Prevalence and Prognostic Value of Myocardial Injury in the Initial Presentation of SARS-CoV-2 Infection among Older Adults

Myocardial involvement during SARS-CoV-2 infection has been reported in many prior publications. Data about this condition in older adults is scarce especially its role in clinical prognosis. We aim to study the prevalence and the clinical implications of acute myocardial injury (MIN) during SARS-CoV-2 infection, particularly in older patients.MethodsLongitudinal observational study where all consecutive adult patients admitted to a COVID-19 unit between March to April 2020 were included. Those patients aged ≥65 were considered as older patients. MIN was defined as at least 1 high-sensitive troponin (hs-TnT) concentration above the 99 th percentile upper reference limit with different sex-cutoff.ResultsAmong the 634 patients admitted during the period of observation 365(58%) had evidence of MIN (hs-TnT>14 pg/mL), and among those 224(61%) were older adults. Individuals with acute MIN were more prone to be older, had more comorbidities, more functional decline at admission, and higher inflammatory parameters. Among older adults, MIN was associated with longer time to recovery compared to those without MIN [13 days(IQR 6-21) vs 9 days(IQR 5-17);p<0.001 repectively. In-hospital mortality was significantly higher in older adults with MIN at admission vs those without MIN [71(31%) vs 11(12%);p<0.001]. In a logistic regression model adjusting by age, sex, severity and Charlson comorbidity index the OR for in-hospital mortality was 2.1 (95% CI:1.02-4.42;p=0.043) among those older adults with MIN at presentation.ConclusionMIN is frequent in individuals with SARS-CoV-2 infection, especially in older adults and in patients with pre-existing comorbidities and with higher inflammatory levels. Older adults with acute myocardial injury had greater time to clinical recovery, more severe presentation of the disease and higher odds of in-hospital mortality.


Introduction
Myocardial involvement has been reported as one of the clinical presentations of individuals with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection (1,2). Cardiological symptoms associated with coronavirus diseased-19  caused by SARS-CoV-2 may include a wide range of presentations from a troponin elevation re ecting an underlying non-ischemic myocardial injury to a myocardial infarction (3,4). The diagnosis of myocardial injury (MIN) in hospitalized patients has signi cantly increased with the widespread use of high-sensitivity troponin (hs-TnT) assay. A precise de nition of MIN is needed in order to distinguish it from a myocardial infarction. Whereas myocardial infarction is a form of MIN that requires clinical or electrical evidence of myocardial ischemia and necrosis, MIN is de ned as at least 1 cardiac troponin concentration above the 99th percentile upper reference limit (5). Although this condition appears to be frequent during the current COVID-19 outbreak (6), the clinical relevance of MIN remains to be determined. It is keystone to understand whether acute MIN during the acute COVID-19 episode is a bystander epi-phenomenon or has a relevant clinical role in the outcome of the infection.
A consistent risk for myocardial injury MIN is age (3) and during the rst COVID-19 wave the largest proportion of patients were over 65 years old (7). All of this makes this group of patients a target population to study the impact of myocardial injury MIN on SARS-CoV-2 infection.
Aging has a well-known detrimental impact on cardiac structure and vasculature. SARS-CoV-2 infection might cause a worsening of clinical conditions through different mechanisms, such as direct myocardial injury and endothelial binding, T cell death, and increased in ammation (3). Considering that during the rst wave the largest proportion of patients suffering of COVID-19 were over 65 years old (7), this group of patients results an interesting target population to study the impact of MIN on SARS-CoV-2 infection.
However, no consistent data looking into this speci c issue has been reported in older adults with SARS-CoV-2 infection. Increasing evidence shows the association between COVID-19 and MIN along with increased mortality and overall a broader signi cance of this form of cardiac injury in COVID-19 outcomes (2). Despite being a prevalent and troublesome disease in this speci c population, there are currently no large studies that deal with the manifestations of COVID-19 in this particular setting.
On the basis of our clinical observation in our cohort along with the available evidence, we hypothesized that SARS-CoV-2 might have early effect on the heart of older adults, presenting mainly as MIN since the very initial stages of infection. Accordingly, we aim to describe the prevalence and the clinical characteristics and prognosis of acute MIN as onset manifestation in a large cohort of older patients with con rmed SARS-CoV-2 infection.

Methods
This is a retrospective observational study conducted at Hospital del Mar in Barcelona, Spain. This hospital provides healthcare to an area up to 500,000 people. During the rst wave of the pandemic the hospital was converted to a COVID-19 hospital with a COVID-19 unit that was equipped with 400 beds for in-hospital admission and with 80-beds for critical care. There is an electronic medical record and a centralized registry of all individuals admitted to the COVID-19 unit. For this study, we included all patients admitted to the COVID-19 unit for ≥ 48 hours between March 9th and April 1st 2020.
Admission criteria to the unit was having a con rmed SARS-CoV-2 infection. This was by having a positive real-time polymerase chain reaction (rt-PCR) for SARS-CoV-2 in nasopharyngeal samples, obtained by trained personnel at hospital admission and ful lling clinical diagnostic criteria (respiratory symptoms such as dyspnea, cough, sore throat, changes in taste/smell; or uni-/bilateral interstitial in ltrates in chest X-Ray).
All subjects aged 65 and above with de nite SARS-CoV-2 were included in the study. As admission protocol all patients admitted to the COVID-19 unit underwent a blood analysis with at least C-reactive protein, D-dimer, hs-TnT. Individuals with other conditions that could induce acute MIN and troponin rise were excluded: severe bacterial sepsis and other severe active infections, individuals with acute myocardial infarction (electrocardiographic changes suggesting acute ischemia and elevated myocardial ischemia markers), active known vasculitis, medications and substance intoxication. All eligible participants ≥65 years were included and divided for comparison purposes between those with MIN and without it. We considered as a reference population the included younger adults (< 65 years old).
Clinical variables, data source, and study outcomes Demographic, clinical and epidemiological data, as well as the information from the episode (laboratory workout, vital signs, treatment), were extracted from the electronic medical record using standardized data collection. Acute myocardial injury was de ned by an hs-TnT value greater than the institutional upper limit of normal (14 pg/mL), without electrocardiographic changes that suggest acute ischemia (5). Clinical severity was assessed at admission with MEWS score (8, 9). Laboratory workups were systematized with an at-admission protocol that included a fasting blood draw with full blood count, electrolytes, renal and liver function, cardiac biomarkers (high-sensitivity troponin T (hs-TnT), N-terminal-proB-type natriuretic peptide (NT-proBNP) and lactate dehydrogenase (LDH)), in ammatory markers (Creactive protein (CRP), interleukin-6 (IL-6), serum ferritin, and coagulation testing (D-dimer). Comorbidity was assessed using Charlson Comorbidity Index (10) and also categorized in no comorbidities, mild (1-2 comorbidities), severe (≥3 comorbidities). A follow-up analysis was done to study the peak in troponin, and in ammatory markers.
Key outcomes included time to clinical stability (de ned as the time the time elapsed since the patient's admission to: oxygen saturation > 94% (FiO2 21%), normalized level of consciousness (baseline), HR < 100 rpm, systolic BP > 90 mm Hg, Temperature < 37.2ºC), admission to a critical care unit, or in-hospital mortality.
Acute myocardial injury was de ned considering sex-speci c cutoff for hs-TnT value greater than the institutional upper limit of normal, that is 9 ng/l for female and 16 ng/l for male, without electrocardiographic changes that suggest acute ischemia.

Ethics considerations
The Institutional Ethics Committee of Hospital del Mar of Barcelona approved the study and due to the nature of the retrospective data review, waived the need for informed consent from individual patients (CEIm 2020/9352).

Statistical analysis
Continuous variables are expressed as medians and interquartile range (IQRs). Categorical variables are expressed as frequencies (percentages). Continuous variables were compared using the Student t-test or the Mann-Whitney U test, as appropriate, and categorical variables using χ2 test or the Fisher exact test, as appropriate.
Spearman correlation was used to test the association between in ammatory markers, and pro-NT-BNP and hs-TnT. A multivariate linear regression analysis was carried out to explore independent factors associated with hs-TnT levels. Those with p < 0.05 in univariate analysis were introduced into the multivariate linear regression model.
To investigate the impact of factors on myocardial injury, logistic regression analyses were carried out to determine odd ratios (ORs) and 95% con dence intervals (CI) for covariates with myocardial injury (hs-TnT > 14) as the bivariate outcome. We considered age and gender, daily habit, comorbidities (pre-existing coronary artery disease, hypertension, diabetes mellitus or cerebrovascular disease), lab tests, and severity of the episode (MEWS score) and days with symptoms before admission for candidate variables in both regression models. An alternative model was tted to study in-hospital mortality in individuals with acute myocardial injury. All statistical analyses were performed using STATA/MP V.14.0, and a twosided p value of < 0.05 was considered statistically signi cant.

Results
Among the 634 patients admitted during the period of observation 365 (58%) had evidence of MIN (hs-TnT>14 pg/mL), and among those 224 (61%) were older adults. The population and subgroups study are shown in gure 1.
Older adults with and without acute myocardial injury Baseline characteristics of older adults with and without acute MIN were different (Table 1). Compared to those without MIN, older adults with cardiac injury were more likely to be older [mean age of 83 years (75-88) vs 74 (68-78); p<0.001] and male sex (48% vs 33%; p=0.02), with a higher prevalence of pre-existing cardiovascular disease (diabetes mellitus, hypertension, pre-existing heart disease) and with a clinical presentation more frequently with consciousness impairment and signi cantly less frequently with cough or fever, and no differences in prior to admission duration of symptoms (table1). Remarkably, older adults with MIN had signi cantly higher mortality rates than those without MIN [71 (31%) vs 11 (12%); p<0.001].  In a logistic regression model adjusting by age, sex, severity and Charlson comorbidity index the OR for in-hospital mortality was 2.1 (95% CI 1.02-4.42; p=0.043) among those older adults with acute myocardial injury at presentation (Table 3).

Discussion
We report worse prognosis among older adults presenting with MIN, even without myocardial necrosis, during the acute episode of SARS-CoV-2 infection. MIN led to a worse prognosis with more severe clinical presentation, prolonged time to recovery, and higher mortality rates among older adults, as had been reported in other series (11,12). We found age, male gender and pre-existing cardiovascular disease to be key factors associated with higher prevalence of MIN. These are also well established risk factors for adverse events in general in outbreaks of respiratory virus infections (13) (14). Nonetheless, when controlling for these possible confounders, MIN was still associated with worse clinical prognosis.
Recently, has been shown that high sensitivity troponin I improves mortality and cardiovascular risk strati cation in older adults beyond traditional risk factors (15). In our study we found similar results in the SARS-CoV-2 infection with hs-TnT.
Based on the accepted de nition of MIN and myocardial infarction (5), we considered as acute MIN cases where there was a signi cant increase in hs-TnT with an absence of electric signs of ischemia or myocarditis. Remarkably, we found this to be as frequent as 50% of our overall cohort, and up to 70% when looking into older adults. In other studies, acute MIN has been reported as less frequent, in up to 12% of patients with . The difference between our study and other prior reports may be explained because we are only considering cases that required hospitalization, which presumably means they have more severe forms of the disease, meanwhile prior studies included mixed series with inhospital and out-patients (16). Another additional reason is that the median age of our cohort was higher compared to previous studies (12).
It would be interesting to see if this myocardial injury also results signi cantly in myocardial dysfunction.
What we have seen in our series is that individuals presenting with myocardial injury at admission had heart failure more frequently, both by clinical and biochemical signs measured by NT-proBNP.
Unfortunately, we were not able to measure ejection fraction during the episode, due to unbelievable pressure on healthcare in the acute moment of the pandemic. All the same, with our data we are inclined to assess that there is an impairment of function as well.
Older adults showed signi cantly higher mortality than those younger counterparts, showing probably that the differences are not only because of MIN but with all the other comorbidities together added to age. However, there is evidence that show that patients with MIN are also more prone to die. There are many hypotheses trying to discern the causes of troponin elevation during SARS-CoV-2 infection, and also the role it plays in the evolution of the disease (6). An interesting hypothesis is that cardiac injury may re ect an ongoing pathological insult due to in ammation or secondary to hypoxemia (19). In our series, we found no differences between older adults with or without MIN in terms of the ratio of arterial oxygen partial pressure to fractional inspired oxygen (PaFi) at admission, moving us away from the theory that it is related to hypoxemia. On the other hand, we did nd a positive correlation between all in ammatory markers that we studied and troponin levels, and have found consistently that individuals with MIN showed signi cantly higher in ammatory markers that those individuals without MIN. It also goes in consonance with the fact that older adults would be a more vulnerable population due to increased underlying in ammation secondary to the ageing of the immune system (3). Moreover, an increased immune response due to SARS-CoV-2 infection is one of the hallmarks of COVID-19. Acute myocardial events could be associated with in ammation triggered by SARS-CoV-2, as it has been suspected in other scenarios such as cytokine release syndrome, which seems to increase troponin after antigen receptor T-cell therapy and cytokine storm (6).
Ageing is associated with an immune alteration that has being characterized. Brie y it consists in the reduction in the number of peripheral blood naïve cells, with a relative increase in the frequency of memory cells. This leads to a dysregulation of immune control and an age increased in ammation called in amm-aging, altogether are considered the hallmarks of immunosenescence (20). An older immune system is de ned by these cells presenting a speci c phenotype and responding abnormally to some external agents (19,20). It has been reported that the acute infection of SARS-CoV-2 leads to an expansion of precisely immunosenescent cells (21). Altogether leads us to entertain the hypothesis that may be an additive effect between the SARS-CoV-2 infection and age, where a more senescent immune system leads to increased tissue damage, including myocardium.
Our study has some limitations since it was conducted in a single center in the rst wave of the COVID-19 outbreak. It has a mid-size sample and we were not able to conduct cardiac ultrasound or MRI, as a consequence, we were only able to de ne acute myocardial injury by troponin elevation without detailing myocardial tissue characteristics and haemodynamic function. Moreover, during this rst wave many older adults with SARS-CoV-2 infection remained in long term care facilities. In that case we can say that our hospital was the reference center for many long term care facilities of the area and a uid drainage of these long term care facilities even in the moments with highest incidence. Moreover, we nd that a key strength of our study is how representative it is of an older adult population, belonging to a large city that has had high incidence of SARS-CoV-2 infection, which makes the results directly relevant to the clinical setting.
We can conclude that MIN was very common in individuals with SARS-CoV-2 infection, especially in older adults and in patients with pre-existing comorbidities and with higher in ammatory levels. We can also conclude that it impacted the clinical outcomes of individuals that experienced it, being associated with greater time to clinical recovery, more severe presentation of the disease and higher odds of in-hospital mortality. The consistent association that we found between in ammation and MIN makes the hypothesis of an in ammatory insult as responsible to heart damage plausible. This is more relevant in a group with higher in ammatory levels due to immune dysregulation linked to ageing and maybe deserves further attention. Ultimately, due to its widespread presence, and its likely role in prognosis, it is advisable that we direct attention to this matter.

Declarations
Ethics approval and consent to participate  Flowchart of the patients included in the study.