COVID-19 Myocarditis and Severity Factors: An Adult Cohort Study

Data Sources

We performed a retrospective study highlighting the clinical and radiographic hallmarks of diagnosed cases as COVID-19 in Yongchuan, Chongqing from Jan 21 to Mar 2, 2020. As an officially nominated hospital for COVID-19 patients, Yongchuan Hospital (affiliated to Chongqing Medical University) has been in charge of the treatment of COVID-19 cases from Yongchuan District, Chongqing municipality (close to Hubei Province). The study was approved by the Ethics Review Committee of Yongchuan Hospital (No. 2020KLS-6). We obtained oral informed consent for involved patients and/or their relatives. All consecutive adult cases (aged over 18 years) diagnosed with COVID-19 were included. All patients were under close surveillance until Mar 2, 2020.

Demographic, epidemiologic, radiologic imaging and laboratory data were collected by intensive medical record reviews of included cases, with double checks and additional inquiries when appropriate. The diagnosis and treatment were conducted strictly conforming to the COVID-19 national program provided by the National Health Commission of China (5^th edition).⁹

Pathogen confirmation with double SARS-CoV-2 tests

Each patient underwent double tests for SARS-CoV-2 nucleic acid detection every 24 hours on his/her nasopharyngeal swab samples. In brief, viral nucleic acid from samples was conducted using viral isolation kits (Daangene, Guangzhou, China). Fluorescence real-time PCR was performed with diagnostic reagent kits (two available kits; Sansure Biotech, Changsha, China and Daangene, Guangzhou, China) on LightCycler 480 Real-Time PCR instrument (Roche, Rotkreuz, Switzerland) at Yongchuan Disease Control Centre. Ct (Cycle threshold) value of no more than 40 (either genes or single gene with double tests) was considered as SARS-CoV-2 positive according to the manufacturer’s instructions. Thermal cycling was performed at 50 ° C for 15min, 95 °C for 15min, 45 cycles of 94 °C for 15s and 55 °C for 45s. The sensitivity of RT-PCR assay kits was 500copies/ml. The primers were: Primer 1 targeting open reading frame 1ab (ORF1ab) of SARS-CoV-2, 5’-CCCTGTGGGTTTTACACTTAA-3’; whereas primer 2 targeting nucleocapsidprotein(N) as 5’-GGGGAACTTCTCCTGCTAGAAT-3.^6-7 Ct value data were collected from Yongchuan Disease Control Centre.

Laboratory tests

Full spectrum laboratory tests were performed at admission and successive appropriate times, including blood count, serum biochemistry, serum inflammatory indicators, myocardial enzymes. Additional tests were performed according to patients’ detailed conditions of background diseases. Lymphocytes sorting analyses were conducted using flow cytometre (Bricyte e6, Mindray, Shenzhen, China) and pertaining assay kits (Mindray, Shenzhen, China).

Chest CT scans

At admission, each patient underwent chest CT scans (Go.top, SIMENS; parameters, 120kv, 50-200mA, automatic modulation, thickness as 5mm) for the evaluation of lung lesions. Subsequently, follow-up chest CT scans were performed when suitable. Diagnostic images were analyzed by experienced radiologists. Pulmonary inflammation index (PII) was estimated according to the criteria proposed by Chongqing Radiologist Association of China,¹⁰ with distribution and size of lesions estimated. As for distribution, one lung segment equals one score with highest score as 20, representing left and right lung lesions. For lesion size, 1 score indicates over 50% lung segment volume invasive, whereas 0 score less than 50%. Accordingly, PII= (Distribution score+ Size score)/40*100%. PII values were calculated twice at three week intervals by two experienced radiologists. Interclass coefficient was used to assess inter-rater reliability.

Electrocardiographic (ECG) examination

At admission, each patient underwent ECG examinations routinely. Each electrocardiogram was evaluated and reported by senior technicians. Abnormal ECGs were collected and analyzed in combination of clinical manifestations and serum myocardial enzyme tests. Repeated ECGs were ordered for patients suspected with myocardial damage.

Statistical analysis

Data were recorded and summarized using spreadsheets. Patients in our cohort were classified into severe and nonsevere groups at admission in combination of national program and the American Thoracic Society guidelines.¹¹ In brief, patients were included in severe group if they meet with the following criteria: 1) Shortness of breath with respiratory rate (RR) >30 times /min; 2) Oxygen saturation <93% at rest; 3)Arterial partial oxygen pressure (PaO2)/oxygen absorption concentration (FiO2)<277mmHg (The result was calculated according to the altitude of the epidemic area, Yongchuan city is about 700 meters above sea level, 1mmHg=0.133 kPa); 3) Pulmonary imaging showed that the lesion had progressed to > 50% at 24 ∼ 48 h; 4) Respiratory failure;5) Shock; 6)Combined with other organ failure. Continuous (means for normally distributed variables and/or interquartile ranges for abnormally distributed variables) and categorical variables (percentage, %) were analyzed using SPSS, version 29.0. Unpaired t test, Fisher exact test, Mann-Whitney U test and chi-square test were performed where appropriate. Multivariable logistic regress model was used to control for confounding factors and distinguish independent risk factors for clinical severity. Risk factors with a univariable p-value<0.05 were eligible for inclusion in the model. Severe group was coded as “1” and nonsevere group was coded as “0”. Variables (p≤0.10) were in the final model, with significant variables identified (p<0.05), via a forward and stepwise protocol. Adjusted odds ratios (ORs, indicating pertaining digital times more likely to present in severe group) and 95% confidence intervals (CIs) were calculated as independent risk factors. Spearman’s correlations were used to assess the links between multiple factors.

Role of the funding source

No funding was received for the clinical cohort study. The corresponding author was in charge of all parts of the study and confirm that issues regarding the accuracy or integrity of the study had been properly scrutinized and settled. The final version of the manuscript was approved by all authors.

Presenting Hallmarks

In total, the study cohort included 84 hospitalized adult patients diagnosed as COVID-19. By March 2, 2020, 68 of the 84 patients had been cured and discharged, and there were still 16 hospitalized patients with no deaths. The median age was 48 years (Interquartile Range [IQR], 42·3-62·5; range, 18-93) with 48 males (57.1%). Amongst these adult patients, 20 cases were classified as severe group at admission; whereas 64 cases as nonsevere group.

The median duration from illness onset to diagnosis was 2 days (IQR, 1-5). Onset symptoms were fever (54[64.3%]), cough (43[51.2%]), expectoration (28[33%]), fatigue (16[19.0%]), anorexia (14[16.7%]), myalgia (11[13.1%]), Dizziness(8[9.5%]), Pharyngalgia(8[9.5%]), Chilly(8[9.5%]), short of breath (7[8.3%]), diarrhea (6[7.1%]), Headache (4[4.8%]), Dyspnea(3[3.6%]), Nausea (3[3.6%]), Vomiting(2[2.4%]) and Fluster(2[2.4%]). Notably, there were 11(13.1%) asymptomatic cases in the cohort, and all were in the nonsevere group(table 1).

The average peak temperature was 37.7°C. Forty cases (47.6%) had family cluster infection; whereas 43 patients (51.2%) had Wuhan linkage.

In comparison with nonsevere patients, the severe cases were older (58.0 [IQR, 48.5-74.5] vs 46.5 [IQR, 39.2-54.8], p<0.001), with higher peak temperatures ([38.3±0.7] vs [37.5±0.8]; p<0.001), and were more prone to symptoms such as fever (19[95.0%] vs 35[54.7%]; p<0.001), expectoration(12[60.0%] vs 16[25.0%]; p=0.002), fatigue (9[45.0%] vs 7[10.9%]; p=0.004), anorexia (11[55.0%] vs 3[4.7%]; p<0.001), myalgia (6[30.0%] vs 5[7.8%]; p=0.019), short breath (5[25.0%] vs 2[3.1%]; p=0.002) and diarrhea (4[20.0%] vs 2[3.1%]; p=0.027).

Thirty-five cases had one or more comorbidities. The most common coexisting diseases were hypertension (12[14.3%]), Hepatic damage (11[13.1%]), diabetes (10[11.9%]), cardiopathy (5[6.0%]), respiratory disease (5[6.0%]) and brain disease (4[4.8%]). Less common coexisting diseases were tumor (1[1.2%]) and renal damage (1[1.2%]). However, there was no significant difference in comorbidities except for diabetes (5[25%] vs 5[7.8%]; p=0.001).

Laboratory indicators

There were significant differences in a number of laboratory indicators between patients in the severe and non-severe groups (table 2). Distinct inter-group laboratory findings included multitude measures listed(figure 1); (1)higher inflammatory markers, CRP (25.4 [IQR,15.4-35.5] vs 6.1[IQR, 1.6-12.6]; p<0.001), ESR (72.1[IQR,46.5-79.6] vs 31.0[IQR, 19.5-66.0]; p=0.002), PCT(0.07 [IQR,0.06-0.09] vs 0.05[IQR, 0.05-0.06]; p<0.001), and IL-6(13.9[IQR, 7.2-22.7] vs 2.8[IQR, 1.7-7.8]; p<0.001); (2)higher indicators of cardiac enzymes such as CK (77.3[IQR, 43.9-109.2] vs 51.1[IQR, 37.0-77.8]; p=0.015); LDH (529.1[IQR, 347.7-580.2] vs 385.5[IQR, 319.9-442.0]; p=0.004); α-HBDH (183.0[IQR, 154.7-224.8] vs 132.0[IQR, 118.4-154.3; p<0.001]), and IMA (75.3[IQR, 72.3-82.9] vs 69.2 [IQR, 65.2-73.6; p<0.001]); (3)higher levels of NEUT(5.0[IQR, 3.6-6.9] vs 3.9 [IQR, 2.9-4.8; p=0.004]), and lower levels of LYM (0.8[IQR, 0.6-1.2] vs 1.4[IQR, 1.1-1.7]; p <.001]); (4) other markers, SF(1104.0[IQR, 774.0-1370.5] vs 368.5 [IQR, 135.3-506.3; p<0.001]), PAB (175.6[IQR, 142.6-209.4] vs 227.5[IQR, 192.4-253.0; p=0.001]), GLB (7.0[IQR, 5.0-8.9] vs 5.7[IQR, 5.4-6.0; p=0.001]), and AST (27.7[IQR, 23.1-41.6] vs 21.8[IQR, 19.0-27.3; p=0.001]).

• Download figure
• Open in new tab

Figure 1. Distribution diagrams of risk factors and laboratory indicators with statistical significance.

A, Heat map demonstrates the two-way hierarchical clustering of cases and Indictors. White strips without red or yellow colors indicate missing data. Yellow strips represent higher Z scores and red strip indicate lower Z scores. Severe and nonsevere groups represent the classification of COVID-19 according to clinical severity on admission. B, C and D, distribution diagrams of age, pulmonary inflammation index and viral load (Cycle threshold value) between severe and nonsevere groups, as identified independent risk factors for COVID-19 severity by multivariate logistic regression analysis. E to N, distribution diagrams of 10 laboratory indicators with statistical significance between severe and nonsevere groups by univariate analysis. *p<0.05.

In addition, in the detection of lymphocyte subsets, total T lymphocyte count, B lymphocyte count, and Inhibitive/Cytotoxic T Lymphocyte count were significantly reduced in the severe group compared with the nonsevere group.

Ct values representing viral load

The Ct values of RT-PCR for two target genes (ORF1ab and N) were recorded in 62 of all patients, with a total of 207 values at corresponding time points (table 3). There were three different types of samples were used to detect SARS-CoV-2 infection, including 190 nasopharyngeal swab, 3 stool and 14 anal swab samples. In order to maintain the uniformity of the samples, the results of the anal swab samples and stool samples were excluded from our study. There were 14 patients with 51 nasopharyngeal swab samples in the severe group, while 48 patients with 139 nasopharyngeal swab samples in the nonsevere group. In the severe group, the median Ct values of RT-PCR for N gene was 33.8(IQR, 27.1-36.6) and for ORF1ab gene was 34.7(IQR, 30.9-37.7). While in the nonsevere group, the median Ct values of RT-PCR for N gene was 38.1(IQR, 34.2-40.2) and for ORF1ab gene was 38.6(IQR, 36.0-41.5). There was significant difference in Ct values either for N gene (p=0.004) or for ORF1ab gene (p=0.002) between the two groups.

Chest CT findings and PII

Totally 84 patients diagnosed with COVID-19 underwent at least two or more CT examinations (table 3). An excellent inter-rater reliability was noted in the assessment of PII values(>0.89).The median of pulmonary inflammation index (PII) was 27.5(IQR, 7.5-42.5). The study showed that the PII values in the severe group was significantly higher than that in the nonsevere group (52.5 [42.5-62.5] vs 20 [5.0-31.6]; p<0.001). The most common CT abnormalities were patchy shadows or ground glass opacities (60[71.4%]), consolidation (41[48.8%]), interlobular septal thickening (58 [69%])(figure 2). Most patients have multiple areas of lung lesions with bilateral lung lesions (60[71.4%]), lower lung distribution (53[63.1%]), mid lung distribution (59 [70.2%]), and upper lung distribution (76[90.5%])(figure 3). Typical CT abnormalities was more likely existing in the severe group than in the nonsevere group with significant difference in patchy shadows or ground glass opacities (19[95%] vs 41[64.1%]; p=0.008), consolidation (17[85%] vs 24[37.5]; p<0.001), and interlobular septal thickening (18[90.0%] vs 40[62.5%]; p=0.020).There were also significant differences in the areas of lung lesions between the two groups. Compared with the nonsevere group, patients in the severe group were more likely to have bilateral lung lesions (20[100%] vs 40[62.5%]; p=0.001), and the lesions was more frequent in the lower lobe (19[95 %] vs 34[53.1%]; p=0.001) and middle lobe (20 [100 %] vs 39[60.9%]; p=0.001) of the lung(figure 4). However, no significant difference was found in the upper lobe of the lung (20[100 %] vs 56 [87.5%]; p=0.189).

• Download figure
• Open in new tab

Figure 2. Representative chest CT images of COVID-19 patients.

A, B, C and D, severe type. E, F, G and H, nonsevere type. A, a middle-aged male patient, chest CT scan performed on 12th days of onset. Multiple diffused ground glass opacities, dense shadow cord strip and nodules in both lung fields were revealed. B, an elderly male patient, chest CT scan performed on 11th days of onset. CT showed bilateral lung multiple diffused ground glass opacities and consolidation, with local interlobular septal thickening and crazy paving appearance. C, a middle-aged female patient, chest CT scan performed on 7th days of onset. CT showed bilateral subpleural multiple diffused ground glass opacities and consolidation. D, a middle-aged male patient, chest CT scan performed on 5th days of onset, multiple diffused ground glass opacities, dense shadow cord strip and nodules in bilateral subpleural lung regions, with bronchiectasis in the subpleural region of the lateral basal segment of the right lower lobe. E, chest CT scan performed on 18th days of onset, a male patient aged 54. Patchy lung infiltrates showed in bilateral subpleural lung regions, with intralesional low-density lesion. F, a middle-aged male patient, chest CT scan performed on 4th days of onset, A few fibrotic foci in the upper lobe of the right lung after clinical cue of TB. G, chest CT scan performed on 6th days of onset, an elderly male patient. CT showed a few fibrotic foci in the right middle lobe medial segment. H, chest CT scan performed on 5th days of onset, a young female patient. CT showed a few fibrotic foci in the medial segment of right middle lobe of lung.

• Download figure
• Open in new tab

Figure 3. Dynamic chest CT scan changes of a nonsevere patient (a young male patient) with COVID-19.

A and E, early-stage of COVID-19, chest CT scan performed on 2th days of onset. Irregular nodular high-density shadow in the basal segment of the left lower lobe. B and F, lung progressions, chest CT scan performed on 5th days of onset. CT showed enlarged lesion area, with patchy slightly high-density shadow and intralesional interlobular septal thickening. C and G, critical period, chest CT scan performed on 9th days of onset. CT showed enlarged lesion area from subpleural area to the hilum of lung, with lung consolidation and bronchiectasis. D and H convalescence phase, chest CT scan performed on 14th days of onset. CT showed decrease of the density focus and consolidation area, with gross fibrotic foci emerging.

• Download figure
• Open in new tab

Figure 4. Dynamic chest CT scan changes of a severe patient (a middle-aged female patient) with COVID-19.

A and E, early-stage of COVID-19, chest CT scan performed on 3th days of onset. CT showed ground-glass opacities and patchy lung infiltrates in bilateral subpleural lung regions with intralesional interlobular septal thickening. B and F, lung progressions, chest CT scan performed on 7th days of onset. CT showed enlarged lesion area, with patchy slightly high density shadow. C and G, critical period, chest CT scan performed on 11th days of onset. CT showed slightly enlarged lesion area and increase of lesion density. D and H, convalescence phase, chest CT scan performed on 20th days of onset. CT showed reduction of lesion and decrease of lesion density, with lesion absorption and fibrotic foci emerging.

Complications during hospitalization

Complications during hospitalization were documented (table 4). There were 47 (56.0%) cases with electrolyte disorders, 37(44.0%) cases with anemia, 36(42.9%)cases with abnormal cardiac enzymes, 30(35.7%)cases with hepatic abnormalities, 27(32.1%)cases with hypoproteinemia, 23(27.4%) cases with gastrointestinal disorders, 22(26.2%) cases with bacterial pneumonia, 20(23.8%) cases with type 1 respiratory failure, and 6(7.1%) cases with renal abnormalities. Except for gastrointestinal disorders, the incidence of all complications in the severe group was significantly higher than that in the non-severe group (p<0.05). Of the 20 patients with respiratory failure, 9(45%) were treated with assisted mechanical ventilation, 7(35%) with nasal catheters and 4(20%) with masks. The median time from onset to respiratory failure was 7.75 days and to mechanical ventilation was 8 days.

Clinical diagnosed myocarditis

Among all 84 patients, we found 13 patients with abnormal myocardial injury indicators combined with abnormal ECG, including 8 severe patients and 5 nonsevere patients (table 5). Four of the 13 patients had a history of cardiovascular diseases, while the other 9 patients had no such history. There were 3 cases with a history of hypertension, 1 patient with hypertension, atrial fibrillation and coronary heart disease who was excluded from our further analysis. The 12 remaining patients had abnormal changes in the myocardial enzymes or cTnI, amongst which 10 cases with high levels of inflammatory markers (except for patient no.9 and 11), and eleven of them exhibiting clinical symptoms of myocardial injury (except for patient no. 9).

Typical abnormal ECG findings were noted in ten patients (except for patient no.2, 6 and 12). Major ECG abnormalities included sinus tachycardia or bradycardia, t-wave changes, atrioventricular block, and anomaly Q wave. Four patients underwent repeated ECG examinations. Patient no. 3 was an elderly female patient in the severe group. The first ECG indicated t-wave changes and sinus bradycardia. However, repeated ECG restored normal after treatment with a decrease in inflammatory markers and cardiac enzyme levels. Patient no. 4 was an elderly male patient in the severe group. His first ECG indicated First-degree atrioventricular block, t-wave changes, left axis deviation and abnormal Q wave in the lower wall, while the abnormal Q and t-wave disappeared after treatments. Patient no. 9 and 10 were in the nonsevere group. Their initial ECGs showed sinus tachycardia and sinus tachycardia associated with t-wave changes respectively, with repeated normal ECGs after treatment.

Based on the most updated diagnostic criteria of viral myocarditis^12-13 (Disease code: H00395), 4 cases(patient no.4, 7, 8 and 10) were clinically diagnosed as SARS-CoV-2 myocarditis in combination of their manifestations, i.e., confirmed SARS-CoV-2 infection, typical ECG abnormalities, and elevated laboratory myocardial damage indicators(figure 5)

• Download figure
• Open in new tab

Figure 5. Representative electrocardiograms of clinical diagnosed SARS-CoV-2 myocarditis.

A, Abnormal ECG of patient no. 4 on the day of admission. It showed low and flat t-wave in II lead, first-degree atrioventricular block and abnormal Q wave in the II, III, and aVF leads. B, ECG of patient no. 4 after treatments. It showed the abnormal Q and t-wave disappeared after treatments, leaving him with only an first-degree atrioventricular block in II, III, and aVF leads. C, Abnormal ECG of patient no. 7 on the day after admission. It showed occasional premature atrial beats in II lead, and inverted t-wave in aVL lead. D, Abnormal ECG of patient no. 10 after admission. It showed sinus tachycardia associated with low and flat t-wave in II lead and sinus tachycardia associated with positive and negative t-wave in aVF lead. E, ECG of patient no. 10 patient returned to normal in II and aVF leads after treatment. F, Abnormal ECG of patient no. 8 after admission. It showed first-degree atrioventricular block in II, V1, V2, and V3 leads.

In order to explore the intrinsic correlations between cardiac enzymes, ECG and cTn I with inflammatory cytokines and lymphocyte levels, we used Mann-Whitney U test to analyze the relevant data (table 6).The results showed that the level of cardiac enzymes was significantly positively correlated with the level of CRP (p=0.004) and PCT (p=0.012). In addition, this study also found that there was a significant correlation between the abnormality of ECG and the level of PCT (p=0.004) and lymphocytes counts (p=0.041). However, we did not find any correlation between cTn I and levels of inflammatory cytokines or lymphocytes.

Multivariate logistic regression analysis

In this study, multivariable logistic regress model was used to identify the independent risk factors for clinical severity of COVID-19 (figure 6). Omnibus Tests of Model Coefficients indicated that the χ² of the multivariate logistic model was 32.595 (p<.050), validating the statistical significance of the model. We tried to include several indicators with statistical differences amongst univariate analyses (except laboratory tests), and the multicolinearity test was carried out according to variance inflation factor. According to the equation fitting degree and the independence of indicators, 6 variables were finally included in multivariable logistic regress model including age, PII, ct value, anorexia, and diabetes. The final results determined that age [OR 2.350; 95% CI (1.206 to 4.580); p =0.012], Ct value [OR 0.158; 95% CI (0.025 to 0.987); p=0.048] and PII [OR 1.912; 95% CI (1.187 to 3.079); p=0.008] were statistically different--independent risk factors for COVID-19(table 7).

• Download figure
• Open in new tab

Figure 6. The extrinsic factors and intrinsic factors of SARS-CoV-2.