Uncertainties, trade-offs and avoidance of harm

Le résumé fourni par la source

The debate around the best second conduit for Coronary Artery Bypass Grafting (CABG) is >30 years old [1] and the controversy is still ongoing. The survival benefit of bilateral internal thoracic artery (BITA) is based almost exclusively on observational studies, which are prone to treatment allocation bias and hidden confounders. In addition, the landmark ART trial at 10-year, comparing single versus bilateral internal thoracic artery was negative [2, 3]. This conundrum of evidence-based medicine remains a major issue: observational studies are prone to selection biases and hidden confounders, whatever the statistical adjustment, but randomized trials, the ‘top of the pyramid’, and respective meta-analysis, face the problem of external validity, i.e. generalizability. A second issue is that evidence-based medicine and its derived clinical guidelines are mostly supported by positive publications, or ad hoc reinterpretations, when they do not fit with a priori knowledge, are not impacted by unpublished negative results and often lack corroborating evidence. All this brings uncertainty to the fore, in the practice of medicine, not to mention the individual technical factor in surgery. In this issue of EJCTS, Chan et al. [4] present a retrospective analysis of 336 321 patients, included in the UK national database who underwent non-emergent, isolated, first-time, coronary artery bypass from 1996 to 2018. The authors report an overall rate of multiple arterial grafting (MAG) of 15.6% with a steady decline since 2001, from 24% in 2010 to 11% in 2018, in virtue of a reduction in the use of the radial artery (RA). The reported adoption of MAG is in line with that previously reported in the USA, Canada and Europe which ranges from 4% to 34% [5]. My country, Portugal, had a BITA rate of 17% in 2020. There are some exceptions. For example, France’s national registry (EPICARD) reports a regular increase in the rate of 2 arteries’ use (data allowing only to know if an artery was used, not the type of artery) from 55% to 80%, between 2007 and 2020 (J. Tomasi and J. L. Debrux, personal communication). Large databases and registries include broadly representative patient populations, rather than the highly selected population of randomized controlled trials and can provide more generalizable knowledge on the effectiveness of therapies. The compulsory national database describes the UK reality over >2 decades. After propensity score matching, the single arterial grafting group was more likely to undergo on-pump CABG (90% vs 69%, P < 0.001), experienced a lower in-hospital mortality (1.1% vs 1.3%, P < 0.001) and incidence of return to theatre for bleeding (2.5% vs 3.0%, P < 0.001). The authors also found that MAG was likely to be performed in younger patients (66.7 vs 62.3, P < 0.001) and males (85% vs 81%, P < 0.001). At its peak, the use of MAG reached 25% in 2001, declining steadily since then. We lack an explanation for the steep decline of BITA, replaced by SITA + RA, between 1996 and 2002, followed by a plateau till 2008, and from then on, a gross equilibrium between the 2 MAG strategies. Was this related to unacceptable rates of sternal wound infections or other complications of BITA, avoided by using the easily harvested RA (which by then, was supported by several randomized trials)? But then why a steep decline in the use of RA during the most recent years? Some insight into these results can be shed by a survey to 97 consultant cardiac surgeons, from 25 centres in the UK [6]. Only 36% and 32% of the interviewed surgeons reported routinely using BITA and RA, respectively. The surveyed surgeons rated (from 1 to 10) the importance of BITA use with an average score of 5.6 and the importance of total arterial grafting with an average score of 4.9. The main factors reported to affect conduit choice were the ‘high-quality evidence’ (or lack of thereof) in 34% of cases followed by ‘consultant experience’ (27%) and ‘competency of harvester’ (16%). Curiously, ‘patient choice’ and ‘theatre time’ usage scored low in the decision-making process. Conduit choice is only one among a myriad of factors influencing CABG outcomes. Country-specific life expectancy, patient co-morbidities, medical therapy, life style, ventricular function, myocardial viability, target vessel quality and completeness of revascularization are some of the most important, along with surgical expertise. The rational for MAG, better long-term graft patency, is sound, but it remains to be proven if it translates into clinical benefit in randomized trials. The ROMA trial might instruct us on this aspect. In fact, MAG adoption depends upon the correct articulation of 3 complementary factors: the surgeon, the patient and the ‘culture’. On the one hand, surgeons’ knowledge of the evidence and its interpretation is key for decision-making. The ART trial may have felt victim to MAG itself as the SITA group received an RA in >20% of cases and the crossover rate from BITA to SIMA was 14%. In fact, the as-treated analysis demonstrated a mortality benefit of 19% with BITA. On the other hand, surgeon expertise, training and consequent familiarity with MAG directly affect conduit choice and outcomes. In the ART trial, almost 9% of surgeons, which were carefully selected for their proficiency in MAG, had an unplanned conversion rate from BITA to LITA [7]. Again, this brings to the fore the issue of generalizability of trials findings. As for the patient, selection is key. Patients <75 years of age, high-grade stenosis and patients with preserved renal function seem to benefit the most from arterial conduits. Avoidance of BITA in patients with high risk of sternum complications—primum non nocere—is critical [7, 8]. Finally, ‘cultural beliefs’, through emphasis on short-term rather than long-term outcomes, acceptance of longer operative times and learning curves, referrals patterns and reimbursement issues may influence decision-making and between countries MAG use differences [2]. The UK national registry also shows that single arterial grafting patients were more likely to undergo on-pump CABG, confirming the notion that the controversy about routine use of arterial revascularization in CABG is not dissimilar from the one regarding off-pump CABG, a technique embraced by a few dedicated surgeons but lacking strong evidence. First and foremost, what this study tells us is that UK surgeons, subject to the highest level of individual scrutiny, have incorporated the ‘first do no harm’ principle in their practice. Trade-offs are part of life. We should not compromise the long-term benefits of MAG, if the evidence for higher MACE-free survival is convincing but without extra early complications, if the goal is a safe, effective and long-lasting operation. It is unfortunate that this study reports only in-hospital results. We miss the key information on event-free MACCE/repeat revascularization rate and more importantly long-term survival data using the UK national statistics office. Until further data enrich our understanding on CABG effectiveness, surgeons will continue to base their practice on the thoughtful blending of critical analysis of the evidence, the specific patient at hand and their own experience. Patients, the cardiac surgery community and society at large would immensely benefit if more countries implemented a compulsory cardiac surgery registry with potential linkage to national vital status data.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.