Vascular response to everolimus-and biolimus-eluting coronary stents versus everolimus-eluting bioresorbable scaffolds – an optical coherence tomography substudy of the EVERBIO II trial

QUESTIONS UNDER STUDY: Head-to-head optical coherence tomography (OCT) data comparing metallic stents with bioresorbable vascular scaffolds (BVS) are lacking. This study assessed vascular healing at 9-month followup after implantation of everolimusand biolimus-eluting stents (EES; BES) and everolimus-eluting BVS. METHODS: OCT was performed in 74 patients enrolled in the EVERBIO II (NCT01711931) trial (23 with EES: 26 lesions, 7 625 struts; 23 with BES: 26 lesions, 6 140 struts; 28 with BVS: 33 lesions, 10 891 struts). OCT images were acquired using the pullback and nonocclusive flushing technique and analysed offline. RESULTS: BVS demonstrated fewer uncovered struts per patient (12 ± 27 [3.8 ± 8.4%] vs 59 ± 55 [21.8 ± 13.7%] in the EES&BES group, p <0.001), and thicker neointimal hyperplasia (BVS 102 ± 44 μm vs EES&BES 66 ± 36 μm, p <0.01). There was no significant difference with regard to malapposed struts (2.1 ± 2.7% in the BVS vs 4.4 ± 8.8% in the EES&BES group, p = 0.41). In a predefined signal intensity scale, quantitative analysis of the “key component” (black) revealed lower intensity in BVS than EES&BES (14 ± 23% vs 13 ± 12%, p = 0.007). Intensity was lower in polylactide-containing stents (BVS&BES) than in EES (15 ± 19% vs 10 ± 10%, p <0.001). CONCLUSIONS: BVS has fewer uncovered struts and presents with a thicker neointimal coverage compared with EES&BES. It is not known whether this improved capping correlates with superior vascular healing. Polylactide-containing stents (BVS and BES) demonstrate lower peristrut intensity compared with EES.


Introduction
The development of drug-eluting stents (DES) has been associated with a significant reduction in the rate of target lesion revascularisation [1].Early generation DES however, suffered a significant rate of late complications (stent thrombosis and/or neoatherosclerosis) [2,3].Newer generations of DES have improved short-and long-term safety.Of these, second-generation everolimus-eluting stents (EES) using a biocompatible durable polymer (fluorinated copolymer) with thin strut (81 µm) and third generation biolimus-eluting stents (BES) using an abluminally coated biodegradable polymer (polylactide) with relatively thick struts (112 µm) are currently considered the safest DES [4][5][6][7][8][9][10][11][12][13].The issue of neointimal proliferation and very late stent thrombosis from lingering polymers and vascular scaffolds has led to the development of completely resorbable stents, among which are the everolimus-eluting bioresorbable vascular scaffolds (BVS).Several optical coherence tomography (OCT) studies have evaluated the vascular healing response to EES, BES [14][15][16] and BVS [17,18].To date, head-to-head OCT data comparing BVS with EES and/or BES are scarce.We therefore sought to assess vascular healing in BVS compared with EES&BES using OCT, 9 months after stent implantation.

Patient population
This study was a substudy of the EVERBIO II trial (Comparison of Everolimus-and Biolimus-Eluting Coronary Stents with Everolimus-Eluting Bioresorbable Vascular Scaffold), a prospective, single centre, assessor-blinded, randomised, superiority trial comparing EES&BES with BVS, the results of which have previously been published [19,20].A total of 240 patients with coronary artery disease were recruited between November 2012 and November 2013 at the University and Hospital Fribourg (Switzer-land).The only exclusion criterion was a reference vessel size of >4.0 mm which precluded BVS implantation.The inclusion period of the substudy extended from January 2013 to June 2014.The first 25-30 consecutive patients in each treatment group willing to undergo additional intracoronary imaging were included in the present substudy.Assessment of OCT outcomes was not blinded.The study complied with the Declaration of Helsinki and was approved by the local ethics committee and all patients gave written, informed consent.The trial is registered in ClinicalTrials.gov,number NCT01711931.

OCT acquisition and analysis
Per protocol OCT acquisition was planned at 9-month angiographic follow-up.After the diagnostic angiography, 200 μg of i.c.nitroglycerin and 2 500-5 000 IU of unfractionated heparin were administrated.OCT was performed with the Optis Illumen system (St.Jude Medical) according to manufacturer guidelines using the Dragonfly™ Duo OCT Imaging Catheter with "54 mm high resolution mode" pullback, the nonocclusive flushing technique and a pullback speed of 25 mm/s.OCT pullbacks were assessed offline using a proprietary software (Lightlab Imaging, St. Jude Medical).Lesions were analysed at cross-sectional level with an interval of 0.5 mm and assessed for strut coverage, malapposition and protrusion by a single analyst (ZK) blinded to patient and lesion presentation.All frames were reviewed by a second analyst (SC) with the final decision based on consensus.Pullbacks were excluded in cases where >30% of the total stent length was not analysable.Thickness of strut coverage was assessed for each individual strut and was measured as the distance between the endoluminal side of the strut in the midpoint of its long axis and the intersection of the lumen contour with the straight line between the endoluminal side of the strut and the gravitational centre of the vessel.Struts were considered uncovered in the case of a partial or complete absence of tissue coverage (<10 μm, minimal axial resolution of OCT).Strut malapposition was defined as a distance Representative images of OCT analysis of BVS and DES.This image shows the analysis of stent/scaffold area.After identifying all struts in a cross-section, stent/scaffold area was delineated by a curvilinear interpolation connecting the middle points of the struts at mid-strut depth.BVS = bioresorbable vascular scaffold; DES = drug-eluting stent; OCT = optical coherence tomography ≥163 μm for BVS (strut thickness 153 μm), ≥122 μm for BES (strut thickness 112 μm) and ≥91 μm for EES (strut thickness 81 μm) based on the consensus derived from the strut thickness plus the minimal axial resolution of OCT.Strut protrusion was defined as strut extension into the lumen for more than 160 μm but with no obvious separation from the vessel wall.Cross-sectional areas of lumen, stent and neointima were measured at intervals of 0.5mm in the stented segment, as well as the luminal areas of the proximal and distal nonstented reference segments.Neointima area was defined as stent area minus lumen area; volumes were calculated using Simpson's rule.Representative images of OCT analyses for both BVS and DES are provided on figure 1.The peristrut low intensity area (PLIA) was analysed quantitavely by measuring the intensity of the "key" component of the CMYK colour model based on raw cross-sectional images, with an interval of 1/10 of the lesion length at the mid-strut depth and at equal distance between two contiguous struts.This quantitative measure reflects the "darkness of the pixels".Peristrut intensity was reported as percentage decrease of intensity units of the "key" component of the CMYK colour model.Further definitions with regard to OCT analysis can be found in the appendix to this article.

Statistical analysis
The sample size calculation can be accessed in the appendix.The substudy is powered on superiority assuming 1% fewer uncovered struts for BVS than for EES&BES (power: 90%, 2-sided alpha: 0.05).Variables were compared between patients treated with ESS or BES (EES&BES) and BVS.Categorical variables are reported as counts and percentages, continuous variables are reported as means and standard deviations.Normality was assessed by means of visual inspection of histograms, computation of QQ-plots and the Shapiro-Wilk test.Categorical variables were compared using chi-square or Fisher's exact tests.Continuous variables were analysed using the Student's t-test or the Wilcoxon rank-sum test according to their distribution.All statistical analyses were performed using dedicated software (Stata version 13, StataCorp LP, College Station, Texas) at a two-tailed significance level of α = 0.05.In the case of multiple testing, the Bonferroni method was employed.A multilevel linear mixed effects model with random effects at patient and lesion level was employed to compare strut coverage and apposition between patients treated with EES/BES and patients treated with BVS.Univariate comparison at strut level was carried out and is provided for illustrative purposes as an appendix (table S3).Hypothesis testing of EES versus BVS and BES versus BVS is provided in the appendix (tables S1-S4) for illustrative purposes.

Baseline patient characteristics
The patient flow chart is depicted in figure 2. A total of 74 patients, 23 patients with 26 EES-treated lesions, 23 patients with 26 BES-treated lesions and 28 patients with 33 BVStreated lesions, were included in the final OCT analysis.
Baseline clinical characteristics among patients undergoing OCT were generally well balanced between the EES&BES and the BVS group and are summarised in table 1. Baseline characteristics of included patients were similar to those of the overall trial population (table S1a; see appendix).

Discussion
This first direct comparative OCT analysis comparing BVS with DES had the following findings: (a) At 9-month OCT

Original article
Swiss Med Wkly.2016;146:w14274 follow-up, BVS demonstrated a higher "capping" effect with fewer uncovered and/or malapposed struts and thicker neointimal hyperplasia, (b) peristrut intensity was, however, significantly lower in BVS than EES&BES.Several trials have previously investigated strut coverage of EES and BES by use of OCT [14][15][16] but only a few subtrials of the ABSORB Cohort B are available for the second-generation everolimus-eluting BVS [17,21,22].To date, there is no reported trial in humans that directly compared OCT findings of BVS to any of the available DES. Gomez-Lara and colleagues compared vascular response to EES and BVS at 1 year by performing a posthoc analysis in 44 unmatched patients from RESOLUTE All Comers and ABSORB Cohort B2 for whom OCT ima- ging was available.They found no difference in neointimal response as assessed from mean neointimal thickness over stent struts (EES 126 vs BVS 136 μm).They further reported a low number of uncovered struts (EES 5.3% vs BVS 4.5%) or malapposed struts (EES 1.1% vs BVS 2.2%, p <0.01) with both devices.In our study, BVS achieved high lesion capping with a low percentage of uncovered struts at 9-month follow-up.These results are in line with OCT findings by Gomez-Lara et al. and other studies assessing strut coverage of BVS [17,21,22]; the ABSORB Cohort B trial reported a mean of 2% uncovered struts at 6 months and 1% uncovered struts at 24 months.When compared with EES&BES, BVS presented significantly fewer uncovered struts in our study.This finding contradicts the results by Gomez-Lara and colleagues and is most likely driven by the high rate of uncovered struts found with EES and BES, which is very different from the available literature.In the post-hoc analysis from RESOLUTE and ABSORB Cohort B, EES was associated with 5.3% uncovered struts.The NEXT OCT substudy reported 3 ± 7% uncovered struts in EES-treated and 9 ± 10% in BES-treated patients at 8-12 months (p <0.001) [16].Tada et al. recently reported no difference in uncovered struts between EES and BES at 6-8 months (588 [15%] vs 479 [17%] unadjusted respectively, p = 0.34) [15].However,, this improved capping does not necessarily reflect better vascular healing.Indeed, the visible covering layer can be formed by loosely organised and possibly prothrombotic elements such as fibrin.Incomplete stent apposition and lack of neointimal strut coverage are thought to be correlated with an increased risk of late stent or scaffold thrombosis and myocardial infarction [23,24].In this trial, the rate of malapposed struts was not significantly different in EES&BES compared with BVS.BVS rates of malapposed struts seen in our study were similar to previously reported data, but

Original article
Swiss Med Wkly.2016;146:w14274 proteoglycan extracellular matrix and inflammatory cell infiltration, indicating delayed arterial healing [26].In the same preclinical study using a porcine model, they reported a 3-fold higher rate of PLIA for early-generation DES than for bare metal stents [26].Tada et al. reported similar rates of PLIA for EES and BES [15].In the present study, peristrut intensity significantly differed between the three platforms studied.There was a significantly lower peristrut intensity found in BVS-and BES-compared with EEStreated lesions.The lower intensity in devices with a degradable polymer coating compared with a durable polymer coating might suggest a prolonged inflammatory process around the degradable polymer coating, which could be a marker of delayed vascular healing.The present study is limited in size with inevitable uncertainty around point estimations.Another issue is the lack of baseline OCT examination precluding any definitive conclusion regarding the cause of the incomplete stent apposition found at 9-month follow-up.A systematic bias might have been introduced by the differences in assessment of malapposition between the metallic stents (estimation of the abluminal border by adding the strut and polymer thickness to the endoluminal border) and the BVS (direct visualisation of the abluminal border and the lumen contour behind it).Furthermore, neointimal hyperplasia may have been overestimated in patients treated with BVS.Due to the reduction of the black box signal (by filling of the strut voids with connective tissue) that induces an abluminal displacement of the endoluminal scaffold strut border, measurement of neointimal hyperplasia becomes systematically larger in BVS than in metallic stents.
Finally, the study was neither powered nor designed to assess the impact of suboptimal stent coverage by neointima, incomplete stent apposition and PLIA on subsequent late clinical events, particularly late stent thrombosis.Notwithstanding, specific strengths include meticulous OCT measurements at longitudinal intervals of 0.5 mm while standard intervals used by OCT investigators to date was 1 mm.Moreover, this was a substudy with a representative sample of the EVERBIO II trial population, an investigator-initiated and funded randomised controlled trial in all-comers.

Conclusions
BVS has fewer uncovered struts and presents with a thicker neointimal coverage compared with EES&BES.It is not known whether this improved capping correlates with superior vascular healing.Polylactide-containing stents (BVS&BES) demonstrated lower peristrut intensity compared with EES.The clinical significance of these findings needs further assessment.

Disclosure statement:
The trial was an investigator-initiated study supported by an unrestricted grant from the Fonds Scientifique Cardiovasculaire (Fribourg, Switzerland).The funding source had no role in the design of the study, data collection, data monitoring, data analysis, data interpretation, or writing of the report.Dr. Cook has received speaker fees/honoraria from Abbott Vascular, Biosensors Int., and Boston Scientific.Dr. Cook receives support from the Swiss National Science Foundation

Figures (large format) Figure 1
Figures (large format)

Table 1 :
[25][15][16][17]21]aracteristics.Swiss Med Wkly.2016;146:w14274EES&BES rates of malapposed struts were considerably higher than reported in the literature[14][15][16][17]21].OCT assessment of neointimal coverage is a useful surrogate for risk stratification of very late DES thrombosis[25].The current study showed better neointimal coverage in BVS-treated than in EES&BES-treated lesions.Whether neointimal coverage is equally important in stratifying the risk for very late scaffold thrombosis is uncertain.It may be that a distinctive pathophysiology and/or different mechanistic phenomena, not yet identified, lead to scaffold thrombosis.The current study is the first to compare PLIA between BVS, EES and BES and to address this issue with quantitative OCT assessment.In a comparative histological observation Teramoto et al. suggested that these areas of low intensity may represent fibrin accumulations surrounded by BES = biolimus-eluting stent; BMI = body mass index; BVS = bioresorbable vascular scaffold; CABG = coronary artery bypass graft; CAD = coronary artery disease; EES = everolimus-eluting stent; MI = myocardial infarction; NSTEMI = non-ST segment elevation myocardial infarction; PCI = percutaneous coronary intervention; SD = standard deviation; STEMI = ST segment elevation myocardial infarction Original article

Table 3 :
OCT analysis at 9-month follow-up.