Bifurcation stenting: the current state of play

Coronary bifurcation lesions are encountered in approximately 15% of percutaneous coronary interventions (PCIs) . Early experience using balloon angioplasty was discouraging due to complications such as acute closure and restenosis in both the main branch (MB) and the side branch (SB) . Although the introduction of bare-metal stents (BMS) in the treatment of bifurcation lesions provided some improvements over balloon angioplasty, results continued to be disappointing, with a 38% rate of target lesion revascularization (TLR) at 6 months and a 32% rate of major adverse cardiac events (MACEs) at 1 year . The beneficial effects of drug-eluting stents (DESs) in reducing restenosis in both simple and complex coronary lesions appear to extend to bifurcation lesions regardless of whether a single- or multiple-stent strategy is employed . However, despite the advances in both stent technology and technique, the treatment of bifurcation lesions continues to pose a challenge to the interventional cardiologist due to a combination of technical complexity and associated morbidity. This review provides a brief summary of the techniques used in the treatment of such lesions and to summarize both the randomized and clinical registry data in the treatment of coronary bifurcation lesions.

1 Modern bifurcation techniques

1.1 Culotte

First described by Chevalier et al. using BMS, the culotte technique results in two layers of stent proximal to the bifurcation, full coverage of the SB ostium and of both branches distal to the bifurcation. The technique is suitable for all angles of bifurcation, but it does leave a double stent layer at both the carina and the proximal part of the bifurcation. Furthermore, rewiring both branches through stent struts may prove both difficult and time consuming.

1.2 Crush, minicrush and reverse crush

The crush technique, first introduced by Colombo et al. as a modified T-stenting technique using DES, ensures uninterrupted patency of both the MB and the SB as well as excellent coverage of the ostium of the SB. Final kissing balloon (FKB) dilatation is now considered mandatory to allow optimal strut contact and drug delivery to the ostium of the SB . The minicrush technique differs from classical crush in the amount of the SB stent protruding into the MB, with protrusion into the proximal end of the SB ostium in the latter, limiting multiple layering of stent struts and allowing for more complete stent endothelialization.

The reverse crush technique is employed when a provisional single-stent strategy becomes suboptimal. Following the placement of a stent in the SB, an appropriately sized balloon is positioned in the MB at the level of the bifurcation, before retracting the SB stent 2–3 mm into the MB and deploying it. If the result in the SB is satisfactory, the deploying balloon and SB wire are removed and the MB balloon is inflated, thus crushing the SB stent. Subsequent steps are similar to those of conventional crush technique.

1.3 Simultaneous kissing stents

The simultaneous kissing stent (SKS) technique is considered most suitable for proximal bifurcation lesions, such as a distal left–main bifurcation lesion with an angle of <90° between the two branches . The technique has the advantage that control of the MB and the SB are not lost at any stage during the procedure and FKB dilatation can be undertaken without the need to recross either stent.

1.4 T-technique

Although less laborious than both culotte and crush, the T-technique invariably leads to inadequate coverage of the SB ostium and has consequently been discontinued in a number of institutions except for either isolated SB ostial lesions or when the result of a provisional single-stent strategy is suboptimal .

2 Randomized studies of single- vs. two-stent strategy in the DES era

The earliest randomized study to evaluate the use of the sirolimus-eluting stents (Cypher, Cordis/Johnson & Johnson, Warren, NJ) in bifurcation lesions was performed by Colombo et al. . In this angiographic-based study, 85 patients with 86 bifurcation lesions were randomized to receive either double (Group A; n =43) or single stenting (Group B; n =43). The mean reference vessel diameter was 2.6 mm in the MB and 2.1 mm for the SB. The bifurcation technique was left at the discretion of the operators, but they were encouraged not to use the culotte technique due to the closed stent design of the Cypher stent and uncertainties regarding double drug dosing. The crossover from Group B to Group A was high at 51.2%, while 4.7% crossed over from Group A to Group B because of inability to deliver the SB stent. Intravascular ultrasound was performed in all cases upon completion of the procedure and at follow-up angiography. The primary study endpoint of binary in-segment restenosis of both the MB and the SB evaluated by angiography at 6 months’ follow-up was 28% in Group A and 18.7% in Group B ( P =.53). There were three cases of stent thrombosis—two affecting the SB and one affecting both the MB and the SB. The authors concluded that the use of the Cypher stent in bifurcation lesions led to low restenosis rates in the MB but high restenosis rates in the SB when an additional stent was used.

The strategy of treating bifurcation lesions was further evaluated by Pan et al. , who compared a simple approach of MB stenting using rapamycin-eluting stents and balloon dilatation of the SB with complex reconstruction of the bifurcation by stenting both vessels. The study randomized 47 patients to the simple strategy (Group A) and 44 patients to the complex strategy (Group B). The mean vessel diameter was 3 mm for the MB and 2.5 mm for the SB. T-stenting was the bifurcation technique of choice. The crossover rate was low, with 2.1% from Group A to Group B and 9.1% from Group B to Group A due to the inability to deliver the SB stent. The 6-month rates of MACE, defined as the composite of cardiac death, myocardial infarction (MI) and the need for target vessel revascularization (TVR), as well as angiographic restenosis, were similar in both groups. This study therefore concluded that a complex strategy in bifurcation lesions offers neither a clinical nor an angiographic benefit over a simple strategy of MB stenting.

The optimal stenting strategy for bifurcation lesions was further assessed in the Nordic Bifurcation Study . This study randomized 413 patients to either a provisional single-stent strategy with the Cypher stent ( n =207) or a double-stenting strategy ( n =206). The diameter of the MB and SB were ≥2.5 and ≥2.0 mm, respectively. The bifurcation technique employed was left to the discretion of the operator. Cardiac biomarkers were measured after 12–18 h postprocedure. Although the primary endpoint of MACE (defined as the composite of cardiac death, MI, stent thrombosis, TVR) after 6 months was similar in both groups (2.9% vs. 3.4%, P >.05), the double-stenting strategy was associated with longer procedure and fluoroscopy times, higher contrast volume usage and higher rates of procedure-related increases in cardiac biomarkers of myocardial injury. Furthermore, planned angiographic follow-up in 74% of the total cohort at 8 months demonstrated no significant differences in the combined angiographic endpoint diameter stenosis >50% in the MB and occlusion of the SB (5.3% vs. 5.1%, P >.05). The rate of stent thrombosis was low and not significantly different in the two groups (0.5% vs. 0.0%, P =1.0).

The emerging theme from these three randomized studies was that routine adoption of a two-stent strategy in the treatment of bifurcation lesions provided neither a clinical nor an angiographic benefit over a provisional single-stent strategy. This was then further supported by three further randomized studies. The Bifurcations Bad Krozingen (BBK) study set out to determine whether stenting of the SB with the Cypher stent could reduce SB restenosis if a technique was applied that avoided nonstented gaps at the SB ostium . In this angiographic-based study, 202 patients were randomized to either provisional T-stenting ( n =101) or routine T-stenting ( n =101). The average diameters of the MB and SB were 3.1 and 2.4 mm, respectively. The crossover from provisional to routine T-stenting was relatively high at 18.8% and from routine to provisional 3%. At 9 months’ follow-up, the primary endpoint of in-segment percentage diameter stenosis of the SB was similar in the two groups (9.4% vs. 12.5%, P =.32). The study also did not show any significant differences in the 12-month clinical outcomes of death (2.0% vs. 1.0%, P =.1), nonfatal MI (1.0% vs. 2.0%, P =1.0), TVR (10.9% vs. 8.9%, P =.64) and stent thrombosis (1.0% vs. 2.0%, P =1.0).

The Coronary Bifurcations: Application of the Crushing Technique Using Sirolimus-Eluting Stents (CACTUS) study set out to further examine which stenting strategy should be used when a Cypher stent is implanted in bifurcation lesions . The CACTUS study randomized 350 patients to either elective crush stenting ( n =177) or provisional T-stenting ( n =173). The reference vessel diameters for the MB and SB were 2.5–3.5 and 2.25–3.5 mm, respectively. In the provisional group, 31% of patients required an additional stent in the SB due to Thrombolysis in Myocardial Infarction flow <3, residual stenosis >50 or coronary dissection. As in previous studies, CACTUS failed to demonstrate any benefits of a routine two-stenting strategy, with no differences in either the clinical endpoint of MACE (defined as the composite of cardiac death, MI or TVR) (15.8% vs. 15%, P >.05) or the angiographic rates of restenosis (4.6% and 13.2% in the MB and SB, respectively, for crush and 6.7% and 14.7% in the MB and SB, respectively, for provisional, P >.05). The rate of stent thrombosis was low and similar between the two groups (1.7% vs. 1.1%, P >.05). Interestingly, in a data subanalysis, the performance of FKB dilatation in both the crush and provisional groups was associated with lower incidences of in-hospital and follow-up MI, TVR, stent thrombosis and angiographic restenosis in both the MB and SB.

The British Bifurcation Coronary Study (BBC ONE) has been the latest randomized study to address the optimal strategy for the treatment of bifurcation lesions . The BBC ONE randomized 500 patients to either a two-stent strategy ( n =250) or a provisional T-stenting strategy ( n =250) using the paclitaxel-eluting stent (Boston Scientific, Natick, MA). The reference vessel diameters for the MB and SB were ≥2.5 and ≥2.25 mm, respectively. The bifurcation technique of choice was either crush or culotte and FKB dilatation was mandatory. Cardiac biomarkers were measured 16–22 h postprocedure. The primary endpoint of MACE (defined by all-cause mortality, MI and target vessel failure [TVF]) at 9 months was significantly greater with the two-stent strategy (15.2% vs. 8.0%, P =.009), driven primarily by an increase in the rates of periprocedural MI (11.2% vs. 3.6%, P =.001). The two-stent strategy was also associated with longer procedure times and higher radiation doses. The rates of stent thrombosis were low and similar between the two groups (2.0% vs. 0.4%, P >.05).

The BBC ONE study is therefore the first study to suggest that a routine two-stent strategy may be associated with adverse clinical outcomes, predominantly based upon incorporation of data on cardiac biomarkers in the primary endpoint, which was not the case in the Nordic Bifurcation Study. Indeed, periprocedural MI has been suggested to be an adverse prognostic marker, although not all studies have supported this concept . Another distinguishing feature of these two studies has been in the choice of stents used—Taxus in BBC ONE and Cypher in—with Cypher having been shown to offer superior results in two prior bifurcation studies . Furthermore, given the benefits of FKB dilatation in reducing both clinical and angiographic endpoints, the differences in rates of FKB dilatation being achieved in CACTUS (90%) and BBC ONE (29%) could also account for the observed differences between these two studies.

Taken collectively, it appears that a simple strategy of stenting the MB would suffice in the treatment of the vast majority of bifurcation lesions. However, it is equally clear that in the presence of a large caliber SB subtending a large amount of myocardial territory and significant length ostial SB disease, the two-stent strategy should be considered as the default strategy. A summary of the randomized studies of single vs. two-stent strategy in the current DES era is provided in Table 1 .

Table 1

Summary of the randomized clinical trials addressing the treatment of coronary bifurcation lesions

Study	Number of patients	Techniques	Primary endpoint	Outcome (%)	Stent thrombosis (%)
Colombo	85	Crush, culotte, T	Angiographic restenosis	18.7 vs. 28 ( P =NS)	3.5
Pan	91	T	Angiographic restenosis	7 vs. 25 ( P =NS)
Nordic	413	Crush, culotte, T	Death, MI, TVR or stent thrombosis	2.9 vs. 3.4 ( P =NS)	0 vs. 0.5 ( P =NS)
BBK	202	T	Angiographic restenosis	23 vs. 27.7 ( P =NS)	2 vs. 1 ( P =NS)
CACTUS	350	Crush	Death, MI, TVR	15 vs. 15.8 ( P =NS)	1.7 vs. 1.1 ( P =NS)
BBC ONE	500	Crush, culotte	Death, MI, TVF	8 vs. 15.2 ( P =NS)	2 vs. 0.4 ( P =NS)

NS, not significant.