Author + information
- †Center for Clinical and Epidemiological Research, Division of Internal Medicine, University Hospital, University of São Paulo, São Paulo, Brazil
- ‡4th Infantry Division Troop Medical Clinic, NATO Role 3 Multinational Medical Unit, Kandahar, Afghanistan
- §Cardiology Service, Department of Internal Medicine, Walter Reed National Military Medical Center, Bethesda, Maryland
- ↵∗Reprint requests and correspondence:
Dr. Márcio Sommer Bittencourt, Center for Clinical and Epidemiological Research, Division of Internal Medicine, University Hospital, University of São Paulo, Avenue Lineu Prestes 2565, Butantã, São Paulo, Brazil CEP 05508-000.
Although considerable discussion has recently permeated the use of coronary artery bypass grafts (CABG) for the initial treatment of stable coronary artery disease (CAD), CABG is currently considered a Class I indication to improve survival in individuals with unprotected left main coronary artery obstruction and 2- or 3-vessel disease with proximal left anterior descending coronary artery disease, as well as to improve symptoms in individuals with significant stenosis amenable to revascularization with unacceptable angina despite medical therapy (1). Still, the long-term efficacy of the surgical procedure is hindered both by the progression of native vessel CAD and by chronic graft failure (2). Fifteen percent of saphenous vein grafts occlude during the first post-operative year, and up to 1 in every 2 patients will develop failure of at least 1 venous graft in the longer term. Furthermore, 15% of left internal mammary artery grafts occlude after 10 years (3). Given that more than 1 CABG procedure is performed per 1,000 U.S. adults per year (4), the clinical and socioeconomic impact of those late CABG complications should not be underestimated. Some studies have previously tried to identify individuals with graft failure or CAD progression to better understand the impact of those findings on clinical outcomes.
A classic study by Lytle et al. (5), including more than 1,000 individuals in a retrospective cohort with a mean follow-up of 83 months, demonstrated that among various predictors, stenosis of the vein graft to the left anterior descending coronary artery was associated with a worse prognosis, whereas grafts to other vessels were not. Another analysis restricted to early (<18 months after CABG) graft failure demonstrated that even nonobstructive early graft failure was a strong predictor of long-term outcomes (6). Nonetheless, a more recent study, including 1,829 individuals who underwent routine invasive angiography after CABG and were followed for a shorter term of 18 months, was unable to detect a significant increase in the hazard risk for cardiovascular (CV) death or myocardial infarction (MI) in individuals with graft failure (7).
Because of the potential risk associated with invasive procedures, several studies have evaluated the role of noninvasive stress imaging tests to stratify risk in this population. A study using thallium-exercise testing on 873 asymptomatic individuals who underwent CABG demonstrated that the detection of any perfusion defect was associated with a 2-fold increase in the risk of death or MI (8). Similarly, a small study including 95 patients who underwent stress echocardiography revealed that a low dose score, which represents the extent of viable myocardium, was independently associated with prognosis after CABG (9). This limited evidence was used to support the potential indications for re-do CABG in selected individuals according to the American College of Cardiology/American Heart Association 2004 Guideline Update for CABG Surgery (10), though more recent guidelines do not clearly address this issue.
More recently, 2 studies have evaluated the role of coronary computed tomography angiography (CTA) as a predictor of outcomes after CABG (11,12). Those studies have developed 2 different scores to accommodate the complexity of the anatomic evaluation after CABG. The first calculated the unprotected coronary territories (UCT), whereas the second combined the number of diseased vessels with the number of UCT. Although both studies were limited by the short-term follow-up of only 20 months, 1 of the studies suggested that UCT was independently associated with an increased rate of CV death or MI, whereas the other suggested the combined score was a better predictor for all-cause mortality (11,12).
In this issue of iJACC, Mushtaq et al. (13) have made a substantial and important contribution to the limited contemporary evidence base evaluating the prognostic significance of anatomic CAD after CABG. This work provides the largest, longest study of the value of coronary CTA to evaluate the long-term prognosis of patients who underwent CABG surgery. The authors should be congratulated, not only for the large sample size, but also for the impressive 98% complete follow-up of more than 700 individuals followed for a mean of 73 months. Consistent with prior studies, the current data clearly demonstrate that both the number of UCT and the combination of number of diseased vessels with the number of UCT are both strong independent predictors of long-term outcomes in this population. Moreover, it has also been revealed that the more complex score was no better than the number of UCT. On the basis of the current results, individuals with 2 or 3 UCT have, respectively, an 8-fold and a 10-fold increase in the hazard risk for future CV death or MI, after extensive adjustment for confounding. One important aspect of the results is that even the results of prior stress testing were included in the multivariable adjustments.
A second important aspect, which should be pointed out, is that the prevalence of any graft disease was 26%. As expected, the prevalence of diseased vein grafts was 37%, whereas only 9% of the left internal mammary arteries were diseased. Interestingly, the vast majority of the diseased segments were occluded (18%), whereas only 8% of the grafts presented with nonocclusive disease.
Some aspects of the current analysis, however, should be interpreted with caution. Although the multivariable model included an extensive list of potential confounders, the potential for residual confounding should always be considered. Some known confounders which were not included in the current analysis include symptoms, indications for coronary CTA, left ventricular function, and chronic kidney disease. Additionally, although the authors included the stress test results in the model, the oversimplification of positive versus negative results underestimates the potential predictive value of those tests. Another important limitation for the external validation of the current cohort is that coronary CTA was only performed in a selected group of post-CABG individuals with some clinical indication for the test (symptoms or prior positive stress test). This selection bias may have resulted in a higher prevalence of graft disease and occlusion, as well as an increased rate of events, which may not be generalizable to other cohorts.
How can we translate the current results into clinical practice? The easy answer is that higher-risk individuals should be treated more aggressively, though individuals in the post-CABG scenario should already have been treated aggressively with the entire CAD secondary prevention armament, and no additional options for medical therapy generally remain. The second, yet more important question is whether the current results support the use of UCT to help discuss the risk–benefit ratio of reintervention versus continued medical therapy for patients with CAD progression post-CABG. One might be tempted to take this even further and consider performing screening CTA of asymptomatic patients after CABG, similar to a consideration for physiological stress testing more than 2 years post-CABG. However, at this stage, the current results are thought provoking with regard to how they could influence treatment decisions, yet they remain observational and do not evaluate the potential benefit of re-intervention. Thus, additional studies such as propensity-matched analysis of CABG registries or, ideally, randomized trials are needed to clarify better the role for CTA after CABG. However, the high prevalence of UCT and its impressively strong association with CV death or MI suggest that coronary CTA should at least be considered as part of the toolbox for the investigation of symptoms in individuals after CABG.
↵∗ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Imaging or the American College of Cardiology.
The opinions and assertions contained herein are the authors' alone and do not represent the views of Walter Reed National Military Medical Center, the U.S. Army, the Department of Defense, or NATO. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Hillis L.D.,
- Smith P.K.,
- Anderson J.L.,
- et al.
- Motwani J.G.,
- Topol E.J.
- Goldman S.,
- Zadina K.,
- Moritz T.,
- et al.
- Lopes R.D.,
- Mehta R.H.,
- Hafley G.E.,
- et al.
- Sawada S.,
- Bapat A.,
- Vaz D.,
- et al.
- Chow B.J.,
- Ahmed O.,
- Small G.,
- et al.
- Small G.R.,
- Yam Y.,
- Chen L.,
- et al.
- Mushtaq S.,
- Andreini D.,
- Pontone G.,
- et al.