Cardiac Troponin (hs-TnT) elevation has been reported in unselected patients hospitalized with COVID-19 however the mechanism and relationship with mortality remain unclear. Consecutive patients admitted to a high-volume intensive care unit (ICU) in London with severe COVID-19 pneumonitis were included if hs-TnT concentration at admission was known. Kaplan-Meier survival analysis performed, with cohorts classified a priori by multiples of the upper limit of normal (ULN). 277 patients were admitted during a 7-week period in 2020; 176 were included (90% received invasive ventilation). hs-TnT at admission was 16.5 (9.0 to 49.3) ng/L, 56% had concentrations >ULN. 56 patients (31.8%) died during the index admission. Admission hs-TnT level was lower in survivors (12.0 (8.0-27.8) vs 28.5 (14.0 to 81.0) ng/L, p = 0.001). Univariate predictors of mortality were age, APACHE-II Score and admission hs-TnT (HR 1.73, p = 0.007). By multivariate regression, only age (HR 1.33, CI: 1.16.to 1.51, p < 0.01) and admission hs-TnT (HR 1.94, CI: 1.22 to 3.10, p = 0.006) remained predictive. Survival was significantly lower when admission hs-TnT was >ULN (log-rank p-value<0.001). Peak hs-TnT was higher in those who died but was not predictive of death after adjustment for other factors. In conclusion, in critically ill patients with COVID-19 pneumonitis, the hs-TnT level at admission is a powerful independent predictor of the likelihood of surviving to discharge from ICU. In most cases, hs-TnT elevation does not represent major myocardial injury but acts as a sensitive integrated biomarker of global stress. Whether stratification based on admission Troponin level could be used to guide prognostication and management warrants further evaluation.
Coronavirus disease 2019 (COVID-19) is often associated with increases in cardiac Troponin concentration, particularly in those with advanced organ involvement. However, the clinical impact of raised cardiac Troponin levels in COVID-19 remains unclear, with recent studies coming to disparate conclusions regarding whether raised cardiac Troponin levels are an independent predictor of poorer patients outcomes. , Furthermore, it is unclear whether the association between cardiac Troponin and mortality is mediated by mechanistically significant myocardial injury as part of the infection and inflammatory response or if the heart is acting as an integrated sensor of global hypoxia and stress. We hypothesized that Troponin elevation does not reflect major myocardial injury but is a marker of global stress and addressed this by examining the prevalence and extent of Troponin elevation and the ability to predict survival in a well-characterized cohort of patients admitted to our intensive care unit (ICU) with advanced COVID-19 pneumonitis, in comparison to existing risk scores and established markers of risk.
Methods
Consecutive patients with COVID-19 pneumonitis, diagnosed according to the interim guidance of the World Health Organization , admitted to the ICU at Guy’s and St Thomas’ Hospital, London, United Kingdom, between 3rd March 2020 and 21 st April 2020 were included in the study if high-sensitivity Troponin-T (hs-TnT) concentration at admission was known.
All biochemistry analyses were performed onsite at dedicated institutional laboratory and Troponin T concentration was measured at admission and serially thereafter, using the Roche Elecsys assay; the limit of blank is 3ng/L, detection 5ng/L; and the coefficient of variation 10% at the limit of quantification; the 99th percentile for the normal population is 13ng/L.
Data were retrospectively collected from electronic medical records and validated through review of all source documentation. The following data were collected: patients’ demographics, medical history, laboratory and echocardiography investigations (outlined below). The APACHE (acute physiologic assessment and chronic health evaluation) II and SOFA (sequential organ failure assessment) scores were calculated on day of admission to ICU. Hyperinflammation was diagnosed by ICU team as per previously published criteria. Duration of ICU admission was calculated using date of death or discharge from ICU. Outcome data were verified from hospital records. Six-month mortality was censored using electronic patient records linked to national health service database. The primary outcome was all-cause mortality. The study was approved by institutional review board for use of de-identified data for COVID-19 related research.
A modified British Society of Echocardiography (BSE) Level 1 transthoracic echocardiogram (TTE) was performed during the index admission using GE E9, E95, and S70 ultrasound machines with a M5SC-D probe (GE Healthcare, Amersham, United Kingdom) or Philips CX50, Affinity and CVx ultrasound machines with an S5-1 or X5-1 probe (Philips Healthcare, Andover, Massachusetts). Retrospective analyses of the echocardiograms for this study were performed by BSE or EACVI accredited echocardiographers. The analyses included visual assessment of left ventricular (LV) and right ventricular (RV) systolic function, linear dimension measurements of the LV and RV, and Doppler analysis and measurements. Echocardiographic analysis was performed according to the joint American Society of Echocardiography (ASE) and European Association of Cardiovascular Imaging (EACVI) guidance.
Normality of data was assessed by the histogram, normal Q-Q plot and Shapiro-Wilk test. Continuous normal data are expressed as mean ± standard deviation (SD) and compared using paired Student t- tests. Non-normal data are expressed as median (interquartile range) and compared using Mann-Whitney test. Categorical variables are presented as frequency (percentages) and compared using chi-square test. Cox regression analysis was performed to identify potential predictors of all-cause mortality. A multivariate logistic regression analysis model with backward stepwise selection of variables (p-entry=0.05, p-exit=0.10) was constructed to identify potential predictors of all-cause mortality. The model included variables with univariate significance of p ≤ 0.15. Receiver operator characteristic (ROC) analysis was performed and areas under the curve (AUC) calculated to determine the relative ability to predict survival; Youden’s index was used to identify the optimal threshold.
Patients were classified by hs-TnT concentration at admission according to multiples of the upper limit of normal (ULN), as <ULN, 1-2 xULN or >2 xULN. Major myocardial injury was defined as peak hs-TnT >20 xULN. Kaplan-Meier curves were used to examine cumulative death rate and differences between groups, tested using a log-rank test. For all analyses, a P value of 0.05 was considered significant, and all p values were 2-sided. All statistical analyses were performed using SPSS version 26.0 (SPSS Inc. Chicago, Illinois) and GraphPad Prism version 9.0 (GraphPad Software, La Jolla, California).
Results
During the study period, 277 consecutive patients with confirmed COVID-19 were admitted to ICU, of whom 176 patients had hs-TnT measured at admission to the unit and comprised the study population (aged 55.1 ± 13.9 years; 71% male). 101 patients were excluded from the study as they had no recorded Troponin levels at admission – 58 were transferred from another ICU for advanced care and 43 did not have hs-TnT performed at admission to our ICU ( Figure 1 ). Patients excluded from the study had a higher rate of venovenous extracorporeal membrane oxygenation (VV-ECMO) (38% vs 3%, p < 0.001; these were largely tertiary/quaternary referrals from other ICUs). Demographic, biochemistry and echocardiographic characteristics of the included (study population) and excluded groups are detailed in Table 1 . Characteristics of the study population are summarized in Table 2 . 90% of the cohort received invasive ventilation and a third of all patients required renal replacement therapy. 56 (31.8%) patients died during ICU admission (“non-survivors” group) and 120 (68.2%) survived (“survivors” group). Those who survived were discharged from ICU 20.5 ± 18.1 days after admission. There was no difference between ventilated and non-ventilated patient in baseline characteristics, hs-TnT level or outcomes. At 6-months from admission, overall mortality was 33% in the entire study population.
| Variable | Overall (n=277) | Included (n=176) | Excluded (n=101) | p Value |
|---|---|---|---|---|
| Baseline Characteristics | ||||
| Age (years) | 55.1 ± 13.9 | 55.1 ± 12.9 | 52.3 ± 15.6 | 0.14 |
| Men | 196 (71%) | 124 (71%) | 72 (71%) | 0.88 |
| Body mass index (kg/m 2 ) | 28.5 (26.3) | 28.5 ± 6.6 | 28.7 ± 5.8 | 0.82 |
| Ethnicity | ||||
| White | 114 (41%) | 74 (42%) | 40 (40%) | |
| Black | 80 (29%) | 50 (27%) | 30 (30%) | |
| Asian | 40 (15%) | 26 (15%) | 14 (14%) | 0.38 |
| Mixed | 6 (2%) | 5 (3%) | 1 (1%) | |
| Other | 15 (5%) | 13 (7%) | 2 (2%) | |
| Not stated | 22 (8%) | 8 (5%) | 14 (14%) | |
| Diabetes Mellitus | 89 (32%) | 60 (34%) | 29 (29%) | 0.36 |
| Hypertension | 126 (46%) | 82 (47%) | 44 (44%) | 0.63 |
| SOFA Score | 6.0 (4.0 – 7.0) | 6.0 (4.0-7.3) | 5.0 (3.0 – 7.0) | 0.11 |
| APACHE II Score | 14.0 ± 5.0 | 14.0 ± 5.0 | 14 ± 5.1 | 0.95 |
| Hyperinflammation | 84 (30%) | 57 (32%) | 27 (27%) | 0.32 |
| Biochemistry characteristics | ||||
| Admission Hs-TnT (ng/L) | 16.5 (9.0-49.3) | 16.5 (9.0-49.3) | – | – |
| Peak Hs-TnT (ng/L) | 49.5 (16.5 – 115.5) | 53.00 (16.0 – 124.0) | 44.0 (24.5 – 111.0) | 0.57 |
| Admission CRP (mg/L) | 160.0 (84.3 – 278.5) | 162.0 (85.0-269.0) | 152.0 (77.0 – 281.5) | 0.51 |
| Peak CRP (mg/L) | 334.5 (226.0 – 414.8) | 344.0 (246.0 – 427.0) | 315.0 (196.0 – 395.0) | 0.50 |
| Admission Ferritin (μg/L) | 1164.0 (708.0 – 2182.0) | 1176.5 (717.8-2106.8) | 1118.0 (689.0 – 2459.5) | 0.96 |
| Peak Ferritin (μg/L) | 1919.0 (999.0 – 3379.3) | 1963.0 (10.45.5 – 3428.5) | 1627.0 (999.0 – 3342.5) | 0.45 |
| Echocardiographic characteristics | ||||
| LVEF, % (n=164) | 57.9 ± 11.1 | 56.7 ± 11.5 | 60.3 ± 9.6 | 0.02 |
| LV Impairment (≥Moderate) (n=164) | 19 (11%) | 17 (14%) | 2 (4%) | 0.20 |
| TAPSE (mm) (n=136) | 19.9 ± 4.7 | 20.0 ± 4.8 | 19.7 ± 4.4 | 0.75 |
| RV Impairment (n=136) | 30 (22%) | 21 (22%) | 9 (22%) | 0.93 |
| IVC (mm) (n=117) | 19.39 ± 4.84 | 19.7 ± 4.7 | 18.7 ± 5.1 | 0.35 |
| Pericardial Effusion (n=164) | 31 (19%) | 21 (18%) | 10 (22%) | 0.78 |
| Organ Support | ||||
| Non-invasive ventilation | 35 (13%) | 23 (13%) | 12 (11%) | 0.78 |
| Invasive ventilation | 238 (86%) | 158 (90%) | 80 (79%) | 0.02 |
| VV-ECMO | 25 (14%) | 4 (3%) | 21 (38%) | <0.001 |
| Renal Replacement Therapy | 89 (32%) | 58 (33%) | 31 (31%) | 0.70 |
| Outcomes | ||||
| Survivors | 191 (69%) | 120 (68%) | 71 (70%) | 0.48 |
| Length of stay (days) | 16.8 ± 15.7 | 18.5 ± 16.8 | 14.6 ± 14.1 | 0.06 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
