key: cord-1038262-xhj0b1lu
authors: De Marzo, Vincenzo; Di Biagio, Antonio; Della Bona, Roberta; Vena, Antonio; Arboscello, Eleonora; Emirjona, Harusha; Mora, Sara; Giacomini, Mauro; Da Rin, Giorgio; Pelosi, Paolo; Bassetti, Matteo; Ameri, Pietro; Porto, Italo; GECOVID study group,
title: Prevalence and prognostic value of cardiac troponin in elderly patients hospitalized for COVID-19
date: 2021-05-28
journal: J Geriatr Cardiol
DOI: 10.11909/j.issn.1671-5411.2021.05.004
sha: 29ab17e3b042204557dfa2c0eb41861314aa887b
doc_id: 1038262
cord_uid: xhj0b1lu

BACKGROUND: Increases in cardiac troponin (cTn) in coronavirus disease 2019 (COVID-19) have been associated with worse prognosis. Nonetheless, data about the significance of cTn in elderly subjects with COVID-19 are lacking. METHODS: From a registry of consecutive patients with COVID-19 admitted to a hub hospital in Italy from 25/02/2020 to 03/07/2020, we selected those ≥ 60 year-old and with cTnI measured within three days from the molecular diagnosis of SARS-CoV-2 infection. When available, a second cTnI value within 48 h was also extracted. The relationship between increased cTnI and all-cause in-hospital mortality was evaluated by a Cox regression model and restricted cubic spline functions with three knots. RESULTS: Of 343 included patients (median age: 75.0 (68.0−83.0) years, 34.7% men), 88 (25.7%) had cTnI above the upper-reference limit (0.046 µg/L). Patients with increased cTnI had more comorbidities, greater impaired respiratory exchange and higher inflammatory markers on admission than those with normal cTnI. Furthermore, they died more (73.9%vs. 37.3%, P < 0.001) over 15 (6−25) days of hospitalization. The association of elevated cTnI with mortality was confirmed by the adjusted Cox regression model (HR = 1.61, 95%CI: 1.06−2.52, P = 0.039) and was linear until 0.3 µg/L, with a subsequent plateau. Of 191 (55.7%) patients with a second cTnI measurement, 49 (25.7%) had an increasing trend, which was not associated with mortality (univariate HR = 1.39, 95%CI: 0.87−2.22, P = 0.265). CONCLUSIONS: In elderly COVID-19 patients, an initial increase in cTn is common and predicts a higher risk of death. Serial cTn testing may not confer additional prognostic information.

Among these CV abnormalities, particular emphasis has been given to an increase in cardiac troponin (cTn) concentrations. During COVID-19, cardiac stress and damage may arise because of a variety of mechanisms, including type 2 ischemia, hypoxia, sepsis and systemic inflammation, pulmonary thrombosis and embolism, cardiac adrenergic hyperstimulation during cytokine storm syndrome, and myocarditis. [2, 8] A rise in cTn may be also due to pre-existing cardiac disease and concomitant comorbidities. [9] Irrespective of the underlying causes, evidence of cardiac injury at the time of admission for COVID-19 has been associated with a more severe clinical course and higher mortality. [2, 6, 8, 10] Limited data also indicate that a rising trend of cTn levels during the hospitalization identifies a subset of COVID-19 patients with worse outcome. [10, 11] Nonetheless, the specific impact of cTn measurement has not been investigated yet in the elderly. This lack of evidence has potential practical implications, since COVID-19 patients older than 60 years of age need hospitalization more often than younger ones, [12] [13] [14] and thereby, are more likely to be tested for cTn.

The aims of this study were to evaluate the relationship between cardiac injury, as demonstrated by cTn elevation either at baseline or on a second measurement within 48 h, and all-cause in-hospital mortality in ≥ 60 year-old patients hospitalized for COVID-19.

This is a retrospective analysis of a prospective registry enrolling all consecutive patients diagnosed with COVID-19 in a hub hospital in Genova, Italy, from February 25 th to July 3 rd , 2020. Genova is the main city of an Italian region with an overall old population (around 1,500,000 inhabitants, with 35.6% being ≥ 60-year-old).

The registry was developed by modifying an established registry of patients with infectious diseases, [15, 16] was approved by the local Ethics Committee (study number 163/2020) and contained anonymized data; all capable subjects gave written in-formed consent to the use of such data for research purposes. SARS-CoV-2 infection was confirmed by reverse transcriptase-polymerase chain reaction (RT-PCR) of pharyngeal swabs or bronchoalveolar aspirates. Laboratory exams and diagnostic procedures were performed as per standard clinical practice.

The study sample consisted of patients aged ≥ 60 years with cTnI measured within three days from the molecular confirmation of SARS-CoV-2 infection. A second cTnI measurement within 48 h from the first one was available for a subset. The 60-year age cut-off was chosen according to the general agreement that ≥ 60 year-old COVID-19 patients represent a distinctive population with specific features. [12] [13] [14] 

For every patient, the following information was retrieved: age, gender, Charlson comorbidities index (CCI), prior myocardial infarction (MI), history of chronic heart failure (CHF), and presence of hypertension, atrial fibrillation (AF), neurological disorder, chronic obstructive pulmonary disease (COPD), diabetes, cancer or chronic kidney disease (CKD), need of non-invasive or invasive ventilation, and admission to the intensive care unit (ICU), as reported in the medical records.

We also assessed the clinical features on admission, including laboratory exams. Plasma cTnI concentration was measured using a sandwich chemiluminescent immunoassay based on LOCI ® technology on Dimension Vista ® 1 500 System. The limit of quantitation (functional sensitivity), which corresponds to the cTnI concentration at which the coefficient of variation is 10%, was < 0.04 μg/L. [17] The upper-reference limit (URL), as defined at the 99 th percentile of the reference interval, was 0.046 μg/L.

All-cause in-hospital mortality was ascertained by review of the medical records.

Categorical variables are presented as frequencies and percentages and were compared by chisquare test or Fisher's exact test. Continuous variables are reported as mean ± SD or median and interquartile range according to their distribution. Normally distributed variables were compared by means of unpaired Student's t test and non-nor-mally distributed ones with the Mann-Whitney U non-parametric test.

For those patients for whom a second cTnI determination was available, the trend between the second and the first measurement was categorized as increase or non-increase, depending on whether the difference between the two values was > 0 or ≤ 0.

Time to all-cause in-hospital death was graphically depicted using the Kaplan-Meier method and compared by log-rank test. Patients were right-censored if they were discharged from the hospital alive or were still hospitalized at the time of data extraction (July 3, 2020). A Cox regression model was used to estimate the hazard ratios (HRs) with 95% confidence interval (CI) of all-cause in-hospital mortality according to cTnI values below or above the URL. The model was adjusted for clinically meaningful covariates that were different between dead and alive patients with P < 0.05.

A potentially non-linear relationship between admission cTnI and all-cause in-hospital mortality was tested by using restricted cubic spline functions with three knots; data were then displayed graphically. As a sensitivity analysis, fitting a proportional sub-distribution hazards regression to the same variables included in the Cox regression model, we performed a competing risk analysis in which discharge from the hospital was treated as a competing risk for all-cause in-hospital mortality.

All analyses were performed with R environment 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria) and packages finalfit, survival, gg-plot2, survminer, rms, and cmprsk.

A total of 1 275 consecutive patients admitted during the study period were included in the registry. Of them, 468 had cTnI measured within three days from the RT-PCR for SARS-CoV-2. Onehundred twenty-five subjects were excluded from the analysis because they had < 60 years of age, leaving a final sample of 343 patients. Their baseline characteristics are shown in Table 1 . Median age was 75.0 (68.0−83.0) years and 119 (34.7%) were men. One-hundred ninety-eight (58.1%) patients had hypertension and 17.6% diabetes. Based on medical history, the prevalence of cardiovascular comorbidities was around 10%. Most patients presented with fever and around half had dyspnoea (Table 1) . Median oxygen saturation (SpO 2 ) and arterial oxygen partial pressure (pO 2 ) were 94.0% (90.0%−97.0%) and 67.1 (55.0−84.3) mmHg, respectively. Inflammatory biomarkers were elevated (Table 1) .

Median cTnI was 0.02 (0.02−0.05) μg/L; 88 (25.7%) patients had a cTnI value above the URL. These latter were older and had more often AF, CKD and a neurological disorders than the subjects with normal cTnI (Table 1) . Although the frequency of dyspnoea was not different between the two groups, patients with cTnI above the URL presented with lower SpO 2 and arterial pO 2 . Concentrations of creatinine, aspartate transaminase (AST), bilirubin, inflammatory parameters, D-dimer, creatine phosphokinase (CPK), and international normalized ratio (INR) were higher in subjects with baseline cTnI above the URL, whilst haemoglobin levels were lower (Table 1) .

Overall, 28.7% and 17.6% patients received noninvasive and mechanical ventilation, respectively; 18.8% were admitted to the ICU. The frequencies of non-invasive and invasive ventilation support, as well as of ICU admission, were non-significantly lower in the group with elevated cTnI (Table 1) .

During a median hospital stay of 15 (6−25) days, 160 (46.6%) patients died (Supplementary Table 1 ). All-cause mortality was higher in patients with increased admission cTnI (65 deaths/88 patients, 73.9% vs. 95 deaths/255 patients, 37.3%, P < 0.001).

Kaplan-Meyer curve showed that patients with increased cTnI survived less throughout the hospitalization than those with normal cTnI (Figure 1 ). The association between baseline cTnI and all-cause in-hospital mortality was confirmed by the adjusted Cox regression model (HR = 1.61, 95%CI: 1.06−2.52, P = 0.039) ( Table 2 ). The sensitivity competing risk analysis yielded results consistent with the Cox regression model (HR for admission cTnI: 1.46, 95%CI: 1.11−2.93, P = 0.043) (Supplementary Table 2 ).

As shown in Figure 2 , the association of admission cTnI with all-cause in-hospital mortality began with concentrations within the range of normality as per manufacturer's indications and was linear until the threshold of 0.3 μg/L, after which a plateau was observed with no further increase in mortality.

A second cTnI measurement within 48 h was available for 191 (55.7%) patients, the median value being 0.02 (0.02−0.05) μg/L; an increasing trend was found in 49 (25.7%) of them. The concentrations of cTnI were higher at both the first (0.03 (0.02−0.06) 

In this study, we show that an initially increased cTn value portends a higher risk of in-hospital mortality in subjects older than 60 years with COVID-19. In the population examined, an increased cTnI concentration within three days from the molecular diagnosis of SARS-CoV-2 infection conferred a 60% higher risk of all-cause in-hospital death. By contrast, a further elevation in cTn over the following 48 h did not provide additional prognostic information. CV involvement is common in COVID-19 and has been pointed out as one of the factors contributing to the dismal prognosis that many patients face. [1] The elevation of markers of myocardial injury, especially cTn, has drawn much attention, since it can be readily assessed and is clinically relevant. [2, 4, 7, 8] Several authors have reported that an increase in cTn concentrations portends a higher risk of in-hospital death. [3] [4] [5] [6] 10, 11, [18] [19] [20] This evidence has been gathered by analyzing various cohorts with important differences, for instance in the severity of COV-ID-19 and, thus, the intensity of treatments, in ethnicity and in the burden of comorbidities. [9] Therefore, it is assumed that the results of these studies can be generalized to all patients hospitalized for COVID-19. [8] However, data about the value of cTn specifically in elderly subjects admitted for COVID-19 are scarce. This lack of information is remarkable, considering the epidemiology of the SARS-CoV-2 pandemic, in which old individuals are the most affected and often need hospitalization. [13, 21, 22] Our results confirm that cTnI elevation is an independent predictor of in-hospital mortality in elderly patients hospitalized for COVID-19, like it is in younger ones. The cohort we investigated was particularly old. In fact, the median age was 75 years, whilst the mean or median age in the other investigations of cTn in COVID-19 was ≤ 70 years (Supplementary Table 3 ). Furthermore, almost 1 in 2 patients died during the hospitalization. A similar proportion of deaths has been described for critically ill COVID-19 Italian patients with a median age of 63 (56−69) years. [23] We believe that the comparable mortality of our patients, who were admitted to the ICU only in about 20% of cases, is explained by the substantially older age. It is notable that, even within this vulnerable population, an initial increase in cTn identified a frailer group with worse prognosis. The mechanisms leading to cTn elevation during COVID-19 are manifold and likely often concomitant. [24] Autopsy studies indicate that myocardial ischemia due to plaque rupture, coronary artery spasm or direct injury, or microthrombi are rare in COVID-19, [25] and the pathogenesis of cardiac damage has been primarily ascribed to other events. [4, 9, 10, 19] First, the heart may suffer from severe hypoxia in the contest of acute respiratory insufficiency. Second, hyperinflammation and sepsis with cytokine storm may directly affect cardiomyocytes, up to causing stress cardiomyopathy or myocarditis. Third, cardiac injury may develop following pulmonary thromboembolism. Finally, SARS-CoV-2 can localize to the myocardium, although an ensuing inflammatory infiltrate has not been demonstrated. [18] The elderly may be more prone to all these events because of the reduced resistance of the aging heart to stressors and of concomitant asymptomatic or overt cardiac disease. Other authors found a continuous relationship between cTn concentrations and mortality. [6] Moreover, in subjects admitted for COVID-19 younger than those evaluated by us, the trend between two measurements of cTn obtained at the beginning of the hospitalization may help better stratifying the risk of death. [21] By contrast, in our cohort, the troponinmortality risk curve plateaued after a relatively low level and the changes in cTn did not refine prognostication. Thus, our analysis suggests that the presence, but not the entity of cardiac injury corresponds to worse outcomes in elderly patients hospitalized for COVID-19, and that multiple determinations of cTn may be of value only if clinically motivated.

The retrospective design is the main limitation of this work. However, most of the literature about the CV complications in COVID-19 is based on data collected retrospectively. Moreover, in Italy the most intense phase of the SARS-CoV-2 epidemic, when the majority of patients was hospitalized, had a relatively brief course, making the conduct of prospective studies very challenging. The possibility of a selection bias must be also acknowledged, since only part of the subjects included in the registry we analysed had cTnI measured.

In conclusion, an initial elevation in cTnI is associated with all-cause in-hospital mortality in elderly patients admitted for COVID-19. Beyond the relatively low threshold of 0.3 μg/L, additional increases in the cTnI concentrations do not confer a higher risk of mortality, indicating that the rise in cTnI in itself, rather than its entity, is marker of worse outcome. An increasing trend in cTnI levels on a second assessment does not give further information about prognosis.

Transmission, Diagnosis, and Treatment of Coronavirus Disease 2019 (COVID-19): A Review

COVID-19 pandemic and troponin: indirect myocardial injury, myocardial inflammation or myocarditis?

Cardiovascular manifestations and treatment considerations in COVID-19

Potential Effects of Coronaviruses on the Cardiovascular System: A Review

Clinical predictors of mortality due to COVID-19 based on an analysis of data of 150 patients from Wuhan, China

Association of cardiac injury with mortality in hospitalized patients with COV-ID-19 in Wuhan, China

Cardiovascular complications in COVID-19

Clinical utility of cardiac troponin measurement in COVID-19 infection

Cardiac Troponin for the Diagnosis and Risk-Stratification of Myocardial Injury in COVID-19: JACC Review Topic of the Week

Prevalence and Impact of Myocardial Injury in Patients Hospitalized With COVID-19 Infection

Longitudinal correlation of biomarkers of cardiac injury, inflammation, and coagulation to outcome in hospitalized COVID-19 patients

Why does COVID-19 disproportionately affect older people?

COVID-19 in older adults

COV-ID-19 and Older Adults: What We Know

From Liguria HIV Web to Liguria Infectious Diseases Network: How a Digital Platform Improved Doctors' Work and Patients' Care

Clinical characteristics, management and in-hospital mortality of patients with COVID-19

Cardiac Troponin Assays: Guide to Understanding Analytical Characteristics and Their Impact on Clinical Care

Cardiovascular Implications of Fatal Outcomes of Patients With Coronavirus Disease 2019 (COVID-19)

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

Association of Troponin Levels With Mortality in Italian Patients Hospitalized With Coronavirus Disease 2019: Results of a Multicenter Study

COVID-19 in older people: a rapid clinical review

SARS-CoV-2 and COVID-19 in older adults: what we may expect regarding pathogenesis, immune responses, and outcomes

Risk Factors Associated With Mortality Among Patients With COVID-19 in Intensive Care Units in Lombardy, Italy

Elevated Troponin in Patients With Coronavirus Disease 2019: Possible Mechanisms

Histopathology and ultrastructural findings of fatal COVID-19 infections in Washington State: a case series

Prevalence and prognostic value of cardiac troponin in elderly patients hospitalized for COVID-19