Do ACE inhibitors increase mortality in CABG patients?

Tuesday, August 18, 2009

by Shelley Wood

Bristol, UK - Authors of a large, observational study say preoperative use of ACE inhibitors may double the risk of dying after CABG surgery [1]. Dr Antonio Miceli (University of Bristol, UK) and colleagues say their findings support a strategy of stopping ACE inhibitors presurgery and then restarting them postoperatively. But not everyone agrees with the results of the new analysis: commenting on the study for heartwire, Dr Harold Lazar (Boston University School of Medicine, MA) argued that there are some patients in whom ACE inhibitors should not be stopped, particularly patients with hypertension and/or diabetes, who have had a recent MI.

The study will be published in an upcoming issue of the Journal of the American College of Cardiology and was published online August 13, 2009.

Miceli et al's research adds a generous dollop of data to an already confusing mélange of studies supporting or rejecting a role for ACE inhibitors in CABG patients. As Miceli et al explain, ACE inhibitors lower blood pressure and have antiatherosclerotic, antithrombotic, and anti-inflammatory effects—all of which paved the way for recommendations to use them in patients with coronary artery disease. But their effect when used preoperatively in CABG patients "is still controversial," they note.

To gauge these effects, Miceli and colleagues reviewed data on preoperative ACE-inhibitor use among more than 10 000 patients who had CABG surgery between April 1996 and May 2008 at the Bristol Heart Institute, UK. In all, 3052 were taking ACE inhibitors prior to their surgery and were matched with a control group using propensity scores.

Miceli et al report that the mortality rate in the ACE-inhibitor group was nearly double that of patients not taking ACE inhibitors: 1.3% vs 0.7%. As well, new-onset atrial fibrillation, renal dysfunction, and use of inotropic support were all higher among patients taking ACE inhibitors preoperatively. Use of ACE inhibitors remained an independent predictor of mortality after multivariate analysis.

Postoperative outcomes

ACE inhibitors (%)
Control (%)
Odds ratio (95% CI)
2.0 (1.17-3.42)
Atrial fibrillation
1.34 (1.18-1.51)
Renal dysfunction
1.36 (1.1-1.67)
Use of inotropic support
1.22 (1.1-1.36)

The authors note that their findings on the link between preoperative ACE-inhibitor use and need for inotropic support largely reflect earlier studies, while the findings regarding kidney disease, atrial fibrillation, and mortality are at odds with at least some of the published literature. The different findings are partially explained by the fact that the current study includes a larger sample size than most of the prior studies and was restricted to patients undergoing CABG, the authors write.

Asked to review the paper for heartwire, Lazar took issue what he identified as a number of limitations of the study (many of which were also acknowledged by Miceli et al in their paper). Lazar pointed to the shortcomings of a propensity analysis, when a randomized clinical trial is the only way to properly address the safety of ACE inhibitors in the setting of CABG. He also questioned the 12-year duration of the study, noting that the kinds of patients being treated for CABG changed substantially over the study period. He also pointed to "missing information" in the paper, such as that regarding dose, type, and duration of ACE-inhibitor therapy; indications for, dose, type, and duration of inotropic support; details on atrial fibrillation and its management in the study cohort; and a breakdown of the cause of death reported in the study. "Were these cardiac deaths, or were they other things that ACE inhibitors have no effect on?" Lazar asked.

Overall, Lazar concluded, "the results are very soft. . . . The use of ACE inhibitors in this study was associated with a higher incidence of death; however, although this may be true statistically, clinically I don't think we're talking about a real, significant end point."

Even the mortality difference between the two groups—a difference of 0.6%—"that's really splitting hairs," he said.

"In my own practice, for an elective operation, we will ask the patient to stop the ACE inhibitor 48 hours before the surgery, and if it's an in-house patient and we can stop it safely, again we'll stop ACE inhibitors 48 hours beforehand. However, in those patients who have had a recent MI with hypertension and those who are diabetic, we will keep the ACE inhibitor going right up to the time of surgery. We will also lower the dose sometimes as well. We feel that gives us the protection of the ACE inhibitor but minimizes the vasodilatation associated with the drug."

But the authors stand by their findings, insisting the best strategy is to stop the ACE inhibitors two days prior to surgery and restart them after the patient is stable again postsurgery. Speaking with heartwire, senior author on the study, Dr Massimo Caputo (University of Bristol), agreed that survival after CABG is higher now than ever before, "so the fact that we found a statistically significant difference in mortality is quite astonishing," he said. But he agrees, in terms of clinical impact, that the adverse effects of ACE inhibitors are even more important for atrial fibrillation and renal dysfunction, "where there is a much clearer difference between the groups."

"Our goal was not to say that ACE inhibitors are bad or should not be used. We're just saying that in the perioperative period, you should stop ACE inhibitors two to five days before and restart them probably two days postoperatively, to reduce the risk of the complications we saw."

And while a large, prospective, randomized trial would be ideal, Caputo said his study fills an important gap and "used the best possible data, to determine the best possible therapy for our patients."


  1. Miceli A, Capoun R, Fino C, et al. Effects of angiotensin-converting enzyme inhibitor therapy on clinical outcome in patients undergoing coronary artery bypass grafting. J Am Coll Cardiol 2009; DOIi:10.1016/j.jacc.2009.07.008. Available at:


Post a Comment