The Prognostic Value of Troponin T for Long-term Outcome After Cardiac Surgery

Hypothesis and purpose:

The primary hypothesis of this study is that the amount of troponin T release after cardiac surgery is associated with mortality and cardiac morbidity within 1 year after surgery. Patients undergoing coronary procedures using cardiopulmonary bypass (CPB) (i.e., CABG, valvular and combined on-pump procedures) will be analysed.

A further purpose of the study is to evaluate the outcome of patients undergoing cardiac procedures at our institution, as this is currently not quantified.

Study design:

After ethical approval, we will conduct our study in two parts. In the first part, we will perform a retrospective cohort study at the University Hospital Basel to assess the 1 year outcome in all patients who have undergone cardiac surgery at our institution between January 1, 2005 and December 31, 2006. In addition, we will identify the best cut-off value of postoperative troponin T for predicting morbidity and mortality 1 year after open-heart surgery.

The second part of the study will be performed between January 1, 2007 and December 31, 2008. We will continue the quality control study and will prospectively test the clinical value of elevated troponin T values, according to the risk model developed in the first part, for predicting for long-term (1 year) morbidity and mortality. The prospective part of the study can include patients operated on as early as January 1, 2007 because the prospective follow-up will take place 12 months after surgery.

Patient selection:

Inclusion criteria: all patients who underwent or will undergo cardiac surgery using cardio-pulmonary bypass at the University Hospital Basel between January 1, 2005 and December 31, 2008 will be eligible.

For the first part of the study, patients who had cardiac surgery between January 1, 2005 and December 31, 2006 will be analysed retrospectively.

For the prospective part of the study, all patients will be studied who will have surgery between January 1, 2007 and December 31, 2008.

Exclusion criteria:

• Cardiac surgery without the use of cardiopulmonary bypass
• Procedures requiring circulatory arrest
• Penetrating cardiac trauma
• Lack of informed consent.

Baseline variables:

Relevant patient data including: history, functional state, medication, EuroSCORE8, and haematological, basic renal, hepatic and metabolic chemistry findings (the routinely performed preoperative laboratory tests in our institution, including preoperative C-reactive protein and troponin T in the presence of a recent (<7 days) ACS will be collected in the retrospective as well as in the prospective cohorts. Pre-operative 12-lead ECG will only be analysed in the prospective cohort.

Peri-operative surgical and clinical variables:

Surgical procedures and techniques, operative data (operating time, blood loss, cardiopulmonary and cross-clamp time, number of arterial and venous grafts and number of distal anastomoses) will be assessed. Early postoperative therapy or events in the intensive care unit, including total time on mechanical ventilation, postoperative inotropic support, blood transfusion and renal replacement therapy, lengths of intensive care unit and hospital stay, re-operation, new onset ventricular arrhythmia and Q-wave PMI will be recorded in both parts of the study study.

Cardiac marker analyses and 12-lead ECG:

Troponin T levels, routinely measured on the first and second postoperative day, will be used for analysis in both parts of the study. Also, the routinely registered 12-lead ECG at discharge will be compared with the preoperative baseline ECG and evaluated for ischemic changes according to the Minnesota criteria for Q- and R-waves as well as for ST-deviation and T-wave changes in the prospective cohort.

Clinical endpoints:

Primary endpoint: All-cause mortality within one year after surgery.

Secondary endpoint: Major adverse cardiac events, defined as cardiac mortality, myocardial infarction, congestive heart failure requiring hospitalisation and need for surgical or percutaneous coronary intervention within 1 year.

The outcome measures will be obtained by reviewing hospital charts and by direct telephone contact with the patients.

Sample Size:

For the retrospective part, we anticipate a sample size of 1000 patients for the planned study period of 24 months. Estimating a 1-year mortality of 6%, we expect 60 adverse events. This number of adverse events will allow for a statistically robust multivariate risk model based on 6 variables. We assume that a sample size of 660 patients will allow for retesting of the model in the prospective part of the study.

Statistical Analysis:

Analyses will be performed for the whole cohort as well as separately for patients with each type of cardiac surgery, i.e., for patients undergoing CABG, aortic valve surgery, mitral valve surgery, combined surgeries (CABG plus valve surgery) and for patients with a recent(<7 days) ACS before surgery.

Continuous data will be presented as mean ± SD and comparison between groups calculated by an unpaired t-test. Categorical data will be presented as absolute numbers and percentages and comparisons calculated by Pearson's χ2 test. Receiver operating characteristics curves will be computed to establish cut-off values of troponin T and its likelihood ratio for the outcome measures.

Multivariate logistic regression will be used to adjust the association of troponin T elevation and outcome for established clinical risk parameters. This risk model, developed in the retrospective first part of the study, will thereafter be retested in the prospective second part.

A p value <0.05 will be defined as pointing to significance. All analyses will be calculated by SPSS Version 14.