High-dose vitamin C ameliorates cardiac injury in COVID-19 pandemic: a retrospective cohort study

Study design and subjects

A retrospective, observational cohort analysis was carried out from February 1 to March 10, 2020. The study subjects consisted of COVID-19 pneumonia with myocardial damage who consecutively admitted to Tongji Hospital of Huazhong University of Science and Technology (Wuhan, China), and were diagnosed as positive for SARS-CoV-2 infection by performing RT-PCR assay on samples from nasal and pharyngeal swab, and were in severe and critically ill condition. Identification of disease condition was achieved in accordance with the criteria in Chinese management guidelines for COVID-19 on February 4, 2020 (trial version 5.0) [8]. The exclusion criteria were listed as follows: (i) patients without complete medical records; (ii) those who were not confirmed to be infected by SARS-CoV-2 up to March 24, 2020; (iii) individuals who lacked determinations of high-sensitivity troponin I (hs-cTnI) and parameters of inflammation on admission and day 21 during hospitalization.

The study complied with the ethical principles of the World Medical Association’s Declaration of Helsinki and received the ethical approval from the Medical Ethics Committee of the hospital. The oral, not written informed consents were obtained from patients underwent HIVC therapy owing to observational nature of the research and isolation precautions of the rapidly evolving pandemic.

Data sources

The screening exam comprised patient’s demographics details and baseline clinical characteristics available in the hospital’s electronic medical records system for analysis. The following information was recorded for each patient: age, gender, smoking status, pulse, blood pressure, comorbidity (i.e., hypertension, coronary heart disease, and diabetes), time from illness onset to admission, as well as presenting clinical symptoms. Simultaneously, we collected the laboratory examinations on complete blood counts, haemoglobin, serum biochemical indicators, N-terminal pro-B-Type natriuretic peptide (NT-proBNP) at baseline. The parameters of inflammation including high-sensitivity C-reactive protein (hs-CRP), interleukin (IL)-6, IL-8, tumor necrosis factor (TNF)-α, and hs-cTnI were obtained at admission and repeated day 21 during hospitalization.

hs-cTnI, NT-proBNP and inflammatory markers determinations were conducted in the central laboratory of Union Hospital. A blood sample was acquired at admission (defined as day 1) and finally beyond day 20 to discharge (defined as day 21 during hospitalization). All procedures were carried out following standard operating procedures.

Treatment regimen

On admission, all subjects were initiated on symptomatic supportive treatments based on the requirements of each individual patient. As suggested in the guideline, [8] symptomatic supportive treatments consisted of antiviral, antibiotics, corticosteroids, immunoglobulin, biologics, mechanical ventilation, rennin-angiotensin system inhibitors, nutritional supplements for myocardium, renal replacement therapy, and antidiabetic therapy prescribed by our medical group according to the hospital protocol, unless contraindicated. Besides symptomatic supportive treatments, a subset of patients with verbal consents also received administration of HIVC infusion as adjuvant therapy during the first 24 hours period after admission. The HIVC treatment protocol was administered as reported previously by our group [9]. Briefly, vitamin C was administered intravenously at the excess dosage of 100 mg/kg every 6 hours for 1 day followed by 100 mg/kg every 12 hours for additional 5 days during hospitalization.

Definitions

The clinical spectrum of COVID-19 ranges from mild, common, severe to critically ill cases. Severe illness was considered in the presence of one or all three of the following criteria: (i) dyspnea with respiratory rate ≥ 30 breaths/min; (ii) arterial oxygen saturation ≤ 93% at a rest state; (iii) the ratio of arterial partial pressure of oxygen/fractional inspired oxygen < 300 mm Hg [8]. Patients who met one of the following conditions were considered to have critically ill pneumonia: (i) respiratory failure and mechanical ventilation required; (ii) sepsis shock; (iii) multiorgan failure [8]. The efficacy evaluation of HIVC therapy on myocardial injury was performed through screening changes on concentrations of hs-cTnI and inflammatory parameters from baseline to day 21 during hospitalization. Myocardial damage was evidenced by blood levels of hs-cTnI > the 99th percentile upper reference limit (26.2 pg/mL) [2, 10]. Ameliorated cardiac injury (ACI) was diagnosed on the basis of the serum levels of hs-cTnI at day 21 during hospitalization below 26.2 pg/mL; non-ameliorated cardiac injury (NACI) was adjudicated on the serum levels of hs-cTnI at day 21 during hospitalization above 26.2 pg/mL, or those of baseline. Hyperinflammation was determined by the elevated levels of inflammatory markers [11].

Statistical analysis

We expressed continuous variables as median with interquartile range (IQR, 25-75th percentile) or categorical variables as number with percentage to characterize the study cohort. When continuous variables conformed to a normal distribution, 2-sample independent group t test was applied to analyze differences between the ACI group and the NACI group; otherwise, the Mann-Whitney U test was conducted. The assumption of normal distribution of the data was determined using the Shapiro-Wilk test. Proportions for categorical variables were compared using chi-square test or Fisher's exact test where appropriate. The independent association between medications and cardiac injury was studied using univariate and multivariate logistic regression analysis accounting for HIVC as well as other symptomatic supportive treatments. The results were described as odds ratio (OR) and 95% confidence interval (CI). A forest plot was created base on logistic regression analysis results. Hypothesis tests were performed using two-tailed test with P value of < 0.05 indicating statistical significance. All statistical analysis was performed using SPSS (version 16.0) and GraphPad Prism 5.0.

Baseline characteristics and laboratory findings

The detailed analysis of baseline characteristics in our study cohort was described in Table 1. A total of 113 confirmed COVID-19 cases were included in the study: 70 in the ACI group and 43 in the NACI group. The median age was 68 years old, and the male ratio was 52/61 (46.0%). There were no significant between-group differences for demographics and baseline clinical characteristics. Meanwhile, similar results were also observed between the two groups with respect to inflammatory markers levels at baseline.

As shown in Figure 1, the serum levels of hs-cTnI displayed no significant difference between the ACI group and the NACI group at baseline [43.7 (30.9-83.3) pg/mL vs 51.4 (33.1-91.8) pg/mL, P = 0.734]. In contrast, the concentrations of hs-cTnI at day 21 during hospitalization appeared significantly lower in the ACI group compared with the NACI group [17.6 (12.2-22.9) pg/mL vs 76.5 (49.2-126.5) pg/mL, P < 0.001].

Effect of HIVC on cardiac injury

Of the 113 hospitalized patients in our study, 62 (54.9%) cases were administered with symptomatic supportive treatments alone and 51 (45.1%) received administration with HIVC in addition to symptomatic supportive treatments. The percentages of subjects with HIVC therapy were 52.8% (n = 37) in the ACI group and 32.5% (n = 14) in the NACI group, respectively; and the proportions were significantly different (χ² = 4.432, P = 0.035; Table 2). Likewise, the percentages of patients receiving biologics (χ² = 4.336, P = 0.049), mechanical ventilation (χ² = 5.230, P = 0.029), and renal replacement therapy (χ² = 4.484, P = 0.045) were also statistically significant between the two cohorts (Table 2).

To investigate the independent effect of HIVC on cardiac injury in more detail, logistic regression analysis was performed with variables entered in the model including HIVC and other medications during hospitalization. The result revealed that administration of HIVC infusion was correlative with the improvement of cardiac injury (OR 2.420, 95% CI 1.022-5.729, P = 0.044; Figure 2) independent of mechanical ventilation (OR 5.322, 95% CI 1.594-17.770, P = 0.007) and renal replacement therapy (OR 5.650, 95% CI 1.139-28.033, P = 0.034). Thus, HIVC seemed to show a beneficial effect in improving myocardial damage among patients with SARS-CoV-2 infection in severe and critically ill condition.

Effect of HIVC on hyperinflammation

We subsequently investigated changes in parameters of inflammation from admission to day 21 during hospitalization. As described in Figure 3, regarding the levels of inflammation parameters on admission, no statistical significances were found between the two cohorts of patients underwent HIVC administration (n = 51) or symptomatic supportive treatments alone (n = 62) (hs-CRP, P = 0.303; IL-6, P = 0.460; IL-8, P = 0.676; TNF-α, P = 0.552). However, the levels of inflammation parameters at day 21 during hospitalization among cases receiving HIVC therapy significantly decreased compared to those administered with symptomatic supportive treatments alone (hs-CRP, P = 0.036; IL-6, P = 0.029; IL-8, P = 0.047; TNF-α, P = 0.037). Furthermore, changes in inflammatory markers levels displayed a tendency of significant decrease at day 21 during hospitalization related to baseline in patients receiving HIVC therapy (all P < 0.001).

Limitations

Limitations to the present study include a single center experience with a relatively small sample size. It was not possible to perform a valid calculation of the sample size due to the critical condition of the ongoing pneumonia in our hospital. Also, the study was a nonrandomized observational study and hence suffered from potential selection and ascertainment biases. The determinations of parameters of inflammation should be very detailed at additional time points during and following treatment, but we only collected hs-CRP, IL-6, IL-8, and TNF-α at baseline and day 21 during hospitalization, it was thus uncertain whether HIVC may have an effect on other inflammatory markers levels. Furthermore, medications such as corticosteroids usage may affect levels of inflammatory markers. Importantly, we had no standardized protocol for HIVC, the dosage of HIVC used in this study may be insufficient for optimal care of SARS-CoV-2 infection. Further research on vitamin C dosages and longer administration time may need optimization.