Cardiac Troponin T and N-Terminal Pro-BNP in Heart Failure
Cardiac Troponin T and N-Terminal Pro-BNP in Heart Failure
This study provides a comprehensive analysis of the prognostic value of hs-cTnT (a marker of myocardial damage), alone or in combination with NT-proBNP (a marker of myocardial stretch), in a real-life cohort of chronic HF patients. Both biomarkers improved risk stratification for death above and beyond a model with eleven well established risk factors.
Our study findings are in agreement with previous reports that assessed the relationship between hs-cTnT and clinical variables. There were remarkable findings in this cohort. First, hs-cTnT levels increased very significantly with the severity of HF (NYHA class), suggesting ongoing myocardial damage and progression of HF in sicker patients. Second, although subgroup analysis should be interpreted with caution, the high hs-cTnT levels observed in hypertensive cardiomyopathy came as a surprise. However, Setsuta et al previously reported that elevated cTnT in hypertensive patients was an important predictor of future cardiovascular events. One possible hypothesis is that subendocardial ischemia caused by hypertensive left ventricular hypertrophy drives myocyte injury, resulting in higher levels of hs-cTnT and ultimately patchy fibrosis. In the general population, even in asymptomatic individuals, high hs-cTnT levels were predictive of future cardiovascular events and correlated with structural heart disease. Finally, the association between cTnT and chronic kidney disease is a consistent finding. Detectable cTnT levels by means of conventional assays in patients with end stage chronic kidney disease are associated with a poor prognosis, even in the absence of coronary heart disease. The clearance and degradation of cTnT remains undefined. However, in a small study, Tsutamoto et al demonstrated a significant correlation between eGFR and serum cTnT levels in HF patients, suggesting that decreased cTnT clearance could contribute to elevated troponin levels in these patients.
The mechanisms of troponin release in HF are not well established, and several processes are likely involved. Although higher troponin levels were observed in patients with HF of ischemic etiology, it has been consistently reported that patients with non-ischemic HF also have elevated troponin levels. Multiple mechanisms may be involved, such as subendocardial ischemia due to increased transmural wall stress and stiffening of the myocardium, myocyte necrosis (induced by ischemia, inflammation, and oxidative stress), myocyte apoptosis, cellular release of proteolytic troponin degradation products, and increased cellular wall permeability because of reversible injury.
Several studies have demonstrated a consistent association between cTnT elevation and prognosis in acute and chronic HF using conventional assays. Latini et al first evaluated the prognostic value of very low cTnT levels using a precommercial version of the hs-cTnT assay in patients enrolled in the Valsartan Heart Failure Trial. Ninety-two percent had detectable hs-cTnT levels, and the risk of death and HF hospitalization increased seven- to eight-fold across increasing deciles of hs-cTnT, and remained strongly associated with these outcomes after adjustment for standard risk predictors and BNP levels. These authors used 12 ng/L (the median value in their population) as the cut-off. Two additional studies that evaluated hs-cTnT in chronic HF, both small studies with limited follow-up, used the upper reference limit of the assay (between 10 and 15 ng/L) to define elevated hs-cTnT levels. In this cohort (a large, prospective, real-life, ambulatory HF population followed for 41 months), the median value of hs-cTnT was 22.6 ng/L. Nevertheless, the optimal cut-off (set at 16 ng/L), was obtained using state-of-the-art statistics by bootstrapping the value that maximized the log-likelihood of the non-adjusted Cox models. This novel approach provided a more precise cut-off for prognostic purposes. To the best of our knowledge, this is the first study in HF that uses this method to select more accurate biomarker cut-off points.
NT-proBNP is well recognized as an important prognostic biomarker in HF. However, beyond natriuretic peptides, the use of other biomarkers for risk assessment is being debated. In this study, the predictive accuracy of hs-cTnT was even higher than that of NT-proBNP according to comprehensive discrimination, calibration, and reclassification analyses. However, the combination of both biomarkers was associated with a substantially higher risk compared with either biomarker alone, reaching a very significant HR of 7.42. Above their respective cut-off points, both biomarkers allowed us to identify a very high-risk subgroup of HF patients with a 5-year predicted survival of 28% (compared with 86% survival for both biomarkers below their respective cut-off points) as assessed by Kaplan-Meier.
There is a risk that the absolute levels of hs-cTnT could have been affected by having been measured from frozen rather than fresh samples. There is little information about long-term stability of frozen hs-cTnT. We have analyzed only one blood sample per patient and cannot comment on the prognostic value of serial determinations. The use of bootstrap method to determine the cut-off points for NT-proBNP and hs-cTnT allows to optimize the prognostic prediction but limits its comparison with other analyses.
Our population was a general HF population treated at a specific and multidisciplinary HF unit in a tertiary hospital, and most patients were referred from the cardiology department, resulting in relatively young men with HF of ischemic etiology and reduced LVEF. As such, the obtained results cannot necessarily be extrapolated to a global HF population.
Hs-cTnT provides significant prognostic information in a real-life cohort of patients with chronic HF. The simultaneous addition of hs-cTnT and NT-proBNP into a model that includes established risk factors improves mortality risk stratification.
Discussion
This study provides a comprehensive analysis of the prognostic value of hs-cTnT (a marker of myocardial damage), alone or in combination with NT-proBNP (a marker of myocardial stretch), in a real-life cohort of chronic HF patients. Both biomarkers improved risk stratification for death above and beyond a model with eleven well established risk factors.
Our study findings are in agreement with previous reports that assessed the relationship between hs-cTnT and clinical variables. There were remarkable findings in this cohort. First, hs-cTnT levels increased very significantly with the severity of HF (NYHA class), suggesting ongoing myocardial damage and progression of HF in sicker patients. Second, although subgroup analysis should be interpreted with caution, the high hs-cTnT levels observed in hypertensive cardiomyopathy came as a surprise. However, Setsuta et al previously reported that elevated cTnT in hypertensive patients was an important predictor of future cardiovascular events. One possible hypothesis is that subendocardial ischemia caused by hypertensive left ventricular hypertrophy drives myocyte injury, resulting in higher levels of hs-cTnT and ultimately patchy fibrosis. In the general population, even in asymptomatic individuals, high hs-cTnT levels were predictive of future cardiovascular events and correlated with structural heart disease. Finally, the association between cTnT and chronic kidney disease is a consistent finding. Detectable cTnT levels by means of conventional assays in patients with end stage chronic kidney disease are associated with a poor prognosis, even in the absence of coronary heart disease. The clearance and degradation of cTnT remains undefined. However, in a small study, Tsutamoto et al demonstrated a significant correlation between eGFR and serum cTnT levels in HF patients, suggesting that decreased cTnT clearance could contribute to elevated troponin levels in these patients.
The mechanisms of troponin release in HF are not well established, and several processes are likely involved. Although higher troponin levels were observed in patients with HF of ischemic etiology, it has been consistently reported that patients with non-ischemic HF also have elevated troponin levels. Multiple mechanisms may be involved, such as subendocardial ischemia due to increased transmural wall stress and stiffening of the myocardium, myocyte necrosis (induced by ischemia, inflammation, and oxidative stress), myocyte apoptosis, cellular release of proteolytic troponin degradation products, and increased cellular wall permeability because of reversible injury.
Several studies have demonstrated a consistent association between cTnT elevation and prognosis in acute and chronic HF using conventional assays. Latini et al first evaluated the prognostic value of very low cTnT levels using a precommercial version of the hs-cTnT assay in patients enrolled in the Valsartan Heart Failure Trial. Ninety-two percent had detectable hs-cTnT levels, and the risk of death and HF hospitalization increased seven- to eight-fold across increasing deciles of hs-cTnT, and remained strongly associated with these outcomes after adjustment for standard risk predictors and BNP levels. These authors used 12 ng/L (the median value in their population) as the cut-off. Two additional studies that evaluated hs-cTnT in chronic HF, both small studies with limited follow-up, used the upper reference limit of the assay (between 10 and 15 ng/L) to define elevated hs-cTnT levels. In this cohort (a large, prospective, real-life, ambulatory HF population followed for 41 months), the median value of hs-cTnT was 22.6 ng/L. Nevertheless, the optimal cut-off (set at 16 ng/L), was obtained using state-of-the-art statistics by bootstrapping the value that maximized the log-likelihood of the non-adjusted Cox models. This novel approach provided a more precise cut-off for prognostic purposes. To the best of our knowledge, this is the first study in HF that uses this method to select more accurate biomarker cut-off points.
NT-proBNP is well recognized as an important prognostic biomarker in HF. However, beyond natriuretic peptides, the use of other biomarkers for risk assessment is being debated. In this study, the predictive accuracy of hs-cTnT was even higher than that of NT-proBNP according to comprehensive discrimination, calibration, and reclassification analyses. However, the combination of both biomarkers was associated with a substantially higher risk compared with either biomarker alone, reaching a very significant HR of 7.42. Above their respective cut-off points, both biomarkers allowed us to identify a very high-risk subgroup of HF patients with a 5-year predicted survival of 28% (compared with 86% survival for both biomarkers below their respective cut-off points) as assessed by Kaplan-Meier.
Limitations
There is a risk that the absolute levels of hs-cTnT could have been affected by having been measured from frozen rather than fresh samples. There is little information about long-term stability of frozen hs-cTnT. We have analyzed only one blood sample per patient and cannot comment on the prognostic value of serial determinations. The use of bootstrap method to determine the cut-off points for NT-proBNP and hs-cTnT allows to optimize the prognostic prediction but limits its comparison with other analyses.
Our population was a general HF population treated at a specific and multidisciplinary HF unit in a tertiary hospital, and most patients were referred from the cardiology department, resulting in relatively young men with HF of ischemic etiology and reduced LVEF. As such, the obtained results cannot necessarily be extrapolated to a global HF population.
Conclusions
Hs-cTnT provides significant prognostic information in a real-life cohort of patients with chronic HF. The simultaneous addition of hs-cTnT and NT-proBNP into a model that includes established risk factors improves mortality risk stratification.
