Questioning ‘Keep it Simple’ in Treatment of Coronary Bifurcation Disease


The favoured approach for coronary bifurcation disease is provisional stenting, which involves stenting the main vessel (MV) and ignoring the side branch unless clinical circumstances warrant placement of a second stent. This approach is based on a number of studies showing that provisional stenting is superior to conventional two-stent approaches. There is reason to suspect, however, that the conventional wisdom regarding provisional stenting does not accurately reflect the risks and benefits of a traditional two-stent approach. Analysis of studies (e.g. Nordic I; Coronary Bifurcations: Application of the Crushing Technique Using Sirolimus-Eluting Stents [CACTUS]; and the British Bifurcation Coronary Study: Old, New and Evolving Strategies [BBC ONE]) shows that provisional stenting frequently has similar long-term outcomes to a conventional two-stent approach in some patient populations. The long-term superiority of provisional stenting in coronary bifurcation disease depends on measuring a periprocedural or post-procedural rise in cardiac enzymes; removing this measure results in similar long-term outcomes between provisional and conventional two-stent approaches. New technologies or techniques will hopefully yield clear, unambiguous improvement in coronary bifurcation stenting.

Disclosure: The author has no conflicts of interest to declare.



Citation:Interventional Cardiology 2011;6(2):150–3

Correspondence: Helen Routledge, Worcestershire Royal Hospital, Aconbury East, Charles Hastings Way, Worcester, WR5 1DD, UK. E:

Acknowledgements: The author would like to thank Jon Van Zile for medical writing services.

Support: The publication of this article was funded by Cappella Medical Devices Ltd.

Open access:

The copyright in this work belongs to Radcliffe Medical Media. Only articles clearly marked with the CC BY-NC logo are published with the Creative Commons by Attribution Licence. The CC BY-NC option was not available for Radcliffe journals before 1 January 2019. Articles marked ‘Open Access’ but not marked ‘CC BY-NC’ are made freely accessible at the time of publication but are subject to standard copyright law regarding reproduction and distribution. Permission is required for reuse of this content.

The preferred treatment approach for bifurcation disease can be generally summed up as ‘keep it simple’, otherwise known as provisional stenting. Based on several contemporary studies, coronary interventionalists have settled on a minimalist approach: starting with a single-stent approach for the main vessel (MV) and ignoring side branch (SB) disease unless clinical circumstances warrant placement of an SB stent.

But is this approach really optimal? A closer look at the prevailing wisdom shows that the optimal solution might not be so optimal after all. In fact, the studies used to support the provisional approach to bifurcation disease have been widely interpreted to demonstrate the superiority of provisional stenting when in fact they demonstrate that the outcomes of provisional stenting are similar in the long term to the more complicated conventional two-stent approaches. Instead, the ‘keep it simple’ conclusion is largely built on a clinical definition that has a questionable effect on long-term prognosis, e.g. periprocedural myocardial infarction (MI) as defined by cardiac enzyme spikes. At the same time, even supporters of simple stenting acknowledge its shortcomings, which can include unpredictable loss of the SB, SB restenosis and target vessel revascularisation (TVR).

Another argument used to recommend simple stenting is its simplicity. Because it uses only one stent, it is not as technically difficult, procedure times are shorter and patients are exposed to less contrast and fluoroscopy. While these are important considerations, they have little effect on long-term outcomes. Based on this view, simple stenting is preferred mainly by default in many patients, while more complicated bifurcation patients are frequently recommended for surgery.

Ultimately, the ideal approach to bifurcation disease is likely one that has not yet been widely available, namely a dedicated SB stent that is incorporated into the original treatment plan, is not technically challenging to deploy and protects the SB lumen. A number of such devices are becoming available, and early data are promising.

The Rationale Behind Provisional Stenting

Coronary artery bifurcation (CAB) disease has remained a stubborn and difficult lesion to confront. CAB disease accounts for up to 30 % of percutaneous coronary interventions (PCIs), but developing a standard approach has been complicated by the varied nature of the lesions themselves. CAB lesions are characterised by:

  • the extent of the disease in the main vessel (MV);
  • the extent of the disease in the ostium of the SB and the involvement of the SB itself;
  • the size of the SB; and
  • the angle of the bifurcation.

To deal with this range of complexities, a number of approaches have been developed, from a single stent in the MV to a variety of multi-stent approaches that involve stenting the SB secondary to the MV. These multi-stent approaches include such procedures as T-stenting, the crush and the culotte, with variations on each.

The multi-stent approaches typically involve placing the SB stent after the MV stent has been deployed and re-crossed with a wire, then inflating a pair of balloons in the deployed stents in a procedure known as a final kissing balloon dilation (FKBD). Depending on which approach is used, there are numerous periprocedural risks associated with this technique, including pushing plaque into the SB during MV stent deployment and distortion of the MV stent during placement of the SB stent.

The consensus that provisional stenting is superior to multi-stent procedures is based on a number of randomised studies – especially the Nordic series, BBC ONE and the CACTUS studies – that compared the various approaches. Results from these studies have generally been interpreted to mean that provisional stenting is at least as effective as multi-stent approaches, while also being faster and technically easier and reducing overall patient radiation exposure because of the reduced procedure time.

As compelling as these reasons may seem, they are not primarily driven by long-term results and are frequently oversimplified into a kind of ‘simple is better’ shorthand. Upon closer examination, however, this prevailing dogma is less convincing that it seems.

Nordic I – Underpowered

The Nordic I study compared simple versus complex stenting of bifurcation disease and concluded that the “simple stenting strategy used in the MV group was associated with reduced procedure and fluoroscopy times and lower rates of procedure-related biomarker elevation.”1 As far as clinical outcomes are concerned, Nordic I’s findings were less than convincing:

  • The study’s primary endpoint of major adverse cardiac events (MACE) was measured at six months, yet there was no statistically significant difference between the MV group and the MV + SB group (2.9 % versus 3.4 %). As the study authors wrote, “There were no significant differences in rates of major adverse cardiac events between the groups.”1
  • The study noted a significant difference in the levels of cardiac enzymes between the two groups, with any rise of ≥ three times the normal limit considered significant, but not included as ‘MACE’ as is convention for periprocedural MI. Pre- and post-procedure enzymes were only obtained in two-thirds of patients (61 % in the MV and 74 % in the MV + SB group). Overall, significant marker elevation was noted in 18 % of the MV + SB group versus 8 % in the MV group. However, this difference in enzyme markers did not yield a clinically significant difference in outcomes: a longer follow-up at eight months showed no significant difference in the rate of restenosis between the two groups.1 At 14 months, the total death rate was actually lower in the MV + SB group (1 % versus 2.4 %), as was the overall thrombosis rate (0.5 % versus 2 %).
  • Finally, the study was “significantly underpowered”, a fact noted by the study authors. The original study design was based on an anticipated primary end-point rate of 30 % in the MV + SB group. The unexpectedly low MACE rate of 3.4 % rendered the findings statistically insignificant. The study authors note they would have to include results from almost 20,000 patients to yield clinically significant results.1

When considering these results, it is particularly important to take patient selection into account. The lesions included were only those in which a single-stent strategy was felt to be an option. The number of bifurcations treated that were not deemed suitable is unknown. In the Nordic I trial, the number of true bifurcations was not singled out, and patients could be enrolled with any SB of ≥2.0 mm in diameter.

Taking all this into account, the actual message of Nordic I was not “simple is always better”; rather, it can be correctly be interpreted as “where either strategy might be employed, the long-term results from simple stenting are similar to multi-vessel stenting, in the absence of a technically easier alternative to SB stenting.”

CACTUS – All Things Being Equal

The Coronary bifurcations: application of the crushing technique using sirolimus-eluting stents (CACTUS) study group compared the two-stent crush technique with the provisional SB stenting in true coronary bifurcations. The study enrolled 350 patients in 12 centres throughout Europe, all of whom had true bifurcations, or stenosis of >50 % in both the main and side branches. The study’s primary angiographic endpoint was in-segment restenosis rate, and the primary clinical endpoint was six-month MACE. Pre-dilation and final kissing procedures were mandatory in both groups.2

The results from the CACTUS study were unambiguous: there was no statistical difference in either angiographic endpoint or clinical endpoint between the provisional group and the MV + SB group. According to the study results:

  • The six-month angiographic restenosis rates were statistically insignificant between the crush group (4.6 % MV; 13.2 % SB) and the provisional group (6.7 % MV; 14.7 % SB).
  • The MACE rate was similar in both groups (15.8 % in the crush group versus 15 % in the provisional group). In the provisional group, 31 % of patients eventually required placement of an SB stent fulfilling one of the pre-determined criteria: residual stenosis of >50 %; dissection of type B or worse; or thrombolysis in MI flow <2).

Again, based on these results, the provisional approach is the de facto choice, but not based on clinical outcomes. Rather, as the study authors noted, it is “less expensive and simpler and can be performed with less contrast and in a shorter procedural time.”

British Bifurcation Coronary Study – A Rise in Cardiac Enzymes

The British bifurcation coronary study: old, new and evolving strategies (BBC ONE) is the most cited study supporting the ‘simple is best’ approach. This relatively large, multicentre study enrolled 500 patients to either an MV-only arm or a ‘complex’ strategy with stents in both the MV and SB. The MV arm included an optional kissing inflation, while the SB + MV arm included mandatory kissing balloon inflation. This branch used the culotte and crush techniques. Eighty-two per cent of the lesions in both study groups were true bifurcations.

The BBC ONE study seemed to yield unambiguous results. The study’s primary endpoint included nine-month follow-up rates of death, MI and target-vessel failure. The results were as follows.

  • Simple approach, all primary endpoint: 8.0 %.
  • Complex group, all primary endpoint: 15.2 %.
  • Simple approach, MI: 3.6 %.
  • Complex group, MI: 11.2 %.
  • Simple approach, in-hospital MACE: 2.0 %.
  • Complex group, in-hospital MACE: 8.0 %.3

The study authors also noted that procedure time and radiation exposure were longer for complex stenting, leading to the conclusion that the “provisional technique should remain the preferred strategy in the majority of cases.”3

While the BBC ONE study was well designed, the caveats of patient selection again apply. Moreover, the clear superiority of the single-stent ‘simple’ strategy is not so obvious.

In the BBC ONE study, periprocedural MI defined as a rise in cardiac enzymes ≥ three times the upper limit of normal (ULN) was the main driver of the difference in the primary end-point. According to the study authors, “The majority of excess events in the complex group were myocardial infarctions” and “most myocardial infarctions occurred either during the procedure or soon afterwards.” In other words, the higher rates of MI among the multi-stent group can largely be attributed to a rise in cardiac enzymes during the procedure.

The real-world value of periprocedural MI as measured by cardiac enzymes on outcomes is far from settled. In the Nordic I trial, there was no significant difference in MACE rates once this figure was excluded. In another large study involving 7,147 patients, Stone et al. found that survival after elective percutaneous intervention was unaffected by any rise in cardiac enzymes less than ≥ eight times the ULN, more than twice the threshold used to diagnose MI in the BBC ONE study.4

Another study examined 1,691 patients with bifurcation lesions and compared long-term ‘real-world’ outcomes of patients who received one or two stents, depending on the operator’s preference. This study did not include periprocedural MI, as measured by cardiac enzymes, in its MACE figures. As a result, the MACE and TLR rates were not significantly different between the one- and two-stent groups over the follow-up period (averaging 22 months).5

Consequences of a Simple Strategy

Based on this analysis, the main difference in outcomes between the various bifurcation studies is more closely related to the definition of MI than to long-term outcomes. Studies such as BBC ONE that include periprocedural elevations in cardiac enzymes find an advantage in the ‘keep it simple’ approach. On the other hand, studies that do not include this measure find little or no difference in long-term MACE rates between simple and multi-stent approaches.

This does not address the other recognised advantage of the ‘keep it simple’ approach: that the procedures are easier to perform and require less contrast and less time under fluoroscopy. Here again, however, there is more to the story.

The idea behind provisional stenting is to ignore the SB whenever possible and hope the vessel is not lost during the procedure and/or there is no clinical penalty for ignoring it. Unfortunately, it is very hard to predict which patients will lose the SB and which will experience symptoms.

In one study hoping to identify factors that would lead to SB stenosis after MV stenting, Gil et al. examined the effect of the bifurcation angle. This group found that the bifurcation angle significantly affected SB stenosis (>30 degrees was predictive of SB stenosis), along with the degree of ostial narrowing (>50 %) and diameter of the SB. In patients who underwent MV stenting, the study group found a 52 % rate of SB stenosis at 11-month follow-up.6

Another group looked at the incidence of SB occlusion among a group of 175 patients who underwent intracoronary stenting. The patients had a total of 224 major SBs covered by stents; of these, 19 % were occluded at nine-month follow-up. The majority of these occlusions occurred during post-stent dilation. Although there were no adverse clinical outcomes associated with loss of the SB, the researchers determined that their results confirmed the ‘snow plow’ effect, by which plaque can be pushed from the MV into the SB during stenting.7

Overall, periprocedural loss of the SB is estimated to be between 8 % and 35 %;6,7 this is significant because loss of the SB is associated with significant chest pain and periprocedural instability resulting in an increased risk of MI. Even among those who do not lose the SB during or immediately after the procedure, a significant percentage will have SB stenosis requiring attention either at the time of the procedure or later. The cross-over rate in the true bifurcations in the CACTUS study for this reason was 34 % and, in the Nordic study, 30 % of patients underwent SB balloon dilation, though (probably because many were small-diameter branches) only 4 % were stented. Registry data suggest that in a non-selected population of bifurcations undergoing PCI largely with a provisional strategy, 28 % of patients require implantation of a second stent.9

Waiting for a Better Solution – A Dedicated Side Branch Stent

Part of the issue with the ‘simple is better’ recommendation lies with the unsettled nature of the data surrounding complex stenting. There are currently at least four major approaches to complicated stenting procedures, with many modifications. None has emerged as the clear favourite. Even the issue of the final kissing inflation is subject to controversy.

Many operators consider a final kissing inflation to be mandatory, but this has been called into question in the Nordic III study. This study examined the value of FKBD in MV stenting and found that a no-FKBD approach resulted in similar short-term clinical outcomes as an FKBD approach while resulting in a shorter procedure and reduced fluoroscopy time.8 But these results only apply to a very select group of patients with results measured at six months. Other studies with a longer follow-up period have found excellent success rates in final kissing inflations in provisional SB T-stenting, with relatively low MACE rates (10.5 % at one year) and a less than 10 % need for repeat revascularisation and low incidence of late stent thrombosis.9

Ultimately, it can be said that the superiority of the ‘keep it simple’ approach is really the result of a lack of well-studied and clearly superior options. In select groups of patients, results are equivalent, but superiority even in these select groups rests mainly only on the measurement of cardiac enzymes during the procedure. Moreover, there remains a significant proportion of patients for whom a second stent is, or at least becomes, necessary.

Currently, none of the multi-stent approaches uses stents that were designed for use in the SB. As a result, patient selection for the single- versus the multi-stent strategy depends on operator bias and technical skill. Some centres enthusiastically pursue various multi-stent strategies while others are more conservative and more willing to risk loss of the SB.

A number of manufacturers are working on dedicated SB stenting systems. A dedicated two-stent approach would incorporate the best of the ‘keep it simple’ approach while mitigating the risk of SB loss. It would be simple to deliver, not require increased time under fluoroscopy and yield clinically superior results compared with today’s bifurcation strategies. Undoubtedly, this will change our approach to bifurcation disease.

As the lead author of the BBC ONE study noted, “Numerous ingenious dedicated systems are in development, and we can expect these to play a major role in the coming years.”10


  1. Steigen TK, Maeng M, Wiseth R, et al., Randomized study on simple versus complex stenting of coronary artery bifurcation lesions: the Nordic bifurcation study, Circulation, 2006;114(18):1955–61.
    Crossref | PubMed
  2. Colombo A, Bramucci E, Saccà S, et al., Randomized study of the crush technique versus provisional side-branch stenting in true coronary bifurcations: the CACTUS (Coronary Bifurcations: Application of the Crushing Technique Using Sirolimus-Eluting Stents) Study, Circulation, 2009;119(1):71–8.
    Crossref | PubMed
  3. Hildick-Smith D, de Belder AJ, Cooter N, et al., Randomized trial of simple versus complex drug-eluting stenting for bifurcation lesions: the British Bifurcation Coronary Study: old, new, and evolving strategies, Circulation, 2010;121(10):1235–43.
    Crossref | PubMed
  4. Stone GW, Mehran R, Dangas G, et al., Differential impact on survival of electrocardiographic Q-wave versus enzymatic myocardial infarction after percutaneous intervention: a device-specific analysis of 7147 patients, Circulation, 2001;104(6):642–7.
    Crossref | PubMed
  5. Gwon HC, Choi SH, Song YB, et al., Long-term clinical results and predictors of adverse outcomes after drug-eluting stent implantation for bifurcation lesions in a real-world practice: the COBIS (Coronary Bifurcation Stenting) registry, Circ J, 2010;74(11):2322–8.
    Crossref | PubMed
  6. Gil RJ, Vassilev D, Rzezak J, et al., Influence of parent vessel - side branch distal angle on restenosis rates after main vessel stenting in coronary bifurcation lesion, Kardiol Pol, 2009;67(1):36–43; discussion 44–5.
  7. Aliabadi D, Tilli FV, Bowers TR, et al., Incidence and angiographic predictors of side branch occlusion following high-pressure intracoronary stenting, Am J Cardiol, 1997;80(8):994–7.
    Crossref | PubMed
  8. Niemelä M, Kervinen K, Erglis A, et al., Randomized comparison of final kissing balloon dilatation versus no final kissing balloon dilatation in patients with coronary bifurcation lesions treated with main vessel stenting: the Nordic-Baltic Bifurcation Study III, Circulation, 2011;123(1):79–86.
    Crossref | PubMed
  9. Routledge HC, Morice MC, Lefèvre T, et al., 2-year outcome of patients treated for bifurcation coronary disease with provisional side branch T-stenting using drug-eluting stents, JACC Cardiovasc Interv, 2008;1(4):358–65.
    Crossref | PubMed
  10. Hildick-Smith D, Flensted Lassen J, Koo B, One or two stents for coronary bifurcation lesions? Euro Intervention, 2010;6(Suppl. J):J64.
    Crossref | PubMed