Author + information
- Received April 22, 2015
- Revision received November 23, 2015
- Accepted December 1, 2015
- Published online March 28, 2016.
- Alexandre Abizaid, MD, PhDa,∗ (, )
- Ricardo A. Costa, MD, PhDa,
- Joachim Schofer, MDb,
- John Ormiston, MBChBc,
- Michael Maeng, MDd,
- Bernhard Witzenbichler, MDe,
- Roberto V. Botelho, MD, PhDf,
- J. Ribamar Costa Jr., MD, PhDg,
- Daniel Chamié, MDg,
- Andrea S. Abizaid, MD, PhDg,
- Juliana P. Castro, PhDg,
- Lynn Morrison, MPHh,
- Sara Toyloy, BSh,
- Vinayak Bhat, PhDh,
- John Yan, MSh and
- Stefan Verheye, MD, PhDi
- aInstituto Dante Pazzanese, São Paulo, Brazil
- bUniversitäres Herz-und Gefäβzentrum, Hamburg, Germany
- cMercy Angiography Unit, Auckland, New Zealand
- dAarhus University Hospital, Aarhus, Denmark
- eCharite Benjamin Franklin Campus, Med. Klinik II, Berlin, Germany
- fInstituto do Coração do Triângulo Mineiro, Uberlândia, Brazil
- gCardiovascular Research Center, São Paulo, Brazil
- hElixir Medical Corporation, Sunnyvale, California
- iZNA Middelheim, Antwerp, Belgium
- ↵∗Reprint requests and correspondence:
Dr. Alexandre Abizaid, Department of Cardiology, Institute Dante Pazzanese of Cardiology, Av. Dr. Dante Pazzanese 500, Ibirapuera, 0401210 Sao Paulo, Brazil.
Objectives This study sought to report the late multimodality imaging and clinical outcomes of the novel poly-l-lactic-acid–based DESolve novolimus-eluting bioresorbable coronary scaffold for the treatment of de novo coronary lesions.
Background Bioresorbable scaffolds are an alternative to drug-eluting metallic stents and provide temporary vascular scaffolding, which potentially may allow vessel restoration and reduce the risk of future adverse events.
Methods Overall, 126 patients were enrolled at 13 international sites between November 2011 and June 2012. The primary endpoint was in-scaffold late lumen loss at 6 months. Major adverse cardiac events, the main safety endpoint, were defined as the composite of cardiac death, target vessel myocardial infarction, or clinically indicated target lesion revascularization. All patients underwent angiography at 6 months. Serial intravascular ultrasound and optical coherence tomography were performed in a subset of patients.
Results The scaffold device success rate was 97% (n = 122 of 126), and procedural success was 100% (n = 122 of 122). The major adverse cardiac event rate was 3.3% (n = 4 of 122) at 6 months and 7.4% (n = 9 of 122) at 24 months, including 1 probable stent thrombosis within the first month. At 6-month angiographic follow-up, in-scaffold late lumen loss was 0.20 ± 0.32 mm. Paired intravascular ultrasound analysis demonstrated a significant increase in vessel, lumen and scaffold dimensions between post-procedure and 6-month follow-up, and strut-level optical coherence tomography analysis showed full strut coverage in 99 ± 1.7%.
Conclusions Our results showed favorable performance of the DESolve scaffold, effective inhibition of neointimal hyperplasia, and for the first time, early luminal and scaffold growth at 6 months with sustained efficacy and safety through 2 years. (Elixir Medical Clinical Evaluation of the DESolve Novolimus Eluting Bioresorbable Coronary Scaffold System—The DESolve Nx Trial; NCT02086045)
Despite the marked efficacy of drug-eluting stents in inhibiting neointimal hyperplasia, late events, including late “catch-up” restenosis and stent thrombosis, still occur (1–3). Conceptually, bioresorbable scaffolds were developed as an alternative to metallic stents to provide temporary vascular support, prevent vessel recoil, and avoid late events such as thrombosis and restenosis, which are thought to be associated with implant site inflammation due to prolonged exposure to the metal and/or drug-carrier components or the antiproliferative agent (4,5). To date, several polymeric and metallic bioresorbable scaffolds have been clinically tested; initial results with poly-l-lactic-acid (PLLA) scaffolds appear promising (4–8). Still, the impact of device design, materials, drug, degradation, and resorption kinetics on lumen remodeling and long-term outcomes remains unclear.
The DESolve novolimus-eluting bioresorbable coronary scaffold (Elixir Medical Corporation, Sunnyvale, California) is a novel PLLA-based bioresorbable scaffold that has demonstrated short bioresorption time (degradation within 1 year and bioresorption within 2 years) and high-expansion capacity without strut fracture (8). In the first-in-man (FIM) evaluation, the DESolve scaffold with myolimus—a sirolimus analog—as the antiproliferative agent, demonstrated feasibility in a small patient sample with single, noncomplex coronary lesions (8). Our objective is to report the procedural, serial multimodality imaging, analysis, and 24-month clinical outcomes of the DESolve scaffold with novolimus—a new sirolimus metabolite compound—in the treatment of diseased coronary vessels.
Study design and population
The DESolve Nx (Elixir Medical Clinical Evaluation of the DESolve Novolimus Eluting Bioresorbable Coronary Scaffold System—The DESolve Nx Trial) trial was a prospective, multicenter, nonrandomized study evaluating the performance, safety, and efficacy of the DESolve scaffold in the treatment of patients with native coronary lesions. Inclusion and exclusion criteria are listed in the Online Appendix. The study complied with the Declaration of Helsinki, and was approved by the local ethics committee at all participating institutions. All patients provided written informed consent before enrollment, and the trial was registered at NCT02086045.
The design and specifics of the DESolve scaffold have been detailed elsewhere (8). In brief, the DESolve scaffold is composed of a PLLA-based polymer with 150-μm strut thickness, coated with a matrix of the drug novolimus and a polylactide-based polymer. The drug is contained in a proprietary bioresorbable PLLA-based polymer, which is from the same polymer family as that of the scaffold backbone. The device thickness and width, including the polymer coating, are 150 μm and 165 μm, respectively. Novolimus, a metabolite of sirolimus, belongs to the family of antiproliferative compounds of macrocyclic lactones and has a similar mechanism of action to sirolimus. Novolimus is applied to the scaffold at a dose of approximately 5 μg per mm of scaffold length; 80% of the drug is eluted over 4 weeks. The polymer coating degrades within 6 to 9 months, and the scaffold degrades within 12 months and resorbs within 24 months (8). There are 2 platinum-iridium markers located on both ends of the scaffold to aide in angiographic placement. The DESolve scaffold uses a balloon-expandable delivery system that is 0.014-inch diameter guidewire and 6-F guide catheter compatible. The device should be stored at 0°C to 8°C.
By protocol, treatment of up to 2 de novo coronary lesions in separate major epicardial vessels was allowed; only 1 target lesion was considered for treatment with the study device. When applicable, the first lesion, designated as non-target, was treated with an approved “limus” drug-eluting stent. If optimally treated, the target lesion was approached. Pre-dilation was mandatory. The DESolve scaffold was available in the 3.0-, 3.25-, and 3.5-mm diameters, and 14- and 18-mm lengths. After lesion access and cross, the scaffold was slowly deployed with 10-s intervals per atm up to 2 atm and then inflated at 2-s intervals per additional atmosphere of pressure until the desired expansion pressure was obtained; final pressure should be maintained for 20 to 30 s. It was recommended, per the instructions for use, not to expand the scaffold beyond 0.5 mm larger than the nominal diameter despite benchtop testing showing the scaffold capable of further expansion without fracture (9). Operators were advised to allow the scaffold to reach body temperature, approximately 60 s, before beginning inflation. This 60-s period began upon insertion of the scaffold system into the guiding catheter and included the time needed to access and cross the target lesion. A total of 4 min was allowed before scaffold removal was recommended if lesion access was not possible. Only 1 study device could be implanted per lesion; however, an additional scaffold could be implanted in case of bailout. Post-dilation with a noncompliant balloon was performed at the operator’s discretion. Post-procedure intravascular ultrasound (IVUS) evaluation was recommended for assessment of optimal scaffold expansion and deployment.
The antithrombotic regimen included a loading dose of aspirin (≥300 mg) and clopidogrel (≥300 mg) if not on long-term use. Procedural heparin or bivalirudin were given per hospital practice; glycoprotein IIb/IIIa inhibitors were used per operator’s discretion. Post-procedure drug regimen was aspirin (≥75 mg) indefinitely and clopidogrel (75 mg daily) for a minimum of 12 months. Pre-procedure laboratory assessments, including cardiac biomarkers, were requested; post-procedure, the same biomarkers were measured within 12 to 24 h, and serially if elevated.
Endpoints and follow-up
The primary (efficacy) endpoint was in-scaffold late lumen loss, as assessed by quantitative coronary angiography (QCA) at 6 months. Clinical endpoints were assessed at 1 and 6 months, and annually to 5 years, and included major adverse cardiac events (MACE), target lesion and target vessel failure, and scaffold thrombosis. Endpoint definitions were similar to other contemporary trials and can be found in the Online Appendix.
An independent clinical research organization was responsible for the overall data collection and analysis, study management, monitoring, and QCA, IVUS, and optical coherence tomography (OCT) core laboratory analyses (Cardiovascular Research Center, Sao Paulo, Brazil). MACE events were independently adjudicated by a clinical events committee. Clinical outcomes were reported by the intention-to-treat (ITT) or the modified-ITT (patients who received the scaffold at the target lesion without major protocol deviations). All patients were assigned to angiographic re-evaluation at 6 months. In addition, serial evaluations with IVUS and OCT were performed post-procedure and at 6 months in a subset of approximately 40 patients at select sites.
Serial angiographic studies were obtained after intracoronary administration of nitroglycerin (100 to 200 μg, unless clinically contraindicated) in 2 orthogonal matching views at baseline (pre-procedure), final (post-procedure) and 6-month follow-up. Qualitative and quantitative angiographic analyses were performed offline by experienced operators blinded to procedural results at an independent angiographic core laboratory as previously reported (8). Angiographic core laboratory protocols appear in the Online Appendix.
IVUS and OCT
Seven of the 13 participating sites were chosen for adequate capability and recognized expertise and experience, and included patients in the IVUS and OCT subsets; the first several consecutive patients enrolled at each of these designated centers were included in the subsets.
This was an observational study designed to provide preliminary observations and generate hypotheses for future larger trials. Thus, the sample size was determined by assessing the minimal number of patients needed to provide reliable and consistent results. For the IVUS/OCT subset, the intended number of patients was approximately 35; 40 patients were included to account for patient loss during imaging follow-up.
Categorical variables are presented as counts and percentages (%). Continuous variables are presented as mean ± SD and 95% confidence intervals. Shapiro-Wilk normality tests were performed to verify normality assumptions; paired Student t tests or nonparametric Wilcoxon tests were performed to compare results between time points at post-procedure and 6 months. All analysis and table results were done with MS Excel (Microsoft, Redmond, Washington), NCSS version 8 (NCSS LLC, Kaysville, Utah), and R software version 3.1.2 (R Core Team, Vienna, Austria). A p value <0.05 was considered significant.
A total of 126 patients were enrolled at 13 clinical sites in Belgium, Brazil, Denmark, Germany, New Zealand, and Poland between November 2011 and June 2012. Table 1 shows the patient demographics, and Figure 1, the trial profile. Overall, the mean age was 62 years, 21% had diabetes, and the majority of patients presented with stable angina. Device success for the ITT population was 97% (122 of 126) because the study scaffold could not be properly delivered in 4 cases; all were subsequently treated with metallic stents. Procedure success for the modified-ITT population was 100%. In 3 cases, the scaffold failed to reach the target site within the protocol-defined time limit (lesions located in distal right coronary artery (RCA) and mid-left circumflex coronary artery, respectively) due to the presence of significant vessel calcification or tortuosity. There was 1 non–Q-wave myocardial infarction among these cases, with no further clinical events reported through 30 days. In the fourth case, scaffold dislodgement occurred while attempting to cross a calcified proximal Ramus branch lesion. Multiple pre-dilations were performed; IVUS assessment revealed a 270° arc of calcium. Despite an adequate result, a 3.0 × 18-mm scaffold could not fully cross the lesion and during removal, the device detached from the balloon. The vessel was rewired (parallel to the scaffold), a pre-dilation performed, followed by implantation of a 3.0 × 24-mm stent crushing the unopened scaffold against the vessel wall. High-pressure post-dilation was performed. This patient remained asymptomatic without further clinical complications through hospital discharge. Six-month angiography showed a patent stent, and the patient remained asymptomatic through 24 months.
Table 2 shows the serial QCA evaluation post-procedure and at 6 months. Final procedure acute recoil was 6.6%. Overall, 93% of patients complied with angiographic follow-up, and results demonstrated a mean in-scaffold late lumen loss of 0.20 ± 0.32 mm. In-segment binary restenosis was 3.5% with 4 cases of focal lesions, of which 3 were considered a “geographic miss” wherein the scaffold did not sufficiently cover the original lesion. Also, there was no thrombus or aneurysm formation, nor exaggerated neointimal hyperplasia formation at the outside edges of the scaffold. Figure 2 illustrates a case with multimodality imaging. The cumulative frequency distribution curve for late lumen loss is shown in Figure 3.
Among those patients in which the device was successfully implanted (modified-ITT, n = 122), there was 1 sudden death within 30 days post-procedure in a patient with a history of acute posterior wall infarction and recanalization with stent implantation in the RCA 6 months before the index procedure. There was no autopsy, and the event was adjudicated as a cardiac death and subacute, probable thrombosis (14). The cumulative MACE rate at 6 months was 3.3%, including 2 target lesion revascularizations (Table 3); both cases involved mainly the 5-mm edges outside of the scaffold. There was 1 target vessel myocardial infarction (non–Q-wave) due to an iatrogenic event, which occurred during the 6-month protocol-required re-evaluation. In this case, follow-up angiography demonstrated a patent scaffold in the mid-RCA, but following IVUS imaging, fluoroscopy showed that the distal radiopaque scaffold marker was not visible at its original site, but was identified more distally with occlusion of the posterior descending artery ostium. OCT was performed and revealed disruption of the scaffold structure at its distal edge. Events between 6 and 12 months included 1 cardiac death, occurring in a patient with multiple comorbidities who developed acute pulmonary embolism and right heart failure on day 362; and 2 target lesion revascularizations, involving the scaffold outside edge in 1 patient, and within the scaffold in the second patient. Between 12 and 24 months, there was 1 cardiac death in a patient presenting with an acute cardiac event, subsequent non-target vessel revascularization on day 454, evolving to acute cardiac failure, pulmonary insufficiency, and death on day 469; the target lesion was patent; and 1 additional target lesion revascularization due to a significant in-scaffold stenosis. Thus, at 24 months, the MACE rate was 7.4%.
Paired analyses for IVUS and OCT were completed in a subset of 40 and 38 patients, respectively (Table 4). By IVUS, there were 3 cases of acute incomplete strut apposition; at the 6-month evaluation, 2 had resolved and 1 persisted, and there was no evidence of late-acquired incomplete strut apposition. In addition, IVUS analysis demonstrated a significant increase in lumen and scaffold area between post-procedure and 6 months. Similarly, OCT imaging showed a significant increase in the in-scaffold mean and minimum area, with a small, yet significant, decrease in lumen dimensions. On the strut-level OCT analysis, the overall number of “discernible” struts with preserved box appearance (8) analyzed immediately post-procedure versus 6 months were 14,893 and 14,261 (p = 0.13). At follow-up, the frequency of covered struts per patient was 98.79 ± 1.69%. The mean thickness of tissue covering the struts was 100.53 ± 30.58 μm. Consistent with the IVUS results, there was no evidence of late-acquired incomplete strut apposition. Lastly, structural discontinuities were identified in 2 of 38 scaffolds (5.3%) and were extremely short with respect to its longitudinal length (0.07 ± 0.25 mm). At 6 months, struts circumferentially misaligned with respect to the adjacent struts, but fully encapsulated by neointimal tissue—a consequence of the ongoing degradation process—were observed in 12 scaffolds (31.6%). Conversely, floating struts, without holding the circumferential geometry of the scaffold, were seen in 5 scaffolds, representing 0.14 ± 0.35% of all analyzed struts.
The present study evaluated the DESolve scaffold and demonstrated a high device success rate, low short- and long-term clinical event rate, and an angiographic in-scaffold late lumen loss at 6 months comparable to currently available scaffolds and drug-eluting stents. The IVUS subset analysis evidenced a phenomenon of lumen enlargement, associated with scaffold growth, between the index procedure and 6 months, whereas the OCT analysis showed 99% strut coverage at 6 months. Thus, in this moderately sized study, the DESolve scaffold appears efficacious and safe in noncomplex lesions, with an added benefit of earlier lumen enlargement than that previously reported for other PLLA-based scaffolds.
The efficacy demonstrated with the DESolve scaffold appears to be comparable with the second-generation ABSORB PLLA bioresorbable vascular scaffold (BVS 1.1, Abbott Vascular, Santa Clara, California). However, the DESolve scaffold has a faster degradation and resorption profile leading to earlier lumen remodeling. Differences in intimal hyperplasia inhibition between devices remain unknown with no head-to-head comparative studies involving serial imaging. In both the DESolve FIM and current study, we found low numbers of events and no safety concerns, including 0% stent thrombosis in the DESolve FIM trial (8) and 1 probable stent thrombosis (0.8%) in the DESolve Nx trial. One patient experienced a non–Q-wave periprocedural MI in each study, but there were no spontaneous MIs reported. Interestingly, in the case developing an “iatrogenic” event, review of the post-procedure OCT imaging noted moderate scaffold malapposition at the scaffold distal end. This lead to the assumption that at 6 months, the guidewire crossed behind the scaffold struts at the site of a persistent malapposition causing scaffold disruption and displacement of the distal markers during advancement of the IVUS catheter. Noteworthy is the lack of operator-reported difficulty nor resistance during guidewire or IVUS catheter advancement through the scaffold. Additional angiography performed at 18 months revealed a patent scaffold and normal flow (Thrombolysis In Myocardial Infarction flow grade 1) in the previously recanalized posterior descending artery, without signs of stenosis or presence of radiopaque marker/scaffold residue.
Our findings evidenced a phenomenon of lumen enlargement between index procedure and 6 months, suggesting an early process of scaffold expansion (15,16) seen later with other PLLA-based bioresorbable scaffold systems (17–19). Coronary artery healing is the ultimate goal after intervention; however, the optimal timeline and overall process remain controversial. A recent subanalysis from the ABSORB trial cohort B2 (using BVS 1.1) including 45 patients with paired IVUS data, showed a significant increase in mean lumen (8%) and scaffold (12%) area between post-procedure and 36 months (18). Compared with the ABSORB scaffold (19), lumen and scaffold enlargement appears to occur at an earlier stage with the DESolve scaffold. IVUS analysis demonstrated a significant increase in lumen (9%) and scaffold (16%) area between post-procedure and 6 months, with no significant change in plaque burden in our study. The early polymer degradation process, as shown in in vitro and in vivo studies (8), with concomitant loss in radial strength, potentially results in the early luminal growth, probably due to ongoing scaffold expansion or scaffold dissolution with subsequent “uncaging.” The scaffold is not self-expanding, but rather, during the degradation process, is free to respond to the positive arterial remodeling process driven by higher wall shear stress, increased blood flow, and lumen narrowing as a result of neointimal proliferation in the implanted region. This increase in scaffold area is a passive as opposed to an active response of the device and is not related to the self-correction phenomenon that is observed post-implantation (8).
Recently published data from the Ghost EU registry suggests that real-world use of the ABSORB scaffold resulted in higher rates (2.1% at 6 months) of definite and probable stent thrombosis than previously reported in studies (20), including the ABSORB II and ABSORB Extend interim analyses, which showed scaffold thrombosis rates of 1% at 1 year (21). In our study, there was only 1 probable scaffold thrombosis (1%) through 2 years. Our findings suggest optimal healing with DESolve at 6 months, but the benefits of 99% strut coverage by OCT are still unknown; also, OCT lacks resolution to discriminate whether these are functioning endothelial cells. The greater fracture resistance and the self-correction ability of the DESolve scaffold may also provide advantages with respect to stent thrombosis, but long-term comparative data will be needed to determine any true differences (9,22).
Despite being the largest fully bioresorbable scaffold trial to date with serial multimodality imaging, our study had limitations. It included mainly stable patients with noncomplex coronary lesions. Therefore, the results cannot be extrapolated to more complex lesions and higher-risk patients. Also, although the event rates were low, the lack of an active comparator precludes any inference of noninferior or superior safety against current-generation drug-eluting stents or scaffolds, albeit the low late lumen loss observed at 6 months positions DESolve in the range of contemporary metallic drug-eluting stents. There was no mandatory IVUS or OCT imaging to characterize lesion and vessel size nor is any very late imaging reported in our study; however, more complete device characterization with long-term invasive imaging is currently ongoing at select sites. This is particularly relevant to late lumen loss after 6 months. Lastly, 24-month follow-up may not be long enough to assess the overall implications of bioresorbable scaffold technology on clinical outcomes (2).
The novel DESolve scaffold demonstrated promising outcomes in the treatment of patients with de novo noncomplex coronary lesions, including high device and procedural success and low rates of adverse events through 24 months. Multimodality imaging assessment demonstrated effective scaffolding of the vessel, efficacy with respect to the inhibition of neointimal hyperplasia, a high percentage of strut coverage without signs of local inflammatory reaction or toxicity, and evidence of early (6 months) lumen enlargement that is achieved through scaffold degradation and resorption. These findings warrant further confirmation in larger and broader patient populations.
WHAT IS KNOWN? To date, several polymeric and metallic bioresorbable scaffolds have been clinically tested and the results with poly-l-lactic-acid (PLLA) scaffolds are especially promising. Still, the impact of device design, materials, drug, degradation, and resorption kinetics on lumen remodeling and long-term outcomes remains unclear.
WHAT IS NEW? The DESolve Nx trial was designed to evaluate the DESolve novolimus-eluting bioresorbable coronary scaffold in de novo coronary lesions using clinical and imaging endpoints over time. The results of the study demonstrate that the DESolve scaffold, with its faster degradation rate and greater expansion capability as compared with currently available polymeric scaffolds may address currently unmet needs.
WHAT IS NEXT? Further evaluation of the DESolve scaffold and other scaffolds in more-complex patient populations is needed to help determine the “workhorse” capability of scaffolds in general or in which patient subsets, scaffolds may be most effective.
Editorial support was provided by Maria Alu, MM, who was contracted for these services by Elixir Medical.
For an expanded Methods section, please see the online version of this paper.
Funded by Elixir Medical Corporation. Dr. Abizaid has received research grants from Abbott Vascular, Reva Medical, Biotronic, and Elixir Medical; and consulting fees from Elixir Medical. Dr. Ormiston serves on the Advisory Board of Abbott Vascular. Dr. Maeng has received research grants from Biosensors International and Boston Scientific. Dr. Witzenbichler has received lecture fees from Elixir Medical. Dr. Botelho has received a research grant from Elixir Medical. Dr. Chamié has received consulting fees from St. Jude Medical. Ms. Morrison, Ms. Toyloy, Dr. Bhat, and Mr. Yan are employees of Elixir Medical. Dr. Verheye has received consulting fees from Elixir Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- first in man
- intention to treat
- intravascular ultrasound
- major adverse cardiac event(s)
- optical coherence tomography
- quantitative coronary angiography
- right coronary artery
- Received April 22, 2015.
- Revision received November 23, 2015.
- Accepted December 1, 2015.
- American College of Cardiology Foundation
- Maeng M.,
- Tilsted H.H.,
- Jensen L.O.,
- et al.
- Stefanini G.G.,
- Kalesan B.,
- Serruys P.W.,
- et al.
- Tamai H.,
- Igaki K.,
- Kyo E.,
- et al.
- Serruys P.W.,
- Onuma Y.,
- Ormiston J.A.,
- et al.
- Verheye S.,
- Ormiston J.A.,
- Stewart J.,
- et al.
- Mehran R.,
- Dangas G.,
- Abizaid A.S.,
- et al.
- Mintz G.S.,
- Nissen S.E.,
- Anderson W.D.,
- et al.
- Tearney G.J.,
- Regar E.,
- Akasaka T.,
- et al.
- Prati F.,
- Guagliumi G.,
- Mintz G.S.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Serruys P.W.,
- Onuma Y.,
- Dudek D.,
- et al.
- Strandberg E.,
- Zeltinger J.,
- Schulz D.G.,
- Kaluza G.L.
- Ormiston J.A.,
- Serruys P.W.,
- Onuma Y.,
- et al.
- Serruys P.W.,
- Chevalier B.,
- Dudek D.,
- et al.