Variability of high-sensitivity cardiac troponin T and I in asymptomatic patients receiving hemodialysis

Variation of high-sensitivity cardiac troponin I and T (hs-cTn) during hemodialysis has been observed. Observational studies demonstrated the increased incidence of adverse cardiovascular events after long compared to short interdialytic intervals. Therefore, we aimed to compare variation of hs-cTnI and hs-cTnT before and after hemodialysis and between short and long interdialytic intervals. We enrolled 200 asymptomatic patients receiving regular hemodialysis. The hs-cTnI and hs-cTnT levels were measured before and after hemodialysis on the day after short and long interdialytic intervals. Mean age was 62.3 ± 14.8 years (Male 55.5%). Prevalence of increased hs-cTnI and hs-cTnT was 34.5% and 99.0%, respectively. The median ± interquartile range of hs-cTnT increased significantly after hemodialysis during short and long interdialytic intervals. However, hs-cTnI level did not increase after hemodialysis during short and long intervals. We found that levels of hs-cTnI and T did not differ between short interdialytic and long interdialytic intervals. We demonstrated higher prevalence of elevated hs-cTnT in patients with regular hemodialysis compared to hs-cTnI. The rise of hs-cTnT was observed immediately after hemodialysis but no significant change of hs-cTnI was noted. Accordingly, hs-cTnI may be preferable as a diagnostic marker in patients with suspected acute myocardial infarction than hs-cTnT.

www.nature.com/scientificreports/ studies have shown that the incidence of adverse cardiovascular events significantly increased during the period after long interdialytic interval compared to short interdialytic interval 14,15 . Accordingly, we also sought to examine the difference of hs-cTn I and hs-cTn T levels between long and short interdialytic interval.

Methods
We enrolled asymptomatic patients diagnosed with ESRD, aged > 18 years, who had undergone regular hemodialysis (two or three times a week) for more than 90 days in the study. We excluded patients who had a recent diagnosis of acute myocardial infarction, heart failure and pulmonary embolism within previous 6 months, had major surgery and trauma within previous 4 weeks, and had coronary and/or valvular intervention within previous 6 months. The hs-cTn I and T levels were measured in all subjects before hemodialysis and after hemodialysis session on the day after short interdialytic interval and the day after long interdialytic interval. With this regard, each patient had four levels of hs-cTn I and four levels of hs-cTn T. The hs-cTn T values were evaluated with electrochemiluminescence immunoassay by using the Cobas e801 system (Roche Diagnostics). The detection limit of hs-cTn T was 3 ng/L, a cut-off point at 99th percentile was 14 ng/L, and a coefficient of variation of less than 10% was at 13 ng/L. The hs-cTn I values were evaluated with chemiluminescence microparticle immunoassay (CMIA) by using the ARCHITECT i2000SR system (Abbott Diagnostics). The detection limit of hs-cTn I was 3.2 ng/L, a cut-off point at 99th percentile was 26.2 ng/L, and a 10% coefficient of variation was 4.7 ng/L. We defined the short and long interdialytic intervals as follows. For the patients who have been receiving thrice weekly hemodialysis, the long interdialytic interval was 2-day interval between hemodialysis sessions. The short interdialytic interval was 1-day interval between hemodialysis sessions. For those who have been received twice weekly hemodialysis, the long interdialytic interval was 3-day interval between hemodialysis sessions. The short interdialytic interval was 2-day interval between hemodialysis sessions.
Clinical data were recorded, including age, gender, duration of hemodialysis, medications, echocardiographic results within 1 year, ultrafiltration volume, comorbidity, and blood chemistry. This study was approved by the institutional research board of Faculty of Medicine, Chiang Mai University (Approval No. 112/2563). The study procedure was performed according to the Declaration of Helsinki. Informed consent was obtained from all participants.
Statistical analysis. Results are expressed as mean ± SD, unless otherwise specified and compared between group with t-test or paired t-test. Results with non-normal distribution are expressed as median (interquartile range) with non-parametric test. The numerical variables were compared within groups with paired t test or Wilcoxon matched paired sign-rank test. Mann-Whitney U test as appropriate. Proportions were compared by Fisher's exact test. Univariate and multivariate linear regression analysis was used to examine the association between potential variables and the change of hs-cTn after hemodialysis. p values < 0.05 was considered statistically significant. Statistical software package IBM SPSS Statistics for Windows, version 23 (IBM Corp., Armonk, NY, USA, https:// www. ibm. com/ produ cts/ spss-stati stics) was used for analysis.
Ethics approval and consent to participate. The Effect of long and short interdialytic interval of chronic hemodialysis on heart rate variability in patients with end-stage renal disease was approved by the ethics committee of the Faculty of Medicine, Chiang Mai University, approval number 112/2563. The investigations were carried out in accordance with the Declaration of Helsinki, including written informed consent of all participants.

Results
Baseline characteristics of the studied population are shown in Table 1. The mean age was 62.3 ± 14.8 years. Male was prevalent in 55.5%. The prevalence of hypertension, diabetes mellitus, and coronary artery disease was 91.0%, 45.0% and 13.0%, respectively. There were 189 (94.5%) patients receiving hemodialysis thrice a week. The rest were hemodialyzed twice a week.
The mean body weight prior to hemodialysis was greater in long interdialytic interval compared to short interdialytic interval (62.8 ± 16.7 kg vs. 62.3 ± 16.8 kg, P < 0.001). Likewise, the mean net ultrafiltration volume and the mean ultrafiltration rate was significantly greater in long interdialytic interval compared to short interdialytic interval. Table 2 shows hemodialysis parameters between short and long interdialytic intervals.
We demonstrated that 198 (99.0%) patients had an increased level of hs-cTn T above 99th percentile of the URL in both short and long interdialytic intervals. On the other hand, only 64 (32.0%) and 69 (34.5%) patients had increased hs-cTn I during short and long interdialytic intervals, respectively (Fig. 1). The histogram of hs-cTn T and I levels before hemodialysis are presented in Fig. 2. The wider range of hs-cTn T levels among interindividual patients was observed compared to hs-cTn I levels.
We also compared levels of hs-cTn between short interdialysis and long interdialysis intervals. The levels of hs-cTn I and T did not differ between short interdialytic and long interdialytic intervals, either pre-hemodialysis or post-hemodialysis level ( www.nature.com/scientificreports/ We performed the univariate and multivariate linear regression to examine the association of the change of hs-cTn levels after hemodialysis and potential variables including age, medications, hemoglobin (Hb) level, vascular access, underlying coronary artery disease (CAD) and ultrafiltration rate (Table 4).
We found that older age was associated with the greater change of hs-cTn T during short and long interdialytic interval. However, age was not associated with the change of hs-cTn I after hemodialysis. In addition, we demonstrated that ultrafiltration rate was an independent factor to predict the change of hs-cTn I during short and long interdialytic interval and the change of hs-cTn T during long interdialytic interval. Furthermore, the lower Hb level was independently associated with the greater change of hs-cTn I during short interdialytic interval.
There were 127 (63.5%) patients receiving beta-blocker and 63 (31.5%) patients receiving angiotensin converting enzyme inhibitor (ACEI) or angiotensin receptor blocker (ARB). Due to the fact that these medications have been shown to protect myocardial ischemia, we explored the effect of these medications on the change of hs-cTn. We demonstrated that the change of hs-cTn T and I was similar in those receiving beta-blocker and those without beta-blocker therapy. Also, patients with and without ACEI/ARB therapy had the comparable change of hs-cTn T and I.  www.nature.com/scientificreports/  There were 26 (13%) patients with history of CAD in the studied population. The presence of CAD was not associated with the greater change of hs-cTn after hemodialysis after multivariate analysis (Table 4). Interestingly, we observed that hs-cTn T levels before and after hemodialysis were significantly higher in patients with history of CAD than those without. In contrast, hs-cTn I levels were not different between those with and without CAD. The data is shown in Table 5.

Discussion
The elevation of hs-cTn T and I. The hs-cTn level is recommended for the diagnosis of acute myocardial infarction but its cutoff level is derived from epidemiological data in general population without ESRD. Several investigators have shown that patients with regular hemodialysis have elevated hs-cTn levels compared to general population [5][6][7]16 . This could be partly explained by the occurrence of microinfarction, heart failure, degenerative changes or other myocardial pathology in patients with ESRD 9 . Previous study has demonstrated that greater number of patients with regular hemodialysis had an elevated cTn T level compared to cTn I level 7,17,18 . In similar, the prevalence of increased hs-cTn T was much higher than hs-cTn I in our studied population. The elevation of troponin levels in dialysis patients could be related to other factors beyond the ischemic cause, such as left ventricular systolic and diastolic dysfunction, left ventricular hypertrophy, myocardial stunning, volume overload, microvascular disease, endothelial dysfunction, and decreased renal clearance [18][19][20] . Table 3. The change of biomarkers before and after hemodialysis in short and long interdialytic intervals. *Compared between pre-and post-hemodialysis, hs-cTn = high-sensitivity cardiac troponin. Data are presented as median level (interquartile range). www.nature.com/scientificreports/ The disparity between the prevalence of increased levels of hs-cTn I and T in patients with ESRD may be explained by the fact that they have different biochemical, genetic, kinetic features and have dissimilar analytical performances 11 . The cellular distribution was different between hs-cTn I and T. It has been shown that cTn T has higher tissue concentration and free cytoplasmic concentrations than cTnI 21,22 . As a result, hs-cTn T may release more early with greater amount than hs-cTn I during myocardial injury. This was in line with previous studies showing that cTn T levels increased in some patients with neuromuscular disease or inflammatory myopathy but no rising of cTn I was noted 12,13 .
Due to the fact that almost all patients in our study had elevated level of hs-cTn T, the cut-off value of hs-cTn T used currently for the diagnosis of acute myocardial infarction would be of no benefit in patients with chronic hemodialysis. Therefore, a significant rise and fall of sequential hs-cTn T levels should be used in favor of single elevated level of hs-cTn T to make the diagnosis of acute myocardial infarction. Interestingly, we demonstrated that only one-third of patients with regular hemodialysis had elevated hs-cTn I. Of importance, we demonstrated that patients with history of CAD had significantly higher level of hs-cTnT compared to those without. However, no significant difference of hs-cTn I level was noted between those with and without history of CAD. From a practical standpoint, hs-cTn I level may be a more favorable biomarker than hs-cTn T level in the evaluation of the patients with suspected acute myocardial infarction, particularly in those with history of CAD.
Previous study has explored the prognostic value of hs-cTn T in pre-dialysis advanced chronic kidney disease patients. They found that those with glomerular filtration rate < 20 ml/min/1.73 m 2 had a 2.5-fold increase in hs-cTn T cutoff level to predict long-term cardiovascular outcome compared to the standard cutoff level used in those with normal renal function. It has been suggested that the cutoff level of hs-cTn T should be 35 ng/L in this group of patient. 23 Nevertheless, the data regarding prognostic values of cTn in patients with ESRD are conflicting. The cTn T has been shown to have better predictor of long-term cardiovascular outcomes than cTn I in patients with chronic hemodialysis in some studies 7,24-26 . However, conventional cardiac troponin assay was used in most of the studies which may have limitation of sensitivity and specificity as compared to high-sensitivity assay, especially in ESRD patients 7,24,26 . Larger studies using hs-cTn to predict long-term outcome in patients with ESRD should be warranted to clarify this issue.

The change of hs-cTn T and I after hemodialysis.
Interestingly, we demonstrated that hs-cTn T level increased significantly after hemodialysis during both short and long interdialytic intervals. Nevertheless, we did not observe the significant change of hs-cTn I after hemodialysis. Our results were similar to previous studies that showed the increase in cTn T, but not cTn I, in patients after hemodialysis 24,27 . On the contrary, other two studies showed that hs-cTn I and hs-cTn T decreased slightly after hemodialysis 9,10 . However, the number of patients in those two studies was small in which only 10 and 20 patients were included.
It is plausible that hs-cTn T may have higher sensitivity to detect minor myocardial injury during hemodialysis compared to hs-cTn I. The hemodialysis has been shown to reduce myocardial blood flow 28 . In accordance, previous study showed that high ultrafiltration rate was associated with increased hs-cTn levels during hemodialysis 29 . High volume depletion during hemodialysis may cause hemodynamic compromise leading to myocardial ischemia. Nevertheless, our result showed that ultrafiltration rate was associated with the change of both hs-cTn T and I level after hemodialysis. The dissimilar biochemical characteristics between these two biomarkers may also partly explain the different findings. It has been described that hs-cTn I can adsorb onto the dialysis membrane because of its hydrophobicity which may result in the lack of hs-cTn I elevation after hemodialysis 11 .
The change of hs-cTn during short and long interdialytic intervals. Several observational studies have shown that the incidence of adverse cardiovascular events significantly increased during the period after long interdialytic interval compared to that after short interdialytic interval 14,15 . Numerous factors have been reported to contribute to the major adverse cardiac events during long interdialytic interval compared to short interdialytic interval. Greater degree of hypervolemic status during long interdialytic interval may induce structural and functional disorders in myocardium, leading to the occurrence of myocardial damage. Electrolyte imbalance and the disorder in autonomic nervous system during long interdialytic interval may result in the increased risk of cardiac arrhythmias. In addition, the increase in oxidative stress, inflammation and abnormal calcium or phosphate metabolism during long interdialytic interval, may play some roles in atherosclerosis of coronary artery [30][31][32][33] . Of interest, we did not find any difference in hs-cTn I and T levels between short and long Table 5. The difference of biomarkers between patients with and without history of coronary artery disease. CAD = coronary artery disease, hs-cTn = high-sensitivity cardiac troponin, LI = long interdialytic interval, SI = short interdialytic interval. Data are presented as median level and interquartile range (IQR). www.nature.com/scientificreports/ interdialytic intervals. Our results suggest that myocardial injury may not be a major factor contributing to worse prognosis during long interdialytic interval. Our study has several limitations. First, the fluid overload and fluid management are significant confounding factors affecting cardiac ischemia via stretching and shrinking cycles. However, we did not assess the fluid status of the patients by any bioimpedance measurements or cardiac biomarkers such as b-type natriuretic peptide (BNP) or N-terminal pro BNP (NTproBNP). We examined only the ultrafiltration volume, the ultrafiltration rate and body weight change which may not be insufficient to evaluate the fluid status of the patients. Second, we did not assess the inflammatory marker in our studied population. This issue should be explored in future studies. Third, we did not include patients with non-dialysis ESRD and those with chronic peritoneal dialysis. Therefore, our findings could not be applied in these population. Future studies are warrant to explore the difference of hs-cTn variability between our studied population and those with peritoneal dialysis and non-dialysis ESRD patients.

Conclusion
We demonstrated the higher prevalence of elevated hs-cTn T in patients with regular hemodialysis compared to hs-cTn I. In addition, the rise of hs-cTn T was observed immediately after hemodialysis but no change of hs-cTn I was noted. Nevertheless, the levels of both hs-cTn I and T did not change between short and long interdialytic intervals. Our results indicate that hs-cTn T may be more sensitive than hs-cTn I to detect minor degree of myocardial injury in patients receiving hemodialysis. However, from a practical standpoint, hs-cTn I may be more favorable as a diagnostic marker in patients with suspected acute myocardial infarction than hs-cTn T.

Data availability
The informed consent given by effect of long and short interdialytic interval of chronic hemodialysis on heart rate variability in patients with end-stage renal disease study participants does not cover data posting in public databases. However, data are available upon request should be sent to kajohnsak.noppakun@cmu.ac.th and are subject to approval by the Faculty of Medicine, Chiang Mai University Ethics Committee.