The selection of grafts used in myocardial revascularization surgery has become a recurrent topic of debate among cardiac surgeons. This is reflected in recent recommendations on graft utilization, previously addressed in another entry in this blog. However, recent studies and the lack of clear evidence leave us questioning which grafts to employ.
The use of the internal mammary artery (IMA) to revascularize the left anterior descending artery (LAD) is the cornerstone of any credible coronary surgery today. Complete arterial revascularization of the left territory has extended as a long-term patency guarantee, thus increasing survival. However, the addition of a second mammary graft remains controversial.
Notably, the use of the right internal mammary artery (RIMA) is recommended as a Class IIa indication in myocardial revascularization guidelines, while the radial artery (RA) is a Class I recommendation when coronary stenosis is severe. These recommendations are primarily based on two clinical trials. On one side, the RAPCO study compared the radial artery to both saphenous vein and mammary artery. In the branch comparing mammary and radial arteries (RAPCO-RITA), the radial artery showed excellent patency and increased 10-year survival, though certain methodological limitations existed due to the small patient number.
On the other hand, the ART study analyzed 5- and 10-year survival in revascularization with one or two mammary arteries, finding no differences between groups. Despite the high expectations generated, some methodological limitations, such as the use of RA in 20% of the patients in the single-mammary group, left the question of RIMA as a second arterial graft unresolved.
In light of whether to endorse these findings and favor the radial artery over RIMA, various meta-analyses have been published pointing in the opposite direction. Among these is the study by Benedetto et al., which aggregates results from numerous observational studies (over 15000 patients) and concludes that the RIMA is superior to the RA in terms of long-term survival. Thus, today, the question of RIMA versus RA remains very much alive.
Urso et al.’s present meta-analysis attempts once again to clarify this age-old dilemma, this time through significant nuances. The authors conduct a systematic review, including only articles with propensity-matched or matched-pair analysis. Of the 51 initially found publications, they select only 12 observational studies including 6450 patients in the double-mammary group versus 9428 in the left mammary and radial artery group. Notably, in 8 of the included studies, mean follow-up exceeded 7 years, and propensity score matching was applied in nearly all. The primary endpoint was long-term survival. For statistical analysis, the authors introduce a novel approach to what has been published thus far: they conduct a global meta-analysis using the inverse-variance method, but also include a Kaplan-Meier meta-analysis with individual patient data.
The initial results using the inverse-variance method favored the use of RIMA as a second arterial graft (HR: 0.84; 95% CI: 0.74-0.95; p=0.04), with no asymmetry or publication bias observed.
The Kaplan-Meier meta-analysis with individual data reflected similar long-term survival in both groups (p=0.31). One-, five-, ten-, and fifteen-year survival was 97.3%, 91.5%, 79.9%, and 63.9% in the RA group versus 97.0%, 91.3%, 80.0%, and 68.0% in the RIMA group. However, Schoenfeld residual analysis indicated a violation of the proportional hazards assumption. To address this, the authors repeated the analysis by dividing the curves into two time intervals: 0-10 years and over 10 years. In this segmented analysis, the proportional hazards assumption held, showing similar survival between the two groups in the first 10 years (HR: 0.99; 95% CI: 0.91-1.09; p=0.93), but increased survival in the double-mammary group after 10 years (HR: 0.77; 95% CI: 0.63-0.94; p=0.01).
The authors suggest that the use of RIMA as a second arterial graft in myocardial revascularization is associated with increased survival after 10 years.
COMMENTARY:
This is an intriguing meta-analysis favoring the use of RIMA over RA starting 10 years post-surgery. The authors employ meticulous methodology and statistical analysis, which introduces time-based survival assessment, a novelty in the literature to date.
Urso et al. have extensive experience in this area, having published numerous meta-analyses in the field of myocardial revascularization, evaluating arterial revascularization strategies, IMA harvesting technique, and revascularization strategy in left main coronary artery disease, among others. In 2019, the authors conducted a first meta-analysis of propensity-matched studies comparing single versus double IMA, finding no benefit in the double-mammary subgroup. After four years of publications on this topic, the population size has increased sufficiently to yield different outcomes in the present study.
As for potential limitations, while the study’s methodology and statistical analysis are robust, we must remember that including observational studies, despite propensity matching, introduces the inherent limitations of such designs. Additionally, the exclusion of methodologically different studies regarding adjustment may introduce other sources of bias.
Given these results, the authors reflect on the possible explanation for higher survival in the double-mammary group only after 10 years. Anatomically, the RA has poorer characteristics than the IMA, which would lead us to expect survival differences across both time intervals; however, no differences are seen in the first 10 years. The authors speculate that early mortality in the double-mammary group due to wound infection complications may dilute its benefits in terms of survival during the initial period.
Despite this, this is an important meta-analysis on a complex and recurring topic in cardiac surgery and represents additional evidence supporting arterial revascularization. While, based on these results, survival following RIMA or RA use is similar in the early years post-intervention; in the long term, double-mammary use reduces mortality. In clinical practice, this may prompt us to reconsider the surgical strategy in younger patients with long life expectancy.
REFERENCE:
Urso S, Sadaba R, González Martín JM, Nogales E, Tena MÁ, Portela F. Bilateral internal thoracic artery versus single internal thoracic artery plus radial artery: A double meta-analytic approach. J Thorac Cardiovasc Surg. 2024 Jan;167(1):183-195.e3. doi: 10.1016/j.jtcvs.2022.03.010.
