First sampled plasma troponin T is associated with one-year mortality in sepsis and septic shock, independent of age, gender, comorbidities, and clinically relevant biomarkers. (3rd October 2022)

• Record Type: Journal Article
• Authors: Loerstad, S; Shekarestan, S; Tehrani, S; Jernberg, T; Astrand, P; Gille-Johnson, P; Persson, J
• Is Part Of: European heart journal. Volume 43(2022)Supplement 2
• Publication Date: 2022-10-03
• Subjects: Cardiology -- Periodicals; Heart -- Diseases -- Periodicals; 616.12005
• DOI: 10.1093/eurheartj/ehac544.1486
• Languages: English
• ISSNs: 0195-668X

Abstract

Abstract: Background: Patients with sepsis and septic shock often have elevated troponin T levels indicating acute myocardial injury. Studies investigating the relationship between troponin T levels as a predictor for mortality in sepsis have been contradictory. Purpose: To investigate if the first in-hospital sampled plasma troponin T value is associated with one-year mortality in sepsis and septic shock independent of comorbidities, Sequential Organ Failure Assessment (SOFA)-score, Clinical Frailty Scale, Charlson Comorbidity Index (including age) and clinical biomarkers (e.g., serum lactate, creatinine, C-reactive protein). Methods: Patients with sepsis and sepsis shock were included if they were admitted to the intermediate and intensive care units at our institution between 1st of March 2012 and 30th of September 2021, and if at least one troponin T value was recorded within the first two calendar days from symptom onset. Variables were collected from medical records. Normal troponin-T was defined as <15 ng/L. Elevated troponin T values were divided into quartiles (Q): Q1 15–35.9 ng/L, Q2 36–61.9 ng/L, Q3 62–125.9 ng/L and Q4 ≥126 ng/L. Kaplan-Meier stratified survival curves were plotted for normal levels and Qs of elevated troponin T. Multivariate Cox regression was used for multivariate analysis. Results: 586 patients were included and 529 (90%) had a first sampled troponin T ≥15 ng/L. 523 (89%) met the Sepsis-3 criteria for septic shock, the remainder met the criteria for sepsis. The mean age was 71 years and 220 (38%) were women. One-year mortality was 45% (n=264). Increasing levels of first sampled plasma troponin T were associated with higher hazard ratios (HR) for one-year mortality compared to normal levels (<15 ng/L): Q1 HR 5.29 (95% confidence interval [CI] 1.9–14.7); Q2 HR 7.47 (95% CI 2.71–20.6); Q3 HR 12.81 (95% CI 4.7–35.0); Q 4 HR 13.82 (95% CI 5.06–37.7) (Figure 1). Troponin T was independently associated with one-year mortality after adjustment for gender, plasma lactate, plasma CRP, prior atrial fibrillation, in-hospital onset atrial fibrillation, SOFA-score, CFS, and CCI (including age) (Table 1). Conclusions: Elevated first sampled plasma troponin T in patients with sepsis and septic shock was associated with higher one-year mortality, independent of age, gender, comorbidities, and clinically relevant biomarkers. There was a risk gradient across the Qs of elevated troponin T with increasingly higher risk of death at one year per higher Q. Further studies are needed to ascertain whether elevated troponin T levels and clinical assessment scores can be used together in the clinical setting to identify patients for more aggressive treatment and cardiac work-up. Funding Acknowledgement: Type of funding sources: Public hospital(s). Main funding source(s): Region Stockholm