Multi-organ Immune-Related Adverse Event Is a Risk Factor of Immune Checkpoint Inhibitor-Associated Myocarditis in Cancer Patients: A Multi-center Study

Background and Objective Immune checkpoint inhibitor (ICI)-associated myocarditis is a fatal immune-related adverse events (irAEs), which is prone to affecting multiple organ systems. Multi-organ irAEs have not been fully studied in ICI-associated myocarditis. Therefore, we aimed to explore the impact of multi-organ irAEs on ICI myocarditis in terms of clinical features, treatment, and prognosis. Methods This was a retrospective study. The clinical data of ICI myocarditis patients were collected from 6 hospitals in China. The risk factors and characteristics of pure myocarditis and multi-organ irAEs were analyzed. The overall survival (OS) after myocarditis was analyzed and univariate and multivariate regression analysis were performed. Results A total of 46 patients were analyzed in this study. Multi-organ irAEs were common (30/46, 65.2%) and prone to severe heart failure. The severe myocarditis was observed in 32 patients (69.6%). When myocarditis occurred, neutrophil to lymphocyte ratio, C-reactive protein, lactate dehydrogenase, interleukin (IL)-6, IL-10, creatine kinase, MB isoenzyme of creatine kinase, and brain natriuretic peptide increased from baseline, but absolute lymphocyte count decreased. Thymoma (B2/B3) was a risk factor for multi-organ irAEs. Heart failure and myocarditis were more severe in patients with multi-organ irAEs and require early corticosteroid therapy (<24 hours). Univariate analysis showed that age ≥ 60 years, myocarditis (grade 3-4), heart failure (grade 3-4), multi-organ irAEs, and severe myocarditis were associated with OS after myocarditis. After adjusting for other factors, heart failure (grade 3-4) was an independent risk factor for immune-related myocarditis (HR: 6.655, 95% CI: 1.539-28.770, p=0.011). Conclusion Patients with ICI-associated myocarditis had multi-organ irAEs with a high incidence of severe myocarditis, mortality, and poor prognosis. Thymoma was prone to those patients with multiple organs involvement. Patients could benefit from early corticosteroid intervention. Heart failure (grade 3-4) was an independent risk factor for OS after myocarditis.

ICI-associated myocarditis is a kind of irAEs with a very low incidence. A previous study pooled 31321 patients receiving ICIs, of which 122 (0.38%) developed ICI myocarditis, with 18 (0.09%) developed severe myocarditis (5). Mahmood et al. reported the incidence of ICI myocarditis was 1.14%, and myocarditis was more likely to occur during combined chemotherapy (34% versus 2%; P<0.001) (6). The combination use of ICI may even induce fulminant myocarditis (7). However, ICI myocarditis is a fatal AE with a lethality rate of 39.7%-50% (5,8). Because of the low incidence, limited studies have explored the risk factors, pathogenesis, treatment and prognosis of ICI myocarditis. A retrospective study involving eight centers showed that diabetes, obstructive sleep apnea, high body mass index (BMI), and receiving angiotensin receptor blockers are risk factors for ICI myocarditis. Another study showed that a history of heart failure and acute coronary syndrome are also high-risk factors for ICI myocarditis (6,9). In addition, a study found that patients with myocarditis have a decrease in ALC and an increase in neutrophil to lymphocyte ratio (NLR), and a decrease in absolute lymphocyte count (ALC) >35% or an increase in NLR >100% can predict subsequent major adverse cardiac events (MACE). All these suggest that the decrease of ALC and the increase of NLR contribute to the diagnosis and risk stratification of patients with myocarditis (10). In addition, Zlotoff et al. founded that the extended duration of QRS leads to an increased risk of MACE (11).
Single-system irAEs have been widely reported, but there are few reports on the clinical patterns and survival effects of irAEs involving multiple organ systems. Previous studies on non-small cell lung cancer (NSCLC) and other tumors have found that the occurrence of some irAEs such as thyroiditis and checkpoint inhibitor pneumonitis (CIP) is associated with improved survival outcomes (12,13). Ganessan et al. founded that multi-organ irAEs were associated with improved overall survival compared with no irAEs or single-organ irAE (3). Previous studies have shown that 47% of 55 ICI myocarditis patients have multi-organ irAEs, indicating that heart-related irAEs are more likely to be affected by multiple organ systems (10). However, multi-organ irAEs have not been fully studied in patients with myocarditis. In China, there are less researches on the risk factors, treatment, and prognostic factors of ICI myocarditis. In order to evaluate the role of multiple organs involvement in ICI myocarditis, we collected data from 6 centers in China to explore the impact of multi-organ irAEs on ICI myocarditis in terms of clinical features, treatment, and prognosis.  Table 1). Informed consent and review of the study were waived by the Institutional Review Board of each hospital. Adverse events occurred while patients were on ICIs or within 30 days after the last dose were attributed to ICIs and were adjudicated as irAEs. These events were organ specific, including cardiovascular, lung, skin, liver and gallbladder and pancreas, endocrine, gastrointestinal, rheumatism, nervous system, laboratory abnormalities and other miscellaneous organ specificities. Cardiologists further reviewed the course of the included patients to determine that the occurrence of myocarditis was attributable to ICIs. Patients with missing clinical data, not receiving ICIs, and other causes of myocarditis were excluded. Diagnosis of ICI myocarditis was performed in either of two ways: 1) histopathology, or 2) clinically suspected myocarditis (according to the European Society of Cardiology guidelines) (14). The diagnosis of other irAEs was based on pathological evidence or multidisciplinary adjudication, or the irAE-based treatment improved clinical outcomes excluding alternative diagnosis (15). Multi-organ irAEs were defined as irAEs that affected more than one organ or system, other than ICI myocarditis.

Data Collection
Data were retrospectively obtained from medical records, including demographic characteristics (age, gender, risk factors related to heart disease, heart-related risk factors), characteristics and treatment information of tumor (cancer type, clinical stage, tumor treatment information), characteristics and prognosis of myocarditis (symptoms, grade of heart failure, grade of myocarditis, survival information, cause of death). The grade of heart failure was based on New York Heart Association (NYHA) classification. The grade of ICI myocarditis was based on the American Society of Clinical Oncology Clinical Practice Guideline (15). Severe ICI myocarditis is defined as grade 3-4 myocarditis. Patients were categorized into low-dose (<500mg/ day) and high-dose (>500mg/day) groups based on initial methylprednisolone-equivalent administered on the first day of treatment. The timing of corticosteroid therapy is defined as the time from the onset of heart-related symptoms or asymptomatic but abnormal cardiac examinations to the initial corticosteroid therapy. Two independent researchers reviewed the data, and disagreement between them has been resolved by consensus.

Statistical Analysis
Continuous data were summarized as median (interquartile range, IQR) or mean ± standard deviation (SD), and categorical data were summarized as frequency (percentage). The difference of continuous variables was compared using t test or nonparametric test. The difference of categorical variables was compared using Chi-square (c2) or Fisher's exact test. Overall survival (OS) was defined as the onset of myocarditis to the date of death from any cause. At the time of analysis, patients who were alive or lost follow-up were considered censored. Kaplan-Meier was used for survival analysis, and the log-rank test was used to assess differences between groups. Cox's proportional regression analysis was conducted to identify significant variables that affect patients' survival after myocarditis and estimate hazard ratios (HR) and 95% confidence interval (CI) for predictors of survival. All tests were two-tailed, and p < 0.05 was considered statistically significant.

Time to Event Onset and Peripheral Blood Changes of ici Myocarditis
Time-event frequency distribution is shown in Figure 1. We depicted the time-event distribution curve of ICI myocarditis. The results showed that immune-related myocarditis mainly occurred within 1 month after receiving ICI and 78.2% within 3 months. We compared the changes in blood test of patients at baseline and onset of myocarditis and found that in addition to the significant decrease in ALC, NLR, C-reactive protein (CRP), lactate dehydrogenase (LDH), interleukin (IL)-6, IL-10, creatine kinase (CK), MB isoenzyme of creatine kinase (CK-MB), and brain natriuretic peptide (BNP) all increased significantly ( Table 2). ALC, NLR, LDH, IL-6, IL-10, CK, CK-MB, BNP changed more than 70% of the patients. Table 3 shows the differences between pure myocarditis and multi-organ irAEs. There was a significant statistical difference in the classification of cancers between the two groups in comparing baseline characteristics. Compared with pure myocarditis, thymoma accounts for more multi-organ irAEs. However, there were no significant differences in age, gender, heart-related risk factors, clinical stage, previous tumor treatment, and immunotherapy. In comparing the characteristics, treatment and outcome of myocarditis, there were significant statistical differences in symptoms, treatment time after myocarditis, arrhythmia, grade of heart failure, grade of myocarditis, initial dose of corticosteroids, and severe myocarditis between the two groups. Multi-organ irAEs were prone to myocarditis-related symptoms and arrhythmia. Moreover, they were more prone to heart failure (grade 3-4) and severe myocarditis, and half of the patients received highdose corticosteroid therapy. There was no statistically significant difference between the two groups in time to receive corticosteroid therapy, blood tests at baseline or onset of myocarditis (Supplementary Table 4).  Table 5). Analysis of prognostic factors was performed in all patients to determine the important variables that affect OS after myocarditis ( Table 4). Univariate analysis showed that patients with age ≥60 years, myocarditis (grade 3-4), heart failure (grade 3-4), multi-organ irAEs, and severe myocarditis have a worse prognosis (Figure 2, Supplementary Table 5).

Survival Analysis and Prognostic Factors in All Patients
We found that the survival analysis results were consistent in myocarditis grade (grade 3-4), heart failure grade (grade 3-4) and severe myocarditis. Incorporating the variables with p<0.1 in the univariate analysis into the multivariate analysis model, we found that myocarditis grade was linearly subordinate to the heart failure grade. Further analysis showed that after adjusting for age and other irAEs in all patients, heart failure (grade 3-4) was an independent risk factor for immune-related myocarditis (HR: 6.655, 95% CI: 1.539-28.770, p=0.011). In the stratification of heart failure grade, the mOS of patients with heart failure (grade 3-4) was 178 days, and1-month and 3-month OS rate was 60% and 40%, respectively. While mOS of patients with heart failure (grade 0-2) was not reached, and1-month and 3-month OS rate was 100% and 62.5%, respectively. In addition, we explored the population characteristics and prognosis of multi-organ irAEs with or without myositis. But there were no statistically significant differences in mortality, higher-grade arrhythmias, or greater degree of myocarditis between the 2 subgroups (Supplementary Tables 5, 6). The effect of corticosteroid therapy on prognosis was explored in all patients ( Figure 3). We found that the dose of corticosteroids had no effect on survival, but earlier corticosteroid therapy (<24 hours) resulted in better survival. Furthermore, we explored the efficacy of corticosteroids usage in different grades of heart failure and organ involvement mode. We found that therewas no significant difference in the onset time of corticosteroids in different grades of heart failure. In the stratification of organ involvement patterns, we found that earlier hormonal intervention has a better prognosis when multiple organs were involved. However, there was no difference in the comparison of corticosteroid usage in the two strata.

DISCUSSION
In the present study, we investigated the impact of multi-organ irAEs on ICI-associated myocarditis and some the risk factors. We found that more than half of ICI myocarditis patients were multi-organ irAEs, and the presence of multi-organ adverse events and severe myocarditis lead to reduced overall survival in patients with ICI induced myocarditis.
Up to date, it seems impossible to predict which patients could be affected by myocarditis after ICI treatment. To our knowledge, there are currently no effective approaches or guidelines to predict ICI-associated cardiotoxicity. Our study was partially designed as a nested case-control study in a cohort comprising ICI myocarditis patients with or without other irAEs. Since this study did not include non-myocarditis patients, it was not possible to predict myocarditis events in patients receiving ICI. Consistent with previous studies, male sex, heart-related risk factors and the combined therapy are more common in patients with myocarditis, and most of them have a history of tumor treatment (6,16). However, inconsistent with previous studies, most patients in our study were less than 60 years old, non-smokers, had no heart-related risk factors and lower BMI (6,9,17). Highly similar to previous studies, our study showed that most ICI myocarditis occurred within 1 month of receiving ICIs treatment, and the incidence within 3 months was 78.2% (5). Most patients with myocarditis have obvious symptoms, mainly related to heart failure. When combined with multi-organ irAEs, the incidence of symptoms is as high as 96.7%. A study showed that the incidence of MACE in patients receiving glucocorticoid therapy within 24 hours after admission for myocarditis was lower than that of patients receiving glucocorticoid therapy within 24-72 hours and >72 hours (18). Therefore, although  symptoms such as ICI myocarditis are not specific, timely medical treatment and corticosteroid therapy after symptoms appear can reduce the incidence of MACE. The effect of multi-organ irAEs on ICI myocarditis is rarely reported. Our research found that age, gender, smoking history, heart disease-related risk factors, tumor treatment history, immunotherapy regimen are not risk factors for multi-organ irAEs in patients with myocarditis. It is worth noting that all 6 thymoma patients (B2/B3) in this study had multi-organ irAEs, and 2 died of myocarditis. Previous studies have reported that patients with thymoma or thymic carcinoma are prone to multi-organ involvement such as myocarditis, myositis, and hepatitis when receiving ICIs (19)(20)(21). Myasthenia gravis (MG), myositis, and myocarditis overlap syndrome is a potentially fatal adverse event of ICIs (22). Johnson et al. found the same selective clonal T-cell populations infiltrating in the myocardium, skeletal muscle and tumors in 2 patients with fatal myocarditis and concomitant myositis (7). In addition, they found high levels of muscle-specific antigens in the tumors of these patients, suggesting that activated T lymphocytes after immunotherapy may attack cardiomyocytes and muscle cells through shared epitopes, leading to the development of autoimmune myocarditis and myositis. Thymoma patients receiving ICIs are prone to develop MG, myositis and myocarditis. Cho et al. reported that 3 cases of 7 patients with thymoma developed grade 3 myocarditis (42.9%), and these patients were of type B2 or B3 (19). In our study, all thymoma patients had grade 3-4 myocarditis. Moreover, 5 of 6 thymoma patients developed MG, and 4 of them (80%) had concurrent myositis. All these suggest that type B2 or B3 thymoma is more prone to severe irAEs. Furthermore, multi-organ irAEs have higher grade of heart failure and myositis, more malignant arrhythmias, and more high-dose corticosteroid therapy. The above evidence shows that multi-organ irAEs in ICI myocarditis are prone to become severe myocarditis. Further survival analysis showed that patients with multi-organ irAEs have a worse prognosis with an mOS of 178 days. So far, few studies have analyzed the risk of survival after ICI myocarditis. Previous studies have shown that ALC reduction>35%, NLR increase>100% and extended duration of QRS are risk factors for MACE (10,11). In our study, we found that multi-organ irAEs, age≥60 years old, myocarditis (G3-4), heart failure (G3-4), and severe myocarditis are risk factors on survival for ICI myocarditis. However, further multivariate analysis found that the grade of myocarditis have a linear relationship with the grade of heart failure. The possible reason is that in our study, the evaluation criteria of the grade of myocarditis were similar to the grade of heart failure. After adjusting for other factors, we found that the severity of heart failure is an independent risk factor that affects the survival of patients after myocarditis. The above results suggest that the classification of heart failure can be used as risk stratification for ICI myocarditis.
Whether multi-organ irAEs can be used as risk stratification factors for ICI myocarditis is worth discussing. As mentioned above, patients with other irAEs are more prone to severe heart failure. Heart failure after ICI myocarditis is caused by various mechanisms, including blockade of immune checkpoints of the myocardium, accumulation of cytokines, and microvasculitis of the heart (23,24). In ICI myocarditis, the increased circulating cytokine levels aggravate the damage of cardiomyocytes (23,25), which may further aggravate heart failure. Our study found that compared with baseline, cytokines increased significantly after the occurrence of irAE, which is consistent with previous studies (10,26). However, it has not been reported whether multi-organ irAEs have higher cytokines than single-organ irAEs. Our study found no difference in cytokine levels between pure myocarditis and multi-organ irAEs when myocarditis occurs. Considering the limited patients, a large sample size study is needed to explore the difference in cytokines between single-organ and multiorgan irAEs. In view of the above evidence, we cannot explain the mechanism of the effect of multi-organ involvement on severe heart failure. However, due to the strong correlation between multi-organ irAEs and more severe heart failure, we believe multi-organ irAEs can be used as risk stratification factors in ICI myocarditis. The administration of corticosteroids plays an important role in improving the prognosis of irAEs (15,27). For the timimg and dose of corticosteroids in ICI myocarditis, the current exploration is still limited. Zhang et al. have shown that high-dose and early corticosteroid intervention (<24 hours) can significantly reduce the occurrence of MACE (18). In our study, we found that earlier corticosteroid intervention but not high-dose corticosteroids improved the prognosis of patients. Further stratified exploration found that patients with multi-organ irAEs benefited from earlier corticosteroid intervention. It is further noted that the timing of corticosteroid therapy in the study is from the appearance of heartrelated symptoms or asymptomatic but abnormal heart examination to the initial corticosteroid therapy, while Zhang et al. is from the diagnosis of myocarditis to the initial corticosteroid therapy. Our research shows that receiving corticosteroid treatment within 24 hours after the appearance of heart-related symptoms or abnormal tests is the golden treatment period, especially for patients with severe myocarditis. In other words, this suggests that doctors should pay attention to patient education, such as emphasizing the early symptoms of myocarditis in the informed consent, so that the patient can contact the specialist or attending physician for diagnosis or elimination as soon as possible. For patients with multi-organ irAEs, who are prone to severe heart failure and worse prognosis, high-dose corticosteroids that reduce MACE should be the first choice. Multidisciplinary consultations should be conducted to develop an individualized treatment plan. Several limitations need to be noted regarding this study. First of all, this is a retrospective study with a small sample. The lack of blood tests makes it difficult to compare the differences between patients with pure myocarditis and multi-organ irAEs. Secondly, this study involved a variety of tumors, and the internal heterogeneity of the samples was large. Finally, a short followup time and too many censored events lead to poor stability of survival analysis results. Further research with a larger patient population are needed to confirm our findings.

CONCLUSIONS
Patients with ICI-associated myocarditis had multi-organ irAEs with a high incidence of severe myocarditis, mortality, and poor prognosis. Thymoma (B2/B3) was prone to those patients with multiple organs involvement. Patients cound benefit from early corticosteroid intervention (<24 hours). Heart failure (grade 3-4) was an independent risk factor for OS after myocarditis.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
Institutional review board/ethics committee approval was obtained from the Institutional Review Board of the First Affiliated Hospital of Guangzhou Medical University (Guangzhou, Guangdong, China). Individual consent for this retrospective analysis was waived.