Author + information
- Received February 28, 2017
- Revision received April 25, 2017
- Accepted May 16, 2017
- Published online August 7, 2017.
- Ziad A. Ali, MD, DPhila,b,∗ (, )
- Todd J. Brinton, MDc,
- Jonathan M. Hill, MDd,
- Akiko Maehara, MDa,b,
- Mitsuaki Matsumura, BSa,b,
- Keyvan Karimi Galougahi, MD, PhDa,
- Uday Illindala, MSe,
- Matthias Götberg, MD, PhDf,
- Robert Whitbourn, MDg,
- Nicolas Van Mieghem, MDh,
- Ian T. Meredith, MBBS, PhDi,
- Carlo Di Mario, MD, PhDj and
- Jean Fajadet, MDk
- aCenter for Interventional Vascular Therapy, Division of Cardiology, New York Presbyterian Hospital and Columbia University, New York, New York
- bCardiovascular Research Foundation, New York, New York
- cStanford University Hospital and Clinics, Stanford, California
- dKing’s College Hospital NHS Trust, London, United Kingdom
- eShockwave Medical, Fremont, California
- fInstitute of Clinical Sciences, Lund University, Skane University Hospital, Lund, Sweden
- gDepartment of Cardiology, St Vincent's Hospital, Melbourne, Fitzroy, Australia
- hVU University Medical Center, Amsterdam, the Netherlands
- iCardiovascular Research Centre, Monash Medical Centre, Clayton, Australia
- jRoyal Brompton Hospital, London, United Kingdom
- kClinique Pasteur, Toulouse, France
- ↵∗Address for correspondence:
Dr. Ziad A. Ali, Center for Interventional Vascular Therapy, Division of Cardiology, New York Presbyterian Hospital and Columbia University, New York, New York 10032.
Objectives This study sought to determine the mechanistic effects of a novel balloon-based lithoplasty system on heavily calcified coronary lesions and subsequent stent placement using optical coherence tomography (OCT).
Background The Shockwave Coronary Rx Lithoplasty System (Shockwave Medical, Fremont, California) delivers localized, lithotripsy-enhanced disruption of calcium within the target lesion (i.e., lithoplasty) for vessel preparation before stent implantation.
Methods We analyzed OCT findings in 31 patients in whom lithoplasty was used to treat severely calcified stenotic coronary lesions.
Results After lithoplasty, intraplaque calcium fracture was identified in 43% of lesions, with circumferential multiple fractures noted in >25%. The frequency of calcium fractures per lesion increased in the most severely calcified plaques (highest tertile vs. lowest tertile; p = 0.009), with a trend toward greater incidence of calcium fracture (77.8% vs. 22.2%; p = 0.057). Post-lithoplasty, mean acute area gain was 2.1 mm2, which further increased with stent implantation, achieving a minimal stent area of 5.94 ± 1.98 mm2 and mean stent expansion of 112.0 ± 37.2%. Deep dissections, as part of the angioplasty effect, occurred in 13% of cases and were successfully treated with stent implantation without incidence of acute closure, slow flow/no reflow, or perforation.
Conclusions High-resolution imaging by OCT delineated calcium modification with fracture as a major mechanism of action of lithoplasty in vivo and demonstrated efficacy in the achievement of significant acute area gain and favorable stent expansion.
Coronary artery calcification adversely affects procedural and clinical success. Whereas calcified lesions in nonculprit vessels of patients presenting with acute coronary syndromes cause fewer future major adverse cardiovascular events (MACE) than lipid-rich lesions (1), moderate to severe calcification in the culprit vessel is a strong predictor of MACE after percutaneous coronary intervention (PCI) (2). Coronary calcification can impair device delivery, damage the drug-eluting polymer (3), and impair stent apposition and inhibit expansion (4), thus predisposing to stent failure. Atherectomy can facilitate delivery of interventional equipment and improve stent expansion; however, it is used infrequently because of unfamiliarity, additional cost, complexity, and paucity of data for clinical benefit (5). We recently described the first use of a lithoplasty balloon in the coronary artery for modification of calcified plaques (6,7). Lithoplasty is a technique based on lithotripsy, an established treatment strategy for renal calculi, in which multiple emitters mounted on a traditional balloon catheter platform create diffusive circumferential pulsatile mechanical energy to disrupt calcified plaque. Herein, we describe the mechanism and efficacy of lithoplasty using optical coherence tomography (OCT).
The current study reports the OCT findings in 31 patients in whom the Coronary Rx Lithoplasty System (Shockwave Medical, Fremont, California) was used to treat calcified, stenotic, de novo coronary lesions before stenting (6). The trial was conducted at 7 hospitals in 5 countries and was designed by the sponsor (Shockwave Medical) as part of the DISRUPT CAD program. An independent clinical events committee adjudicated all MACE. Independent core laboratories analyzed all imaging, including angiography (Yale Cardiovascular Research Group, New Haven, Connecticut) and OCT (Cardiovascular Research Foundation, New York, New York). The institutional review board or ethics committee at each participating center approved the study protocol. The trial was registered at clinicaltrials.gov (NCT02471586).
Patients undergoing planned PCI with a metallic drug-eluting stent for angina (stable or unstable) or silent ischemia with angiographic severe calcification (calcification within the lesion on both sides of the vessel assessed by angiography) as determined by the operators were considered for enrollment. Eligible patients had single target lesions located in a native coronary artery with visually estimated reference vessel diameter (by angiography) of 2.5 to 4.0 mm and length ≤32 mm. Lesions with the following characteristics were excluded from randomization: unprotected left main, planned concomitant use of atherectomy or specialty balloon, chronic total occlusions, and stent within 5 mm of the lesion. Detailed inclusion and exclusion criteria appear in Online Table 1. All patients provided informed written consent before enrollment. The study was conducted in accordance with the Declaration of Helsinki, ISO 14155, good clinical practice guidelines, and applicable laws by all related governmental bodies.
Lithoplasty and PCI
All patients were treated with the Coronary Rx Lithoplasty System. The Coronary Lithoplasty catheter is a single-use, sterile, disposable balloon angioplasty catheter that contains a series of unfocused, electrohydraulic lithotripsy emitters. The emitters convert electrical energy into transient acoustic circumferential pressure pulses that disrupt calcification within vascular plaque. The balloon catheter is connected via a patient cable to the generator, which is pre-programmed to deliver the specified dosage of pulses per treatment.
PCI was performed via femoral or radial access using anticoagulation per operator preference. Dual-antiplatelet therapy and other medications were administered as per local standard of care. The lithoplasty balloon, sized 1:1 to the reference artery ratio, was inflated to 4 atm, and 10 pulses were delivered, followed by further dilatation to nominal pressures. The balloon was then inflated to reference vessel size based on the balloon inflation chart. The procedure was repeated to provide a minimum of 20 pulses in the target lesion, with interval deflation to allow distal perfusion. If the lesion exceeded the 12-mm balloon length, the balloon was repositioned and the lithoplasty repeated. Repositioning of the lithoplasty balloon was required in the majority of lesions to treat the length of calcification. Care was taken to minimize overlap of the lithoplasty treatments; nevertheless, because the lithoplasty balloon length is 12 mm, some overlap was unavoidable. When repositioning was required, the number of lithoplasty treatments, pulses, and inflation time were calculated as the sum of these indexes per lesion. Subsequent stent implantation and PCI optimization were performed at the discretion of the operator.
OCT image acquisition and analysis
OCT was performed with a commercially available frequency-domain OCT system (ILUMIEN OPTIS or C7-XR [St. Jude Medical, St. Paul, Minnesota]). Imaging was attempted at 3 time points: before lithoplasty, after lithoplasty, and at completion of the procedure. A 1.5-mm compliant balloon was inflated at nominal pressure to pre-dilate the lesion if the OCT catheter was unable to pass beyond the lesion. Automated pullback was triggered with intracoronary contrast injection (3 to 4 ml/s, 12 to 14 ml total) with a motorized pullback speed of up to 25 mm/s, a frame rate of 100/s, and maximum scan length of 75 mm. All OCT images were de-identified, digitally stored, and submitted to the imaging core laboratory for independent off-line analysis. OCT images were analyzed with St. Jude Medical Offline Review Workstation software version E.0.2. Calibration was performed for each segment, and every frame was evaluated. All OCT runs were carefully screened for quality, and those with poor visualization because of artifacts or inadequate blood clearance were excluded. Structures were classified according to established OCT reporting standards (8). Reference lumen area was recorded by OCT automated measures or calculated by tracing the luminal contour on the proximal and distal reference segments (the largest area within 5 mm from the edge of lesion). Area stenosis was calculated with the formula (1 − [minimum lumen area/mean reference lumen cross-sectional area]) and expressed as a percentage. Asymmetry index was calculated per lesion as (1 − minimum stent diameter / maximum stent diameter). Eccentricity index was calculated as the ratio of minimum and maximum stent diameter per cross section (9). Calcific plaque was defined as low-attenuation signal with sharply delineated borders (8). Quantitative indexes of calcification were evaluated and measured in each OCT frame. Calcium location was defined as superficial if the luminal leading edge of calcification was located within 0.5 mm from the surface of plaque. Calcification arc was measured with a protractor centered on the lumen. If there was >1 calcium deposit present in a single cross-sectional frame, the arc was defined as the sum of the arcs of each individual calcium deposit for that cross section. Maximum calcium arc was recorded for each lesion. Calcium length was determined by identifying the proximal and distal calcium edges and summed if there were multiple separate calcium deposits. Calcium thickness was determined as the distance between the luminal edge of the calcium and the outer border of the deposit measured throughout the lesion, with the maximum value per lesion recorded. In cases for which the inner border of calcification could be identified but the outer border could not, the maximum calcium depth was defined as the distance from the luminal calcium border to the outermost sharply delineated border. In addition, calcium was measured at 1-mm intervals over the entire length of each lesion and summed and divided by the number of 1-mm-interval frames analyzed to obtain mean calcium arcs. Calcium volume index was then calculated as mean calcium arc × calcium length (° × mm). These values are reported within the text. Calcium severity was stratified into tertiles according to calcium volume index. Matching of the cross-sectional frames between pre- and post-lithoplasty OCT images was performed using fiduciary marks, side branch locations, and plaque shapes. In cases in which OCT was performed with ILUMIEN OPTISI OCT angiography coregistration, the automated coregistration was used for matching. Calcium fracture was identified as a new disruption or discontinuity in the calcium sheet identified on OCT after lithoplasty or stenting. The luminal area was measured by tracing the luminal border on each cross section, which after lithoplasty included marking the lumen contour in continuum with the additional luminal border generated by the fractures, from which the acute luminal gain post-lithoplasty was calculated. To determine the number of fractures per lesion, the fracture lines were traced for continuity frame by frame throughout the lesion and cross-checked with the longitudinal OCT image. Representative examples from sequential OCT acquisition are shown in Figures 1 and 2.
Data were subjected to the Kolmogorov-Smirnov test to determine distribution. Continuous variables are described as mean ± SD compared by Student t tests if parametric and median with interquartile range (IQR) compared by Mann-Whitney U test if nonparametric. When multiple groups were compared, data were analyzed by analysis of variance with Bonferroni post-test. A value of p < 0.05 was considered statistically significant. Statistical analyses were conducted with SAS version 9.4 (SAS Institute, Cary, North Carolina) or higher.
Patients and procedures
Between December 2015 and September 2016, 31 patients at 7 hospitals in 5 countries were treated with lithoplasty and underwent serial OCT imaging. Baseline clinical characteristics are presented in Table 1. Lithoplasty was performed with 2.0 catheters (IQR: 1 to 2 catheters), delivering 4.0 treatments (IQR: 2 to 7 treatments) totaling 94 ± 75 pulses at an inflation pressure of 6.0 atm (IQR: 6 to 6 atm) for 107 ± 56 s. Further procedural details are presented in Table 2.
Although operators only enrolled patients with severely calcified lesions as per their assessment of angiography, the core laboratory analysis identified angiographically severe calcification in 27 lesions (87%) (Table 3). Lesion length on angiography was 21.7 ± 11.6 mm, of which 21.3 ± 10.3 mm displayed evidence of calcification. Using lithoplasty for vessel preparation, stent delivery was facilitated in all cases, followed by stent implantation, with the percent diameter stenosis decreasing from 65.1 ± 14.4% to 13.9 ± 12.5%. After lithoplasty, there were 4 (12.9%) National Heart, Lung, and Blood Institute classification type B or greater dissections that occurred due to angioplasty, all of which were successfully treated with stent implantation. There were no incidences of slow-flow/no-reflow, abrupt closure, or perforation post-lithoplasty. Additional angiographic characteristics are presented in Table 3.
Lesion characteristics on OCT
Pre-dilatation with a ≥1.5-mm balloon was needed in 6 of 31 patients (19%) for delivery of the OCT catheter before lithoplasty, subsequent to which the lithoplasty catheter could be delivered in all lesions without additional pre-dilatation required. Both OCT catheters and lithoplasty balloons could be delivered without pre-dilatation in the remainder of the lesions. On the basis of the image quality assessment by the core laboratory, OCT images were analyzed in 26 patients (83%) pre-intervention, 28 patients (90%) post-lithoplasty, and all 31 patients (100%) after stent implantation. Lesion length by OCT was 31.5 ± 9.7 mm (Table 4). Vessel preparation with lithoplasty led to an increase in minimum lumen area (4.16 ± 1.86 mm2 vs. 2.23 ± 1.11 mm2; p < 0.01), a reduction in area stenosis (39.7 ± 24.2% vs. 66.4 ± 11.3%), and an acute area gain of 2.08 ± 1.65 mm2, with further improvements in all indexes post-stenting (Table 4). After lithoplasty and balloon inflation to reference vessel diameter, stents could be delivered to all target lesions without additional balloon pre-dilatation required. Minimal and mean stent area were 5.94 ± 1.98 mm2 and 8.37 ± 3.17 mm2, respectively. Minimal and mean stent expansion were 79.4 ± 22.7% and 112 ± 37.2%, respectively. Before lithoplasty, the minimum lumen area co-localized with the site of maximal calcification in 7 of 26 lesions (26.9%). After lithoplasty and stent implantation, this number was reduced to 2 of 31 lesions (6.5%; p = 0.04). Final post-PCI asymmetry index was 0.38 (IQR: 0.31 to 0.45), and eccentricity index was 0.72 (IQR: 0.67 to 0.77).
OCT characteristics of calcium modification by lithoplasty and stenting
Before lithoplasty, calcification length on OCT was 20.6 ± 9.9 mm, calcification arc was 130 ± 51.6°, and calcium volume index was 2,753 ± 1,794 mm°. After lithoplasty, calcium fracture was identified in 12 lesions (42.9%), with a trend toward a further increase after stent implantation (17 [54.8%]; p = 0.08) (Table 5). Although fracture depth did not change after stent implantation compared with post-lithoplasty (0.43 ± 0.25 mm vs. 0.42 ± 0.21 mm; p = 0.72), fracture length increased significantly (2.79 ± 4.49 mm vs. 3.36 ± 4.99 mm; p = 0.02), with nonsignificant trends toward increased fracture angle (20.5 ± 19.5% vs. 29.5 ± 33.7%; p = 0.06) and increased number of calcium fractures per lesion (0.0 [IQR: 0.0 to 1.5] vs. 1.0 [0.0 to 2.0]; p = 0.03) observed after stenting. Multiple calcium fractures on a single cross-sectional frame were identified in >25% of treated lesions post-lithoplasty and stent implantation (Table 5).
To assess whether the modifying effects of lithoplasty differed according to calcification severity, we compared calcium modifications between tertiles of calcification (Table 6). Although there was a trend toward greater incidence of calcium fracture in the highest calcification tertile compared with the lowest (77.8% vs. 22.2%; p = 0.057), other metrics of calcium fracture, including the number of calcium fractures per lesion (2 [IQR: 1 to 4] vs. 0 [IQR: 0 to 0]; p = 0.009) and quadrants of calcium fracture (103 [IQR: 20 to 218] vs. 0 [IQR: 0 to 0]; p < 0.0001) were significantly greater. Moreover, the presence of multiple calcium fractures identified on a single cross-sectional frame was highest in the highest tertile (6 [66.7%] vs. 0 [0%]; p = 0.009). Fracture angle was also significantly greater in the middle and highest tertile compared with the lowest tertile. Fracture depth was not statistically different between tertiles (Table 6). Despite an increasingly higher calcification arc and similar baseline percent area stenosis from the lowest to highest tertiles at the site of maximum calcification, similar mean stent expansion was achieved in all tertiles (Table 6).
Here, we describe the effects of lithoplasty on calcified plaques in vivo using high-resolution intravascular imaging by OCT for the first time and report a number of important findings. First, this report supports the overall safety of lithoplasty. There were no major lithoplasty-induced complications or post-PCI sequelae. Second, lithoplasty improved lesion compliance in the presence of severe calcification, augmenting stent expansion throughout the lesions, including at the point of maximal calcification. Third, lithoplasty led to calcium fracture without the need for high-pressure, noncompliant balloon inflation. Fourth, lithoplasty induced circumferential calcium modification, as evidenced by multiple fractures in single cross sections. Lastly, the effects of lithoplasty were more pronounced with increasing severity of calcification. Taken together, these findings demonstrate the mechanism of action of intracoronary lithoplasty in humans for the first time and provide support for its intended clinical utility in the modification of calcified plaque to optimize stent delivery and expansion.
This report highlights the distinctive features of calcified plaque modification by lithoplasty compared with rotational atherectomy (RA) or orbital atherectomy (OA). OCT analysis has demonstrated that RA (10–13) and OA (13–15) modify calcium in the shape of a relatively smooth lumen with a cylindrical shape (groove) that follows the guidewire course (guidewire bias), with relatively small increases in cross-sectional area (15). In regions of tortuosity or eccentric plaque, this could result in tunnel or crater formation with RA or OA, increasing the well-recognized risk of perforation in such vessel anatomy. In contrast, lithoplasty provides circumferential plaque modification, as evidenced by the findings of multiple calcium fractures in single cross sections. Such circumferential modification holds the potential advantage of uniform energy distribution and thus uniform plaque modification, which could reduce asymmetry and eccentricity. Although we found significant asymmetry and eccentricity even after lithoplasty and stent implantation, how this compares to plain old balloon angioplasty, specialty balloons, and atherectomy remains unknown and is the subject of ongoing investigation. Because new clinical data identify asymmetry as a predictor of bioabsorbable vascular scaffold failure (9), lithoplasty holds promise as a new modality of vessel preparation for these novel devices.
Although lacerations and deep craters (particularly in the presence of large lipids or smaller calcium arc with RA and OA) might aid in more uniform stent expansion/apposition (13), they might compromise the structural integrity of the vessel and increase the risk of perforation. In contrast, by inducing circumferential calcium fractures and augmenting compliance throughout the lesion, including at the point of maximum calcification, lithoplasty resulted in enhanced stent apposition and expansion, with stent expansions achieved in complex, calcified lesions that approached those achieved in less complex, noncalcified lesions in a recent contemporary randomized, controlled imaging trial (16). Perforation rates reported in the literature vary from 0.0% to 1.5% with RA (5) and from 0.9% to 1.8% with OA (17), and no perforations were observed in this study.
By delivering local shockwave energy, lithoplasty does not rely on the high-pressure inflation required with scoring or conventional noncompliant balloons to modify calcium. High-pressure balloon dilatation induces fracture in calcified plaques with high calcium arc and low calcium thickness (18), whereas lithoplasty-induced fractures became ever more prominent as the severity of calcification increased. Moreover, as shown in our analysis, shock-wave pulses affect calcium sheets located within the target field regardless of their depth in the vessel wall, which is in contrast to the inefficacy of RA or OA to modify deep-seated calcium (13,14). Lastly, in contrast to plaque abrasion by RA or OA, which generates microparticles that embolize distally, thus impairing microcirculatory function (19), large calcium fragments generated by lithoplasty remained in situ. Indeed, compared with the incidences of 0.0% to 2.5% in contemporary series with RA (20), there were no incidents of slow-flow/no-reflow observed with lithoplasty in the current study.
First, the present study is a nonrandomized, observational cohort study without a control group. Future studies are needed to compare the efficacy and safety of lithoplasty with conventional balloons, cutting/scoring balloons, OA/RA, or excimer laser coronary atherectomy (21) in a head-to-head design, given the limitations of indirect historical comparison (22). Nevertheless, compared with established methods for treatment of severe calcification, we report similar if not improved results as assessed by intravascular imaging. Second, OCT was not performed in all 60 patients enrolled in the DISRUPT CAD program but rather in a selected cohort based on the discretion of the operator. Third, although intravascular OCT is superior to IVUS in its ability to penetrate and delineate calcium volumetrically, its limited depth penetration could have missed deep calcium, and as such, the effects of lithoplasty on this type of calcium remain unclear. Finally, the sample size of our study was small, which makes comparisons subject to type II error.
Here, we present the first description of the mechanism of action of coronary lithoplasty in vivo using OCT. OCT imaging identified calcium fracture along the circumference of the lesions and multiple fractures in a single cross section in >25% of lesions, which led to a mean acute area gain of ≈2.1 mm2. Lithoplasty-induced fractures were independent of calcium depth, with multiple fractures per lesion occurring more frequently as the severity of the underlying calcification increased. Lithoplasty-facilitated PCI resulted in stent apposition and expansion that approached that in contemporary series of drug-eluting stent implantation in less complex and noncalcified lesions, with a safety profile that was comparable to that of other established methods of calcified plaque modification.
COMPETENCY IN MEDICAL KNOWLEDGE: Although calcified lesions in nonculprit vessels of patients presenting with acute coronary syndromes cause fewer future MACE than lipid-rich lesions, moderate to severe calcification in the culprit vessel is a strong predictor of MACE after PCI.
TRANSLATIONAL OUTLOOK: Atherectomy can facilitate delivery of interventional equipment; however, it is used infrequently because of unfamiliarity, additional cost, complexity, and paucity of data for clinical benefit. Lithoplasty is an entirely novel technique based on lithotripsy, an established treatment strategy for renal calculi, in which multiple emitters mounted on a traditional balloon catheter platform create diffusive circumferential pulsatile mechanical energy to disrupt calcified plaque. We describe for the first time, using high-resolution OCT imaging, that lithoplasty is effective for calcific plaque modification in vivo and leads to stent expansion comparable to noncalcified lesions in contemporary trials.
For supplemental data, please see the online version of this paper.
Dr. Ali holds equity in Shockwave Medical; has served as a consultant for St. Jude Medical and ACIST Medical; has received grant support from St. Jude Medical and Cardiovascular Systems, Inc.; and has served on a scientific advisory board for Shockwave Medical. Dr. Brinton is a co-founder of and holds equity in Shockwave Medical and serves on its board of directors. Dr. Maehara has received research grants from Boston Scientific and St. Jude Medical; and has served as a consultant for Boston Scientific and OCT Medical Imaging, Inc. Mr. Illindala is an employee of Shockwave Medical. Dr. Gotberg has served as a consultant for Boston Scientific and Volcano Corporation; has received a research grant from Volcano Corporation; and has served on an advisory board for Medtronic. Dr. Di Mario has received a research grant to his institution from Shockwave Medical. Dr. Fajadet has received educational grants from Terumo, Abbott, Boston Scientific, and Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. P.K. Shah, MD, served as the Guest Editor for this article.
- Abbreviations and Acronyms
- interquartile range
- major adverse cardiovascular event
- orbital atherectomy
- optical coherence tomography
- percutaneous coronary intervention
- rotational atherectomy
- Received February 28, 2017.
- Revision received April 25, 2017.
- Accepted May 16, 2017.
- 2017 American College of Cardiology Foundation
- Xu Y.,
- Mintz G.S.,
- Tam A.,
- et al.
- Généreux P.,
- Madhavan M.V.,
- Mintz G.S.,
- et al.
- Abdel-Wahab M.,
- Richardt G.,
- Joachim Buttner H.,
- et al.
- Brinton T.,
- Hill J.,
- Ali Z.,
- et al.
- De Silva K.,
- Roy J.,
- Webb I.,
- et al.
- Tearney G.J.,
- Regar E.,
- Akasaka T.,
- et al.
- Suwannasom P.,
- Sotomi Y.,
- Ishibashi Y.,
- et al.
- Mestre R.T.,
- Alegria-Barrero E.,
- Di Mario C.
- Karimi Galougahi K.,
- Shlofmitz R.A.,
- Ben-Yehuda O.,
- et al.
- Sotomi Y.,
- Cavalcante R.,
- Shlofmitz R.A.,
- et al.
- Ali Z.A.,
- Maehara A.,
- Genereux P.,
- et al.
- Chambers J.W.,
- Feldman R.L.,
- Himmelstein S.I.,
- et al.
- Maejima N.,
- Hibi K.,
- Saka K.,
- et al.
- Karimi Galougahi K.,
- Bhatti N.,
- Shlofmitz R.,
- et al.
- Tomey M.I.,
- Sharma S.K.
- Reifart N.,
- Vandormael M.,
- Krajcar M.,
- et al.
- Sotomi Y.,
- Serruys P.W.