Abstract
Background
In patients with COVID-19-related fulminant myocarditis (COVID-FM), multisystem inflammatory syndrome (MIS) influences clinical outcomes. Multisystemic inflammation (MSI) based on clinical symptoms correlates with the prognosis of non-COVID-FM diagnosed by endomyocardial biopsy. The effect of MSI on clinically diagnosed non-COVID-FM regarding age and prognosis remains unclear.
Methods and Results
This retrospective cohort study included 736 patients clinically diagnosed with non-COVID-FM. MSI was defined by clinical symptoms at admission aligned with MSI criteria. We compared all-cause death according to MSI status. Baseline balance was adjusted using inverse probability treatment weighting (IPTW). Overall, 275 (37.4%) non-COVID-FM patients met the MSI status. The MSI (+) group was younger, with fewer male patients and complications such as hypertension, diabetes, chronic kidney disease, and dyslipidaemia. When stratified per age, MSI was highest in patients aged <50 years (48.7%). Overall, the probability of all-cause death was lower in the MSI (+) group (30-day: 25.1% vs. 32.1%, P=0.053, 3-year: 36.1% vs. 42.5%, P=0.004). After IPTW, the MSI (+) group, especially in patients <50 years, was associated with significantly lower risk difference and risk ratio for 30-day all-cause death compared to the MSI (−) group, but not 3-year all-cause death.
Conclusions
In patients with non-COVID-FM, MSI often developed at a younger age, and was associated with lower 30-day mortality rate, particularly in patients <50 years.
Key Words: Clinical diagnosis, Fulminant myocarditis, Multisystemic inflammation, Multisystem inflammatory syndrome
Central Figure.
The clinical course of acute myocarditis can vary widely, ranging from mild chest pain or dyspnea without affecting hemodynamics to life-threatening arrhythmias and heart failure with severe cardiogenic shock.1,2
Fulminant myocarditis (FM) is severe myocarditis that is rapidly progressive and complicated with cardiogenic shock, requiring inotropes or mechanical circulatory support (MCS).3 The aetiology of FM can be secondary to infections (most viral infections), systemic autoimmune disorders, medications, and vaccines. Independent of the initial trigger, the subsequent inflammatory response is responsible for myocardial injury, resulting in myocyte necrosis, interstitial oedema, and inflammatory infiltrates.4–6 Coronavirus disease 2019 (COVID-19) infection also induces myocardial injury.7 Children and adults who have been infected with COVID-19 or given a COVID-19 vaccine can develop a life-threatening multisystemic inflammation state called multisystem inflammatory syndrome-children/adult (MIS-C/A).8 Barhoum et al. reported on the differences in clinical presentation, prognosis, and immune response in COVID-19 related FM (COVID-FM) with and without MIS-A and found that in-hospital deaths were fewer in the patients with MIS-A than in the patients without MIS-A.9 Differences in the inflammation phenotype due to the immune response may play an important role.
We previously showed that multisystemic inflammation (MSI), based on the clinical symptoms of the MIS criteria, was associated with prognosis even in non-COVID-FM diagnosed as lymphocytic by pathological examination.10 In the MSI-positive groups, the association was more remarkable in younger aged patients. Pathological diagnosis by myocardial biopsy is difficult in patients with FM because not all hospitals can perform it safely, and the patient’s vital signs are unstable when performing it. There was no significant difference in the severity of myocarditis between patients with and without MSI. Therefore, it was considered that MSI might indicate the inflammatory phenotype due to the immune response rather than the severity of myocarditis.
In Europe, about 20–50% of FM patients undergo a biopsy, whereas a registry of FM patients in the USA demonstrated a biopsy rate of <5%.4 Therefore, some guidelines define the clinical diagnosis or possible status of FM without pathological examination.11,12 Therefore, in this study, we paid attention to the clinical diagnosis of FM.
In addition, because the clinical course of FM progresses rapidly, it would be useful if the prognosis could be predicted by MSI based on clinical symptoms at the time of hospital visit. However, the prognostic relationship between patients with clinically diagnosed non-COVID-FM and MSI and the effect of age on prognosis is unclear. In this study, we aimed to evaluate the association between MSI and clinical prognosis in patients with clinically diagnosed FM, and its variation according to age.
Methods
Data Source and Patient Population
This was a multicentre retrospective cohort study based on the Japanese Registry of FM, conducted in collaboration with 235 facilities in Japan.
To begin, patients with suspected non-COVID-19 infection or vaccination-related myocarditis (≥16 years) were selected using ICD-10 (International Classification of Diseases 10th Revision) codes I40, I41, or I423 in the JROAD-DPC (Japanese Registry of All Cardiac and Vascular Diseases-Diagnosis Procedure Combination) discharge database between April 2012 and March 2017. We categorized FM as presenting with fulminant symptoms that required treatment with intravenous inotropes, MCS, or intravenous inotropes.4 Furthermore, we selected patients with clinically diagnosed FM using the clinical diagnostic criteria of the European Society of Cardiology (ESC)13 or the Japanese Circulation Society (JCS),12 and the definitions of World Health Organization/International Society and Federation of Cardiology (WHO/ISFC) criteria.14 The exclusion criteria were applied to previous studies, and patients who were diagnosed with other diseases, those in whom symptoms had appeared >30 days before admission, or those who had started using catecholamines or mechanical circulatory support after the 14th day of hospitalization were excluded.15
While the patients were hospitalized, individual data and treatment details (demographic data, clinical symptoms, medical history, laboratory, ECG and echocardiographic findings on admission, histological findings of the biopsied myocardium, medications during hospitalization, temporary MCS, and events) were collected depending on the facility. Data were collected and stored using the Research Electronic Data Capture (REDCap; Vanderbilt University). This study followed the Declaration of Helsinki, and the protocol was authorized by the local ethics committees at Nagasaki University and the Japanese Circulation Society (registration numbers 20081719-4 and 10, respectively).
Definition of MSI
MSI was defined using the MIS criteria reported by Vogel et al.8 The Level 1 diagnosis of MIS requires fever for ≥3 consecutive days, ≥2 measures of clinical features (mucocutaneous, gastrointestinal, shock/hypotension, neurological features), a laboratory marker of inflammation (elevated C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), ferritin, or procalcitonin), ≥2 measures of disease activity (elevated B-type natriuretic peptide (BNP) or NT-proBNP, troponin, neutrophilia, lymphopenia, thrombocytopenia, echocardiographic evidence of cardiac involvement or physical stigmata of heart failure, ECG changes consist with myocarditis), and an association with SARS-CoV-2.8 Because laboratory markers of inflammation and ≥2 measures of disease activity are consistent with the FM diagnostic criteria of the ESC or the JCS, and the WHO/ISFC,11,12,14 we focused on clinical symptoms to define MSI in this study. MSI was defined as fever and the recognition of ≥1 of the following symptoms using at the MIS criteria: (1) gastrointestinal symptoms (abdominal pain, vomiting, diarrhoea), (2) shock/hypotension, (3) neurological features (altered mental status, headache, weakness, paraesthesia, and lethargy), and (4) respiratory symptoms (sputum, cough, and rhinorrhoea).3
Outcomes
The primary outcome was 30-day all-cause death. The secondary outcome was 3-year all-cause death.
Other Variables
We collected the following: demographic data; prodromal symptoms; medical history; clinical, laboratory, ECG, and echocardiographic findings on admission; myocardial biopsy status; medications during hospitalization; use of temporary MCS devices; and in-hospital events (shock, ventricular tachycardia [VT], and ventricular fibrillation [VF]). To examine the association between MSI and age, we divided the patients into 3 age groups following the existing Framingham assessment tools: <50 years, 50–64 years, and >65 years.16 We investigated the association of MSI and outcomes in each age group.
Statistical Analysis
Baseline characteristics are summarized using the median and interquartile range (IQR) for continuous variables and counts (%) for categorical variables. The characteristics of the 2 groups were assessed using the Wilcoxon rank-sum test or Fisher’s exact test, as appropriate. In this study, there were only a very small number of cases of heart transplantation, and we did not consider the competing risk. Cumulative incidence probabilities were estimated as 1 minus the Kaplan–Meier survival estimates. Differences between the MSI (+) and MSI (−) groups were assessed using the log-rank test across different age strata.
Some variables were missing (Supplementary Table 1), so we performed multiple imputation (MI) with chained equations, estimated the adjusted risk difference (RD) and risk ratio (RR) using inverse-probability-of-treatment weighting (IPTW).17 We used the imputation model for MI to construct the same covariates as those used in the IPTW model. The effect measures and 95% confidence intervals (CIs) were derived using Rubin’s rule. In the 100 imputed datasets, the standardized mean difference (SMD) was used to check the baseline characteristics between groups. The balance between groups was assessed after weighting, and an absolute standardized difference of <0.1 was considered as evidence of balance. We included baseline demographic, clinical, laboratory, ECG, and treatment variables in the propensity score model for IPTW, because they were considered to be potential confounders. The variables included age, sex, WHO diagnostic category, New York Heart Association (NYHA) class on admission, laboratory markers (e.g., white blood cell count, albumin, CRP level, troponin, kidney, and liver function), cardiac markers (e.g., BNP and creatine kinase [CK]), ECG findings, echocardiographic parameters, preexisting comorbidities, and initial treatments such as IVIG, intravenous immunoglobulin (IVIG), steroids, and MCS. Variables strongly associated with exposure (e.g., fever and respiratory symptoms) were excluded to avoid model overfitting and extreme weights.
All analyses were performed using R version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria). This study was exploratory and no adjustment for multiplicity was performed. Therefore, caution should be exercised when interpreting P values.
Results
Baseline Characteristics
Among the 2,511 patients with suspected myocarditis without COVID-19 infection or vaccination in the JROAD-DPC database, 1,327 (52.8%) had a fulminant presentation. A total of 819 patients were hospitalized in 235 facilities with permission to participate in this study, 736 of whom met the clinical diagnostic criteria for FM from ≥1 societies in the ESC, JCS, or WHO/ISFC. MSI was present in 275 (37.3%) patients (Figure 1). When separated according to age, there was a significant difference in the incidence of MSI in the age groups 16–49 years (48.6%; 145/298), 50–64 years (38.9%; 81/208), and ≥65 years (21.3%; 49/230) (Figure 2). The patients were divided into the MSI (+) and MSI (−) groups.
Figure 1.
Study flowchart. COVID-19, coronavirus disease 2019; ESC, European Society of Cardiology; ISFC, International Society and Federation of Cardiology; ICD-10, International Classification of Disease 10th revision; JCS, Japanese Circulation Society; JROAD-DPC, Japanese Registry of All Cardiac and Vascular Diseases Diagnostic Procedure Combination; MSI, multisystemic inflammation; WHO, World Health Organization.
Figure 2.
Histograms show the incidence of multisystemic inflammation (MSI) in the age groups.
The baseline characteristics of the study participants are shown in Table 1. The median (IQR) age was 56 (39–67) years, and 60.2% (443/736) of the patients were male. The MSI (+) group tended to be younger (median [IQR] age: 48 [34–62] vs. 60 [44–70.0] years, P<0.001) and a smaller proportion of men (54.5% vs. 63.6%, P=0.019) compared to the MSI (−) group. Clinical symptoms, such as fever, headache, respiratory symptoms, and gastrointestinal symptoms, were more common in the MSI (+) group. Comorbidities such as hypertension (20.0% vs. 28.0%, P=0.017), diabetes (6.2% vs. 13.4%, P=0.002), chronic renal failure (CRF) (2.9% vs. 5.0%, P<0.001), and dyslipidaemia (7.3% vs. 16.3%, P<0.001) were less common in the MSI (+) group. The vital signs on admission were higher temperature (median (IQR) temperature: 36.8℃ [36.6–37.9℃] vs. 36.6℃ [36.1–37.3℃], P<0.001) and lower systolic blood pressure (SBP) (median [IQR] SBP: 93 [80–107] mmHg vs. 100 [86–118] mmHg, P<0.001) in the MSI (+) group. There were no differences in the rates of NYHA class III or IV and cardiopulmonary arrest (CPA) at the time of admission. The laboratory data on admission showed that white blood cell (WBC) (median [IQR] WBC: 10,200 (7,225–3,350) /μL vs. 9,200 (6,982.5–12,925) /μL, P=0.019) and CRP level (median [IQR] CRP level: 5.2 (2.0–11.4) mg/dL vs. 3.8 (1.2–10.3) mg/dL, P=0.010) were higher in the MSI (+) group, but there were no differences in myocardial biomarkers such as CK (median (IQR) CK: 750.5 (325.5–1,399.8) IU/L vs. 653.0 (357.8–1400.5)IU/L, P=0.538) and the rate of elevated troponin (troponin >0.02 ng/mL or qualitative test positive) (78.5% vs. 77.7%; P=0.796) between groups. No significant differences were observed in ECG findings or left ventricular ejection fraction (LVEF) on transthoracic echocardiography (TTE); however, pericardial effusion on TTE was observed more frequently in the MSI (+) group (46.9% vs. 39.0%, P=0.043). Endomyocardial biopsy (EMB) was performed in 43.9% of all patients with FM, with no difference between groups.
Table 1.
Baseline Patient Characteristics
| Characteristics | All N=736 |
MSI (+) N=275 |
MSI (−) N=461 |
P value |
|---|---|---|---|---|
| Demographics | ||||
| Age, years | 56 (39–67) | 48 (34–62) | 60 (44–70) | <0.001 |
| Male, n (%) | 443 (60.2) | 150 (54.5) | 293 (63.6) | 0.019 |
| Clinical symptoms | ||||
| Fever, n (%) | 452 (61.4) | 275 (100.0) | 177 (38.4) | <0.001 |
| Neurological features, n (%) | 50 (6.8) | 43 (15.6) | 7 (1.5) | <0.001 |
| Respiratory symptoms, n (%) | 211 (28.7) | 124 (45.1) | 87 (18.9) | <0.001 |
| Gastrointestinal symptoms, n (%) | 203 (27.6) | 116 (42.2) | 87 (18.9) | <0.001 |
| Medical history | ||||
| Hypertension, n (%) | 184 (25.0) | 55 (20.0) | 129 (28.0) | 0.017 |
| Diabetes mellitus, n (%) | 79 (10.7) | 17 (6.2) | 62 (13.4) | 0.002 |
| Chronic kidney disease, n (%) | 31 (4.2) | 8 (2.9) | 23 (5.0) | <0.001 |
| Dyslipidemia, n (%) | 95 (12.9) | 20 (7.3) | 75 (16.3) | <0.001 |
| Clinical findings on admission | ||||
| BMI, kg/m2 | 22.2 (20.0–24.7) | 21.9 (19.8–24.8) | 22.4 (20.0–24.5) | 0.881 |
| Body temperature, ℃ | 36.7 (36.2–37.5) | 36.8 (36.3–37.9) | 36.6 (36.1–37.3) | <0.001 |
| SBP, mmHg | 97 (84–115) | 93 (80–107) | 100 (86–118) | <0.001 |
| NYHA class III or IV, n (%) | 577 (78.4) | 211 (76.7) | 366 (79.4) | 0.322 |
| CPA on admission, n (%) | 19 (2.6) | 10 (3.7) | 9 (2.0) | 0.229 |
| Laboratory data on admission | ||||
| WBC, /μL | 9,690.0 (7,050.0–13,135.0) | 10,200.0 (7,225.0–13,350.0) | 9,200.0 (6,982.5–12,925.0) | 0.019 |
| Hb, g/dL | 13.4 (11.8–14.7) | 13.3 (12.0–14.8) | 13.4 (11.7–14.7) | 0.942 |
| Cre, mg/dL | 1.1 (0.8–1.6) | 1.0 (0.8–1.5) | 1.1 (0.8–1.7) | 0.149 |
| eGFR, mL/min/1.73 m2 | 52.0 (32.0–74.0) | 56.0 (34.5–78.0) | 50.0 (31.0–71.0) | 0.040 |
| CRP, mg/dL | 4.4 (1.4–10.8) | 5.2 (2.0–11.4) | 3.8 (1.2–10.3) | 0.010 |
| Elevated CRP, n (%)* | 686 (93.2) | 268 (97.5) | 418 (90.7) | 0.001 |
| BNP, pg/mL | 579.1 (295.5–121.8) | 553.1 (297.7–931.4) | 608.3 (290.5–1,195.7) | 0.152 |
| Elevated troponin,# n (%) | 574 (78.0) | 216 (78.5) | 385 (77.7) | 0.796 |
| CK, IU/L | 690.5 (347.8–1,400.5) | 750.5 (325.5–1,399.8) | 653.0 (357.8–1,400.5) | 0.538 |
| CK MB, IU/L | 55.0 (25.0–102.0) | 58.0 (23.0–101.0) | 52.0 (26.0–102.0) | 0.875 |
| ECG findings on admission | ||||
| Sinus rhythm, n (%) | 514 (69.8) | 192 (69.8) | 322 (69.8) | 1.000 |
| QRS duration, ms | 120.0 (92.0–140.0) | 114.5 (89.2–137.5) | 120 (84.0–143.2) | 0.107 |
| ST elevation, n (%) | 460 (62.5) | 170 (62.8) | 290 (62.9) | 0.633 |
| CAVB, n (%) | 123 (16.7) | 48 (17.5) | 75 (16.3) | 0.760 |
| Echocardiography | ||||
| LVEF, % | 30.0 (20.0–44.0) | 30.0 (20.0–41.0) | 30.0 (20.0–45.0) | 0.221 |
| LVEF <40%, n (%) | 491 (66.7) | 191 (69.5) | 300 (65.1) | 0.165 |
| Pericardial effusion, n (%) | 309 (42.0) | 129 (46.9) | 180 (39.0) | 0.043 |
| Biopsy | ||||
| Yes, n (%) | 323 (43.9) | 123 (44.7) | 200 (43.4) | 0.759 |
Values are presented as n (%) or median (interquartile range). *Elevated CRP defined as >0.3 mg/dL. #Elevated troponin was defined as >0.02 ng/mL or qualitative test positive. BMI, body mass index; BNP, B-type natriuretic peptide; CAVB, complete atrioventricular block; CK, creatine kinase; CPA, cardiopulmonary arrest; Cr, creatinine; CRP, C-reactive protein; ECG, electrocardiogram; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; LVEF, left ventricular ejection fraction; MSI, multisystemic inflammation; NYHA, New York Heart Association; SBP, systolic blood pressure; VF, ventricular fibrillation; VT, ventricular tachycardia; WBC, white blood cell count.
In-Hospital Events and Treatment
Table 2 shows in-hospital events and treatments. There were no differences in temporary MCS use or in-hospital events between groups. No significant difference was found in the number of in-hospital events. Intravenous steroids were administered more frequently in the MSI (+) group during hospitalization; however, the difference was not statistically significant (19.3% vs. 13.7%, P=0.055). In-hospital all-cause death occurred in 33.4% (246/736) of patients. In-hospital all-cause mortality rate was significantly lower in the MSI (+) group than in the MSI (−) group (27.3% vs. 37.1%, P=0.008). The in-hospital all-cause mortality rates according to age group are shown in Supplementary Table 2.
Table 2.
In-Hospital Events and Treatments
| Characteristics | All N=736 |
MSI (+) N=275 |
MSI (−) N=461 |
P value |
|---|---|---|---|---|
| Temporary MCS devices | ||||
| Intra-aortic balloon pump, n (%) | 192 (26.1) | 70 (25.5) | 122 (26.5) | 0.795 |
| Veno-arterial ECMO, n (%) | 210 (28.5) | 85 (30.9) | 125 (27.1) | 0.274 |
| Ventricular assist device, n (%) | 14 (1.9) | 6 (2.2) | 8 (1.7) | 0.782 |
| In-hospital events | ||||
| Shock, n (%) | 607 (82.5) | 234 (85.1) | 373 (80.9) | 0.161 |
| VT, n (%) | 136 (18.5) | 54 (19.6) | 82 (17.8) | 0.556 |
| VF, n (%) | 72 (9.8) | 24 (8.7) | 48 (10.4) | 0.522 |
| Advanced AV block, n (%) | 187 (25.4) | 73 (26.5) | 114 (24.7) | 0.600 |
| Medication during hospitalization | ||||
| β-blocker, n (%) | 327 (44.4) | 124 (45.1) | 203 (44.0) | 0.818 |
| ACEi/ARB, n (%) | 344 (46.7) | 124 (45.1) | 220 (47.7) | 0.493 |
| MRA, n (%) | 207 (28.1) | 82 (29.8) | 125 (27.1) | 0.446 |
| IV steroids, n (%) | 116 (15.8) | 53 (19.3) | 63 (13.7) | 0.055 |
| IVIG, n (%) | 102 (13.9) | 43 (15.6) | 59 (12.8) | 0.321 |
| Inotropes, n (%) | 708 (96.2) | 267 (97.1) | 441 (95.7) | 0.427 |
| In-hospital all-cause death, n (%) | ||||
| All ages | 246/736 (33.4) | 75 (27.3) | 171 (37.1) | 0.008 |
Values are presented as n (%). ACEi/ARB, angiotensin-converting enzyme inhibitor/angiotensin-receptor blocker; AV, atrioventricular; ECMO, extracorporeal membrane; IV, intravenous; IVIG, intravenous immunoglobulin; MCS, mechanical circulatory support; MRA, mineralocorticoid receptor antagonist. Other abbreviations as in Table 1.
Primary and Secondary Outcomes
The median duration of observation was 229.5 (IQR: 15.7–1,457.7) days. The results are summarized in Table 3. The overall 30-day all-cause mortality rate was 29.4%. The 30-day and 3-year all-cause mortality rates were lower in the MSI (+) group (30-day, 25.1% vs. 32.1%, P=0.053, 3-year: 36.1% vs. 42.5%, P=0.004). Figure 3A–D shows the cumulative incidence probabilities (CIPs) in all ages and subgroups: <50 years, 50–64 years, and ≥65 years. At all ages, the proportion of the 3-year all-cause mortality rate was lower in the MSI (+) group (Figure 3A) (3-year CIPs: 31.3% vs. 43.1%, P=0.004). In terms of subgroups according to age, the CIPs were similarly lower in the MSI (+) group for those <50 years of age (Figure 3B) (3-year CIPs: 19.4% vs. 29.4%, P=0.02), but not in the 50–64 (Figure 3C) (3-year CIPs: 47.7% vs. 41.3%, P=0.2) or ≥65 age groups (Figure 3D) (3-year CIPs: 38.8% vs. 55.9%, P=0.2).
Table 3.
Incidence of Outcomes (30-Day and 3-Year All-Cause Death) and Adjusted RD and RR According to the Presence of MSI and Age Group
| All n (%) |
MSI (+) n (%) |
MSI (−) n (%) |
Adjusted RD1
(95% CI) |
Adjusted RR1
(95% CI) |
|
|---|---|---|---|---|---|
| All ages | |||||
| N=736 | N=275 | N=461 | |||
| 30-day, n (%) | 217 (29.4%) | 69 (25.1%) | 148 (32.1%) | −2.4 (−10.7 to 5.9) | 0.92 (0.69 to 1.23) |
| 3-year, n (%) | 283 (38.4%) | 87 (36.1%) | 196 (42.5%) | −7.2 (−15.8 to 1.4) | 0.82 (0.65 to 1.10) |
| Age <50 years | |||||
| N=298 | N=145 | N=153 | |||
| 30-day, n (%) | 54 (18.1%) | 18 (12.4%) | 36 (23.5%) | −10.5 (−20.2 to −0.9) | 0.51 (0.27 to 0.95) |
| 3-year, n (%) | 75 (25.1%) | 28 (19.3%) | 47 (30.7%) | −9.3 (−21.7 to 3.2) | 0.67 (0.38 to 1.17) |
| Age 50–64 years | |||||
| N=208 | N=81 | N=127 | |||
| 30-day, n (%) | 64 (30.7%) | 32 (39.5%) | 32 (25.2%) | 5.6 (−12.5 to 23.6) | 1.20 (0.65 to 2.20) |
| 3-year, n (%) | 87 (41.8%) | 38 (46.8%) | 49 (38.6%) | −0.6 (−20.6 to 19.4) | 0.98 (0.58 to 1.66) |
| Age ≥65 years | |||||
| N=230 | N=49 | N=181 | |||
| 30-day, n (%) | 99 (43.0%) | 19 (38.8%) | 80 (44.2%) | −18.1 (−66.9 to 30.8) | 0.55 (0.04 to 7.58) |
| 3-year, n (%) | 121 (52.6%) | 21 (42.9%) | 100 (55.2%) | −28.6 (−79.0 to 21.9) | 0.46 (0.02 to 6.17) |
1Adjusted RD and RR were estimated using inverse probability weighting and multiple imputation method. MSI, multisystemic inflammation; RD, risk difference; RR, risk ratio.
Figure 3.
Analysis of the secondary outcomes and cumulative incidence probability (CIP) curves for 3-year all-cause death. CIP curves for 3-year all-cause death and CIPs for each year and age group were compared with and without multisystemic inflammation (MSI). The blue line indicates MSI (+) and the orange line indicates MSI (−). (A) CIP curves for 3-year all-cause death and CIPs for all ages. (B) CIP curves for 3-year all-cause death and CIPs for patients aged <50 years. (C) CIP curves for 3-year all-cause death and CIPs for patients aged 50–64 years. (D) CIP curves for 3-year all-cause death and CIPs for ages >65 years. CI, confidence interval.
IPTW Analysis
The SMDs for patient characteristics, MCS device, in-hospital events, and medications are shown in Figure 4. After MI and IPTW analyses, there were no significant differences in distribution (Figure 4). For all ages, the adjusted RD and RR for 30-day, and 3-year all-cause death tended to be lower in the MSI (+) group (Table 3). The adjusted RD (95% CI) was −2.4 (−10.7 to 5.9) for 30-day all-cause death and −7.2 (−15.8 to 1.4) for 3-year all-cause death. The corresponding adjusted RR (95% CI) values were 0.92 (0.69–1.23) and 0.82 (0.65–1.10), respectively. The adjusted RD and RR for the 30-day all-cause death was remarkably lower in the MSI (+) group at age <50 years (30-day: adjusted RD [95% CI] −10.5 [−20.2 to −0.9], adjusted RR [95% CI] 0.51 [0.27 to 0.95]). In the group aged >65 years, a similar trend to that in those aged <50 years was observed, but the estimation precision for RR and RD was poor. In the middle age group (50–64 years), the same trend was not observed despite having a comparable sample size. The unadjusted RD and RR values are presented in Table 4. Adjusted RD and RR for in-hospital all-cause death are shown in Supplementary Table 2 and similar results were obtained for the adjusted RD and RR for 30-day deaths. The results after adjusting for patient background factors, including fever and respiratory symptoms that were excluded in the IPWT analysis, were as follows: the 30-day all-cause death adjusted RR (95% CI) was 0.72 (0.44 to 1.17), and the 3-year all cause death adjusted RR was 0.66 (0.43 to 1.02) for all ages.
Figure 4.
Standard mean difference in patient characteristics, MCS device, in-hospital events, and medications before and after inverse probability of treatment weighting (IPTW). The blue dots show the unadjusted standard mean difference (SMD), and the red dots show the adjusted SMD with IPTW.
Table 4.
Unadjusted Clinical Outcomes
| Unadjusted RD (95% CI) |
Unadjusted RR (95% CI) |
|
|---|---|---|
| All ages | ||
| In-hospital death, n (%) | −3.67 (−9.24 to 1.90) | 0.74 (0.59 to0.92) |
| 30-day death, n (%) | −2.62 (−8.01 to 2.77) | 0.78 (0.51 to 1.00) |
| 3-year death, n (%) | −4.07 (−9.79 to 1.65) | 0.74 (0.61 to 0.91) |
| Age under 50 years | ||
| In-hospital death, n (%) | −6.31 (−14.77 to 2.15) | 0.53 (0.33 to 0.85) |
| 30-day death, n (%) | −5.41 (−13.43 to 2.61) | 0.53 (0.31 to 0.89) |
| 3-year death, n (%) | −5.55 (−14.37 to 3.26) | 0.63 (0.42 to 0.95) |
| Age 50–64 years | ||
| In-hospital death, n (%) | 5.17 (−5.10 to 15.45) | 1.44 (1.00 to2.08) |
| 30-day death, n (%) | 5.57 (−4.24 to 15.39) | 1.57 (1.05 to2.35) |
| 3-year death, n (%) | 3.24 (−7.55 to 14.04) | 1.22 (0.88 to1.67) |
| Age over 65 years | ||
| In-hospital death, n (%) | −5.64 (−14.0 to 2.72) | 0.58 (0.38 to 0.91) |
| 30-day death, n (%) | −4.74 (−12.7 to 3.21) | 0.59 (0.36 to 0.96) |
| 3-year death, n (%) | −4.92 (−13.6 to 3.78) | 0.67 (0.45 to 0.99) |
Values are presented as RD or RR (95% CI). CI, confidence interval. Other abbreviations as in Table 3.
Discussion
We demonstrated in this study that: (1) 37.3% of patients with clinically diagnosed non-COVID-FM met the definition of MSI and it was more common among young adults; (2) MSI (+) patients with non-COVID-FM were younger, were less often male, had fewer comorbidities, but increased CRP level, and a higher rate of pericardial effusion; (3) patients with non-COVID-FM and MSI tended to have a lower mortality rate than those without MSI, and this trend was particularly pronounced in those with MSI aged <50 years; (4) even after IPTW, both the adjusted RD and RR for 30-day all-cause death was substantially lower in the MSI (+) group at <50 years of age.
EMB is the gold standard in the diagnosis of FM and is important in determining the treatment strategy.18 Early EMB within 2 weeks after intensive care unit admission has been related to improvement in survival free of heart transplantation or left ventricular assist device.19 However, it is often difficult to do a pathological diagnosis using EMB in patients with FM due to rapid disease progression, such as hypotension and organ damage. Therefore, the rate of EMB in patients with FM varies from 34.4% to 44%.4,15 In this study, EMB was performed in 43.9% of clinically diagnosed FM patients, which was comparable to previous reports. Limiting the number of patients to those with FM who had undergone EMB would result in a decrease in the number of patients; therefore, the present study also included patients with clinically diagnosed FM.
Among patients with COVID-FM, 66% met the MIS-A criteria.9 MIS-A, a hyperinflammatory syndrome associated with COVID-19 infection, was more commonly seen in young men between the ages of 19 and 34 years.20 Barhoum et al. reported that the MIS-A positive group had more frequent fever, skin rash, enanthema, pharyngitis, and conjunctivitis, as well as higher CRP, interferon-α2, and interleukin-8 levels.9 RNA polymerase III autoantibodies were present in 54% of the MIS-A-negative group but in none of the MIS-A-positive group.9 The presence of these cytokines might reflect altered immune defences toward viral infections or, alternatively, exaggerated antiviral responses leading to organ damage. One explanation for the high prevalence of MIS-A in young adults is that complications and medications associated with aging can cause immunosuppression, making the development and diagnosis of MIS-A more difficult.21 Our previous study demonstrated an MSI-positive rate of 34% in patients diagnosed with lymphocytic and non-COVID-FM myocarditis using EMB,5 and that proportion was similar to the 37.3% MSI-positive rate observed in clinically diagnosed non-COVID FM. In addition, MSI (+) was also most common among patients with non-COVID-FM in the youngest (<50 years) age group, and the MSI (+) group had higher temperature, CRP level, and more clinical symptoms and significantly fewer complications, such as hypertension, diabetes, or CRF, than the MSI (−) group in our current study. Similar to our previous study, MSI (+) was not associated with the severity of myocarditis, such as elevated cardiac enzymes or reduced EF. Given its demographic and inflammatory characteristics, MSI may represent a hyperinflammatory phenotype analogous to MIS-A, even without COVID-19 infection.
Even with recent improvements in MSC devices, FM has a significantly higher mortality rate than non-FM. Previous studies report in-hospital death or heart transplantation rates of 25.5–40% and a 60-day mortality rate of 27.8%.22,23 In patients with COVID-FM, in-hospital mortality rates have been reported at 13–27.7%.9,24 When stratified according to the presence of MIS-A, the MIS-A negative group has a higher in-hospital mortality rate (31% vs. 4%).9 In our study of non-COVID-FM, the in-hospital, 30-day, and 3-year all-cause mortality rates tended to be lower in the MSI (+) group (in-hospital mortality: 27.3% vs. 37.1%, P=0.008; 30-day mortality: 25.1% vs. 32.1%, P=0.053; 3-year mortality: 31.6% vs. 42.5%, P=0.004). Subgroup analysis according to age showed the MSI (+) group tended to have lower 30-day all-cause mortality rates, especially in patients aged <50 years. The favourable prognostic impact of MSI remained evident even after adjusting for baseline patient characteristics using IPTW, and the present study results were consistent with our previous findings.5 This study found no difference in long-term prognosis at 3 years after MIS presentation. The dissociation between short-term and long-term outcomes may indicate that MSI reflects an acute inflammatory response phenotype that influences early survival, but does not necessarily determine chronic myocardial remodeling or late death. However, data on the impact of MIS-A/C on the long-term prognosis of COVID-FM patients is scarce and further studies are required to clarify whether hyperinflammatory phenotypes influence chronic outcomes.
The diagnosis of myocarditis subtype (e.g., eosinophilic myocarditis, giant cell myocarditis) and severity using EMB is important for both predicting prognosis and determining the treatment strategy;25 however, it is difficult to increase EMB rates immediately because of the hurdles in performing EMB at each institution. Therefore, a simple clinical tool that can easily predict FM prognosis would be valuable in routine practice. The Sequential Organ Failure Assessment (SOFA) score has been reported to reflect the prognosis of FM.26 SOFA scores at intensive care unit admission have been found to be significantly higher in non-survivors.26 Our study showed that in patients with non-COVID-FM, MIS diagnosis focusing on clinical symptoms at admission to the hospital was associated with 30-day deaths primarily among patients younger than 50 years of age. MSI may assist in early risk stratification, particularly in younger patients with non-COVID FM, but should not be interpreted as a causal determinant of prognosis.
Study Limitations
There are several to note. First, some values were missing because this was a retrospective cohort study. The lactate levels at the time of hospital admission and the timing of treatment for FM were not collected. The proportion of missing values was modest across all variables in this study (vital data such as blood pressure data generally had <10% missing values, with the highest being 20% for BNP). The missing values were listed in Supplementary Table 1. MI was performed for missing values, and attempts were made to avoid loss of study sample size.
Second, unadjusted confounding factors may have been included in the results. Using E-values, we quantified how much an unmeasured confounder would need to affect the outcome to nullify the observed association.19 The RR for in-hospital death in patients aged <50 years was 0.49 (95% CI: 0.28 to 0.88), corresponding to an E-value of 3.50. The E-value for the lower bound of the 95% CI (0.88) was 1.54. In other words, for the result to become null, an unmeasured confounder with an RR of ≥1.54 would need to be associated. It is unlikely that such a strong unmeasured confounder exists beyond the variables already adjusted for in our analysis, as the most clinically relevant prognostic factors (comorbidities, severity at admission, and biomarkers) were included.
Third, the diagnosis of MIS requires elevated levels of ferritin, interleukin-6, and procalcitonin, and the timing of the onset of fever symptoms, in addition to clinical findings; however, these data were not collected in this study, and so a complete diagnosis of MIS was difficult. Therefore, the incidence of MSI may have been underestimated in this study. In addition, in this retrospective study, the treatment for FM was left to the discretion of each participating hospital; therefore, the timing and specifics of treatment initiation are not fully understood. The definition of MSI was established later, so we believed it had no impact on decisions regarding in-hospital treatment. In fact, no difference was observed between the MSI (+) and MSI (−) groups for in-hospital treatment.
Finally, the mechanism by which MSI is related to non-COVID-FM prognosis and treatment for MSI related non-COVID-FM are unclear and requires further study.
Conclusions
In this retrospective cohort study of patients with non-COVID-FM, the MSI focused on clinical symptoms that were associated with lower 30-day mortality rate, and this trend was more pronounced in patients under 50 years of age. Further studies are needed to determine the full diagnosis of MIS and the association between MSI and myocarditis severity.
Disclosures
K. Kanaoka has received lecture fees from Eli Lilly Japan, Otsuka Pharmaceutical Co. Ltd, and Johnson and Johnson Co. Y. Saito received all support for the present manuscript from consulting fees from Novartis Pharma K.K., lecture fees from Novartis Pharma K.K., Nippon Boehringer Ingelheim Co ., Ltd., Daiichi Sankyo Co., Ltd., and Bayer Yakuhin, Ltd., payment for expert testimony for Novartis Pharma K.K., travel expenses from Novartis Pharma K.K., and leadership or fiduciary role in other boards, societies, committees (The Japanese Society of Circulation, The ISHR, The Japanese Heart Failure Society, The Japanese Circulation Association). K. Maemura is a member of the Circulation Reports’ Editorial Team.
IRB Information
This study followed the Declaration of Helsinki, and the protocol was authorized by the local ethics committees at Nagasaki University and the Japanese Circulation Society (registration numbers 20081719-4 and 10, respectively).
Data Availability / Discussion
The deidentified participant data will not be shared.
Supplementary Files
Supplementary Table 1. Supplementary Table 2.
Acknowledgments
Y. Saito received all support for the present manuscript from grants (MEXT KAKENHI Grant Number JP11903719 (Grants-in-aid from the Ministry of Education, Culture, Sports, Science), AMED under Grant Number JP21445023, JP20315051 (Practical Research Project for Rare/Intractable Diseases)).
References
- 1. Hong D, Bak M, Park H, Kim HY, Lee S, Kim IC, et al.. Novel risk scoring model to predict the implementation of veno-arterial extracorporeal membrane oxygenation in patients with acute myocarditis. Circ J 2025; 89: 1795–1803. [DOI] [PubMed] [Google Scholar]
- 2. Nagai T, Katsuki M, Amemiya K, Takahashi A, Oyama-Manabe N, Ohta-Ogo K, et al.. Chronic active myocarditis and inflammatory cardiomyopathy: Challenges in diagnosis and treatment. Circ J 2026; 90: 152–161. [DOI] [PubMed] [Google Scholar]
- 3. Ammirati E, Frigerio M, Adler ED, Basso C, Birnie DH, Brambatti M, et al.. Management of acute myocarditis and chronic inflammatory cardiomyopathy: An Expert Consensus Document. Circ Heart Fail 2020; 13: e007405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Veronese G, Ammirati E, Cipriani M, Frigerio M.. Fulminant myocarditis: Characteristics, treatment, and outcomes. Anatol J Cardiol 2018; 19: 279–286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Omori T, Maruyama K, Ohta-Ogo K, Hatakeyama K, Ishibashi-Ueda H, Onoue K, et al.. Clinical and histopathological characteristics of patients with myocarditis after mRNA COVID-19 vaccination. Circ J 2024; 89: 120–129. [DOI] [PubMed] [Google Scholar]
- 6. Tara S, Yamamoto T, Kanaoka K, Oka E, Shiomura R, Matsuda J, et al.. Effects of cardioprotective drugs on 90-day mortality or heart transplantation in patients with fulminant myocarditis. Circ Rep 2024; 6: 322–332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Kato S, Kitai T, Utsunomiya D, Azuma M, Fukui K, Hagiwara E, et al.. Myocardial injury by COVID-19 infection assessed by cardiovascular magnetic resonance imaging: A prospective multicenter study. Circ J 2024; 88: 1450–1458. [DOI] [PubMed] [Google Scholar]
- 8. Vogel TP, Top KA, Karatzios C, Hilmers DC, Tapia LI, Moceri P, et al.. Multisystem inflammatory syndrome in children and adults (MIS-C/A): Case definition & guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine 2021; 39: 3037–3049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Barhoum P, Pineton de Chambrun M, Dorgham K, Kerneis M, Burrel S, Quentric P, et al.. Phenotypic heterogeneity of fulminant COVID-19-related myocarditis in adults. J Am Coll Cardiol 2022; 80: 299–312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kawano H, Ikeda S, Kanaoka K, Sato S, Eto R, Ueno Y, et al.. Multisystemic inflammation influences prognosis in fulminant lymphocytic myocarditis. Circ J 2024; 88: 722–731. [DOI] [PubMed] [Google Scholar]
- 11. Schulz-Menger J, Collini V, Gröschel J, Adler Y, Brucato A, Christian V, et al.. 2025 ESC Guidelines for the management of myocarditis and pericarditis. Eur Heart J 2025; 46: 3952–4041. [DOI] [PubMed] [Google Scholar]
- 12. Nagai T, Inomata T, Kohno T, Sato T, Tada A, Kubo T, et al.. JCS 2023 Guideline on the diagnosis and treatment of myocarditis. Circ J 2023; 87: 674–754. [DOI] [PubMed] [Google Scholar]
- 13. Caforio AL, Pankuweit S, Arbustini E, Basso C, Gimeno-Blanes J, Felix SB, et al.. Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: A Position Statement of the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J 2013; 34: 2636–2648, 2648a–2648d. [DOI] [PubMed] [Google Scholar]
- 14. Richardson P, McKenna W, Bristow M, Maisch B, Mautner B, O’Connell J, et al.. Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the Definition and Classification of cardiomyopathies. Circulation 1996; 93: 841–842. [DOI] [PubMed] [Google Scholar]
- 15. Kanaoka K, Onoue K, Terasaki S, Nakano T, Nakai M, Sumita Y, et al.. Features and outcomes of histologically proven myocarditis with fulminant presentation. Circulation 2022; 146: 1425–1433. [DOI] [PubMed] [Google Scholar]
- 16. Coke LA.. Cardiac risk assessment of the older cardiovascular patient: The Framingham Global Risk Assessment Tools. Medsurg Nurs 2010; 19: 253–254. [PubMed] [Google Scholar]
- 17. Granger E, Sergeant JC, Lunt M.. Avoiding pitfalls when combining multiple imputation and propensity scores. Stat Med 2019; 38: 5120–5132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Caforio ALP, Giordani AS, Baritussio A, Marcolongo D, Vicenzetto C, Tarantini G, et al.. Long-term efficacy and safety of tailored immunosuppressive therapy in immune-mediated biopsy-proven myocarditis: A propensity-weighted study. Eur J Heart Fail 2024; 26: 1175–1185. [DOI] [PubMed] [Google Scholar]
- 19. Huang F, Ammirati E, Ponnaiah M, Montero S, Raimbault V, Abrams D, et al.. Fulminant myocarditis proven by early biopsy and outcomes. Eur Heart J 2023; 44: 5110–5124. [DOI] [PubMed] [Google Scholar]
- 20. Patel P, DeCuir J, Abrams J, Campbell AP, Godfred-Cato S, Belay ED.. Clinical characteristics of multisystem inflammatory syndrome in adults: A systematic review. JAMA Netw Open 2021; 4: e2126456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Tahaghoghi-Hajghorbani S, Zafari P, Masoumi E, Rajabinejad M, Jafari-Shakib R, Hasani B, et al.. The role of dysregulated immune responses in COVID-19 pathogenesis. Virus Res 2020; 290: 198197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Cheng R, Hachamovitch R, Kittleson M, Patel J, Arabia F, Moriguchi J, et al.. Clinical outcomes in fulminant myocarditis requiring extracorporeal membrane oxygenation: A weighted meta-analysis of 170 patients. J Card Fail 2014; 20: 400–406. [DOI] [PubMed] [Google Scholar]
- 23. Ammirati E, Veronese G, Brambatti M, Merlo M, Cipriani M, Potena L, et al.. Fulminant versus acute nonfulminant myocarditis in patients with left ventricular systolic dysfunction. J Am Coll Cardiol 2019; 74: 299–311. [DOI] [PubMed] [Google Scholar]
- 24. Guglin ME, Etuk A, Shah C, Ilonze OJ.. Fulminant myocarditis and cardiogenic shock following COVID-19 infection versus COVID-19 vaccination: A systematic literature review. J Clin Med 2023; 12: 1849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kawai H, Izawa H, Yanase M, Yamada A, Takahashi H, Ozaki Y, et al.. Re-evaluation of intravenous steroid therapy for histologically confirmed fulminant myocarditis. Int J Cardiol 2025; 427: 133108. [DOI] [PubMed] [Google Scholar]
- 26. Lee YI, Chung S, Yang JH, Sung K, Kim D, Choi JO, et al.. Extracorporeal membrane oxygenation for fulminant myocarditis: increase of cardiac enzyme and SOFA score is associated with high mortality. J Clin Med 2021; 10: 1526. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Table 1. Supplementary Table 2.
Data Availability Statement
The deidentified participant data will not be shared.