Radial artery vs. internal thoracic artery or saphenous vein grafts: 15-year results of the RAPCO trials

Abstract

The standard coronary artery bypass grafting operation over the past 30 years has used the left internal thoracic artery to the left anterior descending artery. However, the optimal graft for the second most important coronary target has remained debated. To answer this question, the Radial Artery Patency and Clinical Outcomes program (RAPCO) was designed as two separate randomized controlled trials—in which the radial artery was compared with either the right internal thoracic artery (RAPCO-RITA) or a saphenous vein (RAPCO-SV), as the conduit for grafting the second most important coronary target. Patient numbers in each trial were determined by a priori two-tailed power calculations. In the RAPCO-RITA, 394 patients were randomized to receive either radial artery or right internal thoracic artery—if they were younger than 70 years old, or younger than 60 years old if also diabetic. In the RAPCO-SV, 225 patients were randomized to receive either the radial artery or saphenous vein—if 70 or more years of age, reduced to 60 years of age if also diabetic. All included patients received the standard left internal thoracic artery graft to the left anterior descending coronary artery. A careful standardized procedure was mandated for the harvesting, protecting, and subsequent patency maintenance of the radial artery grafts. The primary outcome for this analysis was the 15-year event rate of composite major adverse cardiovascular events (MACE), defined as all-cause death, myocardial infarction, or coronary revascularization. Secondary outcomes included the individual component event rates of the primary composite outcome. The primary analysis was by intention-to-treat principle. Fifteen years after randomization, only one patient in each trial was lost to follow-up. There were significantly less clinical events for the radial artery, when compared with either the right internal thoracic artery (26% reduction) or saphenous vein (29% event reduction in favour of the radial artery). The reduction in events was primarily driven by a reduction in all-cause mortality. This suggests that the radial artery should be favoured as the second most important graft, after the usual left internal thoracic artery to the left anterior descending artery, unless there is a specific contraindication. ‘The 15-year clinical results of the Radial Artery Patency and Clinical Outcomes randomized trial comparing radial artery to right internal thoracic artery or saphenous vein grafts.’ (EURHEARTJ-D-22-03079) The standard coronary artery bypass grafting (CABG) operation uses the left internal thoracic artery to the left anterior descending artery (LITA-LAD).1 However, the optimal graft for the second most important coronary target remains debated.2 The Radial Artery Patency and Clinical Outcomes (RAPCO) program (NCT00475488) was designed as two separate randomized trials in which the radial artery (RA) was compared with either the right internal thoracic artery (RITA)—as a free graft—or the saphenous vein (SV), as the conduit for grafting the second most important coronary target. We herein present the 15-year clinical outcomes of the RAPCO program. Ethics approval was granted by the Austin Hospital Human Research Ethics Committee (H95/086 and H2006/02690), and all patients gave written informed consent. Patients with left ventricular ejection fraction > 35%, at least one non-LAD vessel with a proximal stenosis of at least 70% and a diameter of at least 1.5 mm, were included. Patients were randomized to receive either the RA or the RITA (RAPCO-RITA) if they were younger than 70 years, or younger than 60 years if diabetic. Patients were randomized to receive either the RA or the SV (RAPCO-SV) if older than 70 years, 60 years if diabetic. All patients received a LITA-LAD. The primary outcome for this analysis was the 15-year event rate of the composite major adverse cardiovascular events, defined as all-cause death, myocardial infarction, or repeat revascularization. Secondary outcomes included the individual component event rates of the primary composite outcome. The primary analysis was by intention-to-treat principle. In the RAPCO-RITA, there were no inter-group differences at baseline and only 1 patient was lost to follow-up. The 15-year MACE rate was 39.4% in the RA group and 48.5% in the RITA group [hazard ratio (HR) 0.74, 95% confidence interval (CI) 0.55–0.97, P = 0.04, Figure 1A]. Kaplan–Meier estimates of mortality at 15 years were 22.2% and 30.1% in the RA and the RITA groups, respectively (log rank P = 0.06; HR 0.69, 95% CI 0.47–1.02). No differences between the RA and the RITA were found in myocardial infarction (13.1% vs. 15.3%; competing risk pseudo-HR 0.76, 95% CI 0.45–1.29, Gray–Fine’s P = 0.31), or in repeat revascularization (18.7% vs. 21.4%; competing risk pseudo-HR 0.79, 95% CI 0.51–1.23, Gray–Fine’s P = 0.30). The treatment effect for the primary outcome showed no heterogeneity in pre-specified key clinical subgroups (age, sex, and diabetes), with consistency in sensitivity analyses based on a multivariate model adjusted for age, sex, diabetes, and previous myocardial infarction. The as-treated analysis was qualitatively consistent with the intention-to-treat analysis (HR 0.76; 95% CI 0.56–1.03). Incidence of the primary outcome during the 15-year follow-up. Panel A: RAPCO-RITA (radial artery vs. right internal thoracic artery comparison). Panel B: RAPCO-SV (radial artery vs. saphenous vein comparison). The primary outcome is the 15-year event rate of composite major adverse cardiovascular events (MACE), defined as all-cause death, myocardial infarction, or repeat revascularization. HR, hazard ratio; CI, confidence interval In RAPCO-SV, there were no intergroup differences at baseline and only 1 patient was lost to follow-up. The 15-year MACE rate was 60.2% in the RA group and 73.2% in the SV group (HR 0.71, 95% CI 0.52–0.98, P = 0.04; Figure 1B). Kaplan–Meier estimates of mortality at 15 years were 52.2% and 63.4% in the RA and the SV groups, respectively (log rank P = 0.08; HR 0.74, 95% CI 0.52–1.04). No differences between the RA and the SV were found in myocardial infarction (11.5% vs. 16.1%; competing risk pseudo-HR 0.59, 95% CI 0.29–1.21, Gray–Fine’s P = 0.15), and in repeat revascularization (10.6% vs. 16.1%; competing risk pseudo-HR 0.57, 95% CI 0.27–1.19, Gray–Fine’s P = 0.13). The treatment effect for the primary outcome was consistent in key clinical subgroups (age, sex, and diabetes) and also in sensitivity analyses based on a multivariate model adjusted for age, sex, diabetes, and previous myocardial infarction. The as-treated analysis was consistent with the intention-to-treat analysis (HR 0.71; 95% CI 0.52–0.99). This is the first single randomized trial to report superior clinical results of the RA compared with both the RITA and the SV for CABG. The clinical outcome differences in CABG clinical trials are not generally seen in the early years after randomization. In the RAPCO program, the 5-year interim safety and efficacy analyses showed no differences in outcomes between the different conduits.3 At 10 years, there was statistically significant better graft patency for the RA in both studies, with the suggestion of a MACE reduction when comparing the RA with the RITA, but with no significant MACE differences seen between the RA and the SV.4 It is possible that, due to the limited number of events at 10 years, these clinical outcomes were not robust. Thus, the 15-year data from the RAPCO program are pivotal in confirming the long-term clinical benefits of RA grafting over other CABG conduits within a single trial structure, rather than relying on the methodologically limited reports of propensity-controlled pooled data,5 or patient-level meta-analyses.6,7 Our findings suggest that a paradigm shift in conduit use would seem necessary to optimize outcomes in contemporary coronary surgery. The uptake of multi-arterial grafting remains slow. Between 2004 and 2015, the North American Society of Thoracic Surgeons registry reported that 10.6% of patients had more than one arterial graft with only 6.5% having the RA and 4.9% a RITA.8 Barriers to implementation of routine RA conduit use include perceived complexity of peri-operative RA handling. Importantly, RAPCO mandated a strict protocol for graft harvesting and post-operative pharmacologic treatment.4 It should be noted that in RAPCO, the RITA was used as an aortocoronary graft to maximize technical homogeneity between groups, whilst in the current clinical practice, it is commonly used as an in situ graft. Whether this difference impacts on the generalizability of RAPCO findings is unclear, as there is evidence that graft configuration does not affect RITA outcomes.9,10 This trial has limitations. RAPCO was a single-centre trial, affecting the generalizability of the reported results. In addition, surgical practice has evolved over the follow-up duration. Finally, the trial was powered for 10-year outcomes, rather than the 15-year clinical outcomes. In conclusion, we found that using the RA to graft the second most important coronary target, in patients undergoing coronary surgery, improved long-term clinical outcomes compared with either the free RITA or the SV. This suggests that the RA should be favoured to complement the LITA-LAD graft unless there is a specific contraindication. The authors publish on behalf of all who have contributed to the RAPCO study: Participating Surgeons Prof. Brian Buxton, A/Prof. George Matalanis, Prof. Jai Raman, Mr Alexander Rosalion, A/Prof. Siven Seevanayagam, Mr Jullien Gaer, Mr Justin Negri, and Dr Masashi Komeda Anaesthesia and Critical Care Prof. Rinaldo Bellomo, Dr Laurie Doolan, and A/Prof. Larry McNicol Interventional Cardiologists Dr John Brennan, Dr Robert Chan, Dr David Clark, Dr Ronald Dick, Dr Anthony Dortimer, A/Prof. David Ecclestone, A/Prof. Omar Farouque, Dr Dharsh Fernando, Dr Mark Horrigan, Dr Anthony Jackson, Dr Leslie Oliver, Dr Nilesh Mehta, Dr Voltaire Nadurata, Dr Nim Nadarajah, Dr George Proimos, Dr Michael Rowe, Dr Ben Sia, and Dr Christopher Webb Clinical Cardiologists Dr Nagesh Anaveker, Prof. Peter Barlis, A/Prof. Paul Calafiore, Dr Boniface Chan, Dr John Cotroneo, Prof. David Hare, Dr Jennifer Johns, Dr Elizabeth Jones, Dr Paul Kertes, Dr David O’Donnell, Dr Stephen Sylviris, and Prof. Andrew Tonkin Coronary CT Radiologists Dr Robert Fabini, Dr Leighton Kearney, Dr Ruth Lim, Dr Maurice Molan, Dr Gerard Smith, Dr Chris Wellman, and Dr John Eng Statistical and Data Management Prof. Ian Gordon, Ms Margaret Shaw, and Ms Sandra Gerbo Garry W. Hamilton, MBBS, FRACP (Conceptualization [equal]; Data curation [lead]; Investigation [supporting]; Methodology: [supporting]; Writing—original draft [equal]; Writing—review & editing [equal]), Jaishankar Raman, PhD, FRACS, Simon Moten, FRACS, George Matalanis, FRACS, Alexander Rosalion, FRACS, Arnaldo Dimagli, MD, Siven Seevanayagam, FRACS, Mario Gaudino, MD, and David L. Hare, DPM, FRACP All requests for data sharing will be considered by the investigators. This study was completed with assistance of grants from the National Heart Foundation of Australia (Buxton BF, Hare DL. G09M4392), Johnson and Johnson (Medical, Australia), and the Sir Edward Dunlop Medical Research Foundation at the Royal Australasian College of Surgeons.

Topics & Concepts