Respiratory syncytial virus infections and myocardial injury: the GEMINI Initiative

The pulmonary outcomes with respiratory syncytial virus (RSV) infection are well known,1 but data on the cardiac sequelae are sparse. Knowing the cardiac risks of RSV infection will help clinicians, patients, and policy makers weigh the potential benefits of RSV vaccination in individuals at low pulmonary risk. While a systematic review of studies published to July 2024 highlighted the paucity of evidence about cardiac risks with RSV infection,2 six studies published since then have reported cardiac complications in 5% to 37% of patients hospitalized with RSV infection.3–8 However, four of these studies were restricted to older patients, only three3,6,8 included patients with other respiratory virus infections as a comparator, and only one7 reported cardiac biomarker results (with 30% of 471 patients having elevated troponins and 61% elevated natriuretic peptides [NP]). Recently, we reported that although clinically recognized cardiac complications developed in 6% of adults hospitalized with COVID-19, 43% had elevated troponins and 72% elevated NP.9 Here, we used similar methods as described in that paper9 (a retrospective cohort study of electronic health records from 29 hospitals in Ontario, Canada between April 2015 and March 2023) to examine frequency of elevated cardiac biomarkers and new clinical cardiac diagnoses in patients hospitalized with RSV infection (using COVID-19 as the comparator since the risks of myocardial sequelae with COVID-19 are now well recognized).2 We only included the first hospitalization for any patient in the study period and excluded nosocomial infections or patients with both infections recorded during the same hospitalization. We defined hospitalizations on the basis of the most responsible diagnosis: International Classification of Diseases 10th Revision Canadian version [ICD-10-CA] codes U07.1 or U07.2 for COVID-19 (a case definition with 98% sensitivity and 99% specificity)9 and ICD-10-CA diagnosis codes J12.1, J20.5, J21.0, or B97.4 for RSV infection (99% specificity).10 We also included hospitalizations where RSV or COVID-19 was a secondary diagnosis but the most responsible diagnoses were symptoms (cough, dyspnoea, malaise, fever) or potential manifestations (viral pneumonia, sepsis, respiratory failure, delirium) of either infection. Our outcomes of interest during the index hospitalization are listed in Figure 1 and we examined the association of outcomes with RSV vs COVID-19 after adjusting for the covariates listed in the figure legend and employed generalized estimating equations with an exchangeable working correlation structure to account for within-hospital clustering. Study outcomes during the index hospitalization. CI, confidence interval; ICU, intensive care unit; IQR, inter-quartile range; NP, natriuretic peptide; RSV, respiratory syncytial virus; ULN, upper limit of normal. Canadian Cardiovascular Society definition for probable heart failure: BNP >400 pg/ml; NT-proBNP >900 pg/ml if aged 50–75 or >1800 pg/ml if older than 75 years. *Adjusted effects for the outcomes were calculated using Poisson regression models (for all outcomes except length of stay, which was analysed using a negative binomial model) and baseline covariates for risk adjustment included patient age, sex, long-term care residence, year of admission, neighbourhood income quintile, GEMINI modified Laboratory Acute Physiology Score, clinical comorbidities (the Charlson comorbidity index score as a summary metric for comorbidity adjustment but prior myocardial infarction, prior chronic obstructive pulmonary disease, prior atrial fibrillation, or prior heart failure as separate covariates in the sensitivity analysis mentioned in the text), and admitting hospital as clusters (to account for within-hospital correlations) within generalized estimating equation models We found that the 1683 patients hospitalized with RSV were older, and were more likely than patients hospitalized with COVID-19 to: be female, have higher Charlson comorbidity scores, come from higher income neighbourhoods or long-term care facilities, and to have pre-existing cardiac or chronic pulmonary disease (although less than one-quarter had prior cardiac diagnoses). At the time of admission, hospital capacity strain was similar in both groups and both had similar likelihood of having their serum troponin (82.3% vs 84.2%) or NP (20.0% vs 21.6%) measured (Table 1). Characteristics of adults admitted for COVID-19 or RSV from April 2015 to March 2023 RSV, respiratory syncytial virus; SD, standard deviation; SMD, standardized mean difference. For categorical variables with more than two levels, the maximum pairwise SMDs are reported. By convention, SMDs >0.10 are considered significant. aData are slightly different than reported in McAlister et al.9 due to inclusion of some patients excluded from that study due to influenza co-infection. The data for the COVID-19 patients are reproduced with permission from McAlister et al and Elsevier.9 Although patients hospitalized with RSV had shorter lengths of stay and were less likely to require intensive care or die during the index hospitalization than those with COVID-19 (Figure 1), they were more likely to have new clinical diagnoses of atrial fibrillation (7.8% vs 3.4%, adjusted risk ratio [aRR] 1.35, 95% confidence interval [CI] 1.13–1.61) and heart failure (4.9% vs 2.4%, aRR 1.30, 95% CI 0.87–1.95) than patients hospitalized with COVID-19. In a sensitivity analysis, we found that the size of the effect estimates were attenuated but not negated after adjusting for prior cardiac diagnoses in lieu of the Charlson comorbidity score. Amongst those who had them measured, patients with RSV were just as likely to have elevated cardiac biomarkers as patients with COVID-19: 83.7% vs 72.3% for NP (aRR 1.05, 95% CI 0.98–1.12) and 44.2% vs 42.7% for troponin (aRR 0.94, 95% CI 0.90–0.97). A sensitivity analysis in a sub-cohort of 1649 RSV and 3298 COVID-10 patients matched by propensity score (for having biomarkers measured) and using matched strata rather than hospital as the clustering variable revealed nearly identical results as the primary analysis. The magnitude of biomarker elevations were similar with RSV or COVID-19 hospitalizations (Figure 1). Elevated troponin levels were more common in patients with RSV who died (68.6% vs 41.4%, P < .001) or had new cardiac diagnoses (59.5% vs 42.0%, P < .001). NP elevations were also more common in RSV patients who died (92.9% vs 82.4%, P = .13) or had new cardiac diagnoses (89.1% vs 82.4%, P = .27), although they fell short of statistical significance as fewer patients had NP measured. Although our study included all patients admitted with RSV or COVID-19 to the participating hospitals, there are some limitations. For one, the observed rate of cardiac diagnoses likely underestimates the true burden of cardiac involvement as not all patients had biomarkers measured (and 42% of troponins were low sensitivity assays), not all patients had electrocardiograms or cardiac imaging done, and our study only evaluated outcomes during the index hospitalization. Second, as RSV vaccine was only available in Ontario after the timeframe we studied, none of the patients were vaccinated against RSV. Third, as with any observational study, we acknowledge there is always the possibility of selection bias in whom had biomarkers measured and it is likely testing practices vary between clinicians, institutions, and over time. Our study found that new cardiac diagnoses and elevated cardiac biomarkers were at least as common with RSV as with COVID-19 and thus, just as with COVID-19, RSV should be recognized as a cardiovascular risk and not just a respiratory pathogen. We believe the discrepancy between the frequency of biomarker elevations and clinically recognized diagnoses suggest that clinicians may need to increase their index of suspicion for concomitant cardiac involvement in patients hospitalized with RSV, just as they do with COVID-19. At the very least our findings can be viewed as hypothesis generating and highlight the need for further research to investigate the long-term significance of cardiac biomarker elevations in patients with viral respiratory infections. Future studies on the epidemiology of RSV infection and the efficacy of RSV vaccines should include routine screening for evidence of cardiac complications which will help inform decisions about the potential merits of RSV vaccination for individual patients. This study was supported by GEMINI, a research programme based out of Unity Health Toronto. The development of GEMINI’s data platform has been supported by several funding partners, which can be found listed at https://geminimedicine.ca/partners/. All analyses, results, conclusions, and opinions presented in this paper are solely those of the listed authors and are independent of GEMINI’s funding sources. No endorsement by GEMINI’s funding sources is intended, nor should it be inferred. Amol Verma is supported by the Temerty Professorship of AI Research and Education in Medicine from the University of Toronto and is a part-time employee of Ontario Health. Nothing to declare for other authors. Data from this manuscript can be accessed upon request to Amol Verma ([email protected]) to the extent that is possible in compliance with local research ethics board requirements and data-sharing agreements. No project specific funding. Research ethics approval was obtained from all participating GEMINI hospitals, as per the study protocols approved by Unity Health Toronto’s REB (SMH REB: 20-216, CTO ID: 3344) and in full compliance with PHIPA and TCPS2. Not applicable.