Coronary bifurcation lesions with involvement of a significant side branch are a frequent occurrence. The Rapamycin-Eluting Stent Evaluated At Rotterdam Cardiology Hospital (RESEARCH) study showed that systematic use of drug-eluting stents (DES) resulted in an increase from 8% to 16% in the number of bifurcation lesions being stented over a one-year period. 1 This high incidence of coronary bifurcation can be explained by the propensity for atherosclerosis to develop at branch points because of turbulence and high shear stress.
The incidence of bifurcation disease is also frequently under-appreciated for many reasons - bifurcation lesions may be hard to visualise when the ostium is hidden under the main vessel or overlapping branches, often because of an angulated take-off or when the operator focuses on the main branch only, for example during primary myocardial infarction (MI) intervention or when the side branch is considered too small. Moreover, many cases that were not planned as bifurcation angioplasty, because only one branch was initially involved, ultimately turn into a true bifurcation case during the intervention. These 'unexpected' bifurcation lesions (also known as 'false bifurcations'), are located adjacent to a branch point and, in such cases, the side branch almost always becomes involved in the course of the angioplasty due to axial or circumferential plaque redistribution.
Consequently, operators must accurately visualise the configuration of a diseased coronary bifurcation before the procedure in order to be able to delineate an optimal strategy. In daily practice, a strategy of bifurcation lesion treatment is decided upon in instances when treatment of the main branch is likely to compromise a side branch considered significant by the operator.2-4
Before the Era of Stenting
Prior to the introduction of coronary stents, despite gradual equipment improvements and enhancements in technical approaches, such as kissing-balloon inflation,5-9 the immediate and mid-term results of balloon angioplasty for bifurcation disease were relatively disappointing and the presence of a bifurcation lesion with a significant branch constituted a frequent indication for coronary artery bypass graft (CABG) surgery. 'Debulking' techniques, such as directional10-12 or rotational13,14 atherectomy, which were in vogue in the early 1990s, did not significantly improve outcomes of bifurcation angioplasty.
The Bare Metal Stent Era
As early as 1992, it became clear that coronary stenting combined with dual anti-platelet agents would soon become the treatment of choice in the majority of coronary lesions. Nevertheless, the results of early experiences with first-generation stents in bifurcation lesions were relatively poor.15-24 More recent studies using second- or third-generation stents have shown that coronary bifurcation stenting has gradually become an alternative treatment to coronary surgery and that it was associated with acceptable immediate and mid-term results.25-31
Currently, in experienced centres using bare metal stents (BMS), the angiographic success rate is more than 95% for the main branch and more than 88% for both branches, with a major adverse cardiac event (MACE) rate at six to 12 month follow-up of 18% to 35% including a target vessel revascularisation (TVR) rate of 12% to 28%.32 Many approaches involving techniques of varying complexity, using two and even three stents, have been proposed (culotte, simultaneous kissing stent and skirt stent).15-21,23,27,32 The simplest of these approaches using BMS (provisional side branch T-stenting with final kissing-balloon inflation), rapidly proved to be associated with the best mid-term outcome.26,28-30,32,33
Lessons from the Bench
The most important lessons stem from the finding that the opening of struts towards the side branch (see Figure 1) results in stent deformation at the main branch level opposite the ostium of the side branch.28,34,35 This contributed to major strategic changes in the approach to bifurcation treatment. Indeed, final kissing inflation, by re-apposing the stent against the vessel wall, may reduce the risk of stent thrombosis as well as the occurrence of restenosis, especially when DES are implanted.
The second finding is that the opening of struts towards the side branch results in at least partial coverage of the side branch ostium; given the fact that in the majority of coronary bifurcation cases there is no side branch involvement at the beginning of the procedure (false bifurcation) or only ostial involvement (true bifurcation), placement of a single stent followed by final kissing inflation constitutes an optimal treatment of the lesion. Whatever the technique used, a thorough knowledge of these phenomena is essential for delineating the optimal treatment strategy, in appropriately using final kissing inflation in cases when one of the stent struts is opened towards the side branch and in paving the way for development of new dedicated systems.
Lessons from Clinical Trials and Experience
One of the main technical issues when approaching a bifurcation is to determine whether to use one stent or two stents. Most studies conducted in order to address this question are observational and non-randomised and do not provide definitive answers. The common finding of all these studies is that, whatever the technique used, implantation of a single stent compared with two is associated with a lower event rate, including a lower re-intervention rate, both in-hospital and at follow-up (see Figure 2).
The study by Al Suwaidi24 shed considerable light on the outcome of the techniques used by showing that the excess of adverse events in recipients of two stents was exclusively observed when the culotte technique was used, whereas T-stenting resulted in similar outcomes between recipients of one or two stents. The authors found the same trends in a retrospective study involving more than 1,000 patients. Indeed, the increase in the reintervention rate was directly related to the complexity of the technique used.36 In order of importance, the lowest intervention rate was associated with the placement of a single stent, followed by T-stenting starting with the main branch, T-stenting starting with the side branch, V-stenting and culotte stenting (see Figure 3).
The 'jailed wire' technique has become, quite deservedly, very popular. This technique involves stenting the main branch while a wire remains in the side branch, temporarily 'jailed' The principal advantage of this technique is that a wire placed in the side branch changes the angulation between the main branch and the side branch which in turn facilitates access to the side branch through the struts of the stent with the main branch wire.37 The second advantage is that some degree of antegrade flow and patency is maintained in the side branch in the majority of cases. It also facilitates the visualisation of the origin of the side-branch ostium in the occurrence of side branch occlusion following stent placement in the main branch. One of the predictors38 of side branch occlusion during the procedure is a Y-shaped angle of the bifurcation. This technique should therefore be used in almost all cases because in cases of a T-shaped angle, it can favourably modify side branch access, whereas in cases of Y-shaped angle it serves to maintain the flow or as a good marker of the side branch in case of occlusion.
There are very few disadvantages to the jailed wire technique. Guide-wire rupture is very rare and in the author's experience has only occurred when jailed hydrophilic wires were used. When the main branch stent is deployed with more than >1 balloon-artery diameter, withdrawing the jailed guidewire may result in inadvertent forward advancement of the guiding catheter with subsequent damage to the proximal left or right coronary artery. Deployment of the stent with a balloon-artery ratio of 0.9 to one is therefore recommended. Guide wire exchange can then be undertaken and stent deployment is completed with final kissing-balloon inflation.
The occurrence of dissection in either branch of the bifurcation during pre-dilatation should be avoided as this could create access difficulties, particularly to the side branch. The size of the pre-dilatation balloon should therefore be carefully selected according to the size of each branch distal to the bifurcation. For the same reason, there is no advantage in pre-dilating with kissing-balloon inflations.
Secondly, in patients with acute coronary syndromes, the target lesion is usually a soft plaque that can be treated by direct stenting of the main branch using the 'jailed' wire technique followed by kissing-balloon inflation. Avoiding pre-dilatation can reduce the risk of side branch dissection and therefore increase the rate of using only one stent.
Thirdly, the length of the pre- and post-dilatation balloon-catheters should be carefully selected in order to avoid 'geographical miss'. This consideration is of even greater importance in the era of DES.
The development of bifurcated stents has been virtually abandoned. The current technological limitations of this type of stent with two branches lie in its poor profile, poor flexibility and its limited rotational ability to be in phase with the two branches of the bifurcation. Other non-bifurcated dedicated devices providing easy side branch access throughout the procedure, good side branch ostial scaffolding and allowing provisional side branch T-stenting (see Figure 4) are currently being developed. The MultiLink Frontier stent (Guidant Corporation) has been evaluated in a pilot study of 105 patients. Technical success was obtained in 91% of cases and procedural success in 93%. All patients had coronary angiogram at follow-up. At six months the cumulative MACE rate was 17% (including a target lesion revascularisation (TLR) of 13%). There were no deaths and no stent thrombosis. A new generation of DES multilink frontier stent will be evaluated in 2005.
A comparable approach has been developed by Advanced Stent Technologies with the SLK-View ™ stent and more recently the Petal stent which is currently being evaluated. A new dedicated stent and delivery system stent from Invatec is also under evaluation in a small pilot European dual centre study. A self-expanding nitinol stent (from Devax) designed for bifurcation lesions employs a reverse tapering conical design to spread the carina, theoretically decreasing plaque shift into the side branch to avoid the necessity for side branch stenting. This device is currently under evaluation in a European multicentre registry.
DES for Bifurcation Lesions
The next advance in the interventional treatment of bifurcation lesions will involve the prevention of restenosis with DES. The first important study was a small multicentre pilot study with two objectives - firstly to assess the safety and feasibility of using the Cypher stent in the treatment of bifurcation lesions and secondly to randomly assess the use of one (main vessel only) compared with two (main vessel plus side branch) stents. A total of 86 patients were included. A very high cross-over rate was observed in the group randomised to only one stent in the main branch (52%).
The question as to whether one or two DES should be used cannot be answered on the basis of this study. The study showed a low rate (less than 6%) of main branch restenosis with the Cypher stent (regardless of the strategy) and an acceptable rate of side branch restenosis (22% in the two stents group compared with 14% in the balloon + stent group, p=not significant). To date, the reason for the excess of restenosis in the side branch has not been elucidated, though possible explanations may include geographical miss, lack of stent apposition at the carina and role of the technical approach.
In a more recent retrospective study, Ge et al.,41 compared two treatment types with sirolimus-eluting stents in 174 consecutive patients with bifurcation lesions (stenting only one branch in 57 or stenting both branches in 117). They found the same trend with DES (see Figure 5) as observed with BMS (see Figure 2) in favour of using one stent (incidence of target vessel revascularisation 5.4% compared with 11.1%, p=0.51 and cumulative MACE (18.9% compared with 23.3%, p=0.76) at nine months.
With the advent of DES, new techniques such as Crush42 or simultaneous kissing stent43 have been proposed in order to simplify the technical approach and reduce the risk of side branch restenosis. Recent data presented during the American College of Cardiology (ACC) scientific sessions did not demonstrate any benefit associated with these techniques. An excess of restenosis was even evidenced in the absence of final kissing (see Figure 6) as well as an increased risk of stent thrombosis (see Figure 7).
Coronary bifurcation disease is frequent and can be adequately treated using an appropriate stenting strategy. The advent of DES has generated lengthy debates and a consensus has been reached as to a rational and safe approach applicable in more than 70% of cases. In cases where there is no lesion in the side branch or a purely ostial lesion, stenting the main branch with a jailed wire in the side branch followed by provisional T-stenting of the side branch after guide wire exchange appears to be the most rational and successful strategy, provided that final kissing-balloon inflations are systematically performed.
In the presence of more complex lesions, namely when the side branch lesion extends beyond the ostium by more than 2 or 3mm, a second stent is required in the majority of cases. In such cases, the most appropriate technique between the Crush, simultaneous kissing stent or T-stenting techniques is still the subject of much debate. The data generated by the bench studies and preliminary clinical studies seem to favour the T-stenting technique starting with the main branch.
New dedicated DES and delivery systems will be available in the near future. These new devices will undoubtedly simplify the technical approach and possibly solve the problem of side branch restenosis.