Abstract Aims Patients undergoing percutaneous coronary intervention (PCI), with a history of coronary artery bypass grafting (CABG), may be at increased risk for mortality and repeat revascularization, compared with patients without prior CABG. In this post-hoc analysis of the ORBIT II trial, safety and efficacy of coronary orbital atherectomy (OA) to modify severe coronary artery calcium, prior to stent placement, was evaluated in subjects based on history of CABG. Methods and results Comorbidities: diabetes, dyslipidemia, hypertension, and history of myocardial infarction (MI) were more prevalent in the CABG group. The in-hospital major adverse cardiac event (MACE) rate, defined as a composite of cardiac death, MI (CK-MB > 3 × ULN), and target vessel revascularization (TVR), was higher in the CABG group (16.9% vs. 8.5%, p = 0.04), driven primarily by a higher incidence of MI (16.9% vs. 8.0%, p = 0.03); however, Q-wave rates were low at 1.5% vs 0.5%, (p = 0.38). There was no significant difference in rates of cardiac death (6.2% vs. 2.7%, p = 0.17) and TVR (7.9% vs. 5.5%, p = 0.47). Conclusions Low rates of TVR, cardiac death, and Q-wave MI, suggest OA treatment to facilitate stent delivery is successful and provides durable outcomes in subjects with and without prior CABG. Condensed abstract Patients with history of CABG have extensive coronary artery disease. Those who undergo PCI may be at increased risk for mortality and repeat revascularization, compared with patients without prior CABG. This post-hoc analysis of ORBIT II trial evaluated safety and efficacy of coronary OA to modify severe coronary artery calcium, prior to stent placement, based on subject history of CABG. The MACE rate was higher in the CABG group, driven by higher incidence of MI; however, Q-wave rates were low. OA treatment to facilitate stent delivery is successful, but higher incidence of non-Q-wave MI in CABG patients warrants further study. 1 Introduction For patients who undergo coronary artery bypass grafting (CABG), the need for repeat revascularization is common. By 10 years post-CABG, 12% of patients undergo repeat revascularization. Severe coronary artery calcification is frequently found in patients who undergo CABG—nearly 33% of subjects in the SYNTAX CABG registry had heavily calcified lesions . Further, severe lesion calcification is associated with increased mortality related to myocardial infarction (MI) in patients undergoing CABG . Patients with a history of CABG are typically excluded from percutaneous coronary intervention (PCI) trials . PCI of severely calcified coronary lesions often results in poor outcomes: procedural failure, coronary dissection and thrombosis, MI, restenosis, death, target lesion revascularization (TLR), and other major adverse cardiac events (MACE) . The impact of CABG on patient outcomes has not previously been studied for PCI involving severely calcified native coronary arteries. The ORBIT II trial assessed the safety and efficacy of the coronary Orbital Atherectomy System (OAS) in preparing de novo, severely calcified coronary lesions for stent placement. In this post-hoc analysis, the ORBIT II data were evaluated to investigate the effect of subject history of CABG on clinical outcomes. 2 Methods The ORBIT II trial design was previously published. Briefly, ORBIT II was a prospective, single arm, multi-center study designed to demonstrate that in the treatment of de novo, severely calcified coronary lesions: (1) the Diamondback 360® Coronary OAS (Cardiovascular Systems, Inc. [CSI]; St. Paul, MN) is safe and (2) the OAS successfully facilitates stent delivery. No patients underwent orbital atherectomy of bypass grafts. The study enrolled 443 subjects at 49 U.S. institutions from May 25, 2010 to November 26, 2012. Key inclusion criteria were: (1) target vessel must be a native coronary artery; (2) target lesion must have fluoroscopic or intravascular ultrasound (IVUS) evidence of severe calcium deposit at lesion site; (3) target vessel reference diameter ≥ 2.5 mm and ≤ 4.0 mm; and (4) target lesion ≤ 40 mm in length. Severe calcium was defined as (a) presence of radiopacities noted without cardiac motion prior to contrast injection, involving both sides of the arterial wall, with calcification length of ≥ 15 mm and extension partially into the target lesion, or (b) presence of ≥ 270° of calcium at one cross section via IVUS. Key exclusion criteria included: (1) acute MI (STEMI or non-STEMI: CK-MB > Upper Limit of Normal (ULN)) within 30 days prior to index procedure; (2) subject diagnosed with chronic renal failure (creatinine (CR) > 2.5 mg/dl) unless under hemodialysis; (3) evidence of current left ventricular ejection fraction ≤ 25%; (4) more than one lesion requiring intervention unless the lesions are staged; (5) in-stent treatment; and (6) target lesion in ostial location, bifurcation, or has a ≥ 1.5 mm side branch. There were no required/mandated medications per the study protocol, and post-OAS/pre-stent percutaneous transluminal coronary angioplasty (PTCA) was optional. The ORBIT II trial was conducted per Good Clinical Practice (GCP) and applicable Code of Federal Regulations (CFR), and was approved by each institutional review committee. Procedural success was defined as success in facilitating stent delivery with a final residual stenosis < 50% and without in-hospital MACE. MACE was defined as a composite of cardiac death, target vessel revascularization (TVR), and MI (CK-MB > 3 × ULN with or without new pathologic Q-wave). An independent Clinical Events Committee adjudicated adverse events and persistent slow flow, persistent no flow, and abrupt closure complications. An Angiographic Core Laboratory (Cleveland Clinic Foundation, Cleveland, OH) analyzed procedural angiograms and reported final minimum lumen diameter, final percent residual stenosis, and the presence and type of dissections and perforations. Angiographic success was defined as success in facilitating stent delivery with a residual stenosis of < 50% without severe angiographic complications, defined as persistent slow flow, persistent no flow, abrupt closure, Type C-F dissections, and perforations. Subjects were followed in clinic at 30 days and by phone or in clinic at 1 year post-procedure. The coronary OAS manufactured by CSI is a catheter-based device indicated to facilitate stent delivery in patients with coronary artery disease, who are acceptable candidates for stenting due to de novo, severely calcified, coronary artery lesions. The OAS tracks and rotates over the ViperWire® (CSI) guidewire and modifies coronary plaque on the vessel wall by using a diamond-coated crown. The crown's orbital diameter expands laterally via centrifugal force and sands away the hard components of the plaque, allowing the soft components of the plaque and vessel wall to flex away from the crown. In the ORBIT II trial, two OAS configurations were used: Pneumatic (1.25 mm, 1.50 mm, 1.75 mm. 2.00 mm crowns) and Electric (1.25 mm and 1.50 mm crowns). Statistical analyses were performed with either the SAS Software System (SAS Institute Inc., Cary, NC) or R (R Core Team-2012). Subject demographics, pre- and post-procedure lesion characteristics, procedural characteristics, and outcome variables were summarized using descriptive statistics for continuous variables (presented as mean ± SE) and frequency tables or proportions for discrete variables. Data were compared using Wilcoxon rank-sum test for continuous parameters and Fisher's exact test for categorical parameters. Kaplan Meier methods were used to obtain estimates of the 30-day and event rates; comparisons were made using Cox proportional hazards models. p-Values < 0.05 were considered statistically significant. 3 Results Of the 443 subjects in the ORBIT II study, 65 subjects had a history of CABG ( Table 1 ). This CABG group contained a greater proportion of subjects with history of diabetes mellitus, dyslipidemia, hypertension, and MI than did the group with no prior history of CABG. Table 1 Baseline subject characteristics. CABG = coronary artery bypass grafting. Values are n (%) or mean ± SE. ⁎ p-Values from Wilcoxon rank-sum test (continuous parameters) and Fisher's exact test (categorical parameters). a Unless otherwise indicated. Subjects with prior CABG had smaller minimum lumen diameter (0.4 ± 0.0 mm vs. 0.5 ± 0.0 mm, p = 0.002) and larger mean diameter stenosis (87.4% ± 1.0% vs. 83.8% ± 0.5%, p = 0.004) than those without a history of CABG ( Table 2 ). The OAS was inserted in 65 (100%) subjects with prior CABG and 375 (99.2%) subjects with no-prior-CABG ( Table 3 ). The mean number of stents used post-OAS, was significantly higher in subjects with prior CABG (1.4 ± 0.1 vs. 1.2 ± 0.0, p = 0.02), as was final procedure stenosis (8.9% ± 2.2% vs. 3.9% ± 0.7%, p = 0.01). Table 2 Baseline lesion characteristics. ACC/AHA = American College of Cardiology/American Heart Association; CABG = coronary artery bypass grafting; OAS = orbital atherectomy system. Values are n/N (%) or mean ± SE. ⁎ p-Values from Wilcoxon rank-sum test (continuous parameters) and Fisher's exact test (categorical parameters). Table 3 Procedural characteristics. CABG = coronary artery bypass grafting; OAS = orbital atherectomy system. Values are n/N (%) or mean ± SE. ⁎ p-Values from Wilcoxon rank-sum test (continuous parameters) and Fisher's exact test (categorical parameters). a Unless otherwise indicated. Although the angiographic success rate and the incidence of severe angiographic complications were similar in both groups, there was an increased rate of severe dissection (Type C-F) in subjects with prior CABG (7.7% vs 2.6%, p = 0.05) ( Table 4 ). The in-hospital MACE rate was also higher in the CABG group (16.9% vs. 8.5%, p = 0.04) ( Table 5 ). This was driven primarily by the higher rate of MI (16.9% vs. 8.0%, p = 0.03); however, Q-wave rates were low (1.5% vs 0.5%, p = 0.38). When utilizing the current Society for Cardiac Angiography and Interventions (SCAI) definition of MI, there was no significant difference in rates of MI between the two groups (3.1% vs 1.9%, p = 0.63). Table 4 Severe angiographic complications. CABG = coronary artery bypass grafting. Values are n (n/N %). ⁎ p-Values from Fisher's exact test. The CABG patients had higher rates of 1-year MACE (27.7% vs. 15.0%, p = 0.01) and MI (18.5% vs. 9.3%, p = 0.03) ( Table 5 and Fig. 1 A, C ). However, there was no significant difference in rates of cardiac death (6.2% vs. 2.7%, p = 0.17) or TVR (7.9% vs. 5.5%, p = 0.47) between the two groups at 1 year ( Fig. 1 B, D). The rate of TLR at 1 year was also similar (6.3% vs. 4.4%, p = 0.52) in both groups. Fig. 1 Time-to-event curves through 365 days in ORBIT II subjects with and without history of coronary artery bypass grafting (CABG). (A) Major adverse cardiac events, (B) cardiac death, (C) myocardial infarction, and (D) target vessel revascularization. Table 5 Major adverse cardiac events through 1-year. CABG = coronary artery bypass grafting; MACE = major adverse cardiac events; MI = myocardial infarction; SCAI = Society of Coronary Angiography and Intervention; TLR = target lesion revascularization; TVR = target vessel revascularization. ⁎ Kaplan-Meier rate estimates. a p-Values from Fisher's exact test for in-hospital outcomes. p-Values from Cox Proportional Hazards Model for 30-day and 1-year MACE outcomes. 4 Discussion This post hoc analysis of the ORBIT II trial revealed patients with a history of CABG who underwent orbital atherectomy of severely calcified native coronary lesions was associated with a significantly higher MACE rate in-hospital and at 1-year, driven by a higher rate of non-Q-wave MI. This outcome may be explained by the significantly greater percentage of diabetes mellitus, dyslipidemia, hypertension, and prior MI in the CABG group. A retrospective analysis of the ACUITY (Acute Catheterization and Urgent Intervention Triage Strategy) trial revealed that patients with a history of CABG experience a significantly greater incidence of MACE upon undergoing subsequent treatment . The increase in MACE may be explained by substantial differences in baseline comorbidities in the CABG group typically seen in patients with calcified lesions (including advanced age, diabetes, chronic renal insufficiency, and more extensive coronary artery disease), as evidenced by the higher pre-PCI percent stenosis. The TWENTE trial and registry evaluated 1-year outcomes following PCI with second-generation drug-eluting stents in patients with previous CABG relative to patients without previous CABG. For native lesions, repeat revascularization rates were similar in the TWENTE registry between these two groups (5.1% vs. 2.3%, p = 0.08). Of subjects in the TWENTE trial with a history of CABG, only those whose subsequent PCI was performed on degenerated vein grafts showed a significant increase in TVR rates (95.4% in vein grafts vs. 18.5% in native lesions, p = 0.002). An analysis of the j-Cypher registry, which enrolled patients who underwent PCI with sirolimus-eluting stents, reported that subjects with prior CABG had a significantly increased risk of TLR at 5-year follow-up, compared to those without previous CABG (adjusted hazard ratio 1.25, 95% confidence interval 1.06–1.47, p = 0.01). In the ORBIT II study, TLR rates were low and similar between the CABG and non-CABG groups at 1-year follow-up (6.3% vs. 4.4%, p = 0.52) and at 3-year follow-up (9.8% vs. 7.4%, p = 0.46). The increased TLR rate in the prior-CABG patients in the j-Cypher analysis was driven by those who had saphenous vein graft intervention. In addition to TLR, the CABG subjects also had a significantly higher risk of cardiac death, MI, and definite stent thrombosis compared to those subjects who had PCI of the native coronary arteries. In these two registries, clinical outcomes of the prior-CABG population undergoing PCI with first generation (j-Cypher) or second generation (TWENTE) drug-eluting stents were superior only with a native coronary artery target. The ORBIT II protocol inclusion criteria required the target vessel to be a native coronary artery; therefore, any comparison with other trials must be nuanced. In j-Cypher and TWENTE, when PCI was performed on a native coronary artery, the prior-CABG subjects had outcomes similar to the no-prior-CABG subjects. In ORBIT II, the higher MACE rates (in-hospital and 1-year) observed in the CABG group were driven by a higher rate of MI. When the SCAI definition (which relates only to in-hospital MI) was used, the rate of MI in the entire ORBIT II cohort was 2.0%, compared with 9.3% using the ORBIT II definition. Additionally, there was no significant difference in rates of SCAI-defined MI between the CABG and non-CABG groups (3.1% vs 1.9%, p = 0.63), respectively. Considering the increased sensitivity of the ORBIT II definition of MI, it is possible that the reported cases of MI were not clinically significant. Further, only 1 of the 11 MI events (at all three time points: in-hospital, 30-days, and 1-year) in ORBIT II was Q-wave MI. Finally, there was one incident of stent thrombosis, which occurred in-hospital in a patient without prior CABG. Study limitations include the lack of a control arm in the ORBIT II trial. This analysis was not designed to assess differences in outcomes based upon history of CABG a priori. Differences in several baseline characteristics between the two groups may have affected clinical outcomes. Thus, a randomized controlled trial is warranted to directly compare OAS with other PCI approaches. To address this, the ECLIPSE (Evaluation of Treatment Strategies for Severe CaLcIfic Coronary Arteries: Orbital Atherectomy vs. Conventional Angioplasty Technique Prior to Implantation of Drug-Eluting StEnts) trial was recently launched (ClinicalTrials.gov Identifier: NCT03108456). 5 Conclusion Patients with a history of CABG, who then undergo PCI, may be at increased risk for mortality and repeat revascularization, compared with patients who have not undergone CABG. In this post-hoc analysis of the ORBIT II trial, safety and efficacy of coronary OA to modify severe coronary artery calcium, prior to stent placement, was evaluated in subjects in these two populations. The data show a reasonable safety profile for the Diamondback 360® Coronary OAS (CSI; St. Paul, MN), irrespective of patient history of CABG. Q-wave MI rates were low at 1.5% vs 0.5% (p = 0.38), and there was no significant difference in rates of cardiac death (6.2% vs. 2.7%, p = 0.17) and TVR (7.9% vs. 5.5%, p = 0.47) in patients with or without prior CABG, respectively. The low rates of TVR, cardiac death, and Q-wave MI, suggest OA treatment to facilitate stent delivery is successful and provides durable outcomes in subjects in both of these populations.