key: cord-0909753-y4r9nfu9
authors: Li, Jing-Wei; Han, Tian-Wen; Woodward, Mark; Anderson, Craig S.; Zhou, Hao; Chen, Yun-Dai; Neal, Bruce
title: The impact of 2019 novel coronavirus on heart injury: A systemic review and Meta-analysis
date: 2020-04-16
journal: Prog Cardiovasc Dis
DOI: 10.1016/j.pcad.2020.04.008
sha: 3dc4e491aacc2c606aed389e9c8382ac190ad3b7
doc_id: 909753
cord_uid: y4r9nfu9

Abstract Background Evidence about COVID-19 on cardiac injury is inconsistent. Objectives We aimed to summarize available data on severity differences in acute cardiac injury and acute cardiac injury with mortality during the COVID-19 outbreak. Methods We performed a systematic literature search across Pubmed, Embase and pre-print from December 1, 2019 to March 27, 2020, to identify all observational studies that reported cardiac specific biomarkers (troponin, creatine kinase–MB fraction, myoglobin, or NT-proBNP) during COVID-19 infection. We extracted data on patient demographics, infection severity, comorbidity history, and biomarkers during COVID-19 infection. Where possible, data were pooled for meta-analysis with standard (SMD) or weighted (WMD) mean difference and corresponding 95% confidence intervals (CI). Results We included 4189 confirmed COVID-19 infected patients from 28 studies. More severe COVID-19 infection is associated with higher mean troponin (SMD 0.53, 95% CI 0.30 to 0.75, p < 0.001), with a similar trend for creatine kinase–MB, myoglobin, and NT-proBNP. Acute cardiac injury was more frequent in those with severe, compared to milder, disease (risk ratio 5.99, 3.04 to 11.80; p < 0.001). Meta regression suggested that cardiac injury biomarker differences of severity are related to history of hypertension (p = 0.030). Also COVID19-related cardiac injury is associated with higher mortality (summary risk ratio 3.85, 2.13 to 6.96; p < 0.001). hsTnI and NT-proBNP levels increased during the course of hospitalization only in non-survivors. Conclusion The severity of COVID-19 is associated with acute cardiac injury, and acute cardiac injury is associated with death. Cardiac injury biomarkers mainly increase in non-survivors. This highlights the need to effectively monitor heart health to prevent myocarditis in patients infected with COVID-19.

The systematic review is reported following the Meta-analysis Of Observational Studies in J o u r n a l P r e -p r o o f cardiac injury according to severity of COVID-19, or mortality according to cardiac injury, were included. Studies of the same hospital with period within range of other studies, or with data that could not be reliably extracted were excluded.

TH and JL scrutinized the titles and abstracts of the 2961 reports identified and excluded clearly irrelevant studies. Two authors (TH and JL) then did independent review of the full reports for the remaining 553 studies and extracted data independently from the 28 with relevant information.

(eFigure 1 in supplementary material) Any disagreements were resolved by a third author (YC). When estimates were only presented graphically, we used the software g3data (version 1.51, www.frantz.fi/software/g3data.php) to extract estimates.

For analysis of cardiac injury biomarkers, first standard mean differences (SMD) were calculated for all biomarkers using mean (SD) values and group size, and pooled using random effects models. Then, separate meta-analysis was carried out, with SMD using for troponin (troponin I and T, or high sensitivity troponin I and T), and weighted mean differences (WMD) for other biomarkers. Summary relative risks (RRs) with 95% CIs were estimated for the association between acute cardiac injury and death. The degree of heterogeneity was assessed using the I 2 -index. An I 2 statistic was considered to reflect low likelihood (0%-25%), moderate likelihood (26%-75%), and high likelihood (76%-100%) of differences beyond chance, as was a P value, from a Q test of heterogeneity, of less than or equal to 0.05 6 . If the results were homogeneous (I 2 <50% and P>0.05), fixed effects models were utilized, whereas if these results were heterogeneous (I 2 ≥50% or P≤0.05), then random effect models were used. If only median and interquartile range (Q25, Q75) were reported, then we assumed the median was equal to the mean and that the standard deviation (SD) was (Q75-Q25)/2. We also conducted meta-regression to assess whether severity differences in cardiac injury biomarkers were modified by patient characteristics: age, sex, smoking, diabetes, hypertension, history of cardiovascular diseases, J o u r n a l P r e -p r o o f coronary heart disease, cerebrovascular diseases, chronic obstructive pulmonary diseases, chronic kidney disease, and severity definitions of COVID-19. In this meta-regression analysis cardiac injury biomarkers are selected with priority troponin>CK-MB>myogloblin>NT-proBNP for each study, for example if one study reported troponin and CK-MB, we used troponin as the outcome and used the standard mean difference model to combine the results. Evidence of publication bias was examined using Egger's regression test for funnel asymmetry, in addition to visual inspection of the funnel plots.

Combined means of hsTroponin I and NT-proBNP were calculated as the sum, over studies, of the mean value of the biomarker and n is the number of participants. Data were summarised using inverse variance weighted meta-analysis and a 2-sided P value of less than or equal to 0.05 was deemed significant. Statistical analysis was performed with Stata, version 15.1. Fourteen studies reported data from patients in Wuhan, where the epidemic first emerged. Those with more severe disease were somewhat older, fewer women than men, and had higher prevalence of coexisting disorders. (Table 1) J o u r n a l P r e -p r o o f levels were evident in survivors, NT-proBNP in survivors was only available in one small study 32 ( Figure 5 ).

In this systematic review and meta-analysis, we found that there is an increased risk of acute cardiac injury associated with more severe COVID-2019 infection and this acute cardiac injury is associated with death. Patients with a history of hypertension seem to suffer more from cardiac damage.

Our results are in line with a previous meta-analysis of COVID-19 on cardiac troponin I, which found troponin I significantly increased in COVID-19 patients with severe disease compared to those with milder infection. 38 NT-proBNP show that cardiac injury biomarkers rise above normal by the midpoint of hospitalization and spike immediately before death, which seems to be most seen in severe cases. This pattern of elevation suggests that cardiac damage may already exist before multiple organ dysfunction syndrome.

In a large report containing 44,672 confirmed cases from China, approximately 81% of COVID-19 infections were mild cases which did not require hospitalization, with case-fatality rates of 49% in critical cases and 0% in mild ones. 40 In our analysis more severe infection with COVID-19 is associated with 14 times higher mortality risk compared to mild infection, and acute cardiac injury mostly indicated by abnormal cardiac biomarker levels, is associated with 4 times higher mortality risk. In one of the included studies abnormal troponin I was associated with an 80 times higher risk of in-hospital death, although the sample size was small. 20 It is possible that direct cardiac damage or cardiac shock partially explain death.

Based on these results, we are in support of previous suggestions that longitudinal measurement of cardiac damage biomarkers is needed during hospitalization stay for SARS-CoV-2 infection, which J o u r n a l P r e -p r o o f may help to identify a subset of patients with cardiac injury, and thereby predict the progression of COVID-19 towards an unfavorable outcome. 38 Moreover, studies shall also be planned to determine whether or not cardiac supportive measures, such as echocardiographic testing and heart failure treatments such as mechanical ventilation, inotropic agents, and vasopressors, are beneficial in patients with significant elevation of cardiac injury biomarkers.

Our meta-analysis has several potential limitations. Firstly, there was obvious heterogeneity among studies regarding definitions of severity of COVID-19, acute cardiac injury and biomarkers to detect cardiac injury although no significant publication bias were observed (eFigure 6). Secondly, this meta-analysis was conducted for studies that failed to describe all relevant characteristics of individual patients and it was hard to adjust for potentially confounding factors such as age, gender and use of treatments, such as renin angiotensin aldosterone system inhibitors. Finally, all included studies were retrospective and there are risk of bias in the data collected.

The severity of COVID-19 is associated with acute cardiac injury, and the latter is associated with death. Patients with a history of hypertension seem to suffer more from this kind cardiac damage. Future studies are needed identify whether cardiac supportive measures and heart failure treatments are beneficial in severe patients infected with COVID-19.

J o u r n a l P r e -p r o o f 

The clinical characteristics of moycardial injury in servere and very severe patients with 2019 novel coronavirus disease. SSRN pre-print

Clinical and radiographic features of cardiac injury in patients with 2019 novel coronavirus pneumonia. medRxiv preprint

Epidemiological and Clinical Features of

COVID-19) in Changsha, China. SSRN pre-print

Clinical characteristics of 77 novel coronavirus 2019 infected patients with respiratory failure in the terminal stage in Wuhan. SSRN pre-print

Heart injury signs are associated with higher and earlier mortality in coronavirus disease 2019 (COVID-19). medRxiv preprint

Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads and lung injury

Risk Factors Associated With Acute Respiratory Distress Syndrome and Death in Patients With Coronavirus Disease

Sanchis-Gomar F. Cardiac troponin I in patients with coronavirus disease 2019 (COVID-19): Evidence from a meta-analysis