Cardiac Troponin T Before and After Kidney Transplantation: Determinants and Implications for Posttransplant Survival


Corresponding author: Fernando G. Cosio,


Pretransplant cardiac troponin T(cTnTpre) is a significant predictor of survival postkidney transplantation. We assessed correlates of cTnT levels pre- and posttransplantation and their relationship with recipient survival. A total of 1206 adult recipients of kidney grafts between 2000 and 2010 were included. Pretransplant cTnT was elevated (≥0.01 ng/mL) in 56.4%. Higher cTnTpre was associated with increased risk of posttransplant death/cardiac events independent of cardiovascular risk factors. Elevated cTnTpre declined rapidly posttransplant and was normal in 75% of recipients at 3 weeks and 88.6% at 1 year. Elevated posttransplant cTnT was associated with reduced patient survival (cTnT3wks: HR = 5.575, CI 3.207–9.692, p < 0.0001; cTnT1year: 3.664, 2.129–6.305, p < 0.0001) independent of age, diabetes, pretransplant dialysis, heart disease and allograft function. Negative/positive predictive values for high cTnT3wks were 91.4%/50% respectively. Normalization of cTnT posttransplant was associated with reduced risk. Variables related to elevated cTnT posttransplant included pretransplant diabetes, older age, time on dialysis, high cTnTpre and lower graft function. Patients with delayed graft function and those with GFR < 30 mL/min at 3 weeks were more likely to have an elevated cTnT3wks and remained at high risk. When allografts restore sufficient kidney function cTnT normalizes and patient survival improves. Lack of normalization of cTnT posttransplant identifies a group of individuals with high risk of death/cardiac events.




cardiac troponin


pretransplant cardiac troponin T


3-week posttransplant cardiac troponin T


1-year posttransplant cardiac troponin T


delayed graft function


fasting plasma glucose


impaired fasting glycemia


left ventricular hypertrophy


major adverse cardiac events


new onset diabetes after transplant.


Cardiovascular (CV) morbidity and mortality are highly prevalent in patients with chronic kidney disease (CKD) and kidney transplant recipients. As such, assessments of CV status is an integral part of pretransplant patient evaluation with the dual purpose of (1) identifying patients at risk for death or CV events posttransplant and (2) reducing CV risk with pretransplant interventions. Unfortunately, it is not clear whether our current approaches to CV assessment achieve either of these two purposes [1]. In part, this failure may be due to the fact that current approaches focus on coronary artery disease which is only partially responsible for the high CV mortality of CKD patients and transplant recipients [2]. Indeed, patients with CKD have multiple cardiac pathologies including left ventricular hypertrophy (LVH) and/or dilatation, diastolic dysfunction, heart failure and arrhythmias, each of which is associated with significant CV morbidity and mortality [3-5].

Elevation in the cardiac biomarker troponin T (cTnT) has been shown to be a marker of myocardial ischemia as well as of multiple other cardiac pathologies that are frequently present in patients with CKD [6]. In addition, cTnT is an excellent predictor of survival in patients with CKD [7-9]. Therefore, it is perhaps not surprising that the relationship between increased cTnT level and survival in CKD patients is multifactorial and cannot be attributed to a single pathophysiology such as coronary artery disease [9-11]. More specifically, cTNT levels relate poorly with severity of coronary artery disease in patients with CKD [12].

Our group previously showed that elevated cTnT pretransplant (cTnTpre) is associated with significant mortality and CV risk in kidney transplant candidates on the waiting list [9] and posttransplantation [11]. These associations were independent of the results of other cardiac tests used to assess CV risk. As a result of these findings, cTnT has been incorporated in our program as an integral part of the evaluation of kidney transplant candidates and recipients. For this purpose, an elevated cTnT is not used to exclude patients from transplantation but rather to more accurately assess cardiac risk and guide cardiac evaluation. To extend previous observations, in these analyses we aimed to (1) reassess the relationship between cTnTpre and posttransplant patient survival/CV risk; (2) examine the changes in cTnT after kidney transplantation and the variables that might relate to these changes and (3) evaluate the relationship between posttransplant cTnT levels and posttransplant survival and cardiovascular events.


Patient population

The study cohort included 1206 adult recipients of kidney transplants performed at Mayo Clinic, Rochester, MN, from January 2000 to October 2010 who had cTnT levels measured pretransplant and/or posttransplant at the time of protocol visits. Measurements of cTnT levels became part of our routine protocol for candidate evaluation in 2004. Cardiac troponin T measurements prior to that date were performed from stored serum samples, frozen at −70°C. Recipients of other allografts prior to, simultaneous or after the kidney transplant were excluded from the analysis. The institutional review board approved this study and the collection of demographic, clinical and laboratory information using the institution's electronic databases.

Cardiovascular disease assessment

The protocol for CV disease assessment pretransplant was previously reported [11]. In brief, patients who met any of the following criteria were considered high risk for CV disease: diabetes, age greater than 59 years, history of major cardiac events (see later) or dialysis for greater than 2 years. High-risk patients had cardiac stress echocardiogram and in patients with stress-induced ischemia and/or ejection fraction < 50%, cardiology referral was obtained and coronary angiogram was pursued at the discretion of the cardiologist. Starting in mid-2004 cTnT was incorporated into the pretransplant CV assessment protocol for all candidates. In patients who remained on the waiting list for more than a year, cTnT levels were measured annually and the most recent result prior to the transplant was used for analysis. In addition, cTnT measurement was incorporated into the laboratory assessment of patients presenting for their protocol visit at 3 weeks posttransplant (cTnT3wks) and annually thereafter. Cardiac troponin T was measured using an electrochemiluminescent immunoassay (Roche Diagnostics fourth generation) and results were analyzed in four groups: < 0.01 ng/mL, 0.01–0.03, 0.04–0.1 and >0.1 ng/mL) per previous publications [8]. In our institution, the 99th percentile value for cTnT is < 0.01 ng/mL that is considered to be normal [13]. The cardiac evaluation protocol described above has not been modified based on cTnT levels. However, appreciating the risk associated with an elevated cTnT level pretransplant has encouraged the selection committee to perform a complete cardiac evaluation in some patients if not previously done. A cTnT-guided cardiac evaluation protocol is being considered in our program.

Variable definitions

Abnormalities in fasting plasma glucose (FPG) levels posttransplant were classified according to accepted criteria [14] including normal when FPG < 100 mg/dL; impaired fasting glycemia (IFG) when FPG was between 101 and 125 mg/dL; and new onset diabetes after transplant (NODAT) when FPG was greater than 125 mg/dL in more than two measurements. In this study FPG levels are reported at 3 weeks and at 1 year posttransplant as the average of FPG levels obtained between 2 and 4 weeks or between months 11 and 13, respectively. All patients requiring initiation of hypoglycemic agents after kidney transplantation were considered to be diabetic regardless of FPG levels. Major adverse cardiac events (MACE) pre or posttransplant were defined as in previous studies [11], including a history of acute myocardial infarction or coronary artery revascularization (bypass grafting or percutaneous intervention).

Renal allograft function was assessed by serial serum creatinine measurements and estimation of glomerular filtration rate (eGFR) using the MDRD equation [15]. Delayed graft function (DGF) was defined as requirement of temporary dialysis within a week posttransplantation. Proteinuria was assessed by 24 h urine collection at 3 weeks and 1 year posttransplant.


Induction immunosuppression included antithymocyte globulin in 954 of the 1206 patients (79.1%), anti-CD25 antibodies in 149 (12.4%) and alemtuzumab in 97 (8.0%) patients. Six patients (0.50%) did not receive induction therapy. Posttransplant, 1050 of the 1206 patients (87.0%) received maintenance immunosuppression with tacrolimus, mycophenolate mofetil and corticosteroids. Twenty-nine patients (2.41%) received cyclosporine instead of tacrolimus and 31 (2.6%) sirolimus. Five patients (0.4%) received a combination of both sirolimus and tacrolimus. Ninety-seven patients (8.0%) received tacrolimus and mycophenolate mofetil with no maintenance corticosteroids.

Data analysis

Continuous variables were reported as means with standard deviation or median with range or mean with 95% confidence intervals. Binary and categorical variables were reported as total number and proportion. Continuous data were analyzed using Student's t-test if normally distributed or Wilcoxon Rank Sum test as appropriate. Comparisons between more than two groups of data were done by ANOVA for normally distributed and by the Kruskall–Wallis test for skewed data. Proportions were compared by the chi-square. The primary endpoint of these analyses was a composite of death with a functioning graft or MACE posttransplant. The patient follow-up was censored at the time of last follow-up, graft failure or at death. Survival analyses were performed using Kaplan–Meier plots and Cox proportional hazard models.


Patient characteristics

As shown in Table 1, patients were more commonly male (58.8%) and Caucasian (93.3%). Glomerular disease was the most common cause of kidney failure (34.7%). Fifty-nine percent of recipients required dialysis before kidney transplantation and the majority received a living donor kidney (83.0%). Pretransplant MACE was diagnosed in 252 patients (20.9%). Over a follow-up period of 46.1 ± 26.9 months, 135 grafts were lost not due to patient death (11.2%), 95 patients (7.9%) expired and 77 (6.7%) suffered nonfatal MACE.

Table 1. Baseline characteristics (N = 1206)
Patient characteristicsResults
Recipient age (years)51.8 ± 13.7
Recipient gender (% female)497 (41.2)
Recipient race (% Caucasian)1119 (93.3)
Diabetes (%)324 (26.9)
Primary renal disease419 (34.7)
Glomerular disease419 (34.7)
Diabetes207 (17.2)
Polycystic kidney disease175 (14.5)
Hypertension/vascular122 (10.1)
Other205 (17.0)
Unknown78 (6.47)
Pretransplant dialysis (%)711 (59.1)
Dialysis months—(median, range)15.7 (0.23, 285.5)
Donor age (years)44.0 ± 12.7
Donor gender (% female)670 (55.6)
Donor type (% living donors)1001 (83.0)
Pretransplant MACE252 (20.9)
Follow-up months46.1 ± 26.8

Pretransplant cTnT (cTnTpre) and survival posttransplant

Pretransplant cTnT levels were measured according to the protocol in 1099 patients [mean 0.066 ng/mL (<0.01–4.17)]. The distribution of cTnTpre levels in this cohort is shown in Figure 1A. It should be noted that cTnTpre levels were normal (<0.01 ng/mL) in 479 patients (43.6%). Elevated cTnTpre levels related to an increasing proportion of patients reaching the combined endpoint of death or MACE posttransplant (HR = 1.510 (1.276–1.787), p < 0.0001) (Figure 1B). Compared to patients with normal cTnTpre, the risk of the combined endpoint increased progressively with increasing cTnTpre levels: cTnT between 0.01–0.03 ng/mL (HR = 2.787, 95% CI (1.305–5.951), p = 0.008); cTnT between 0.04 and 0.1 (HR = 2.930 (1.346–6.380), p = 0.007); and cTnTpre >0.1 (HR = 6.651 (3.140–14.088), p < 0.0001). The relationship between cTnTpre and event-free survival remained significant beyond 1 year posttransplant in patients with cTnTpre>0.1 (HR = 4.064 (1.612–10.248), p = 0.003). In contrast, in patients with lower levels of cTnTpre this relationship was not significant beyond 1 year. Elevated cTnTpre related to both reduced patient survival (HR = 1.455 (1.177–1.798), p = 0.001) and to higher incidence of MACE posttransplant (1.703 (1.324–2.192), p < 0.0001) when both outcomes were analyzed separately. Given the high proportion of preemptive and living donor transplants in this cohort it was relevant to note that the relationship between cTnTpre and posttransplant event-free survival was similar in recipients of nonpreemptive transplants (N = 711, HR = 1.333 (1.090–1.631), p = 0.005) and in recipients of deceased donor kidneys (N = 205, HR = 1.463 (1.037–2.063), p = 0.030).

Figure 1.

(A) Distribution of cTnTpre in 1099 patients including normal range (<0.01 ng/mL) (white area), cTnTpre 0.01–0.03 (light grey area); cTnTpre 0.04–0.1 (darker gray area) and cTnTpre >0.1 (black area). (B) Event-free survival in patients with increasing levels of cTnTpre, including normal cTnT (N = 479) (—); cTnT levels 0.01–0.03 ng/mL (N = 270) (—); cTnT 0.04–0.1 (N = 230) (−+−+−) and cTnT >0.1 (N = 120) (……) (log rank p < 0.0001).

Change in cTnT level after transplant

Cardiac troponin T measurements at 3 weeks (cTnT3wks) were available in 524 patients and at 1 year (cTnT1year) in 697. In order to assess the change in cTnT levels after transplantation, we analyzed a subgroup of patients who had cTnT measured at all time points (pretransplant, 3 weeks and 1 year posttransplant) (N = 429) (Figure 2). The proportion of patients with normal cTnT levels increased from 45.9% pretransplant to 75.0% at 3 weeks and 88.6% at 1 year. Similarly, the magnitude of cTnT levels declined from pretransplant (mean 0.059, 95% CI [0.042–0.076]) to 3 weeks (mean 0.0396 [0.006–0.073]) and declined further from 3 weeks to 1 year (mean 0.0142 [0.01–0.018]) (both p < 0.0001 paired Wilcoxon). The decline in cTnTpre was particularly striking in patients with diabetes in whom only 16% had normal cTnTpre while 51.4% and 79.2% had normal cTnT at 3 weeks and at 1 year, respectively. Overall, 78% of patients with cTnTpre >0.03 ng/mL had posttransplant cTnT3 wks < 0.03.

Figure 2.

Distribution of cTnT levels pretransplant (open bars) and at 3 weeks (stripped bars) and 1 year posttransplant (black bars) (N = 429). Numbers of patients include pretransplant, cTnTpre < 0.01 ng/ml (N = 197); cTnTpre 0.01–0.03 (N = 105); cTnTpre 0.04–0.1 (N = 85) and cTnTpre >0.1 (N = 42). At 3 weeks, cTnT3wks < 0.01 (N = 318); cTnT3wks 0.01–0.03 (N = 75); cTnT3wks 0.04–0.1 (N = 23) and cTnT3wks >0.1 (N = 8). At 1 year, cTnT1year < 0.01 (N = 380); cTnT1year 0.01–0.03 (N = 34); cTnT1year 0.04–0.1 (N = 12); cTnT1year >0.1 (N = 3).

Patient survival and posttransplant cTnT

At 3 weeks posttransplant 30% (157 of 524) of recipients had elevated cTnT3wks. Elevated cTnT3wks levels related to increased risk of the combined endpoint beyond 1 month posttransplant (Figure 3A). Quantitatively, compared to patients with cTnT3wks≤0.03 ng/mL, the 13% of patients with cTnT3wks >0.03 ng/mL had markedly increased risk of the combined outcome (HR = 5.575 [3.207–9.692], p < 0.0001). The negative and positive predictive values of cTnT3wks (using a cut-off at 0.03) were 91.4% and 50% respectively. Posttransplant survival related to the change in cTnT after the transplant (Figure 3B). Thus, compared with recipients with low cTnT pre and at 3 weeks (defined as <0.03), recipients with high cTnTpre but low cTNT3wks had statistically similar survival while patients with high cTnTpre that remained high at 3 weeks had markedly reduced survival. In these analyses, 11 patients had normal cTnTpre but elevated cTnT3wks, 4 of whom suffered MACE but 100% survived. Due to the small sample size these patients were not included in the analysis.

Figure 3.

Kaplan–Meier plots of patient survival/cardiac events censored. (A) Relationship between cTnT3wks and survival (log rank p < 0.0001). The groups include normal cTnT level (N = 392) (—); cTnT levels 0.01–0.03 ng/mL (N = 96) (—); cTnT 0.04–0.1 (N = 45) (−+−+−); and cTnT >0.1 (N = 27) (……). (B) Survival after 1 month posttransplant in patient with low cTnTpre (defined as <0.03 ng/mL) and low cTnT3wks (N = 265) (—); recipients with high cTnTpre but low cTnTwks (N = 123) (—) and in patients with high cTnT both pre and at 3 weeks (N = 30) (−+−+−) (log rank p < 0.0001).

Table 2 displays univariate analyses of the variables associated with reduced event-free survival beyond 1 month posttransplant. In addition to these variables, it should be noted that in patients without pretransplant DM the development of hyperglycemia 3 weeks posttransplant related negatively to event-free survival. By multivariate analysis, reduced event-free survival related to older recipient age (HR = 1.522 [1.374–1.686], p = 0.012), longer time on dialysis (HR = 1.488 [1.083–2.046], p = 0.014), history of MACE pretransplant (HR = 2.064 [1.033–4.126], p = 0.04) and elevated cTnT3wks (HR = 3.847 [2.09–7.08], p < 0.0001). The type of immunosuppression and allograft function at 3 weeks did not relate to survival.

Table 2. Relationship between the risk of the combined outcome and pre- and posttransplant clinical variables at 3 weeks and at 1 year posttransplant
VariableUnivariate analysis of variables at 3 week posttransplant HR (95% CI), pUnivariate analysis of variables at 1 year posttransplant HR (95% CI), p
  1. a

    Calculated for every 10 years increase in age.

  2. b

    Months on dialysis were considered in these groups: no dialysis (N = 346, 37.4%), dialysis for 1 year (N = 259, 28%), dialysis for 2 years (N = 126, 13.6%) and dialysis for more than 2 years (N = 194, 21%).

  3. c

    Analyzed as normal glucose, IFG or NODAT;

  4. d

    log transformed values.

Recipient agea1.551 (1.480–1.625), p < 0.00011.620 (1.536–1.709), p < 0.0001
Female gender0.650 (0.443–0.955), p = 0.0280.605 (0.392–0.934), p = 0.023
Diabetes1.527 (1.051–2.218), p = 0.0261.483 (0.976–2.254), p = 0.065
Pretransplant dialysis monthsb1.327 (1.139–1.546), p < 0.00011.207 (1.015–1.435), p = 0.033
Heart events pretransplant1.926 (1.368–2.882) p < 0.00011.900 (1.252–2.882), p = 0.003
Hyperglycemiac1.404 (1.025–1.923), p = 0.0351.591 (1.143–2.215), p < 0.0001
Delayed graft function2.923 (1.871–4.565), p < 0.00011.123 (0.904, 1.415), p = 0.3012
Deceased donor1.447 (1.183–1.769), p < 0.00011.177 (0.991, 1.408), p = 0.0637
Serum creatinine1.404 (1.025–1.923), p = 0.0351.812 (1.232–2.663), p = 0.003
cTnTd5.575 (3.207–9.692), p < 0.00013.664 (2.129–6.305), p < 0.0001

One year posttransplant 17% of patients had elevated cTnT1year which was associated with reduced event-free survival beyond 1 year posttransplant (Figure 4) (HR = 3.664 [2.129–6.305], p < 0.0001). For these analyses all patients with cTnT1year>0.03 ng/mL were combined due to the small number of patients. Three- and 5-year event-free survival were 96.8% and 91.7%, respectively, for patients with normal cTnT1year (N = 500), 92.2% and 69% for cTnT1year between 0.01 and 0.03 (N = 54) and 86.2% and 52.5% for cTnT1year >0.03 (N = 37). The negative and positive predictive values of cTnT1year (using a cut-off at 0.03) were 91.3% and 34.6%, respectively. Table 2 displays the variables related to event-free survival beyond 1 year on univariate analysis. By multivariate analysis, reduced survival beyond 1 year related to older age (HR = 1.336 [1.236–1.445], p = 0.026), longer time on dialysis pretransplant (HR = 1.357 [1.059–1.739], p = 0.016) and elevated cTnT1year (HR = 2.657 [1.476–4.781], p < 0.0001) were independently related to the combined endpoint.

Figure 4.

Relationship between cTnT1year and survival in the following groups of recipients: normal cTnT level (N = 540) (—); cTnT levels 0.01–0.03 ng/mL (N = 57) (—); cTnT >0.03 (N = 42) (−+−+−) (log rank p < 0.0001).

Variables related to elevated posttransplant cTnT levels

Because elevated cTnT levels posttransplant were independently associated with increased risk we next sought to identify variables that may predict persistent elevations in cTnT posttransplant. Elevated cTnT3wks related to pretransplant variables (older age, diabetes, time on dialysis, history of MACE and high cTnTpre) and posttransplant variables associated with reduced graft function (deceased donor, DGF and eGFR3wks) (Table 3). Compared to patients with immediate graft function, those with DGF were less likely to have normal cTnT3wks (74.5% vs.35.8%, p < 0.0001). Furthermore, lower levels of eGFR3wks related to higher cTnT3wks but that relationship was not linear. Thus, cTnT3wks was normal in 74% of patients who achieved any level of eGFR3wks >30 mL/min/1.73 m2 but it was normal in only 30% of patients with GFR3wks < 30 (p < 0.0001). Posttransplant hyperglycemia at 3 weeks also related to higher cTnT3wks (Table 3). This relationship was such that elevated cTnT3wks was found in 17.0% of patients with normal glucose, 22.8% of patients with IFG and 45.0% of patients with NODAT (p < 0.0001).

Table 3. Variables related to elevated cTnT at 3 weeks and 1 year posttransplant (univariate logistic regression)
VariablecTnT3 weeks OR (95% CI), pcTnT1 year OR (95% CI), p
  1. a

    Calculated for every 10 years of age.

  2. b

    Per 1 mL/min increase.

  3. c

    Excluding patients with diabetes prior to the transplant.

Recipient agea1.271 (1.264–1.278), p < 0.00011.419 (1.328–1.516), p < 0.0001
Diabetes2.824 (1.622–4.915), p < 0.00012.722 (1.813–4.089), p < 0.0001
Pretransplant MACE3.755 (1.923–7.335), p < 0.00013.584 (1.858–6.916), p < 0.001
Dialysis pretransplant3.013 (2.016–4.504), p < 0.00011.739 (1.147–2.636), p = 0.009
Pretransplant cTnT9.197 (2.131–39–68), p = 0.0035.864 (2.072–16.59), p = 0.001
Deceased donor1.467 (1.172–1.838), p = 0.0011.471 (1.174–1.844), p = 0.001
Delayed graft function5.229 (2.8 (-9.570), p < 0.00014.372 (2.532–7.548), p < 0.0001
eGFR3wks, eGFR1yearb0.976 (0.964–0.989), p < 0.00010.960 (0.946–0.975), p < 0.0001
Glucose1 mo, Glucose1yearc1.028 (1.014–1.041), p < 0.00011.023 (1.002–1.025), p = 0.023
Proteinuria at 1 year2.376 (1.394–4.050), p = 0.001

By multivariate analysis elevated cTnT3wks was associated with DM (OR = 3.480 [2.065–5.866], p < 0.0001), pretransplant dialysis (OR = 2.554 [1.509–4.321], p < 0.0001) and the risk was lower in patients with eGFR3wks>30 mL/min/1.73 m2 (OR = 0.279 [0.111–0.703], p = 0.007). The relationship between cTnT3wks and older recipient age was of borderline significance (p = 0.054). The following variables did not relate independently to cTnT3wks: pretransplant cardiac disease, donor demographics, recipient gender, LVH, posttransplant hyperglycemia, immunosuppression, beta blockers or statin use.

The variables predictive of an elevated cTnT1year (Table 3) were similar to those associated with an elevated cTnT3wks. In addition, proteinuria at 1 year related to elevated cTnT1year. In non-DM recipients elevated cTnT1year was noted in 5% of patients with normal fasting glucose levels, 11.4% with IFG and 20% with NODAT (p < 0.0001) at 1 year. By multivariate analysis elevated cTnT1year related independently to older recipient age (OR = 1.699 [1.562–1.848] for every 10 years, p < 0.0001), DM (OR = 1.993 [1.064–3.733], p = 0.031), elevated cTnTpre (OR = 5.981 [1.989–17.986], p = 0.001) and proteinuria (OR = 2.380 [1.198–4.725], p = 0.013). The risk of elevated cTnT1year was decreased in patients with higher eGFR1year (OR = 0.973 [0.952–0.995], p = 0.017). In patients without pretransplant DM, the association between IFG and NODAT with elevated cTnT1year (OR = 2.501 [1.424–4.392], p = 0.001) was statistically independent of recipient age, dialysis, cTnTpre and eGFR1year.


These studies confirm and extend previous results showing a strong association between cTnT levels measured prior to kidney transplant and patient survival and MACE after transplant. Moreover, our results show for the first time that cTnT levels normalize within the first 3 weeks posttransplantation and that normalization of cTnT levels correlated with reduction in cardiovascular risk. In contrast, patients who did not normalize cTnT posttransplant continued to have a significant risk of mortality and MACE. Contrary to our previous study [11] these results showed that even relatively small increases in pretransplant cTnT levels (0.01–0.03 ng/mL) are associated with increased risk of death/cardiac events posttransplant. This new finding is likely the result of the larger sample size and the longer follow-up period available in this study. Our results show that the relationship between elevated cTnTpre and reduced posttransplant survival does not extend beyond 1 year posttransplant except in patients with the highest levels of cTnTpre (>0.1 ng/dL). The fact that the risk of an adverse outcome abates beyond 1 year in patients with cTnTpre≤0.1 ng/mL is intriguing and suggests that the heightened risk of patients with high cTnTpre is at least in part mitigated by transplantation. Similarly, among patients with elevated cTnT pretransplant, normalization of cTnT posttransplant was associated with excellent survival while persistently elevated cTnT3wks related to both an increased risk of death and MACE. Thus, elevated cTnTpre levels are modifiable and the change in cTnT posttransplant from high to low correlated with reduction in CV risk.

These analyses showed that restoration of kidney function by transplantation is associated with rapid normalization of cTnT levels in a large proportion of patients. Previous studies provided contradictory information regarding the changes in cTnT that follow kidney transplantation perhaps due to small sample sizes [16, 17]. The parallelism between cTnT normalization and restoration of GFR is consistent with the hypothesis that cTnT (molecular weight 37 kD) and/or perhaps fragments of that molecule that may be measured by the troponin assay may be cleared by the kidney [6, 18]. Changes in cTnT levels could also be due to impairment of the normal degradation process of the cTnT molecule in patients with CKD resulting in disruption of clearance mechanism. However, reduction in GFR alone is not sufficient to cause an elevation of cTnT. For example, these results showed that 44% of patients with end-stage renal disease have normal cTnT levels. Then, how can we explain that (a) cTnT levels increase only in some patients with CKD, (b) that elevated cTnT levels are related to patient survival and (c) that normalization of cTnT levels within days of a successful transplant is associated with a reduction in risk? We suggest that the elevation in cTnT levels observed in some patients with CKD reflects myocardial injury and reduced cTnT clearance. This postulate is consistent with the observation that cTnT is elevated preferentially in patients with CKD and high CV risk such as in patients with diabetes and atherosclerotic renal disease [9, 19, 20]. In addition, patients with CKD and elevated cTnT have demonstrable cardiac injury [21] and there is a strong association between cardiac injury with concomitant cTnT elevation and CV outcomes in CKD [6, 8, 9, 22].

Of interest, the rapid reduction in cTnT levels that occurs after successful transplantation is associated with improved survival. This observation however argues against the notion that the myocardial injury associated with high cTnT in CKD patients is due mainly to large vessel coronary artery disease as these lesions likely do not reverse soon after transplantation. Therefore we can only speculate about possible physiologic changes that occur after transplantation that may result in reduced CV risk. First, it is possible that restoration of kidney function improves myocardial perfusion by improving endothelial cell function [23]. Second, restoration of kidney function may result in correction of intravascular volume that likely result in decrease in myocardial strain resulting in lowering of cTnT. Lastly, it is known that the renal failure milieu causes multiple perturbations in skeletal and likely cardiac muscle protein turnover and may be directly cardiotoxic [24, 25]. Normalization of cTnT posttransplant may be an indication of improved myocardial health by removal of the cardiotoxic effects of renal failure. Over the longer term, there is remodeling of the cardiac muscle [26] that also likely improves survival.

The strong association between elevated cTnT posttransplant and survival suggests the hypothesis that interventions that improve myocardial health as evidenced by lower cTnT level may result in improved survival. The search for variables related to elevated cTnT suggest potential targets for preventive or therapeutic intervention one of which is avoidance or minimization of dialysis pretransplant as dialysis relates to marked increase in cTnTpre [9] and persistent elevation posttransplant. It is well established that time on dialysis pretransplant relates to posttransplant survival [27, 28] and that preemptive transplantation is beneficial [29-31]. Moreover, our results show a strong association between kidney allograft function and cTnT levels. This observation is consistent with previous associations shown between allograft function and posttransplant survival [32-36]. There is renewed interest in using allograft function as an endpoint in therapeutic trials in transplantation. Although this endpoint is fraught with interpretative difficulties [37], from the point of view of patient survival and CV risk measuring cTnT provides information about the possible impact of lower GFR on CV risk and this biomarker may be a useful parameter to assess the success of interventions designed to improve graft function. Finally, patients who develop posttransplant hyperglycemia are at risk of having high cTnT levels and indeed high CV risk [38, 39]. The use of cardioprotective medications did not correlate with normalization of cTnT posttransplant. This result should not be interpreted as lack of efficacy but a reflection of the fact that these medications, during the time period of the study, were not used systematically but rather preferentially in patients with the highest perceived cardiac risk.

A few limitations of our studies warrant discussion. First, patients in our transplant program have several unique characteristics reflecting the geographic location (high proportion of Caucasians) and the goals of the program (high proportion of living donor and preemptive transplants). However, these characteristics do not alter the results of these analyses as the relationship between cTnT and survival remained significant in nonpreemptive transplants and in recipients of deceased donor organs. Second, significant proportion of patients in this cohort did not have cTnT measured at all time points. However, this problem is minimize as the overall sample size is quite large. Finally, these data are observational, therefore we can only speculate about the mechanisms underlying the relationships observed. Nonetheless, these results are valuable as a springboard for future studies using cTnT as a target for intervention.

These results have several clinical implications. First, we suggest that elevated cTnTpre should not be used to exclude patients from transplantation but rather to better quantify risk and perhaps guide cardiac work-up. For example, normal cTnTpre levels have a 98% negative predictive value, i.e. are associated with remarkably low posttransplant risk [11]. Thus, intensive cardiac work up and/or interventions in these patients are unlikely to improve an already excellent outcome. In contrast, candidates with high cTnTpre, in the absence or in the presence of other CV risk factors, have high risk and should therefore be the focus of investigative and therapeutic efforts. Second, these studies showed that elevations in posttransplant cTnT identify a cohort of recipients with very high risk. Therefore it is reasonable to suggest that these individuals should be investigated for causes of cTnT elevation and that their cardioprotective regimen needs to be reviewed and perhaps intensified [40]. Conversely, those individuals with normal cTnT levels posttransplant have excellent survival (91% negative predictive value for normal cTnT3wks or cTnT1year) (see Figure 3B). Lastly, since the number of patients with an elevated cTnT at 1 year posttransplant is low, we suggest that measurement of cTnT beyond 3 weeks posttransplant should be performed preferentially in patients with other identifiable risk factors such as NODAT.

Kidney transplantation is associated with remarkable improvement in patient survival compared to dialysis [41]. Still, approximately 3% of kidney transplant recipients die every year [42], a significantly higher rate than in the general population. Cardiac troponin T appears to be an excellent identifier of patients at high risk and, as these data suggest, may be a reasonable endpoint to identify whether cardioprotective interventions are effective in reducing risk.


The authors would like to thank the Mayo Clinic transplant coordinators for their tireless efforts in patient follow-up and data collection.


The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.