Review of the TRACE Study

A clinical trial of the angiotensin-converting-enzyme inhibitor trandolapril in patients with left ventricular dysfunction after myocardial infarction

NEJM 1995;333:1670-6.

Background Up to this point in history, a series of trials had been conducted using ACEi’s in post-MI patients. A small to moderate short-term benefit had been shown when the drugs were started immediately (GISSI-3 and ISIS-4) and much greater long-term benefits were demonstrated when the drugs were started 5-11 days, on average, following AMI in patients with LV dysfunction and congestive heart failure (SAVE and AIRE).

The SAVE and AIRE trials, however, were more selective and it was not clear how representative they were among all potentially eligible patients. Thus, TRACE authors sought to re-test the hypotheses tested in SAVE and AIRE with a focus on generalizability of trial procedures and results. Specifically, the Trandolapril Cardiac Evaluation Study (TRACE) sought to test the hypothesis that trandolapril would reduce all-cause mortality in post-MI patients with LV dysfunction when used in the majority of consecutively screened, potentially eligible patients.

Patients Consecutive patients ≥18 years of age who were hospitalized with a confirmed AMI were screened between day 2 and 6 after the onset of symptoms. All screen eligible patients underwent echocardiography and those with a wall motion index of ≤1.2, which corresponds to an EF ≤35%, were considered for enrollment. The key exclusion criteria included an absolute or relative contraindication to an ACEi or a definite need for an ACEi, severe uncontrolled diabetes, a serum sodium <125 mmol per liter or a serum creatinine >2.3 mg/dl.

Baseline characteristics The average age of patients was 68 years and 72% were men. Approximately one third of patients had a prior MI, 13% had diabetes, 23% had hypertension and smoking status was not listed. The average wall motion index was 1.0. Two thirds of patients had a Q wave MI (anterior 47% and inferior 19%).

The mean time to randomization was 4.5 days. Forty-five percent of patients received thrombolysis. The average blood pressure and heart rate were 120/70 mmHg and 76 beats per minute, respectively. At the time of randomization 16% of patients were receiving a beta blocker and 28% digoxin. Before randomization, 60% of patients had been classified as Killip class ≥2 and at the time of randomization it was 21%.

A total of 6,676 consecutive patients experienced an AMI of whom 2,606 had a wall motion index of ≤1.2. There was an inverse relationship between wall motion index and mortality. In patients with scores ≥1.3, 40% had signs of CHF and the 1-year mortality rate was 12%. Among patients with scores ≤1.2, 74% had signs of CHF and the 1-year mortality was 34%.

Of the 2,606 eligible patients, 859 (33%) were excluded. The most common reasons for exclusion included need for mandatory ACE inhibition [6%], cardiogenic shock [4%], death during screening [3%], renal failure or a single kidney [2%], intolerance of the test dose of trandolapril [1%], lack of consent [8%], or other reasons [8%].

Altogether, 1,749 (67%) of patients with a wall motion index score ≤1.2 were enrolled.

Procedures Eligible patients were given a test dose of 0.5 mg of trandolapril, which led to the exclusion of 1% of patients. These patients were not included in the ITT analysis. Double-blind medication was started between day 3 and day 7 after AMI. Patients were randomly assigned to receive 1 mg of trandolapril once daily or matching placebo. After two days, the dose was increased to 2 mg once daily. After four weeks, the dose was again increased, to 4 mg once daily. If the highest dose was not tolerated, patients could continue with a dose of 2 mg or 1 mg once daily, but the drug was withdrawn if a dose of 1 mg once daily was not tolerated.

Outpatient visits were scheduled one and three months after the infarction, with subsequent visits every three months. Echocardiography was repeated after 3, 6, and 12 months.

The original protocol specified that treatment would continue for at least 12 months. When the results of the SAVE study were published in 1992, showing no survival benefit until after almost one year of treatment with ACE inhibitors, the steering committee decided (without any knowledge of the results of the study) to extend the closing date to 24 months after the last random assignment.

Endpoints The primary study endpoint was all-cause mortality. Secondary endpoints were death from a cardiovascular cause, sudden death, progression to severe heart failure, recurrent MI, and change in wall motion index.

The investigators estimated they would need a sample size of 1,500 patients to detect a 25% relative reduction in the risk of death with 80% power and 1-sided alpha of 2.5%. This was based on an estimated death rate of 30% at 12 months in the placebo group; however, the steering committee increased the sample size to 1,860 patients to allow for the possibility of a lower-than-expected placebo mortality rate.

In the spring of 1992 the overall mortality of randomized patients followed for 1 year was 24%. Inclusion of patients was therefore terminated at the end of June 1992 at the point where 1,749 patients had been randomized.

Results The final analysis included 1,749 patients; 876 in the trandolapril group and 873 in the placebo group.

Information on the percentage of patients discharged on various doses of the study drug are not provided.

Compared to placebo, trandolapril significantly reduced all cause death by 22% [(35% vs 42%; 95 percent confidence interval, 0.67 to 0.91 p = 0.001)}. The mortality curves diverged early (Kaplan–Meier estimate of mortality at one month 9% vs 11%) and continued to diverge throughout the follow-up period. Trandolapril also significantly reduced secondary endpoints, including death from CV causes, sudden death, and progression to severe heart failure but it did not significantly reduce reinfarction.

Examination of subgroups showed no evidence of treatment effect heterogeneity for all cause mortality, but again, similar to the SAVE and AIRE trials, the size of TRACE limits subgroup testing.

Premature withdrawals from study drug, not including death, occurred in 37% of patients in the trandolapril group compared to 36% in the placebo group. The most common reason for withdrawal was need for treatment with an open-label ACEi and this occurred more in the placebo group. Withdrawal due to cough, hypotension and reduction in kidney function were rare in both groups but slightly more common in patients on trandolapril compared to placebo.

Conclusions In the majority (67%) of consecutively screened patients with AMI complicated by left ventricular dysfunction, trandolapril significantly reduced death over at least 2 years of follow-up with a number needed to treat of approximately 14 patients.

Results from TRACE strengthen support for ACEi in post-MI patients, and the trial has high external validity. It not only tested the intervention in two-thirds of potentially eligible patients but was highly transparent about why patients were excluded.

A clinician looking to apply the procedures used in TRACE to the management of patients in clinical practice would not have to guess whether or not their patient would have been included. This is rare in clinical research and the investigators should be applauded for their efforts.

Investigators studying ACEi in post-MI patients have triangulated the population of patients who benefit from this therapy. In our opinion, TRACE provides the final piece to the puzzle. There is no doubt about the clinical efficacy of these drugs in the overwhelming majority of post-MI patients and higher risk patients stand to benefit the most.