duration of GCS treatment was 42 days (range: 14-60 days). Three patients were treated with tocilizumab and one patient with infliximab. All patients with grade 3-4 and 2 patients with grade 2 were also treated with IVIG. Pacemakers were implanted in 3 patients with 3°AVB. The mortality rate of 19 cases was 21.1% (4/19), and the mortality rate of grade 4 myocarditis was 42.9%(3/7). The longer course before GCS therapy (p = 0.023), the higher NTproBNP (p = 0.030) and the lower LVEF value (p = 0.014) were found in the death group compared to the survival group. Logistic regression analysis showed that LVEF is an independent risk factor. The myocarditis prognostic model was condstructed as follows: Y = 8.794-18.002*LVEF, the prognostic critical value of LVEF was 57.5%, the area under the ROC curve of the prognostic model was 0.894 (95%CI: 0.735̃1.000), the sensitivity was 0.769, and the specificity was 1.000. Conclusions: We describe the largest single-center immune myocarditis cohort to date. Paying attention to the symptoms related to myositis is helpful to find more grade 1-2 myocarditis. cTnI is sensitive to myocarditis, and NTproBNP is of great significance to grade 3 ̃ 4 cases. LVEF < 57% may be an independent risk factor for poor prognosis.