Original Contribution

Lower Repeat Revascularization Rates Among Patients With Prior Coronary Artery Bypass Graft Surgery are Due to Lack of Adequate Target Vessels

Anirudh Kumar, MD1,3;  John F. Wagener, MD1,3;  Daniel Wojdyla, MS1;  W. Schuyler Jones, MD1,2;  Manesh R. Patel, MD1;  Sunil V. Rao, MD1,2

Anirudh Kumar, MD1,3;  John F. Wagener, MD1,3;  Daniel Wojdyla, MS1;  W. Schuyler Jones, MD1,2;  Manesh R. Patel, MD1;  Sunil V. Rao, MD1,2

Abstract: Background. Studies comparing percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG) surgery in patients with multivessel coronary artery disease (CAD) have shown lower repeat revascularization rates in patients who undergo CABG. The reason remains unclear. Methods. We identified patients with multivessel CAD who received CABG or PCI enrolled in the Duke Databank for Cardiovascular Disease (2003 to 2012). We compared the incidence of major adverse cardiovascular and cerebrovascular events (MACCE) between the two groups. Clinically performed follow-up angiograms for CABG patients were reviewed to determine adequacy of intervenable targets. Results. A total of 1555 patients were included: 861 underwent PCI and 694 underwent CABG. Patients with index PCI were more often female, African-American, presented with ST-elevation myocardial infarction (MI), and had previous MI; they were less often diabetic and had less heart failure or proximal left anterior descending disease. The adjusted hazard ratio of MACCE for CABG vs PCI was 0.68 (95% confidence interval, 0.58–0.80; P<.001). The adjusted odds ratio for repeat revascularization for CABG vs PCI was 0.45 (95% confidence interval, 0.28–0.72; P<.001). Fifty-seven patients with index CABG were found to have ≥1 occluded graft on subsequent angiography without repeat revascularization; 48 patients (6.9%) had inadequate targets for intervention. Conclusion. Among patients with multivessel CAD, repeat revascularization rates are lower among CABG patients compared with PCI patients. However, a high proportion of CABG patients with occluded grafts on repeat angiography lack targets for repeat revascularization. This may partially explain the disparity in repeat revascularization rates and suggests that future comparison studies should additionally assess angiographic outcomes.

J INVASIVE CARDIOL 2018;30(2):51-55. Epub 2017 November 15.

Key words: multivessel coronary artery disease, angiographic outcomes, long-term outcomes

Current American College of Cardiology/American Heart Association guidelines recommend that coronary artery bypass graft (CABG) surgery should be favored in patients with multivessel coronary artery disease (CAD) and complex coronary lesions and anatomy.1 The data for these recommendations are predicated on randomized controlled trials and large observational studies demonstrating CABG to have lower major adverse cardiac or cerebrovascular events (MACCE), including all-cause death, stroke, myocardial infarction (MI), and repeat revascularization compared with percutaneous coronary intervention (PCI).2-8 This lower MACCE rate, however, was largely driven by a lower rate of repeat revascularization, with similar rates of all-cause death, stroke, and MI between CABG and PCI in the randomized trials. Two trials comparing the treatment modalities in patients with significant left main disease have come to differing conclusions regarding MACCE including MI; however, both indicated a higher risk for repeat revascularization with PCI.9,10 The reason for this difference in repeat revascularization is not known, but it has been cited as a limitation of currently available PCI techniques.

On the other hand, a potential explanation for this difference is that progression of native CAD in patients who undergo CABG results in no targets amenable for repeat revascularization, suggesting that a more equitable comparison between CABG and PCI is to examine rates of graft failure. Accordingly, we retrospectively examined a cohort of patients with multivessel CAD who underwent PCI and CABG and hypothesized that among patients who underwent index CABG and presented with at least one graft occlusion on repeat angiography, the lack of repeat revascularization was due to inadequate native coronary targets.


Data source and study sample. We undertook a retrospective review of the Duke Database for Cardiovascular Disease identifying consecutive patients with multivessel CAD between August 1, 2003, and November 30, 2012. The Duke University Medical Center institutional review board approved this study. The informed consent requirement was waived.

The study sample consisted of patients 18 years of age or older with angiographically confirmed multivessel CAD with stenoses of >70% of the vessel diameter in major epicardial vessels in the territories of at least three coronary arteries or the left anterior descending artery and one other vessel, and received PCI or isolated CABG within 30 days after initial diagnostic cardiac catheterization. Revascularization strategy for each patient was determined based on the patient’s background and comorbid diseases in a Heart Team conference. The study excluded: patients who were lost to follow-up; patients who had left main stenosis ≥50%, congenital heart disease, or moderate/severe valvular disease; and patients who had prior PCI, CABG, valve, aortic, or great vessel surgery.

Study definitions and outcomes. Demographic information, baseline characteristics, clinical presentation, and occurrence of MACCE were compared between those who underwent initial PCI and those who underwent initial CABG. We then identified patients who underwent initial CABG and had subsequent coronary angiography demonstrating at least one occluded bypass graft without repeat revascularization. Two senior interventional cardiologists (SVR, WSJ) reviewed the angiograms of these patients to determine if there was a native coronary vessel in which PCI was technically feasible and appropriate. Disagreements were adjudicated by a third expert (MRP).

Statistical analysis. Demographic and clinical characteristics were summarized for patients in the PCI and CABG groups using medians and interquartile ranges (IQR; 25th-75th percentile) for continuous variables and frequencies and percentages for categorical variables. The two groups were compared using Chi-square tests for categorical variables and Wilcoxon tests for continuous variables. Kaplan-Meier estimates were derived for overall MACCE and all-cause death; cumulative incidences accounting for competing risk due to mortality were estimated for MI, stroke, and repeat revascularization. The association between index revascularization strategy and MACCE was assessed using a multivariable Cox regression model, adjusting for baseline covariates. Continuous variables were tested for linearity and included as restricted cubic splines if needed. The associations were summarized using hazard ratio (HR) with 95% confidence interval (CI). The proportional hazards assumption regarding the two cohorts was assessed using scaled Schoenfeld residuals, and alternative models were explored. Revascularization within 30 days of repeated angiography was compared between index revascularization groups using a logistic regression model. Results are presented as adjusted odds ratio (OR) with 95% CI. A two-sided P-value ≤.05 was considered statistically significant. SAS software, version 9.4 (SAS Institute), was used for all statistical analyses.


A total of 39,109 patients during the selected time period were reviewed; 37,554 patients were excluded mostly due to lack of multivessel CAD as defined above or a history of prior cardiothoracic surgery (Figure 1). A total of 1555 patients were included in this study (Table 1). Of these, 694 underwent initial CABG and 861 underwent initial PCI. The median ages were 63 years (IQR, 54-73 years) and 63 years (IQR, 55-70 years) in the CABG and PCI cohorts, respectively. The median follow-up period was 5 years (IQR, 2.7-7.3 years). Compared to patients with index CABG, patients with index PCI were more often female, African-American, presented with ST-elevation MI, and had previous MI; they less often were diabetic, less often had congestive heart failure, less often had proximal left anterior descending disease, and were less often revascularized in the setting of stable angina. There were no intergroup differences in age, ejection fraction, body mass index, systolic blood pressure on presentation, incidence of hypertension, cerebrovascular disease, peripheral vascular disease, chronic kidney disease, chronic obstructive pulmonary disease, or current or previous use of tobacco. 

The 10-year rate of overall MACCE was 54.1% for CABG vs 67.3% for PCI (P<.001), while the all-cause death rate was 38.6% for CABG vs 45.1% for PCI. The estimated 10-year cumulative incidence of MI was 8.5% for CABG vs 14.1% for PCI, while incidence of stroke was 9.6% for CABG vs 8.8% for PCI, and incidence of repeat revascularization was 12.0% for CABG vs 24.7% for PCI (Table 2). Figure 2 presents Kaplan-Meier estimates of MACCE from time of index revascularization, stratified by CABG vs PCI. Using a multivariable Cox regression model adjusting for baseline covariates, HR for first occurrence of MACCE for CABG vs PCI was 0.68 (95% CI, 0.58-0.80; P<.001). The proportional hazard assumption was not met (P=.01) and a piecewise Cox regression model was derived to allow different hazard ratios in the first and subsequent years. The adjusted HR for MACCE within the first year after index CABG vs PCI was 0.59 (95% CI, 0.46-0.74; P<.001) and after the first year was 0.78 (95% CI, 0.63-0.98; P=.03) (Table 3). 

A total of 164 patients (23.6%) in the CABG arm and 295 patients (34.3%) in the PCI arm underwent repeat angiography (Population II) (Figure 3). In the index CABG group, 95 patients received medical treatment and 69 patients underwent repeat revascularization. Of the 95 patients who underwent index CABG and received subsequent medical therapy, 57 patients (8.2%), 48 of whom presented with acute coronary syndromes, were found to have ≥1 occluded graft. Forty-eight of these patients (6.9%) were found to have no targets amenable to revascularization. Of those with index PCI who had repeat angiography, a total of 102 received medical treatment while 193 underwent repeat revascularization. Adjusted OR for repeat revascularization for index CABG vs index PCI was 0.45 (95% CI, 0.28-0.72; P<.001). 


The main findings of our study are as follows: (1) CABG is associated with a lower risk for MACCE, including repeat revascularization; (2) among patients with index CABG who received medical management despite subsequent angiography demonstrating at least one occluded graft, 84.2% had no targets for revascularization. These data suggest that the use of “repeat revascularization” as an endpoint in trials comparing PCI with CABG may not appropriately assess the durability of drug-eluting stents with that of saphenous vein grafts. 

Revascularization for multivessel CAD has remained a controversial topic, with a wealth of data demonstrating an increase in MACCE with PCI, primarily driven by greater repeat revascularization. Most recently, the BEST trial demonstrated increased MACCE among patients with multivessel CAD treated with PCI with everolimus-eluting stents as opposed to CABG with similar rates of death, MI, and stroke between the two groups.11 Similar findings were noted in a non-randomized study of patients with multivessel CAD and chronic kidney disease, defined as glomerular filtration rate ≤60 mL/min/1.73 m2, when managed with PCI with everolimus-eluting stents vs CABG.12 Among patients with diabetes mellitus and multivessel CAD, the FREEDOM trial demonstrated superior clinical outcomes, including mortality and MI, in patients with multivessel CAD and diabetes mellitus who underwent CABG compared with those who underwent PCI with drug-eluting stent implantation.13 Subsequent studies comparing the two modalities in diabetic patients aged ≥65 years and presenting with moderate-risk and high-risk non-ST elevation acute coronary syndromes demonstrated similar outcomes between CABG and PCI.14,15

The premise that patients receiving CABG are less likely to undergo subsequent revascularization because of progression of underlying native CAD has not been previously explored. In our study, we demonstrated that a high percentage of patients undergoing CABG were not candidates for repeat revascularization due to progression of native CAD leading to no adequate targets. This unique finding represents an important perspective to the ongoing debate regarding the utilization of PCI and CABG for multivessel CAD. With numerous randomized and cohort studies demonstrating similar rates of death, MI, and stroke between those receiving PCI vs CABG, particularly among non-diabetics, the need for repeat revascularization has been the primary driver behind increased MACCE and guideline recommendations favoring CABG over PCI. The concept that patients undergoing CABG do not undergo repeat revascularization not because it is unnecessary, but rather because it is not technically feasible, represents a significant change of thinking in the current clinical paradigm. If considering patients with CABG who become “incompletely revascularized” as treatment failure, incorporation of them into MACCE may result in less disparate outcomes between PCI and CABG. Published data suggest that failure rates of both grafts and stents are quite high.16,17 As such, assessing the rates of graft and stent failure may facilitate a more fair comparison between PCI and CABG. Future studies are needed to further analyze this scenario.

Patients in this study who underwent PCI included revascularization via plain old balloon angioplasty, bare-metal stents, and first-generation drug-eluting stents. Second-generation drug-eluting stents, in comparison, have thinner struts and more biocompatible polymer properties that reduce inflammation, promote faster vessel healing, and reduce the risk of restenosis and stent thrombosis.18 In addition, thinner struts potentially result in fewer side-branch occlusions and periprocedural MIs.19 The advent of third-generation drug-eluting stents and completely biodegradable polymeric stents is currently being evaluated and may potentially result in improved outcomes. 

Study limitations. First, although the results of our study are compatible with those demonstrated in prior randomized trials, this study is a retrospective cohort study with a relatively small number of included patients at a single center. Second, heterogeneity between those receiving PCI and CABG had a potential effect on the results of this study, although a Cox regression analysis was performed to minimize this effect. Third, as noted above, technical advances in the fields of intervention (particularly third-generation and biodegradable drug-eluting stents) and multi-arterial CABG allow for further future investigations. Finally, our study is observational and thus may be subject to unmeasured confounding variables.


Among patients with multivessel CAD who undergo CABG and are found to have occluded grafts on repeat angiography, a high proportion lack adequate targets for repeat revascularization. This may help to explain the disparity in repeat revascularization rates between PCI and CABG. Future studies comparing PCI and CABG should also assess angiographic outcomes, such as stent and graft occlusion, in addition to clinical outcomes. 


1.    Levine GN, Bates ER, Blankenship JC, et al. 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. Circulation. 2011;124:e574-e651.

2.    BARI Investigators. The final 10-year follow-up results from the BARI randomized trial. J Am Coll Cardiol. 2007;49:1600-1606.

3.    Hueb W, Lopes N, Gersh BJ, et al. Ten-year follow-up survival of the Medicine, Angioplasty, or Surgery Study (MASS II): a randomized controlled clinical trial of 3 therapeutic strategies for multivessel coronary artery disease. Circulation. 2010;122:949-957.

4.    The SoS Investigators. Coronary artery bypass surgery versus percutaneous coronary intervention with stent implantation in patients with multivessel coronary artery disease (the stent or surgery trial): a randomized controlled trial. Lancet. 2002;360:965-970.

5.    Mohr FW, Morice MC, Kappetein AP, et al. Coronary artery bypass graft surgery versus percutaneous coronary intervention in patients with three-vessel disease and left main coronary disease: 5-year follow-up of the randomised, clinical SYNTAX trial. Lancet. 2013;381:629-638. 

6.    Serruys PW, Ong AT, van Herwerden LA, et al. Five-year outcomes after coronary stenting versus bypass surgery for the treatment of multivessel disease: the final analysis of the Arterial Revascularization Therapies Study (ARTS) randomized trial. J Am Coll Cardiol. 2005;46:575-581.

7.    Weintraub WS, Grau-Sepulveda MV, Weiss JM, et al. Comparative effectiveness of revascularization strategies. N Engl J Med. 2012;366:1467-1476.

8.    Deb S, Wijeysundera HC, Ko DT, Tsubota H, Hill S, Fremes SE. Coronary artery bypass graft surgery vs percutaneous interventions in coronary revascularization: a systematic review. JAMA. 2013;310:2086-2095.

9.    Stone GW, Sabik JF, Serruys PW, et al. Everolimus-eluting stents or bypass surgery for left main coronary artery disease. N Engl J Med. 2016;375:2223-2235. 

10.    Makikallio T, Holm NR, Lindsay M, et al. Percutaneous coronary angioplasty versus coronary artery bypass grafting in treatment of unprotected left main stenosis (NOBLE): a prospective, randomized, open-label, non-inferiority trial. Lancet. 2016;388:2743-2752.

11.    Park SJ, Ahn JM, Kim YH, et al. Trial of everolimus-eluting stents or bypass surgery for coronary disease. N Engl J Med. 2015;372:1204-1212.

12.    Bangalore S, Guo Y, Samadashvili Z, Blecker S, Xu J, Hannan EL. Revascularization in patients with multivessel coronary artery disease and chronic kidney disease.  J Am Coll Cardiol. 2015;66:1209-1220.

13.    Farkouh ME, Domanski M, Sleeper LA, et al. Strategies for multivessel revascularization in patients with diabetes. N Engl J Med. 2012;367:2375-2384.

14.    Ben-Gal Y, Mohr R, Feit F, et al. Surgical versus percutaneous coronary revascularization for multivessel disease in diabetic patients with non–ST-segment–elevation acute coronary syndrome: analysis from the acute catheterization and early intervention triage strategy trial. Circ Cardiovasc Interv. 2015;8:e002032.

15.    Naito R, Miyauchi K, Konishi H, et al. Comparing mortality between coronary artery bypass grafting and percutaneous coronary intervention with drug‑eluting stents in elderly with diabetes and multivessel coronary disease. Heart Vessels. 2016;31:1424-1429. 

16.    Hess CN, Lopes RD, Gibson CM, et al. Saphenous vein graft failure after coronary artery bypass surgery: insights from PREVENT IV. Circulation. 2014;130:1445-1451.

17.    Raber L, Wohlwend L, Wigger M, et al. Five-year clinical and angiographic outcomes of a randomized comparison of sirolimus-eluting and paclitaxel-eluting stents: results of the Sirolimus-Eluting Versus Paclitaxel-Eluting Stents for Coronary Revascularization LATE trial. Circulation. 2011;123:2819-2828.

18.    Kolandaivelu K, Swaminathan R, Gibson WJ, et al. Stent thrombogenicity early in high-risk interventional settings is driven by stent design and deployment and protected by polymer-drug coatings. Circulation. 2011;123:1400-1409.

19.    Popma JJ, Mauri L, O’Shaughnessy C, et al. Frequency and clinical consequences associated with sidebranch occlusion during stent implantation using zotarolimus-eluting and paclitaxel-eluting coronary stents. Circ Cardiovasc Interv. 2009;2:133-139.

From the 1Department of Medicine and Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; 2Department of Medicine, Durham Veterans Affair Medical Center, Durham, North Carolina; and 3Heart and Vascular Institute, Cleveland Clinic Foundation, Cleveland, Ohio.

Funding: This analysis was supported by The Duke Clinical Research Institute.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. None of the authors report conflicts of interest regarding the content herein.

Manuscript submitted July 2, 2017, provisional acceptance given July 20, 2017, final version accepted August 1, 2017.

Address for correspondence: Anirudh Kumar, MD, Heart and Vascular Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44114. Email: Anirudh.Kumar23@icloud.com