Coronary Bifurcation Stenting: Review of Current Techniques and Evidence

Background:

Coronary bifurcation stenting constitutes 20% of all PCI performed. Given the extensive prevalence of bifurcation lesions, various techniques have sought to optimally stent the bifurcation to improve revascularization while also decreasing rates of stent thrombosis and lesion recurrence. Advanced techniques, such as planned two-stent approaches, have been shown to have improved outcomes but also require fluoroscopy and procedure time, posing an economic argument as well as a patient-outcome one.

Objective

Because of the many strategies posited in the literature, it becomes essential to objectively evaluate evidence from randomized controlled trials and meta-analyses to help determine the optimal stenting strategy.

Methods

We reviewed the clinical evidence on the efficacy of coronary bifurcation stenting.

Results

In this paper, we review the most recent randomized controlled trials and meta-analyses on the efficacy of various stenting techniques and advances in stenting technologies published to gauge the current state of understanding and chart where the field is heading.

Conclusion

Bifurcation stenting is a maturing problem in the field of interventional cardiology that is adapting to the needs of the patients and advances in technology.

1. INTRODUCTION

The hydrodynamic nature of a bifurcation in coronary arteries makes it a susceptible site for atherosclerotic plaque formation. When this occurs, management requires novel techniques that address the potential for plaque formation in both the main vessel and the side branch. There are many classification systems designed to categorize the types of coronary bifurcation lesions [1] (CBLs). Fig. (1) summarizes the Medina classification system [2] and the classification system defined by the SYNTAX study [1]. Four more classification systems, including the Duke, Sanborn, Safian, and Lefevre systems, are also used [1]. In all cases, classification is dependent on the location of the lesions in the proximal and distal main vessel and side branch and is determined in most cases via angiography. True CBLs are defined as lesions where both side branch and main vessel have been significantly stenosed ≥ 50%. In these cases, challenging maneuvers and techniques must be utilized for stenting. Management of CBLs constitutes 20% of all percutaneous coronary intervention (PCI) performed [3, 4]. Given the large prevalence of these procedures, it also becomes essential to understand the direction the field of PCI for bifurcation lesions is heading. Additionally, given the lack of clear clinical guidelines for the use of technologies, it is relevant to identify the pertinent clinical evidence that may influence technology adoption and integration into routine practice. In this review of the literature, we review the evidence behind the current standard of care. We also elaborate on the most recent meta-analyses and randomized controlled trials published in the past 5 years on the stenting techniques and advances in stenting technologies and describe emerging techniques that chart the future for the management of bifurcation stenting.

Fig. (1).

A representation of the Medina classification [2] (A) and the SYNTAX classification [1] (B) commonly used in characterizing coronary bifurcation lesions.

2. TREATMENT ALGORITHM FOR BIFURCATION LESIONS

The standard of care algorithm in the management of CBLs depends on various factors, such as lesion location, the need to preserve SB, angulation, and size of SB (Fig. 2). In the Manual of Percutaneous Coronary Intervention [5], Brilakis describes the clinical algorithm to determine the preferred stenting strategy (Fig. 2A). First, a determination is made whether the side branch needs to be preserved. This determination can be made based on the size and the function of the vessel. Typically branches <2 mm in diameter do not need to be preserved.

Fig. (2).

Decision-making algorithm for optimal stenting strategy adapted from Brilakis [5] (A) and Colombo and Stankovic [1] (B).

Next, it must be decided what the likelihood of occlusion of the side branch is when the main branch is stented. This determination can be made using the Medina classification [2], lesion morphology, the severity of side branch disease, and bifurcation angulation. For cases of a low likelihood of SB occlusion, a provisional strategy is considered the gold standard [1, 5]. This involves placing a wire in the side branch, placing a crossover stent covering the bifurcation lesion, and performing a balloon angioplasty if the side branch is compromised. If the balloon angioplasty fails, a secondary procedure must be performed based on the angulation of the side branch. In the case that the side branch is 70˚-90˚, TAP stenting is preferred. If the side branch is <70˚, Reverse crush, or culotte stenting is preferred. For cases with high likelihood of SB occlusion, a planned two-stent strategy is used. If lesions are localized only in the distal MV and SB (Medina 0.1.1), V-stenting is performed. In many emergency PCI cases, simultaneous kissing stenting can be performed. In non-emergent cases, the next determination is often made depending on the SB angulation. When the side branch is 70˚-90˚, TAP stenting is preferred. In the case that the side branch is at <70˚, Double kissing (DK) crush, or culotte stenting is preferred.

Other more generalized algorithms have also been elucidated. Colombo and Stankovic propose the algorithm shown in Fig. (2b), which starts upon the determination that the lesion is a true bifurcation lesion. If true, the determination is made whether the side branch is suitable for stenting. If the side-branch disease is focal and localized within 3 mm of the ostium of the side branch, a provisional strategy is recommended. If not, a planned two-stent approach is suggested by the authors.

3. A COMPARISON OF STENTING TECHNIQUES

3.1. Provisional Stenting Compared to a Planned Two-stent Approach

The provisional stenting strategy involves first stenting the main branch and then treating the side branch via stenting or balloon dilation, if necessary [6]. Stenting of the side branch in provisional stenting can be implemented using T-stenting [7], T and small protrusion (TAP) stenting [8], culotte [9], or the reverse/internal mini-crush technique [10], as shown in Table 1 [1, 7, 8, 10-16]. This varies from a planned two-stent approach, which requires the systematic use of 2-stents in a complex and coordinated fashion [17]. The simple provisional stenting strategy is frequently associated with reduced procedure and fluoroscopy times, resulting in lower rates of procedure-related biomarker elevation [6] compared to its two-stent counterpart techniques, which lend themselves to being considered the gold standard for CBLs [17]. Despite the reduction in costs for patients with provisional stenting strategies [6], it becomes necessary to evaluate the effectiveness of the techniques to ensure that the quality of care is not compromised. For this reason, various randomized controlled trials and meta-analyses have been performed over several years to answer this question.

Table 1.

Summary of provisional stenting strategy techniques.

Stenting Technique	Provisional Single Stenting [ 1 ]	T-stenting [ 1 , 7 , 11 ]	TAP Stenting [ 1 , 8 , 12 ]
Summary of stenting strategy procedure	Placement of two wires (MB and SB) Pre-dilation of MB or both branches, when needed Stenting of the MB Stent optimization with proximal optimization technique (POT) Recrossing with a wire into the SB Performance of FKBI (Final kissing balloon inflation) with moderate pressure (8 atm) in the SB, until the balloon is fully expanded Final POT, if result in the SB is adequate after FKBI Placement of a second stent in SB when result is unsatisfactory.	Advance second stent into the SB (following dilation of the MB stent struts) Position stent at ostium of SB minimizing any possible gap. Perform second kissing balloon inflation.	A second stent is advanced into the SB to minimally protrude (1-2mm) into the MB A balloon is advanced into the MB The SB stent is deployed at (12+ atm), MB balloon is simultaneously inflated at 12+ atm. Both balloons are deflated and removed.
Stenting Technique (continued)	Culotte Stenting [1, 9, 13, 14, 15]	Reverse/internal mini-crush stenting [1, 10, 16]
Summary of stenting strategy procedure (continued)	After MB stenting, second stent is advanced into SB protruding into the MB to overlap with proximal part of the MB stent and expanded following removal of MB wire. MB is rewired through the stent struts and dilated. Kissing balloon inflation is performed.	After stenting the MB, a second stent is advanced into the SB, and left in position without deployment. A balloon sized to the diameter of the MB, but shorter than the already-deployed stent, is advanced in the MB, positioned at the level of the bifurcation. SB stent is withdrawn approx. ≤ 3 mm into MB and deployed. Deploying balloon is removed, and angiography is performed to verify the absence of distal dissection and need of additional stent. SB struts are recrossed with wire and balloon, sized to SB reference diameter and inflated at high pressure (12-20 atm). FBKI is performed.

In a comparative, randomized superiority trial (Tables 2 [18-27] and 3 [14, 28-33]) published in 2020, Kumsars et al. [28] enrolled 450 patients to study whether a simple stenting or complex stenting approach improved outcomes for patients. The primary endpoint of the trial was major adverse cardiac events (MACE) at 6 months and after a follow-up of two years. They also tested the efficacy of newer generation drug-eluting stents, comparing the outcomes for the simple stent strategy versus the complex double stent strategy. While their results for their primary endpoints indicated a trend towards improved outcomes for the complex stent approach, they found that there was no significant difference between MACE for simple and complex strategies. Their data for the subgroup for drug-eluting stents had a similar conclusion. The finding from this clinical trial is consistent with previously published RCTs in the literature.

Table 2.

Meta-analysis outcomes for coronary bifurcation lesions.

Name	# of Studies	Comparisons	Conclusions
Stenting Techniques
Ford et al. 2018 [18]	9	Provisional single-stent strategy compared to an up-front two-stent approach	Provisional single-stent strategy is associated with a reduction in all-cause mortality in long-term follow-up (>1 year) with no difference in adverse major cardiac events.
Zhong et al. 2018 [19]	5	Kissing balloon inflation compared to non-kissing balloon inflation strategies for one-stent approaches	Both had similar clinical outcomes; however, the kissing balloon-inflation strategy had reduced the incidence of side branch restenosis and increased the risk of main branch restenosis.
Lv et al. 2018 [20]	6	Modified double-stent strategy compared to provisional stenting	The modified double-stent strategy is associated with improved clinical and angiographic outcomes in the short term but has increased the risk of stent thrombosis and cardiac death in late-term outcomes.
Chen et al. 2019 [21]	7	Crush stenting compared to culotte stenting	No significant differences in target lesion revascularization but better outcomes in DK Crush.
Huang et al. 2019 [22]	6	Crush stenting technique compared to provisional stenting	Crush stenting associated with significantly lower adverse events and repeated revascularization with no changes in mortality, MI, stent thrombosis.
Gioia et al. 2020 [23]	21	Provisional stenting, T stenting, T and protrusion, crush, culotte and DK-crush strategies compared	DK-Crush associated with fewer adverse cardiovascular events. Benefits of 2-stent techniques observed over provisional stenting when a side branch lesion length exceeds >10 mm.
Zhou et al. 2018 [24]	14	Stent thrombosis rates in singe-stent and double-stent strategies	Early definite stent thrombosis rates are reduced when a single-stent strategy is used
Stenting Technologies
Gao et al. 2014 [25]	9	For patients receiving drug-eluting stents, comparing a simple strategy and complex strategy.	A complex strategy using drug-eluting stents is comparable to a simple strategy in terms of outcomes. For large side branches, a complex strategy had lower rates of reintervention.
Mohamed et al. 2019 [26]	5	1-year outcomes comparing dedicated bifurcation stents to conventional stent strategies	Dedicated bifurcation stents may be associated with similar 1-year clinical and angiographic outcomes compared to conventional strategies
Edo Tondas et al. 2020 [27]	6	Comparing safety and efficacy in preventing side branch occlusion rates of conventional jailed balloon technique to modified jailed balloon technique for bifurcation lesions	Modified jailed balloon technique is potentially better than conventional jailed balloon methods in terms of side branch loss, dissection, and major adverse cardiovascular events.

Table 3.

Notable RCTs in the past 5 years (2016-2021) comparing stenting techniques.

Name	Year	# of Patients	Comparison	Outcomes
Kumsars et al. (Nordic-Baltic Bifurcation study IV) [28]	2020	450	Provisional side branch stenting compared to two-stent strategy for the treatment of true coronary bifurcation lesions involving a large side branch	There were no differences in outcomes between the two-stent technique compared with the provisional main vessel stenting technique after 2 years.
Ferenc et al. (BBK II angiographic trial) [14]	2016	300	Culotte stenting compared to TAP stenting with bifurcation lesions with the need for side-branch stenting	Culotte stenting was associated with a significantly lower incidence of angiographic restenosis compared to TAP stenting.
Zheng et al. [29]	2016	300	Crush stenting compared to culotte stenting for complex bifurcation lesions	Crush and culotte methods have the same 12-month clinical and angiographic results. Culotte stenting has lower restenosis rates and favorable flow patterns.
Chen et al. (DK Crush II Study) [30]	2017	370	Double Kissing (DK) Crush Technique compared to provisional stenting technique for coronary artery bifurcation lesions	DK Crush technique had lower rates of revascularization and clinical outcomes, and lower complication rates over a 5-year follow-up period.
Chen et al. (DK Crush V Study) [31]	2019	482	Patients with unprotected left main distal bifurcations were randomized to DK Crush or provisional stenting	Provisional stenting had higher rates of TLF, stent thrombosis over 3-year follow-up. DK Crush decreased rates of MI, and improved TLR.
Jin et al. [32]	2019	89	Balloon-stent kissing technique compared to jailed wire technique	The balloon-stent kissing technique was associated with lower rates of side-branch occlusion, side-branch post-processing, and lower degrees of postoperative proximal main-branch residual stenosis, and side-branch ostial stenosis. Also associated with lower rates of post-operative complications.
Hildick-Smith et al. (EBC TWO study) [33]	2016	200	Provisional T-stenting compared to systematic 2-stent culotte strategy in large-caliber true bifurcations	No difference was found between the provisional T-stent strategy and 2-stent culotte strategy in a composite endpoint of death, myocardial infarction, and target vessel revascularization

Recent meta-analyses evaluate the single stent provisional strategy in comparison to the planned two-stent approach, shown in Table 2. Ford et al. [18] performed a meta-analysis of 9 RCTs with 3,265 patients, evaluating long-term outcomes, defined as greater than 1 year, for provisional single stenting and up-front 2-stent strategies. In that study, they reported that provisional single stenting was associated with lower total mortality using prespecified sensitivity analysis in addition to a lower risk ratio of 2.94% compared to 4.23%. Additionally, they found that there was no statistical difference between MACE, myocardial infarctions (MI), target lesion revascularization (TLR), or stent thrombosis within groups. This finding was also confirmed by Nairooz et al. in 2017 [34], who published a meta-analysis of 8 studies, including 2778 patients. Consistent with the findings of Ford et al., they also report that provisional stenting had improved mortality and no difference between rates of MACE, MI, TLR, and stent thrombosis. Interestingly, they note that MACE risk increases in provisional stenting, specifically for patients presenting with acute coronary syndrome. Zhou et al. [24] performed another meta-analysis evaluating 14 RCTs looking specifically at stent thrombosis rates, a notable complication of PCI intervention for bifurcation lesions. Using an intention-to-treat and as-treated analysis, they confirmed that single stenting had significantly lower rates of stent thrombosis, specifically lower rates in early definite stent thrombosis. Long-term thrombosis risk was specifically evaluated by Zimarino et al. [35], who performed a meta-analysis of 12 studies (5 RCTs, 7 observational studies) comparing late thrombosis following single (via a simple provisional strategy) or double drug-eluting stents. Their findings show that the double stenting strategy resulted in a higher risk of stent thrombosis (RR: 2.31, 95% CI: 1.33-4.03) and MIs (RR: 1.86, 95% CI: 1.34-2.60). While Nairooz et al. and Ford et al. evaluated long-term outcomes, another meta-analysis by Lv et al. [20] evaluated short-term outcomes for provisional stenting and modified double stenting. They evaluated 6 studies (1683 patients) using outcomes in a <6-month window and >6-month window as primary endpoints. They also evaluated in-stent restenosis as an angiographic endpoint. They found that the modified double-stent strategy was associated with a significantly lower risk of cardiac death and lower rates of MI, TLR, and MACE in <6-month outcomes, while rates are similar for >6-month outcomes. While previous studies suggest no difference in the long-term comparing these two strategies, it is possible that the advantage of the double stenting approach lies in the shorter time window. Furthermore, Lv et al. evaluated a modified double-stenting strategy. The meta-analysis done by Lv et al. did not consider the effect of different stenting strategies. Additionally, they did not stratify patients based on the type of stent used, duration of antiplatelet therapy, or if DK crush or culotte technique was used. This suggests that there is heterogeneity in outcomes between the double-stenting approach depending on the specific technique used that future studies should seek to interrogate and compare via substratification.

3.2. A Comparison of Crush and Culotte Stenting

The original Crush technique was described by Colombo et al. in 2003 [11]. The technique was designed to prevent the recurrence of restenosis at the ostium of the side branch, a problem that was elucidated for drug-eluting stents in a 2002 clinical trial [36]. The authors posited that incomplete coverage of the ostium of the side branch was the most probable cause of restenosis. The technique introduced the first stent into the side branch before expansion, then the second stent into the main branch to fully cover the bifurcation. The side branch stent is then positioned such that the proximal marker of the side branch stent is 4-5 mm proximal to the carina of the bifurcation. Upon expansion, this allowed for more robust stenting of the complete lesion. In some ways, this technique mirrored the culotte technique, which also involves protrusion of the side branch stent into the main branch [9].

The crush stenting technique and culotte stenting technique have been compared in various randomized controlled trials. The Nordic Stent Technique Study [37] was a large multicenter randomized controlled trial comparing the efficacy of crush and culotte stenting techniques for bifurcation lesions. After 6 months, 424 patients that were randomized to treatment with crush or culotte stenting were evaluated for their primary endpoints, which included MIs, MACE, and stent thrombosis. A subset of 324 patients was also evaluated angiographically after 8 months. At 6 months, there was no significant difference between MACE, MIs, and stent thrombosis. However, angiographically, there was a significant reduction in in-stent restenosis following culotte stenting. In 2013, Kervinen et al. [13] reported the long-term efficacy of the same Nordic Stent Technique Study. After 36 months, it was found that mortality, MACE and stent thrombosis outcomes were comparable between both treatment groups. A more recent study (Table 3) by Zheng et al. [29] confirms the same findings. 300 patients were randomized to crush and culotte treatments and evaluated for primary endpoints – cardiac death, MI, stent thrombosis, and TLR – after 12 months. They also measured restenosis rates and surface integrals of time-averaged wall shear stress, a measure for flow patterns. Their findings indicate that there were no significant differences between MACE rates in the two groups. However, they also found that rates of restenosis, a secondary endpoint, were greater in the crush group (12% vs. 6%, respectively). In addition, they found that culotte stenting produced more favorable flow patterns through the stent. This suggests that while crush and culotte stenting may not have significantly different clinical outcomes, culotte stenting may be more advantageous for reducing in-stent restenosis.

The culotte technique has also been compared to TAP stenting for de-novo CBLs. In the BBK II angiographic trial, [14] 300 patients were randomly assigned to culotte stenting or TAP stenting using drug-eluting stents. The primary endpoint was maximal percent diameter stenosis of the lesion at 9-month follow-up. TLR and target lesion failure (a composite variable designed to include measures of cardiac death, MI, TLR) were also assessed. It was found that culotte stenting had lower stenosis than TAP stenting (21 ± 20% vs. 27 ± 25%, respectively). Target lesion failure, the 1-year incidence of TLR, and binary restenosis rates were statistically insignificant but trended towards improved outcomes for the culotte stenting group.

Several meta-analyses, as summarized in Table 2, compare crush and culotte stenting techniques. Meta-analytic findings observed by Chen et al. in 2019 [21] remain consistent with the most recent RCTs comparing crush and culotte stenting techniques. In this study, 3 RCTs and 4 observational studies with long-term follow-up, for a total of 2211 patients (1281 treated with crush, 930 treated with culotte), were evaluated for TLR, MACE, stent thrombosis, cardiac death, and MI. The authors found no significant difference in TLR and MACE (RR = 0.76). Cardiac death, MI, TVR were also similar for both groups. Stent thrombosis was found to be lower in patients treated with Crush. However, the authors note that the equivalent success observed by both techniques of intervention was biased by a study that evaluated the DK Crush method, as opposed to the classic crush technique, which observed a much higher heterogeneity. This suggests that culotte stenting, when compared to the traditional crush technique, may be superior. A separate meta-analysis by Huang et al. [22]. Compared crush techniques to provisional stenting techniques. They compared the outcomes of 6 studies with 2220 patients (1085 vs. 1135 for crush and provisional stenting, respectively), with follow-up times ranging from 6 months to 5 years. Their findings show a significantly lower MACE (OR: 0.73), target vessel revascularization (OR: 0.62), and target lesion revascularization (OR: 0.62) for crush techniques. However, all-cause mortality, cardiac death, MI, and stent thrombosis rates were not found to be statistically different between the two groups. Given the benefit of preventing adverse cardiac events, crush stenting was considered to be a better strategy than simple provisional stenting.

3.3. The Double Kissing (DK) Crush Technique

As a part of the standard crush technique described previously, final kissing balloon inflation is performed to decrease the likelihood of restenosis in the side branch ostium [38]. This has also been validated by measuring the improvement in clinical and angiographic outcomes after deployment of the classic crush technique [39]. The DK Crush technique was designed as an adaptation of the classic crush technique to further perfect the final kissing balloon inflation to further improve outcomes. Clinical studies show that for specifically unprotected left main CBLs, the DK crush technique improved clinical and angiographic outcomes when compared to the classic crush technique [40]. In this technique, a stent is first loaded into the side branch and inflated. After removing the guidewire and the side branch balloon, a balloon in the main vessel is inflated. This resulted in the protruding side branch stent being crushed against the main vessel wall. The side branch is subsequently re-wired. A second stent in the main vessel is then deployed to further crush the side branch stent. Final kissing inflation is then performed by the wire in the side branch [41]. Ormiston et al. performed micro-computed tomography in bench deployments of various crush techniques. They found that the improvement in angiographic outcome in the double kissing technique may be explained by the greater stent strut coverage in the ostium of the bifurcation lesion, thereby reducing the risk of restenosis [42]. Thus, while this technique has been noted to be very challenging and more time-consuming [43], many have editorialized that the improvement in clinical outcomes may be worth the difficulty in the deployment of the technique [44, 45].

The DK Crush technique has been validated in two recent randomized controlled trials (Table 3). Chen et al. [30] published the DKCRUSH-II Study in 2017, designed to evaluate the DK Crush technique. 370 patients that had a Medina 1,1,1 or 0,1,1 de novo bifurcation lesions were randomized to either the DK Crush stenting strategy or provisional strategy and followed for 1 year and 5 years. After these time points, MACE, defined as MI, cardiac death, and TVR rates, was evaluated. Angiographic outcomes were also evaluated. During a 5-year follow-up period, there was a trend toward improved MACE rates in patients treated with the DK Crush technique. Additionally, TLR was found to be improved in the DK Crush group. They also found that complex CBL had improved TLR rates when a complex stenting strategy was used, as opposed to a simple strategy. Another RCT [31], titled the DKCRUSH-V study, evaluated the efficacy of the DK Crush technique for Left Main Distal CBL. In this study, they randomized 482 patients with unprotected left main distal CBL to the DK Crush technique and provisional stenting. At a follow-up time of 3 years, target lesion failure, target vessel myocardial infarction, and stent thrombosis rates were evaluated. The findings indicated that the DK Crush technique was associated with substantially better outcomes than the provisional technique, with lower rates of target lesion failure (8.3% to 16.9% respectively) and decreased rates of myocardial infarction (5.8% vs. 1.7% respectively). Stent thrombosis rates were also substantially decreased from 4.1% in the provisional group to 0.4% in the DK Crush group. These findings suggest the superiority of the DK Crush technique to the provisional stenting group.

Two meta-analyses (Table 2) confirm the findings of the aforementioned RCTs. A large network meta-analysis [36] comparing the outcomes of twenty-one randomized clinical trials, including 5,711 patients, investigated 5 bifurcation PCI techniques – provisional, T-stenting/T-and-protrusion, Crush, Culotte, and DK Crush. Across all groups, MACE rates were substantially less in the DK Crush group (OR: 36), driven primarily by a reduction in TLR. When comparing other stenting strategies, there was no significant difference in MACE rates. In cases of side branch lesions greater than 10 mm, the 2-stent technique, primarily the DK Crush technique, offered superior outcomes when compared to provisional stenting. This finding was also confirmed by Chen et al. [21] in their meta-analysis comparing crush and culotte stenting, as previously described. To re-summarize their findings briefly, they found that the DK Crush technique demonstrated much improved rates of TVR, MACE, and stent thrombosis.

3.4. Balloon Angioplasty Techniques

One of the challenges in the one-stent provisional strategy for CBLs is the risk of side-branch occlusion and increased rates of target lesion revascularization. According to the most recent estimates, 51% of CBL stented with the provisional approach necessitate the stenting of the side branch [46]. In addition, another clinical trial by Chaudhry et al. found that side-branch compromise occurs in 16% of patients undergoing provisional stenting, which dramatically increases the risk of periprocedural MI and renal failure [47]. The standard technique used for provisional stenting is the jailed wire technique (JWT) – which lends itself to challenges such as trapping after main branch stenting or fracture during withdrawal [48]. For this reason, new ballooning and wire techniques have been designed to address this challenge and reduce rates of TLR seen with the standard provisional approach. The “balloon-stent kissing” technique (BSKT) was initially described by Jin et al. as a modification to the jailed balloon technique (JBT) in a seminal paper in 2013 [49]. In this technique, the authors utilized an inflating balloon with suitable pressure in the side branch during the main branch stent deployment. First, the side-branch balloon is inflated, then the main branch stent is immediately deployed at this pressure. The balloon in the side branch prevents the observed carina and plaque shift by occupying the side branch ostium during the main branch stent deployment [49].

Recent RCTs, as shown in Table 3, demonstrate the advances in balloon angioplasty techniques. In the study published by Jin et al. in 2013, they also described the clinical outcomes for patients for whom the BSKT technique was utilized. In a retrospective analysis, 126 patients were identified who underwent PCI treatment for de novo CBL. Of these patients, 60 patients successfully received the BSKT strategy in both emergent and elective clinical settings. In 2019, Jin et al. published a randomized controlled trial of 89 patients with 90 true CBL that were randomized to the novel BSKT technique or the standard jailed-wire technique (JWT) [32]. The primary endpoint compared between the two groups included operative success rate, occurrence of complications, postoperative quantitative coronary angiography, and incidence of perioperative and long-term MACE. The BSKT group was found to have significantly lower rates of side branch occlusion and TLR (0% vs. 15.6%, respectively) and side branch post-processing (8.9% vs. 26.7%, respectively). Additional findings also showed a lower rate of restenosis in the main-branch and side branch ostial stenosis for the BSKT group. MACE rates were also significantly improved in BSKT (0% vs. 13.3%). Given these findings, they concluded that the BSKT technique is a safe and effective technique that improves upon the JWT to deliver better outcomes for side-branch protection during PCI of CBL.

The modified BSKT (m-BSKT) technique seeks to add another optimization step to the existing technique. This technique, published by Qu et al. in 2019 [50], proposed that the routine usage of the proximal optimizing technique after rewiring could correct the mispositioning of the stented MV segment while maintaining arterial circularity. They evaluated the m-BSKT technique in a population-matched cohort study of 120 consecutive patients. M-BSKT patients were compared to their provisional stenting counterparts. Outcomes were reported immediately after the procedure and after 12 months of follow-up. Their findings interestingly demonstrated no significant difference between this modified technique and provisional stenting, which suggests a need for a randomized controlled trial, in addition to a direct comparison between the m-BSKT and the BSKT to evaluate whether the optimization step adds clinical value.

Given the improvement described by Jin et al. for the BSKT technique, Zhong et al. [19]. Sought to answer whether kissing balloon inflation (KBI) should be a routine component of the one stent approach (Table 2). They conducted a meta-analysis of 5 randomized trials, including 1,264 patients, with a primary outcome of cardiac death. Other secondary endpoints evaluated included stent thrombosis, MI, TLR, TVR, and restenosis of the main vessel and side branch. Consistent with Jin et al., they found that the KBI strategy was associated with a reduction in side-branch restenosis (OR: 0.44) but unexpectedly a higher risk of main vessel restenosis (OR: 2.96). Despite these changes, there were no differences between other primary or secondary endpoints. A separate systematic review by Tondas et al. evaluates modified-JBT (m-JBT) broadly, including BSKT, against the conventional JBT [27]. They confirmed the findings of Jin et al., describing how the m-JBT technique has fewer instances of side branch loss, dissection, and MACE.

4. ADVANCES IN STENTING TECHNOLOGY

4.1. Advances in Drug-Eluting Stents (DES)

Coronary stents were first introduced to the field of interventional cardiology in the mid-1980s. The need for stents was largely driven by unacceptable rates of acute vessel closure and restenosis using balloon angioplasty alone [51]. The use of metal stents became more commonplace after studies demonstrated the benefit of dual antiplatelet therapy administration in addition to stenting procedures [52]. Despite the improvement in outcomes, concerns about the rates of subacute thrombosis and frequent occurrence of in-stent neointimal hyperplasia remained [53]. The advancements in DES technology have allowed interventionists to overcome such challenges [54-56]. However, despite the advent of DES, they have not been universal. Reports describe DES usage range from 35-75% of all PCI [51], as the technology continues to face challenges relating to stent thrombosis and inadequate coverage of the bifurcation [53, 57]. We review recent RCTs that test new functional designs of DES, also summarized in Table 4 [58-63].

Table 4.

Notable RCTs in the past 5 years (2016-2021) comparing novel stenting technologies.

Name	Year	# of patients	Comparison	Outcomes
Yamawaki et al. (BEGIN trial) [58]	2019	226	2-link Biolimus-eluting stents compared with 3-link Everolimus-eluting stents in de novo true coronary bifurcation lesions	There were no significant differences between 3-link and 2-link design in terms of complications and angiographic outcomes.
Walsh et al. (CELTIC Bifurcation Study) [59]	2018	170	3-connector XIENCE stent compared to 2-connector SYNERGY	Both stents had a comparable performance for complications and desirable outcomes.
Bennett et al. 2018 [60] (COBRA II study) & Bennett et al. 2016 [61]	2018, 2016	40	Dedicated self-expanding Biolimus A9-eluting stent (Axxess, Biomatrix) compared to a culotte strategy using Everolimus-eluting stents (Xience)	No comparable difference in clinical coverage at 9 months and excellent clinical outcomes at 5 years.
Gil et al. [62] (PLOBOS II trial)	2016	202	Comparison of BiOSS LIM stent compared with a regular drug-eluting stent, in addition to evaluation of the final kissing balloon inflation technique.	Target lesion vascularization and cumulative major adverse cardiac events rates are comparable between both stent groups. Binary restenosis rates were significantly lower in the final kissing balloon inflation subgroup of the BiOSS group.
Yamawaki et al. [63]	2017	113	Provisional and routine kissing-balloon technique after main vessel crossover stenting using a second-generation drug-eluting stent	Provisional-FKI had increased rates of side branch restenosis and myocardial ischemia in the SB territory in the short term, but comparable long-term outcomes

In the BEGIN trial, a multicenter open-label randomized study in 2019 [58], the 2-link Biolimus A9-eluting stent (2-link Nobori BES) was compared with the durable polymer-based cobalt-chromium Everolimus-eluting stent (3-link Xience EES) in treating CBLs. This was the first randomized trial on second-generation DES in the management of CBLs. 226 patients were randomized to two groups, with 111 in BES and 115 in DES groups. Single cross-over stenting with final kissing inflation for aggressive treatment of the side branch was the main strategy used. They were followed, and results were interpreted at 8 months and 12 months. This trial has shown that at the end of 12 months, MACE was not different between the two groups. Follow-up angiographic measurements of the proximal major vessel (MV), bifurcation core, distal MV, and side branch were performed, which showed MLD at SBO, DS, in-segment restenosis, and late lumen loss was similar in both groups. Subgroup analysis in LMCA CBLs showed no difference in clinical outcomes at 8 months. They concluded that at 8 months and 12 months follow up, 2-link BES was not superior to 3-link EES despite aggressive treatment of the side branch.

Walsh et al. also performed a prospective randomized trial at nine centers in Europe in 2018 [59], called the CELTIC study. They treated 170 patients with a Medina 1,1,1 lesion with the two-stent culotte technique using the 3-connector XIENCE (81 patients) and 2-connector SYNERGY (80 patients) DES. After a follow-up of 9 months, major adverse cardiovascular or cerebrovascular events (MACCE) were reported at 5.9%, which was lower than the Nordic Stent Technique study and DKCRUSH-III study with the assumption of the same definition of MIs. Angiographic restenosis rate was reported to be 11%, higher than other studies on the culotte stenting technique. The transradial approach is seen to offer less support for complex PCI procedures; however, this study reported an extremely high success rate utilizing final kissing balloon inflation. Overall, they reported good outcomes for the transradial approach using the culotte stenting strategy with either the Xience or SYNERGY DES for appropriate symptomatic patients - with a Medina 1,1,1 lesion with a large side branch. This study was not compared to the provisional technique, so in this case, the results cannot demonstrate the superiority of the culotte technique to the provisional technique.

The current standard of care for the treatment of coronary bifurcation lesions (CBL) is the provisional side branch (SB) approach. There were no established guidelines for SB stenosis after main vessel stenting, whether to do a balloon or stent. All previous studies before PROTECT-SB investigated side branch treatment with the one stent strategy. Yamaawaki et al. in PROTECT-SB focused on MI burden at 8 months and long term with different side branch intervention strategies, which will help reflect long-term outcomes. Investigators have conducted a randomized open-label, prospective trial in four centers in Japan. They randomized 120 patients into two groups of provisional FKI group (57) and a routine FKI group (56) between February 2010 and 2012. The primary endpoint was to assess ischemic burden at 8 months after using different SB intervention strategies like provisional FKBI versus routine aggressive FKI used to reduce SB stenosis to <50% after stenting the MV using second-generation DES. The primary endpoint included the frequency of TVI after 8 months. MACE at 3 years was a composite of MI, TLR, and cardiac death. Dipyridamole-stress MPS with 99mTc and follow-up CAG were performed after 8 months, irrespective of the symptoms. MACE at 3 years was similar between the two groups, and MACE-free survival was also similar. Assessment for moderate to severe ischemia at 8 months using MPS images in the target vessel (TV) also was similar between the two groups. The rate of SB-occlusion and deterioration immediately after stenting was higher in the provisional FKI group (18 vs. 4%; 11 vs. 2%). Provisional FKI is a simple procedure with a lower need for 2 stent conversion, but it was associated with higher restenosis and ischemia in SB at 8 months.

Meta-analytic findings for DES from previous RCTs performed remain consistent with findings from recent RCTs. A meta-analysis was done by Gao et al. [25] using nine high-quality (based on Jadad scores) randomized trials to compare simple and complex stenting strategies for DES. They reported no significant differences in cardiac death, stent thrombosis, target lesion revascularization (TLR), target vessel revascularization (TVR), main vessel, and side branch restenosis (MVR, SBR). Early (during or within 1 month) and follow-up MI (within 6 months) was found to be lower in the simple strategy odds ratio (OR) of 0.53 and 0.60, respectively. In a subgroup analysis done for patients with a large side branch, no difference was noted in TLR and side branch restenosis rates, and true subgroup analysis showed similar results. However, in the same subgroup analysis, the simple strategy had a higher risk of TVR, OR of 2.27, and MVR, OR of 2.56; true subgroup analysis also showed similar results with OR of 2.04 and 2.06. The DK crush technique subgroup analysis complex strategy was associated with favorable outcomes. Overall, it can be inferred that the utilization of a complex strategy DK Crush technique using DES with patients with true bifurcation lesions with large side branches can be beneficial. To further clarify the risk of stent thrombosis with DES, Zhou et al. [24] performed a meta-analysis of 14 studies comparing the simple one-stent and the complex two-stent strategy utilizing DES for the management of CBLs. As previously described, they demonstrated that the simple one-stent strategy had improved rates of stent thrombosis. However, they did not do a formal subgroup analysis examining the role of the DK Crush technique. Thus, even though a complex strategy for DES may demonstrate more favorable MACE, TLR, and side-branch restenosis rates, the risk of stent thrombosis remains elevated in the DES-complex strategy group.

Options to treat coronary artery disease have advanced from bare-metal stents to newer-generation DES, and there has been considerable success in decreasing the risk of early and late stent thrombosis compared to 1^st generation DES [64]. The ideal DES has yet to be developed, and some of the areas for future studies are target lesion related events years after implantation due to neo-atherosclerosis, development of material that stimulates early reendothelization, development of biodegradable polymer materials allowing to mitigate in-stent restenosis due to low-grade inflammatory response to the polymer over time [65]. The duration of DAPT is also an important clinical question as many of these patients come with an increased risk of bleeding. Biodegradable polymers help in decreasing the duration of DAPT.

4.2. Dedicated Bifurcation Stents

Current stenting strategies for bifurcations utilize stents that are also used for non-bifurcation lesions. Given the various unique challenges afforded by CBL, the utilization of dedicated bifurcation stents (DBS) has also been discussed by interventionists. The use of such stents would help overcome challenges with SB protection, multiple layers, distortion of stents, SB access, crossing through the side of the stent, and gaps in the scaffolding [1]. In theory, this should reduce procedure times and decrease contrast use while also adapting to the unique bifurcation anatomy.

The use of DBS in bifurcation lesions was studied in the BiOSS Expert^® stent (Balton, Warsaw, Poland) and one with the Tryton Side Branch Stent (Tryton Medical, Durham, NC, USA). A randomized multicenter open-label trial was conducted by Gil et al. [62] to compare any regular drug-eluting stent (rDES) with the dedicated bifurcation sirolimus-eluting stent BiOSS LIM for the treatment of bifurcation lesion in POLBOS II (POLish Bifurcation Optimal Stenting) trial. They also reported the effect of FKBI on clinical outcomes. The trial was conducted between December 2012 and 2013 at four centers in Europe, with 202 patients randomly assigned to BiOSS LIM group (102) and rDES (100). Patients who received rDEs were randomized again to FKBI (49) and no FKBI (51). A provisional strategy was used in both groups with FKBI dilatation at the end. Patients were followed for 12 months, and angiography was performed at 12 months unless clinically indicated earlier. The success rate of the BiOSS LIM stent was high, and its use is safe and effective. 12-month cumulative MACE was lower in the BiOSS group but did not reach a statistically significant p-value. QCA analysis showed a significant increase in MLD and a decrease in distal stenosis percentage in the MV and MB at 12 months. No change was noted in the angle between MV and SB. Subgroup analysis of FKBI vs. no FKBI showed FKBI was related to a higher rate of SB stenting and longer fluoroscopy time. The BiOSS group who had FKBI reported having a significantly lower rate of restenosis (5.9% vs. 11.8%). The p-value was not significant for the rDES group for restenosis. The POT (proximal optimization) technique was associated with a more pronounced effect on TLR in both groups than FKBI (BiOSS group: 5.3% vs. 12.5%, and rDES group: 2.9% vs. 25%). Lower LLL (late lumen loss) was reported in both the BiOSS group and the rDES group when FKBI and POT were both applied. Lower TLR was reported in the BiOSS LIM subgroup when used for distal LM stenosis. Lower TLR and LLL reported with BiOSS stents suggested the use of these stents in bifurcations with a large difference between MV and MB can be beneficial. The study has shown the importance of POT in BiOSS LIM stents in TLR rates. The stepped design of BiOSS stents is such that theoretically, FKBI and POT-like effects would come. So interventionalists do not perform these when BiOSS LIM is used. However, this study has shown that POT had a pronounced positive effect on TLR rates (5.3% with POT and 12.5% without POT in the BiOSS group; 2.9% with POT vs. 25% without POT in the rDES group).

The sparse evidence for DBS was thoroughly examined by a meta-analysis of 5 RCTs by Mohamed et al. [26], also summarized in Table 2. Out of 1,249 patients included for analysis, 627 were treated with DBS. In general, DBS had comparable rates of MACE, all-cause mortality, cardiac mortality, MI, definite stent thrombosis, and cumulative and clinically driven TLR to conventional stenting strategies for 1-year follow-up times. They additionally note the need for a higher quality of evidence, arguing that their findings were limited by higher risks of bias, indirectness, and imprecision. Larger-scale randomized trials with longer follow-up times need to be performed to adequately assess the role of DBS in the treatment of CBLs.

5. FUTURE DIRECTIONS

Further advances in stenting technology, including improved designs of drug eluting stents, improved intravascular ultrasound tools, and upcoming clinical trials, may elucidate the optimal PCI methodologies for patients with CBL. At the time of writing this manuscript, four relevant clinical trials (Table 5) may contextualize the optimal strategy for PCI. The TIP TAP-1 Trial (NCT03714750) is a randomized controlled trial of the Reverse TAP stenting strategy to the DK Crush strategy, utilizing optical coherence tomography to evaluate post-stent angiographic outcome [12]. The primary endpoint of this study is the percentage of stent expansion in the side branch, while the secondary endpoints include procedural and angiographic success, procedural time, fluoroscopy time, and OCT parameters [12]. Another clinical trial (NCT04192760) seeks to compare the DK Crush strategy to culotte stenting. While this has been compared through indirect comparison using network meta-analytic methodologies [23], this would be the first randomized controlled trial directly comparing these two techniques. The Szabo Technique Trial (NCT03714802) compares modified T-stenting with the Szabo technique to conventional T-stenting. The Szabo 1-stent and 2-stent techniques have both been validated previously for short-term outcomes in cohort studies [66]. The insProvisional Trial (NCT03780192) evaluates the performance of the INSPIRON Sirolimus-eluting stent, a third-generation stent designed with a cobalt-chromium alloy and thin struts [67], placed with the provisional strategy.

Table 5.

Ongoing clinical trials.

Name	Registration ID	Description
TIPTAP-1 Trial	NCT03714750	Randomized controlled trial of Reverse TAP compared to DK Crush stenting strategies for CBL.
BBK-3 Trial	NCT04192760	Randomized controlled trial of culotte compared to DK-Crush strategy for non-left main CBL.
Szabo Technique Trial	NCT03714802	Randomized controlled trial of modified T-stenting with Szabo technique compared to a conventional T-stenting strategy.
insProvisional Trial	NCT03780192	Randomized controlled trial evaluating the performance of the INSPIRON Sirolimus-eluting stent using a provisional strategy.

Additionally, advancements in intravascular ultrasound (IVUS) may procedurally change the management of CBL. Presently, angiography is not only used to guide the catheterization procedure but also to ascertain the efficacy of the stenting strategy during follow-up. Recently, Gao et al. published the 3-year outcomes of the ULTIMATE Trial, which compared IVUS to angiography-guided drug-eluting stent implantation [68]. Findings showed significantly lower rates of TVR and TLR using IVUS compared to angiography-guided stent placement (4.5% vs. 6.9%).

Bioresorbable scaffolds (BRS) are another area where technology in bifurcation stenting is actively being developed. BRS are a new paradigm that allow for temporary vessel support and drug delivery without indefinite caging of coronary vessels and theoretically allow for a reduction in the long-term complications present in metallic stents [69, 70]. Despite the promising validation of this technology in bench deployment, results in randomized controlled trials have been suboptimal [71], as rates of definite scaffold thromboses for many BRS have been unacceptably high [72]. The most recent Fantom Encore® scaffold was found to have high rates of single strut fracture in bench testing, leading the authors of the study to conclude that the device would not be suitable for conventional bifurcation strategies [73]. However, some deployments of BRS strategies have shown promise on a limited scale. The Magmaris scaffold has shown positive outcomes in the BIOSOLVE-IV registry with outcomes scheduled at 12 months [74], with excellent device and procedural success and good safety profiles in low-risk populations. Additionally, BRS may also be more advantageous and preferable for specific stenting strategies. A randomized deployment of BRS in the ABC trial found rates of malapposition to be lower in the provisional stenting approach and higher rates of complications following a complex stenting strategy [75]. The theoretical advantages that BRS provides are vast, and continued investigation into the efficacy of emerging technologies will be essential for the advancement of the field [64, 65].

CONCLUSION

In this review of the state of understanding for coronary bifurcation lesion stenting, we discuss recent randomized controlled trials published in the past 5 years and meta-analyses that evaluate the performance of different stenting strategies. We also discuss advances in stent development, including novel drug-eluting stents on the market and the performance of dedicated bifurcation stents. We also review ongoing clinical trials and advances in technologies, like IVUS, that may help with the diagnosis and management of bifurcation lesions.

LIST OF ABBREVIATIONS

MV

Main vessel

MB

Main branch

SB

Side branch

CBL

Coronary Bifurcation Lesions

RCTs

Randomized controlled trials

MACE

Major adverse cardiac events

MI

Myocardial infarction

TLR

Target lesion revascularization

TVR

Target vessel revascularization

PCI

Percutaneous coronary intervention

TAP

T-and-small protrusion stenting

FBKI

Final Balloon Kissing Inflation

JBT

Jailed balloon technique

BSKT

Balloon Stent Kissing Technique

TLF

Target lesion failure

IVUS

Intravascular Ultrasound

DES

Drug Eluting Stents

BRS

Bioresorbable Scaffolds

LMCA

Left Main Coronary Artery

DBS

Dedicated Bifurcation Stents

AUTHORS’ CONTRIBUTIONS

Aedma SK and Naik A contributed equally to this work and are co-first authors. Naik A and Aedma SK designed, performed the collection of the data, edited, and wrote the paper; Naik A, Aedma SK, and Kanmanthareddy A contributed to the critical revision and editing of the paper.

CONSENT FOR PUBLICATION

Not applicable.

FUNDING

This study was not funded by any funding sources or grants, and thus there are no disclosures.

CONFLICT OF INTEREST

No potential conflicts of interest or funding sources.