Introduction
Orthotopic heart transplantation (HTX) is still the gold standard therapy for end-stage heart failure. Unfortunately, there is an ongoing donor shortage which limits the number of HTX and leads to an increasing number of patients on HTX waiting lists [
1]. In addition, the treatment of end-stage heart failure is continuously improving so that the patients undergoing HTX are getting older and have more comorbidities with an increased risk for postoperative complications [
2,
3]. To optimize perioperative risk stratification and early re-estimation of risk, the development of risk prediction models and the identification of perioperative prognostic factors became more and more important [
4,
5,
6]. Numerous risk stratification tools have been developed in the past years, but their clinical use is often limited by insufficient predictive values [
5]. In this context, the Index for Mortality Prediction After Cardiac Transplantation (IMPACT) score was introduced as validated tool for prediction of 1-year mortality after HTX [
7,
8]. However, some studies report poor-to-moderate discrimination for mortality in their cohorts [
9‐
11].
Previously biomarkers were established as another possibility to support perioperative stratification and early re-estimation of risk. A sensitive biomarker for myocardial cell injury is high-sensitivity troponin T (hsTnT) [
12]. Postoperative troponin release has been investigated extensively in cardiac and non-cardiac surgery and is associated with adverse events [
13,
14,
15]. Recently, Devereaux et al. investigated the prognostic value of high-sensitivity troponin I (hsTnI) in patients undergoing cardiac surgery and showed that levels of hsTnI were independently associated with mortality [
16]. The role of hsTnT as a prognostic factor after HTX, however, is not clear and recent literature is ambiguous [
17]. Therefore, we conducted this analysis to evaluate whether hsTnT is a suitable marker for risk stratification and prognosis after HTX.
Methods
This retrospective single-center cohort study was approved by the University of Duesseldorf’s ethics committee (reference number: 4567) and complies with the International Society for Heart and Lung Transplantation (ISHLT) ethics statement. Data were extracted from the local prospective HTX database. All patients had given written informed consent to be registered in this database. Reporting of this work corresponds to the “Strengthening the Reporting of Observational Studies in Epidemiology” (STROBE) guidelines [
18].
Patient population
All patients aged ≥ 18 years who underwent HTX at the University Hospital Duesseldorf, Germany, in a time period from September 2010 to August 2021 were considered for inclusion. Patients with missing data regarding survival, DAOH and hsTnT measurements, as well as patients without completed 1-year follow-up were excluded from analysis. In all patients HTX was conducted using bicaval technique and traditional cold storage was used for donor organ preservation.
High-sensitivity troponin T measurements
Main exposure was postoperative hsTnT measured in ng/L after HTX at different time points. At our institution hsTnT is routinely measured preoperatively, within the first 12 h, day 1, day 2 and day 3 after HTX. Measurements were performed in the central laboratory of the University hospital Duesseldorf.
IMPACT score calculation
The risk index for mortality prediction after cardiac transplantation (IMPACT) is a score validated to predict 1-year mortality after HTX from preoperative recipient risk factors. This score assigns varying points for 10 variables: age, serum bilirubin, creatinine clearance, dialysis, sex, heart failure etiology, preoperative infection, race, circulatory support and type of ventricular assist device. The score was calculated for each HTX patient as described before with a maximum of 50 points [
7,
8].
Outcomes
The primary outcome of this study was mortality during the first year after HTX. Days alive and out of hospital (DAOH) at 1 year after HTX was the secondary endpoint of this study. Calculation of DAOH was conducted by summing up all days of hospitalization in the first year after HTX and subtracting them from 365 days, as described before [
19‐
21]. In case of mortality, the number of days the patient did not survive and of days spent in hospital were subtracted from 365 days.
Statistical analysis
Statistical analysis was performed using IBM SPSS© software version 25.0 (Armonk, NY, USA), GraphPad Prism© version 8.02 (La Jolla, California, USA), MedCalc® Statistical Software version 20.114 (MedCalc Software Ltd, Ostend, Belgium) and R Statistical Software (v4.1.2; R Core Team 2021). Patients characteristics with continuous variables were presented as mean ± standard deviation (SD) or as median and interquartile ranges (IQR, 25–75%), as appropriate. Categorical variables were presented as numbers (n) with corresponding percentages (%) in brackets. Fisher’s exact test or unpaired t-tests were used to test for differences between dichotomous or continuous variables between groups defined by survival status. For analysis of the primary endpoint, receiver operating characteristic (ROC) analyses were performed for hsTnT levels within 12 h, 24 h, 48 h and 72 h after HTX. Cutoff values for troponin levels were determined by Youden index. The cutoff of the hsTnT time point with the strongest discrimination for 1-year mortality in ROC analysis was added to a logistic regression model, with adjustment using the continuous IMPACT score. The net reclassification improvement (NRI) and the net absolute reclassification improvement (NARI) of the mortality prediction model by adding the postoperative troponin cutoff were assessed. Discrimination (ROC-AUC) of the models with and without postoperative hsTnT was quantified and compared using Delong test. To compare net benefit of using these models to detect patients’ risk for 1-year mortality, a decision curve analysis was performed for both models.
For analysis of DAOH patients were classified by hsTnT cutoff. DAOH was compared using non-parametric Mann–Whitney U test. Association of continuous hsTnT elevation per 100 ng/L with DAOH was adjusted by the continuous IMPACT score using multivariable linear regression. For all statistical tests, a p < 0.05 was considered significant.
Discussion
The present study suggested that early hsTnT levels at 48 h after HTX are independently associated with mortality and DAOH after HTX. Moreover, this study showed that the addition of hsTnT improved risk prediction for mortality and DAOH over IMPACT score.
Referring to the current literature, the prognostic value of troponin after HTX is underexplored. A recent systematic review by Liu and colleagues identified only three studies including a total of 372 patients that investigated the association between elevated troponin levels and mortality. As these studies revealed significant heterogeneities, the authors decided not to perform a meta-analysis [
17]. Labarrere et al. investigated the value of persistent troponin I (TnI) levels in 110 HTX patients during first year after HTX. They found that persistent TnI levels greater than 0.5 ng/ml were associated with development of coronary artery disease and graft failure. However, as patients were only included if they had survived the first year after HTX, association of postoperatively elevated TnI and early mortality was not investigated [
22]. Another study investigated the prognostic value of postoperative hsTnT for 1-year mortality in 141 HTX patients. They identified that elevated hsTnT levels at 6 weeks after HTX were highly associated with 1-year mortality. Again, association of early postoperative levels of hsTnT was not investigated by the authors [
23]. The last study by Franeková et al. demonstrated an association of hsTnT levels at 10 days after HTX and 1-year mortality [
24]. However, an earlier postoperative assessment of risk might be favorable as it might change clinical practice for patients at risk.
Postoperative troponin release has been extensively studied in non-cardiac surgery before [
25,
26]. In these studies, early postoperative troponin elevation above the upper limit of normal was associated with major adverse events like mortality. In cardiac surgery however, this upper limit of normal troponin concentration is frequently exceeded by myocardial trauma due to surgery, with not necessarily higher risk for mortality [
13]. Therefore, Devereaux et al. defined new cutoffs for association of troponin and 30-day mortality after cardiac surgery corresponding with an hsTnI level 218 times the upper reference limit [
16]. Our recent study now adds data for association of early hsTnT with 1-year mortality after HTX. Patients who died within the first year after HTX had significantly higher troponin values at each timepoint of measurement. These findings were independently associated when adjusted for the IMPACT score. IMPACT score showed weak-to-moderate association with 1-year mortality in our cohort. This goes in line with previous reports, describing similar AUC in ROC analysis [
9,
11]. Addition of hsTnT level at 48 h improved the discrimination ability of IMPACT score. The net benefit using the combined model was also higher to identify patients at risk for 1-year mortality. Recently, similar risk prediction models and decision curve analyses for mortality were presented as effective to guide palliative care consultation [
27,
28]. Additionally, hsTnT levels were independently associated with low DAOH. This is an important finding, as this complements the current knowledge on more patient-centered outcomes in the field of end-stage heart failure and HTX surgery [
21,
29,
30].
In the Eurotransplant area, the responsible parties discuss the implementation of a cardiac allocation score (CAS) which is supposed to optimize the allocation of the limited donor organs. In the lung transplantation setting, a similar score already exists (Lung allocation score (LAS)) which is also used to prioritize waiting list candidates 12 years and older based on a combination of waitlist urgency and post-transplant survival. Based on the data of this study, postoperative hsTnT may also be included into such a score for early re-estimation of postoperative risk and prognosis. Further studies should focus on potential interventions depending on hsTnT values that might be able to prevent complications and finally to reduce mortality after HTX. These may include intensified monitoring or standardized protocols for strict follow-up of these patients.
Strengths and limitations
Strengths of the presented data include first, standardized troponin measurement at multiple time points with high data completeness; second, complete 12-month follow-up. Further, we did not only address mortality but also DOAH, a more patient-centered endpoint to quantify life impact [
19]. We are aware of the following limitations. First, as a single-center study, sample size and number of events was limited. However, only 2 variables (troponin and IMPACT) were included into the logistic model that can therefore be considered robust. Second, we cannot exclude that any external hospitalization took place. However, HTX patients are very closely connected to our center so that we consider the risk of misclassification bias was very limited. Further, although DAOH is a measure of life impact, we did not collect data on quality of life, another relevant patient-centered outcome. Finally, we chose hsTnT at 48 h after HTX as primary biomarker for our analysis as it showed the numerically strongest discrimination for 1-year mortality. However, ROC-AUC did not significantly differ from values at 24 h and 72 h. In this context an earlier timepoint like 24 h after HTX might be favorable for early re-estimation of postoperative risk in the clinical setting. Therefore, optimal cutoff and sampling time point should be investigated in a larger cohort.
The generalizability of these findings may be hampered by the single-center design. However, characteristics such as 1-year mortality were in line with the current literature.
Conclusion
Early hsTnT levels after HTX surgery are independently associated with poor 1-year survival and reduced DAOH. Therefore, early hsTnT concentrations might be useful for early risk reassessment to tailor postoperative therapy or decision-making in the intensive care unit.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.