Towards a lesion-specific approach to percutaneous coronary intervention for bifurcation lesions

This editorial refers to ‘Multicentre, randomized comparison of two-stent and provisional stenting techniques in patients with complex coronary bifurcation lesions: the DEFINITION II trial’†, by J.J. Zhang et al., on page 2523. Practice guidelines and position papers recommend provisional side-branch stenting as the default strategy for percutaneous coronary intervention (PCI) of bifurcation lesions.1 This recommendation is largely based on a patient-level pooled analysis of two large trials, showing improved 5-year survival with the provisional single-stent approach as compared with systematic dual stenting.2 A subsequent study-level meta-analysis involving eight randomized studies with 2778 patients extended these findings.3 It showed that the risk of myocardial infarction after provisional side-branch stenting was lower than that after systematic dual stenting, and also confirmed the survival benefit of the provisional stenting approach.3 This paradigm has been challenged. The DKCRUSH-II study in true bifurcation lesions was the first to show a reduction in target lesion revascularization (TLR) with the two-stent approach as compared with provisional side-branch stenting that prevailed during 5-year follow-up.4 Post-hoc analysis even revealed a reduction in target lesion failure (TLF).4 Subsequently, the DKCRUSH-V study focused on true left main bifurcation lesions.5 Compared with provisional T-stenting, the two-stent approach reduced the risk of TLF at 1 year (primary endpoint), with increasing benefit during 3-year follow-up.5 This included a significant reduction in target vessel myocardial infarction and TLR. The Nordic–Baltic Bifurcation IV study in true bifurcation lesions showed consistent results, albeit without reaching statistical significance.6 Based on these studies, in particular DKCRUSH-V, recent guidelines opened the door for a systematic two-stent approach in left main bifurcation lesions.1 When comparing the early studies that established the concept of provisional side-branch stenting with those promoting a systematic two-stent strategy, it is conspicuous that in the former studies first-generation drug-eluting stents (DES) were predominant, whereas new-generation thin-strut DES were prevalent in the latter studies.3 Notably in the Nordic–Baltic Bifurcation IV study, the trend towards a benefit of the two-stent approach was more pronounced with new-generation DES than with first-generation DES.6 In addition, recent technical refinements, such as T- and protrusion (TAP), Culotte, or double-kissing (DK) crush techniques, facilitated a more meticulous optimization of side-branch coverage in contemporary studies, which may have contributed to better outcomes with dual stenting. Another important aspect is lesion selection. Even with the default strategy of provisional side-branch stenting, a substantial number of side branches are stented; ∼5–20% in randomized studies and ∼40% in registries. Thus, clinical experience already suggests that a one size fits all approach is not feasible. In complex true bifurcation lesions with long, high-grade stenoses of the side branch, particularly when heavily calcified or associated with awkward angles, the chances of achieving an optimal acute and long-term result are low with a single-stent approach. The more recent studies suggesting a benefit of systematic dual stenting exclusively addressed true bifurcation lesions, whereas most of the pivotal earlier studies included mixed lesion subsets.3 Also, where reported, the length of the side-branch lesion to be covered with a stent was mostly short in these early studies.3 In summary, advances in stent design, technical refinements, and differential lesion selection may explain the apparent discrepancy between the earlier studies supporting the concept of provisional side-branch stenting and the more recent studies showing benefit of systematic dual stenting. It is the merit of the group of Chen and co-workers that they have systematically worked-up the impact of lesion characteristics on the choice of stenting technique for bifurcation lesions. They built on two large registries of patients with true bifurcation lesions to develop criteria for lesion complexity that predict the risk of major adverse cardiac events (MACE) during 1-year follow-up.7 These criteria, summarized by the DEFINITION score, were derived from a training cohort of 1550 patients and subsequently confirmed in a validation cohort of 3660 patients. The study suggested a differential treatment effect of the stenting strategy depending on bifurcation complexity as defined by the DEFINITION score.7 With complex lesions, dual stenting was associated with a lower 1-year incidence of MACE, cardiac death, and myocardial infarction than provisional side-branch stenting. Nevertheless, the impact of the single or double stenting technique on MACE and myocardial infarction did not reach statistical significance.7 In the DEFINITION II study reported in the current issue of the European Heart Journal, a multinational research consortium led by Chen and co-workers prospectively tested the hypothesis that the DEFINITION score is able to identify those complex bifurcation lesions that benefit from systematic dual stenting.8 This study included 653 patients with a complex true bifurcation lesion randomly assigned to systematic dual stenting or provisional single stenting. At 1-year follow-up, systematic dual stenting significantly reduced the incidence of TLF from 11.4% in the provisional group to 6.2% (P = 0.019). This comprised a significant reduction in target vessel myocardial infarction and in TLR, whereas the incidence of cardiac death was similar between the two groups (2.5% vs. 2.1%, P = 0.77).8 While the techniques for systematic dual stenting are fairly standardized, the approach to provisional single stenting is highly variable. In DEFINTION II, the study protocol mandated very strict criteria for handling of the side branch. Thus, >28% of the side branches were left without angioplasty—despite a ≥10 mm lesion length with ≥70% diameter stenosis—and bail-out side-branch stenting was performed in 22% of the lesions. We cannot exclude that the provisional single-stenting arm would have achieved a somewhat better result with a more liberal approach to side-branch optimization. Nevertheless, the crossover rate in the provisional single-stenting arm was of the same order or even higher than that in the early studies showing a benefit of provisional single stenting.3 Assessment of haemodynamic indices by pressure wires was not part of the study protocol, and intravascular ultrasound (IVUS) was performed in only about a quarter of the patients. Functional assessment might have facilitated a better choice of side branches to be treated. Yet, in DKCRUSH-VI, such a benefit could not be demonstrated.9 Likewise, the use of IVUS did not appreciably interact with the treatment effect of systematic dual stenting in DEFINITION II.8 Thus, despite some potential limitations, DEFINITION II presents sound evidence that patients with complex bifurcation lesions as defined by the DEFINITION score benefit from systematic dual stenting instead of provisional single stenting. This message is further supported by retrospective analyses from DKCRUSH-II and -V as well as by registry data.4,5,7 A gap in evidence is the role of stenting technique. Most of the evidence for systematic dual stenting in complex bifurcation comes from DK crush stenting. In DEFINITION II, although any technique for dual stenting was allowed, DK crush was predominant (78%), followed by Culotte stenting (18%).8 There may be differences in outcome depending on the two-stent technique: in DKCRUSH-III,10 DK crush was superior to Culotte stenting in left main bifurcations, and in BBK-2 Culotte stenting was superior to TAP stenting in a cohort with mostly non-left main bifurcations.11 When interpreting these findings, it has to be kept in mind that the operators of the DKCRUSH studies had outstanding expertise in the DK crush technique. It remains to be shown whether similar results can be obtained by other advanced techniques for bifurcation stenting performed by dedicated operators. The ability to achieve full stent expansion with complete coverage of both branches may be more important than a specific stenting technique. Further studies are needed to clarify this issue. The best approach to non-complex true bifurcation lesions remains another gap in evidence. In DEFINITION I, DKCRUSH-II, and DKCRUSH-IV, there were numerical differences consistent with a benefit of systematic dual stenting in non-complex true bifurcations.4,5,7 A significant interaction between lesion complexity and treatment effect of stenting technique has not been established in any of these studies. Given the lower long-term risk with non-complex vs. complex bifurcation lesions, a statistically significant and clinically relevant absolute risk reduction by systematic dual stenting will be hard to establish. Improved DES designs and advances in stenting technique have broken the ground for new studies challenging the concept of provisional single stenting as the default strategy for PCI of bifurcation lesions. An increasing number of studies suggest that complex bifurcation lesions need complex stenting. DEFINITION II establishes evidence-based criteria for this strategy.8 We now have substantial evidence that bifurcation lesions that meet the DEFINITION criteria should be treated systematically with a two-stent approach.4,5,7,8 The best two-stent technique is less well established. It appears that optimal coverage of the carina is crucial—with the best evidence for DK crush. Based on studies with contemporary stents and stenting techniques, a systematic two-stent approach may also be preferable in non-complex true bifurcations.4,5,7 Yet, the long-term risk with non-complex true bifurcation lesions is lower than that with complex bifurcation lesions. Thus, the potential absolute benefit of systematic dual stenting may be small—if it exists at all. In non-true bifurcation lesions, there is currently no reason to deviate from the default strategy of provisional single stenting. Current evidence for a lesion-specific approach to PCI in bifurcation lesions (left main and non-left main). *According to DEFINITION score.8 Current evidence for a lesion-specific approach to PCI in bifurcation lesions (left main and non-left main). *According to DEFINITION score.8 Conflict of interest: F.-J.