Sensitive troponins--which suits better for hemodialysis patients? Associated factors and prediction of mortality

Abstract

Background: In hemodialysis patients, elevated plasma troponin concentrations are a common finding that has even increased with the advent of newly developed sensitive assays. However, the interpretation and relevance of this is still under debate.

Methods: In this cross-sectional study, we analyzed plasma concentrations of sensitive troponin I (TnI) and troponin T (TnT) in stable ambulatory hemodialysis patients (n = 239) and investigated their associations with clinical factors and mortality.

Results: In all of the enrolled patients, plasma TnI or TnT was detectable at a median concentration of 14 pg/ml (interquartile range: 7-29) using the Siemens TnI ultra assay and 49 pg/ml (31-74) using the Roche Elecsys high sensitive TnT assay. Markedly more patients exceeded the 99th percentile for TnT than for TnI (95% vs. 14%, p<0.0001). In a multivariate linear regression model, TnT was independently associated with age, gender, systolic dysfunction, time on dialysis, residual diuresis and systolic blood pressure, whereas TnI was independently associated with age, systolic dysfunction, pulse pressure, time on dialysis and duration of a HD session. During a follow-up period of nearly two years, TnT concentration above 38 pg/mL was associated with a 5-fold risk of death, whereas elevation of TnI had a gradual association to mortality.

Conclusion: In hemodialysis patients, elevations of plasma troponin concentrations are explained by cardiac function and dialysis-related parameters, which contribute to cardiac strain. Both are highly predictive of increased risk of death.

Figure 1. Depiction of the plasma troponin I and T concentrations in the cohort.

A Box-and-Whiskers-Plot of the plasma troponin I and T concentrations indicating the median, interquartile range and the range between the 2.5^th and 97.5^th percentile. Solid lines represent the 99^th percentile of the assay . B Distribution of the plasma troponin I and T concentrations in the cohort (n = 239). C Box-and-Whiskers-Plot of the plasma troponin I and T concentrations expressed as multiples of the 99^th percentile . ^*** indicates a significant difference (p<0.0001) between the troponin T and troponin I concentrations.

Figure 2. Plot showing the influence of single predictors in the multivariate model for troponin T (A) and troponin I (B).

^# p<0.10, ^* p<0.05, ^** p<0.01, ^*** p<0.001.

Figure 3. Agreement between predicted and observed log troponin I and T concentrations.

Plot of the predicted and observed troponin values for troponin T (A) and troponin I (B) as a result of the models shown in table 4. The diagonal line represents the line of identity.

Figure 4. Survival curves of the cohort stratified according to classes of systolic LV function and tertiles of age, troponin T and troponin I.

Figure 5. Hazard ratio of mortality according to tertiles of troponin T (<37; 38–67; >67 pg/ml), troponin I (<10; 10–22; >22 pg/ml), age (<63; 63–75; >75 years) and classes of systolic LV function (1 = normal, 2 = mildly reduced; 3 = moderately and severe reduced).

Error bars represent lower and upper 95% confidence interval. The hazard ratio of the first tertile or class, respectively, was set to 1. ^** p<0.01, ^*** p<0.001.