Transatlantic editorial: the use of multiple arterial grafts for coronary revascularization in Europe and North America

The use of multiple arterial grafts (MAGs) for coronary artery bypass grafting (CABG) is encouraged by all existing guidelines and recommendations from professional societies (Table 1) [1–3]. However, it should be noted that the evidence basis in support of the use of MAGs is not definitive.

Table 1:

Summary of current guidelines on the use of multiple arterial grafts for coronary artery bypass grafting

Guidelines	Summary
ACC/AHA [1], 2011	When anatomically and clinically suitable, use of a second IMA to graft the left circumflex or right coronary artery (when critically stenosed and perfusing LV myocardium) is reasonable to improve the likelihood of survival and to decrease reintervention (class IIa, LOE B)
STS [2], 2016	As an adjunct to LITA, a second arterial graft (RITA or RA) should be considered in appropriate patients (COR IIa, LOE B) Use of BITAs should be considered in patients who do not have an excessive risk of sternal complications (COR IIa, LOE B) As an adjunct to LITA to LAD (or in patients with inadequate LITA grafts), use of a RA graft is reasonable when grafting coronary targets with severe stenoses (COR IIa, LOE B) The RGEA may be considered in patients with poor conduit options or as an adjunct to more complete arterial revascularization (COR IIb, LOE B) Use of arterial grafts (specific targets, number and type) should be a part of the discussion of the heart team in determining the optimal approach for each patient (COR I, LOE C)
ESC/EACTS [3], 2018	Additional arterial graft should be considered in appropriate patients (class IIa, level B) The use of the RA is recommended over the saphenous vein in patients with high-grade coronary artery stenosis (class I, level B) BIMA grafting should be considered in patients who do not have a high risk of sternal wound infection (class IIa, level B) Multiple arterial grafting should be considered using the RA for high-grade stenosis and/or BIMA grafting for patients who do not have an increased risk of sternal wound infection

ACC: American College of Cardiology; BIMA: bilateral internal mammary artery; BITA: bilateral internal thoracic artery; COR: class of recommendation; EACTS: European Association for Cardio-Thoracic Surgery; ESC: European Society of Cardiology; IMA: internal mammary artery; LAD: left anterior descending; LITA: left internal thoracic artery; LOE: level of evidence; LV: left ventricle; RA: radial artery; RGEA: right gastroepiploic artery; RITA: right internal thoracic artery; STS: Society of Thoracic Surgeons.

On one hand, a large body of observational evidence suggests that patients who receive MAGs for CABG have better postoperative outcomes and superior life expectancy [4, 5]. In addition, several randomized trials have reported that the midterm patency rate of arterial conduits, the radial artery (RA) in particular, is higher than that of the saphenous vein (Table 2) [6–17].

Table 2:

Summary of randomized control trials comparing graft conduits for coronary artery bypass grafting

Trials	Total patients (N)	Comparison	Mean follow-up (years)	Main findings
Arterial Revascularization Trial (ART) [6]	1548	BITA vs SITA	10	Intention-to-treat analysis: mortality for BITA vs SITA: HR 0.96, 95% CI 0.82–1.12 As-treated analysis: mortality for BITA vs SITA: HR 0.81, 95% CI 0.68–0.95
Petrovic et al. [7]	200	RA vs SV	8	No significant difference in survival between RA vs SV (P = 0.979) No difference in the composite of cardiac death, non-fatal myocardial infarction and repeat revascularization between RA and SV (P = 0.509)
Radial Artery Patency and Clinical Outcomes (RAPCO) [8]	649	Group 1: RITA vs RA (age<70) Group 2: RA vs SV (age >70)	5.5	Patencies were similar for either conduit in each group (log-rank P = 0.06 and P = 0.54, respectively)
Radial Artery Patency Study (RAPS) [9]	510	RA vs SV	7.7 ± 1.5	Frequency of functional graft occlusion was lower in RA vs SV: 28 of 234 (12.0%) vs 46 of 234 (19.7%; P = 0.03) Frequency of total graft occlusion significantly lower in RA vs SV: 24 of 269 (8.9%) vs 50 of 269 (18.6%; P = 0.002)
Stand-in-Y [10]	815	LITA + RA LITA + SV BITA	2.0 ± 0.81	Survival at 2 years was not different among groups (P = 0.59) Cardiac event-free survival was better in patients receiving 2 arterial grafts (P < 0.0001), even among elderly patients (P = 0.02)
Song et al. [11]	60	RA vs SV	1	No difference in postoperative mortality, patency and survival between RA and SV
Goldman et al. [12]	757	RA vs SV	1	No significant difference in graft patency at 1 year after CABG for RA vs SV (adjusted OR 0.99, 95% CI 0.56–1.74; P = 0.98)
SAphenous VEin versus Right Internal Thoracic Artery as a Y-Composite Graft trial (SAVE-RITA) [13]	224	RITA vs SV	5.1 ± 0.4	No difference in graft occlusion rate (P = 0.910) between RITA vs SV No difference in overall survival (P = 0.439) and freedom from major adverse cardiac and cerebrovascular event rates (P = 0.354) at 5 and 8 years between the 2 groups
Muneretto et al. [14]	160	RA vs SV	1.3 ± 0.2	SV was an independent predictor for graft occlusion (OR 1.11, 95% CI 1.06–1.16) and angina recurrence (HR 2.45, 95% CI 1.67–3.19)
Radial artery versus Saphenous Vein Patency randomized trial (RSVP) [15]	142	RA vs SV	5.6 ± 0.8	At 5 years, 98.3% of RA grafts and 86.4% of SV grafts were patent (P = 0.04) Graft narrowing occurred in 10% of patent RA grafts and in 23% of patent SV grafts (P = 0.01)
Glineur et al. [16]	210	RITA vs SV vs RGEA for the right coronary system	3	SV was associated with superior graft functionality compared with the other conduits (OR 6.1, 95% CI 2.4–15)
Gaudino et al. [17]	120	RA vs RITA vs SV in patients with in-stent restenosis	4.3 ± 0.7	SV grafts had a lower patency rate than arterial grafts and a high failure rate in patients with previous in-stent restenosis (P = 0.001)

BITA: bilateral internal thoracic arteries; CABG: coronary artery bypass grafting; CI: confidence interval; HR: hazard ratio; LITA: left internal thoracic artery; OR: odds ratio; RA: radial artery; RGEA: right gastroepiploic artery; RITA: right internal thoracic artery; SITA: single internal thoracic artery; SV: saphenous vein.

On the other hand, observational studies comparing different surgical interventions are open to treatment allocation bias [18] so that hidden confounders, rather than true biological effect, may explain the observed differences (i.e. the patients with the likeliest long-term longevity receive MAGs). In addition, although a pooled analysis of 6 randomized trials showed a lower risk of adverse cardiac events at 5 years of follow-up with the use of the RA instead of the saphenous vein for CABG [19], the intention-to-treat analysis of the large Arterial Revascularization Trial (ART) found no difference in survival between patients receiving a single internal thoracic artery (SITA) and bilateral ITAs (BITAs) at 10 years of follow-up [6].

ART reported a high (14%) crossover rate from the BITA group to the SITA graft group and a frequent use (>20%) of the RA in the control group. Both of these factors may have diluted the treatment effect of BITA grafting [20] and worked in favour of the null hypothesis. Indeed, a post hoc non-randomized comparison of patients who received MAG vs a single arterial graft in the ART showed MAG was associated with a significant clinical benefit [6].

Currently, a second large trial, Randomized comparison of the Outcome of single vs Multiple Arterial grafts (ROMA), comparing the use of MAGs (using a second ITA or RA) vs 1 ITA and vein is underway [21]. The results of ROMA are expected in 2025; however, we here aim to report the attitude of the surgical community towards the use of MAGs in North America and in Europe.

PREVALENCE OF THE USE OF MULTIPLE ARTERIAL GRAFT

Data on the use of MAG are relatively limited for both North America and Europe (Table 3) [22–30]. In a review of The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery database including 1,493,470 patients undergoing CABG between 2004 and 2015, Schwann et al. [22] reported a rate of MAG of 11.3%, with a significant decrease from 16% in 2004 to 9% in 2015. In another report from the STS database focused on low-risk patients most likely to benefit from MAG (age <70 years, no or mild chronic lung disease, body mass index <30 kg/m² and absence of diabetes), LaPar et al. [31] reported a BITA use rate of 6%.

Table 3:

Main series reporting the use of multiple arterial grafting in North America and Europe

Study	Registry	Variables	Percentage
USA and Canada
Schwann et al. [22], 2018	The Society of Thoracic Surgeons Adult Cardiac Surgery Database	%MAG	11.3
Goldstone et al. [23], 2018	California State Registry	%MAG	9.9
Samadashvili et al. [24], 2019	New York’s Cardiac Surgery Reporting System and New York’s Vital Statistics	%MAG	20.0
Chikwe et al. [25], 2019	New Jersey Cardiac Catheterization Registry, New Jersey Discharge Data Collection System	%MAG	14.0
Guru et al. [26], 2006	Cardiac Care Network database (Canada)	%MAG	12.1
Pu et al. [27], 2017	Cardiac Services British Columbia Registry (Canada)	%MAG	27.8
Rocha et al. [28], 2018	CorHealth Registry (Canada)	%MAG	22.4
Europe
Janiec et al. [29], 2017	SWEDEHEART Registry	%MAG	3.9
Holm et al. [30], 2019	European Multicenter Study on Coronary Artery Bypass Grafting Registry	%BITA	34.2
Austrian Society for Cardiac and Thoracic-Vascular Surgery	Austrian Society for Cardiac and Thoracic-Vascular Surgery Registry	%MAG	16.7

BITA: bilateral internal thoracic artery; MAG: multiple arterial graft; RA: radial artery; SWEDEHEART: Swedish Web-system for Enhancement of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies.

Goldstone et al. [23] analysed a state-maintained clinical registry including all 126 non-federal hospitals in California between 2006 and 2011 and found the percentage of MAG use was 9.9%, decreasing significantly from 10.7% to 9.1% over the study period. In the mandatory clinical New York State registry, a MAG rate of 20% was reported among patients who underwent CABG between the years 2005 and 2014 compared with 14% in the mandatory New Jersey clinical registry between 2005 and 2012 [7, 32].

Guru et al. [26] studied CABG patients in Ontario using the Cardiac Care Network database and reported an overall MAG rate of 12.1% between 1991 and 2003, having increased from 3.0% to 26.5%. Rocha et al. [28], using the same registry, reported MAG in 22.4% of patients between 2008 and 2016. In another Canadian study, Pu et al. [27] studied patients in the province of British Columbia from 2000 to 2014 and reported a MAG rate of 27.8%, decreasing from 29% to 25% over the study period.

In Europe, Holm et al. [30] reported a BITA grafting rate of 34.2% in patients undergoing CABG in Finland, France, Germany, Italy, Sweden and the UK from 2015 to 2017 using the European Multicentre Registry on Coronary Artery Bypass Grafting. Janiec et al. [29] reported a MAG rate of 3.9% using the Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies registry. The Austrian Society for Cardiac and Thoracic-Vascular Surgery reported an overall MAG rate of 16.7% in the period from 2009 to 2018, increasing from 13.4% in 2009 to 25.0% in 2018. Recent data from Germany’s Institute for Quality Assurance and Transparency in Healthcare for the years 2015 and 2016 show a BITA rate of 24.4% and 25.5%, respectively, and a use of the RA at approximately 9%.

POTENTIAL REASONS FOR THE SLOW ADOPTION OF MAG BY THE SURGICAL COMMUNITY

One route to understanding the reason for low rates of multiarterial revascularization in the USA and Europe is to analyse the practice of individual surgeons. The national rates of multiarterial revascularization conceal wide extremes of practice variation between individual providers [24, 25]. For example, in a recent analysis of 26,124 patients who underwent CABG in the USA between 2005 and 2012, the prevalence of MAG was 14%: however, many surgeons performed MAG in fewer than 10% of patients, while a small minority of enthusiasts performed MAG in most patients (Fig. 1) [23]. Even wider practice variation was described within the ART, in which unplanned conversion rates from bilateral to SITA grafts ranged from 1.9% for 1 outlier to 100% in almost one-third of surgeons, suggesting ample variability in the level of confidence with the BITA procedure (Fig. 2) [6, 33].

Figure 1:

Rate of multiarterial revascularization by individual surgeon, stratified by annual coronary bypass case volume in a US mandatory clinical registry. (From Chikwe et al. [25].)

Figure 2:

Rate of unplanned conversion from bilateral to single internal mammary artery by individual surgeon, stratified by number of patients randomized in the Arterial Revascularization Trial. ( © (2019) Massachusetts Medical Society. Reprinted with permission from Taggart et al. [6].)

These extremes of practice may be explained in terms of evidence, expertise and incentives. The absence of compelling evidence was cited by almost one-third of UK surgeons when asked in 2018 what issues limited use of BITA grafts in clinical practice [34]. As discussed, although a large body of observational data suggests that a second arterial conduit is associated with improved survival, graft patency and freedom from reintervention, the findings of randomized studies have been equivocal [4–6, 17, 18, 20–28, 31].

For those surgeons convinced by the available evidence-base, an important factor limiting use of MAG may be technical expertise. The extent of and reason for this gap in practice are not apparent from simply analysing surgeon case-volume and revascularization strategy. In 2 US registry studies, for example, no correlation was found between surgeon case-volume and MAG use: high volume surgeons were just as likely to prefer a single arterial graft strategy as their low-volume counterparts [24, 25].

Prospectively collected data on intraoperative decision-making influencing MAG vs single arterial graft would be very beneficial. In this context, the data on unplanned conversion from BITA to a SITA graft in the ART are particularly important because it prospectively details the technical and clinical challenges that precluded BITA grafting by a group of surgeons specifically selected for their expertise in and enthusiasm for that approach [33]. The most common reason, reported in 31% of cases, was that the conduit was ‘unsuitable’; specifically ‘damaged’ in more than half such cases, with ‘poor flow’ in one-third and ‘too short’ in the rest [33]. It is notable that similar problems were reported in fewer than 1% of patients allocated to a SITA graft. Successful deployment of the right ITA requires greater technical facility, especially with composite and sequential grafts. Notably, this does not apply to the use of the RA. In fact, in the quoted analysis of the STS database, Schwann et al. [22] reported a clear volume-to-outcome relationship for BITA but not RA grafting.

Finally, incentive is the third factor potentially driving the wider use of MAG and, specifically, the acceptance of the potential disadvantages of this strategy such as higher risk of sternal or forearm complications, bleeding, longer operative times and the additional training and planning associated in particular with BITA use [34, 35]. Increased sternal wound infection is a significant disincentive in the US health care setting, especially among diabetic patients, because providers may not receive reimbursement for the additional care associated with wound complications. It is important to note that even among diabetic patients, however, observational data suggest a mortality benefit for BITA over SITA grafting [32].

Also, 30-day outcomes, including complications and patient satisfaction, are commonly publicly reported and may form part of managed care contracting criteria. There may be real concern that referrals could be targeted elsewhere on the basis of 30-day outcomes rather than putative late survival benefit.

STRATEGIES TO INCREASE MULTIPLE ARTERIAL GRAFT

On this basis, there are, in our opinion, 4 strategies that could potentially increase the appropriate use of MAG:

• Foster a cultural shift from the view of CABG as an entry-level operation for learners towards a specialist operation for experts. Update curricula and operative requirements for professional certification to include current best practice in MAG with associated advanced training courses.

• Revise consensus guidelines to emphasize the role of targeted referral to expert providers offering a high likelihood of MAG and simultaneous low mortality in appropriate patients.

• Expand publicly reported CABG benchmarks, such as the STS quality bundle, to include MAG (or documentation of its contraindication) for young patients and compliance with evidence-based prophylactic sternal infection care bundles.

• Add appropriateness of intervention to public reporting for surgical revascularization.

CONCLUSION

In conclusion, a substantial body of evidence suggests that MAG may be associated with improved survival, graft patency and freedom from reintervention. However, lack of clear level I evidence (in particular for BITA grafting), variable surgical expertise, and disincentives linked to short-term outcomes impact public reporting and reimbursement and have led to extreme variation in the use of MAG even in the most appropriate patients. Disseminating consensus guidelines updated to reflect current best practice, especially in younger patients, and expanding quality benchmarks to include appropriateness of the revascularization strategy may help in addressing this knowledge and practice gaps.