Author + information
- Received November 27, 2007
- Accepted December 10, 2007
- Published online June 1, 2008.
- ↵⁎Reprint requests and correspondence:
Dr. Antonio Colombo, EMO Centro Cuore Columbus, Via Buonarroti 48, 20145 Milan, Italy.
The percutaneous treatment of coronary bifurcations has moved past an important milestone in that the 1- versus 2-stent debate appears to have been resolved. The provisional approach of implanting one stent on the main branch should be the default approach in most bifurcations lesions. Selection of the most appropriate strategy for an individual bifurcation is important. Some bifurcations require 1 stent, whereas others require the stenting of both branches. Irrespective of whether a 1- or 2-stent strategy is chosen, the results after bifurcation percutaneous coronary intervention (PCI) have dramatically improved. Dedicated bifurcation stents are an exciting new technology that may further simplify the management of bifurcation PCI and change some of these concepts.
Approximately 15% to 20% of percutaneous coronary interventions (PCIs) are performed to treat coronary bifurcations (1,2); PCIs in these circumstances are renowned for being technically challenging and historically have been associated with lower procedural success rates and worse clinical outcomes than when used to treat nonbifurcation lesions. In addition, there has also been a large amount of uncertainty and debate as to the most appropriate strategy when treating bifurcations. This debate has predominantly stemmed from a lack of randomized data, which may explain why therapeutic strategies have been largely based on the personal clinical experiences of highly skilled operators practicing in high-volume centers (3). However, in the last few years, significant improvements have occurred in our understanding and treatment of bifurcations: 1) introduction of drug-eluting stents (DES); 2) numerous randomized-controlled trials specifically in bifurcations; 3) a more selective usage of 2 stents as intention-to-treat; 4) the acceptance of a suboptimal result in the side branch (SB), i.e., one stent only on the main branch (MB), when treating bifurcations involving a minor SB; 5) the better performance of any 2-stent technique (high pressure post-dilation, kissing inflation, and possibly intravascular ultrasound); and 6) the understanding that a different approach and result should be accepted for bifurcations involving the left main compared with other bifurcations.
As a result of these changes, the outcomes after bifurcation PCI have improved considerably and, currently, approximate those of nonbifurcation PCI (4). In this review, we highlight what we currently know about the treatment of bifurcation disease and, based on this knowledge, we offer our recommendations and a practical approach to bifurcation intervention.
DES Versus Bare-Metal Stents (BMS)
The effectiveness of DES in reducing restenosis and revascularization in less-complex lesions has been extended to the coronary bifurcation. The only randomized data that compares DES and BMS comes from a subanalysis of the SCANDSTENT (Stenting Coronary Arteries in Non-Stress/Benestent Disease) trial, which examined a total of 126 patients with bifurcation lesions treated with sirolimus-eluting stents (SES) or BMS (5). In 55% of the SES cohort and 53% of the BMS cohort, stents were implanted in both branches of the bifurcation. Implantation of SES was associated with significant reductions in restenosis rates at the MB (4.9% vs. 28.3%, p < 0.001) and SB (14.8% vs. 43.4%, p < 0.001), as well as major adverse cardiac events (MACEs) during the 7-month follow-up period (9% vs. 28%, p = 0.009).
Similarly, registry studies have shown marked reductions in MACE and target lesion revascularization (TLR) rates compared with historical BMS controls. These reductions occurred irrespective of whether a 1-stent (MACE: 5.4% vs. 38%; TLR: 5.4% vs. 36%) or 2-stent (MACE: 13.3% vs. 51%; TLR: 8.9% vs. 38%) strategy was used (6,7). As a result, DES have become the preferred stent platform for the treatment of coronary bifurcations. However, BMS are still indicated when there are: 1) contraindications to prolonged dual antiplatelet therapy; 2) in the setting of bifurcation stenting in acute myocardial infarction due to concerns about a higher risk of stent thrombosis (ST) (8); or 3) in short lesions in the MB of non-true bifurcations.
1- Versus 2-Stent Strategy
There are now 5 randomized studies (1,9–12) available that compare a provisional approach of implanting 1-DES in the MB only versus a 2-DES approach of implanting a DES on both the MB and SB of the bifurcation, the results of which are summarized in Table 1. It is apparent from these data that routine stenting of both branches offers no clear advantage over a provisional strategy of stenting the MB only with balloon angioplasty of the SB, with regard to restenosis in the MB or SBs or in repeat bifurcation revascularization. A 2-DES approach is associated with procedures that are longer, with more fluoroscopy time and contrast volumes, and a greater rate of procedure-related biomarker release (1).
Thus, the contemporary treatment of bifurcation lesions has finally overcome a major crossroad: 1 versus 2 stents. There are bifurcation lesions that require 1 stent as a default treatment with a second stent implanted on the SB if a suboptimal result has been obtained and an optimal result is needed; most bifurcation lesions can be placed in this group. Likewise, there are bifurcation lesions in which 2 stents (MB and SB stenting) need to be implanted as intention-to-treat because of the characteristics of the lesion and the distribution of the SB. The distinction between these strategies is that, in the 1-DES approach, the operator may be willing to accept a suboptimal result in the SB provided that Thrombolysis In Myocardial Infarction (TIMI) flow is normal and the SB has limited clinical relevance regarding territory of distribution.
However, how do we define what a suboptimal result in the SB is? It is important to note that a major difference between the 5 randomized trials in Table 1 was the definition of a suboptimal result in the SB. This definition has a major impact both on the crossover rate from a 1-DES to a 2-DES strategy and the restenosis rate in SBs treated with a provisional strategy. In the Sirius Bifurcation study (9), a residual stenosis of >50% in the SB was considered unacceptable, which explains the very high crossover rate of 51.2%. In contrast, in the Nordic study (1), the residual SB stenosis was irrelevant, and the SB had just to remain open with TIMI flow grade >0. This information clarifies why the greatest (19.2%) SB restenosis rate with a 1-DES approach was observed in this study. Although it may be satisfactory to accept a suboptimal result with TIMI flow grade 1 in a small obtuse marginal branch, such a result is not acceptable when treating a distal left main bifurcation or a bifurcation involving a large diagonal branch. In examining the recent CACTUS (Coronary Bifurcation Application of the Crush Technique Using Sirolimus-Eluting Stents) (12) and Bad Krozingen (11) bifurcation studies, a more realistic figure is that in 20% to 30% of bifurcations treated with a provisional strategy, a second stent will have to be implanted on the SB.
Additionally, there appears to be increasing evidence that the attempt to get an optimal angiographic result with minimal residual stenosis in the SB may not be physiologically important. This concept is especially important in smaller SBs, where the majority of angiographically significant SB lesions are demonstrated to be not functionally significant by fractional flow reserve (FFR) analysis (13). Koo et al. (13) performed FFR measurements on 94 jailed SB lesions after stent implantation on the MB. No lesion with a ≥50% and <75% stenosis had a FFR <0.75. Among 73 lesions with >75% stenosis, only 20 lesions were functionally significant. Furthermore, smaller SBs are less likely to result in angina if a residual stenosis is left untreated or if restenosis occurs (14,15). However, this should not diminish the importance of protecting SBs with guidewires to prevent their closure, because it has been shown that SB compromise is not inconsequential. Occlusion of SBs >1 mm can be associated with 14% incidence of myocardial infarction (16), and SB (≥2 mm) compromise during a provisional approach can be associated with a large periprocedural myocardial infarct (17). In the end, the operator needs to decide how important an optimal result is in that particular SB for that particular patient.
An important observation from the randomized data in Table 1 is that the clinical outcomes, in particular revascularization rates, after bifurcation PCI have improved considerably. In the Nordic study, TLR was extremely low and was based solely on clinical adjudication, which was performed 2 months before angiographic follow-up, possibly neutralizing the oculostenotic reflex on repeat revascularization. The apparent disparity between restenosis rates (1-DES = 22.5%; 2-DES = 16% for the bifurcation overall) and repeat revascularization (1-DES = 1.9%; 2-DES = 1.0%) observed in this study may suggest that majority of these SB restenoses were clinically silent and possibly irrelevant. These improved outcomes in bifurcation PCI were also confirmed in the ARTS II (Arterial Revascularization Therapies Study II) study (4), in which outcomes at 1 year were similar in bifurcation and nonbifurcation lesions that were treated with SES implantation (major adverse cardiac and cerebrovascular event 11.0% vs. 13.3%, p = 0.46; revascularization 7.8% vs. 9.0%, p = 0.66). Both of these studies also confirm that outcomes in bifurcation PCI are very good if only clinically driven angiographic follow-up is performed.
The most important studies in which a 2-DES technique was used to treat a bifurcation are summarized in Table 2 (18–26). Niemela et al. (26) have performed the only randomized study comparing 2 different 2-DES techniques (Culotte vs. Crush). The study showed no difference in clinical outcomes at 6 months but was limited by the short follow-up period, lack of angiographic follow-up, and significantly less final kissing inflation (FKI) in the crush group. Thus, there are insufficient data to recommend one technique over another based solely on lower event rates. The decision is based rather on the anatomy of the bifurcation and the familiarity and competence of the operator with a specific technique. However, it is apparent that optimal performance of 2-DES techniques is important and improves outcomes. An example is the importance of FKI in reducing late loss and restenosis, especially at the SB, which has been repeatedly demonstrated and has now become standard in the performance of all 2-DES techniques (18,19). There are other important technical factors that may contribute to optimizing outcomes when performing 2-DES techniques, such as high-pressure SB inflation, the use of noncompliant balloons, selection of correct balloon size for FKI, and the use of intravascular ultrasound (IVUS).
An Individualized Approach to Bifurcation PCI
Bifurcations vary not only in anatomy (plaque burden, location of plaque, angle between branches, diameter of branches, bifurcation site) but also in the dynamic changes in anatomy during treatment (plaque shift, dissection). As a result, no 2 bifurcations are identical, and no single strategy exists that can be applied to every bifurcation. Thus, the more important issue in bifurcation PCI is selecting the most appropriate strategy for an individual bifurcation and optimizing the performance of this technique.
Although there are currently at least 6 different classifications of bifurcation lesions (27–32), the most commonly used one is the Medina classification (Fig. 1). Practically, the most important distinction is to divide bifurcation lesions into: “true” bifurcations (i.e., Medina 1.1.1, 1.0.1, 0.1.1) where MB and SB are both significantly narrowed (>50% diameter stenosis) and “non-true bifurcations,” which include all the other lesions involving a bifurcation. Non-true bifurcations should always be treated with a 1-DES strategy. Other important elements to consider that are not inherent in the bifurcation classifications include: the extent of disease on the SB (limited to the ostium or involving the vessel beyond the ostium), its size (>2.5 mm of reference diameter), angle, and distribution. The previous assumption stating that most true bifurcation lesions need 2 stents whereas the other bifurcation lesions require one stent is probably obsolete. Current approaches may require 2 stents in approximately 30% of the true bifurcations unless located on the left main coronary artery, where this percentage may be as great as 50%. These concepts may change when dedicated bifurcation stents become widely available.
The objective of bifurcation PCI is to end the procedure with both branches open and an optimal result in the MB. This statement is completely different if the SB is an important vessel sometimes as important as the MB. Among the questions the operator has to answer at the beginning of the procedure are: 1) How important the SB is for that patient and for that specific anatomy? 2) The size and extent of disease in the SB. If the SB has ostial disease, this condition is different from disease extending beyond the ostium. In true bifurcations, we like to consider a special group in which the disease of the SB extends from the origin to 10 to 20 or more mm distally which, in our experience, occurs in approximately 30% of bifurcations that we treat.
A practical approach to bifurcation PCI can be summarized here. Two wires should be placed in most bifurcations, and the SB wire should be “jailed” in the majority after the deployment of the stent on the MB. This approach is important in protecting the SB from closure as the result of plaque shift and/or stent struts during MB stenting. The jailed SB wire also facilitates rewiring of the SB (if SB post-dilation/stenting or FKI is needed, or if the SB occludes) by widening the angle between the MB and SB (33,34), by acting as a marker for the SB ostium, and by changing the angle of SB take-off. Interestingly, in the French multicenter TULIPE (Provisional T-stenting for Coronary Bifurcation Lesion Prospective Evaluation) study, the absence of this jailed wire was associated with a greater rate of reinterventions during follow-up (34). There is no need to remove the jailed wire during high-pressure stent dilation in the MB. It is preferable to avoid jailing hydrophilic guidewires because there is a risk of removing the polymer coating. Accurate handling of the guiding catheter to prevent migration into the ostium of the coronary vessel will allow removal of the jailed wire.
Two stents as “intention-to-treat” should be the technique when the disease in the SB extends beyond the ostium and when the diameter and territory of distribution are relatively large. There are no solid data to support the supposition that 2 stents are more thrombogenic than one, that is, provided correct stent placement has been performed and compliance with antiplatelet therapy is maintained. In all other conditions, SB provisional stenting should be the procedure of choice.
Our individualized approach to treating a true bifurcation is dictated by the SB. It is summarized in Figure 2 and detailed in the following sections.
SB is not suitable (too small) for stenting or clinically irrelevant AND has ostial or diffuse disease
The strategy proposed in these bifurcations is the “keep it open” strategy, which is performed as follows:
1. Wire both branches.
2. Dilate MB if needed but not SB.
3. Stent MB and leave wire in the SB.
4. Perform post-dilation of the MB with jailed wire in the SB.
5. Do not rewire SB or postdilate SB.
This “jailed wire” strategy allows protection of a SB that may not require treatment but where the need to maintain patency is important. This strategy can be used as a stand-alone technique or as part of the provisional strategy when the operator may need to eventually dilate or stent the SB. This approach of just “keep it open” for the SB was the strategy used in the provisional stenting group of the Nordic Study.
SB is suitable for stenting AND SB has minimal disease or disease at the ostium only
In these bifurcations the “provisional” strategy is used. This strategy is quick, safe, easy to perform, and has been shown to be associated with results that are comparable with a more complex approach. A 6F guide catheter is generally used but, if Xience V (Abbott Vascular Devices, Redwood City, California) or Promus (Boston Scientific, Natick, Massachusetts) are to be implanted, a 7-F guide is preferred. The provisional approach is performed as follows:
1. Wire both branches.
2. Predilate the MB and the SB as required; many SBs without significant disease do not require predilation.
3. Stent the MB, leaving the SB wire in place. If the angiographic results in MB and SB are satisfactory, the procedure is complete and the SB wire jailed behind the MB stent struts can be removed gently.
4. Rewire SB and then remove jailed wire. In our experience, recrossing into the SB through the MB stent struts is usually possible using the Rinato-Prowater wire (Asahi Intecc Co. Ltd., Nagoya, Japan/Abbott Vascular Devices) and in extremely difficult cases the ACE fixed wire balloon (Boston Scientific). In difficult situations, we have also successfully used the Pilot 150 (Abbott Vascular Devices/Guidant Corporation, Santa Clara, California) or the Miracle 3 or 4.5 gm (Asahi Intecc Co. Ltd./Abbott Vascular Devices) wires. The jailed wire in the SB should always be left in place as a marker until complete re-crossing has been done.
5. SB balloon dilation and FKI. FKI is mandatory if the SB is dilated through the MB stent struts to correct MB stent distortion and expansion (34,35).
6. If the result remains unsatisfactory (suboptimal result, plaque shift with >75% residual stenosis or TIMI flow grade <3, in a SB ≥2.5 mm) or SB balloon dilation is complicated by a flow-limiting SB dissection, then SB stenting should be performed. A second stent in the SB may only be required in 20% to 30% of cases (Table 1).
There is uncertainty as to whether FKI is mandatory when a provisional approach is used and there are no data to support it. Theoretically, and from benchmark studies, FKI has the advantage of opening stent struts that potentially can scaffold the SB ostium and, thus, facilitate future access to the SB. There is also concern that stenting across a bifurcation without opening the stent struts into the SB results in “malapposed” struts across the SB ostium that are not endothelialized. The use of FKI should only be performed in bifurcations in which the SB is suitable for stenting should dissection occur.
SB is suitable for stenting AND has disease extending beyond its ostium
A 2-stent technique is suggested in these cases. We recommend using an 8-Fr guiding catheter.
1. Wire both branches.
2. Dilate MB and SB if needed.
3. Perform crush, culotte, V-stent, simultaneous kissing stents, or T-stenting.
4. If crush is performed, rewire SB and perform high-pressure SB dilation (2-step kiss).
5. Use FKI always.
A detailed description of the various 2-stent techniques has been previously published (36). Apart from familiarity with a technique, the angle of the SB from the bifurcation is an important determinant of the technique chosen. In bifurcations with angles close to 90°C, T-stenting provides complete coverage of the SB ostium but should be avoided with more acute angles where the culotte or crush techniques are a better choice.
The Role of IVUS in Bifurcation PCI
Intravascular ultrasound may provide valuable insights into our understanding of bifurcation PCI, as illustrated by an IVUS study of the Crush technique (37). The use of IVUS demonstrated that incomplete crushing (i.e., incomplete apposition of the 3 layers of MB and SB stent struts) and ostial SB stent underexpansion are common and often not suspected angiographically. As a result of this study and our own observations, we modified our technique by using noncompliant balloons to perform high-pressure SB inflation before FKI, thereby performing a 2-step kiss.
Unfortunately, other than this study, few other experiences have been published regarding the role of IVUS in bifurcation lesions, and what follows come entirely from the direct experience of the authors during the performance of a prospective registry (INSIDE) presented by Ricardo Costa at Transcatheter Cardiovascular Therapeutics 2007 (38). The authors of this registry enrolled 31 patients with 35 bifurcations in which IVUS-guided stenting was performed on the MB and/or SB. Angiographic follow-up has been now completed in most of the patients. This study found that an optimal IVUS result at the end of the procedure was the main factor associated with the absence of restenosis on the MB or SB. We also noticed that the need for 2 stents was mainly confined to lesions with a large plaque burden at the baseline IVUS.
An interesting finding from this early experience was that an optimal angiographic result on the SB was frequently not supported by an adequate IVUS evaluation. An optimal IVUS result was finally obtained after dilation with a larger balloon and at greater pressure. These observations remain preliminary because of the small number of patients evaluated. Only a prospective study with set criteria may be able to fully evaluate this area. Despite these limitations, the authors strongly believe that IVUS guidance should be used when implanting stents in bifurcation lesions involving a large amount of myocardium at jeopardy.
In considering adjunctive procedures such as rotational or directional atherectomy, there are no data to support the superiority of these devices in bifurcation lesions. Plaque removal with directional atherectomy, in particular, may seem a rational choice in bifurcation lesions. Unfortunately, in the AMIGO (Atherectomy before Multi-Link Improves Lumen Gain and Clinical Outcomes) trial, a dedicated, randomized trial in which researchers compared BMS versus. directional atherectomy before stenting, the results failed to show any superiority for directional atherectomy, even in the subgroup of lesions involving a bifurcation (39). However, it may be that such niche technology only has a favorable cost-benefit ratio when used selectively and appropriately, as in the PERFECT (PrE Rapamycin-eluting stent FlExi-CuT) registry (40,41).
Stenting bifurcation lesions with DES are associated with an increased risk of ST. This risk is not greater when 2 stents versus 1 stent is used (1,42,43). In some reports, the use of 2 stents has been associated with an increased risk of ST in the setting of acute myocardial infarction (8). A study by Hoye et al. (18) reported a ST rate of 4.3% with the crush technique. However, of the 10 patients who had a ST, only 2 were documented angiographically, and 4 had discontinued dual-antiplatelet therapy within 7 months of the procedure (18). The Nordic study is reassuring in that only one patient had a definite ST, and this patient was treated with 1 stent. In the ARTS II study, 5 cases of ST (1.5%) occurred in a total of 465 bifurcations in 324 patients treated with DES. Four of these were subacute ST, with 3 of the bifurcation lesions having had a poor angiographic result at the end of the procedure and the only case of late ST occurring in a nonbifurcation lesion (4). Thus, there is currently no convincing evidence to suggest that we should refrain from using DES in bifurcations or that a 2-stent strategy is associated with a greater risk of ST. Despite these statements, we should take into consideration the fact that implanting 2 stents always demands more attention and expertise to obtain the best result in both MB and SB.
Dedicated Bifurcation Stents
Thus far, the main advantage of most dedicated bifurcation stents is to allow the operator to perform the procedure on a bifurcation lesion without the need to rewire the SB. Dedicated bifurcation stents can be broadly divided into 2 categories:
1. Stents for provisional SB stenting that facilitate or maintain access to the SB after MB stenting and do not require recrossing of MB stent struts (e.g., Petal, Boston Scientific; Invatec, Invatec S.r.l., Brescia, Italy; Antares, Trireme Medical Inc., Pleasanton, California; Y-med Sidekick, Y-med Inc., San Diego, California; Nile CroCo, Minvasys, Genevilliers, France; Multi-link Frontier, Abbott Vascular Devices/Guidant Corporation). These stents allow placement of a second stent on the SB if needed.
2. Stents that usually require another stent implanted in the bifurcation (e.g., Sideguard, Cappella Inc., Auburndale, Massachusetts); Tryton, Tryton Medical, Newton, Massachusetts; Axxess Plus, Devax, Irvine, California). The Tryton and Sideguard are designed to treat the SB first and require recrossing into the SB after MB stenting for FKI. The Axxess Plus is the exception as it is implanted in the proximal MB at the level of the carina and does not require re-crossing into the SB but may require the additional implantation of 2 further stents to completely treat some types of bifurcation lesions.
The Axxess Plus stent is the first of these dedicated bifurcation stents designed to elute an antirestenostic drug (biolimus A9). The Axxess Plus stent is a self-expanding, nickel-titanium, conically shaped stent that is placed at the level of the carina. Grube et al. (44) have recently published the results of the prospective multicenter single-arm Axxess Plus trial in which 139 patients were enrolled. The Axxess stent was successfully implanted on the MB in 93.5% of cases, with 80% of the patients receiving an additional stent to the MB or SB. At 6 months' follow-up, the in-stent late loss was 0.09 ± 0.56 mm, and the overall TLR was 7.5%, which confirmed the feasibility and efficacy of a dedicated bifurcation stent. A number of the other dedicated bifurcation stents are currently undergoing “first in human,” studies and the results of clinical studies using these devices are eagerly awaited. These dedicated bifurcation stents are a promising innovation in the percutaneous management of bifurcation disease but larger studies with control groups demonstrating their clinical applicability and benefit are required before they are widely incorporated into daily practice (3).
The percutaneous treatment of coronary bifurcations has moved past an important milestone in that the 1-stent versus 2-stent debate appears to have been resolved. The provisional approach of implanting one stent on the MB should be the default approach in most bifurcations lesions. Selection of the most appropriate strategy for an individual bifurcation is important. There are some bifurcations that require 1 stent, whereas others require the stenting of both branches. Irrespective of whether a 1- or 2-stent is strategy chosen, the results after bifurcation PCI have dramatically improved. Dedicated bifurcation stent are an exciting new technology that may further simplify the management of bifurcation PCI and change some of these concepts.
- Abbreviations and Acronyms
- bare-metal stent
- drug-eluting stent
- fractional flow reserve
- final kissing inflation
- intravascular ultrasound
- major adverse cardiac event
- main branch
- percutaneous coronary intervention
- side branch
- sirolimus-eluting stent
- stent thrombosis
- Thrombolysis In Myocardial Infarction
- target lesion revascularization
- Received November 27, 2007.
- Accepted December 10, 2007.
- American College of Cardiology Foundation
- Steigen T.K.,
- Maeng M.,
- Wiseth R.,
- et al.
- Myler R.K.,
- Shaw R.E.,
- Stertzer S.H.,
- et al.
- Tsuchida K.,
- Colombo A.,
- Lefevre T.,
- et al.
- Yamashita T.,
- Nishida T.,
- Adamian M.G.,
- et al.
- Ong A.T.,
- Hoye A.,
- Aoki J.,
- et al.
- Colombo A.,
- Moses J.W.,
- Morice M.C.,
- et al.
- Ferenc M.,
- Buettner H.J.,
- Bestehorn H.P.,
- et al.
- Colombo A.
- Koo B.K.,
- Kang H.J.,
- Youn T.J.,
- et al.
- Dauerman H.L.,
- Higgins P.J.,
- Sparano A.M.,
- et al.
- Hoye A.,
- Iakovou I.,
- Ge L.,
- et al.
- Ge L.,
- Airoldi F.,
- Iakovou I.,
- et al.
- Burzotta F.,
- Gwon H.C.,
- Hahn J.Y.,
- et al.
- Ge L.,
- Iakovou I.,
- Cosgrave J.,
- et al.
- Niemela M.,
- Kervinen K.,
- Erglis A.,
- et al.
- Popma J.,
- Leon M.,
- Topol E.J.
- Spokojny A.M.,
- Sanborn T.M.
- Safian R.D.
- Iakovou I.,
- Colombo A.
- Costa R.A.,
- Mintz G.S.,
- Carlier S.G.,
- et al.
- Costa R.A.
- Colombo A.
- Tsuchikane E.,
- Aizawa T.,
- Tamai H.,
- et al.
- Kuchulakanti P.K.,
- Chu W.W.,
- Torguson R.,
- et al.