Endothelial progenitor cell capture stents: will this technology find its niche in contemporary practice?

This editorial refers to ‘Genous™ endothelial progenitor cell capturing stent vs. the Taxus Liberté stent in patients with de novo coronary lesions with a high-risk of coronary restenosis: a randomized, single-centre, pilot study’†, by M.A.M. Beijk et al. on page 1055 Drug-eluting stents (DES) have significantly reduced rates of restenosis; however, elution from their surface of cytotoxic and cytostatic drugs, together with the presence of non-erodable polymers, is associated with impaired endothelialization, allergic reactions, inflammation, and vascular dysfunction, factors which have all been implicated in the most prominent safety concern with DES today —stent thrombosis (ST).1–3 The development of newer DES with polymers that are either more biocompatible, biodegradable, or completely absent, together with completely biodegradable stents has been an important step forward to minimize the potential risk of ST. Despite this, however, the rapid restoration of a functional endothelium appears, in theory at least, to be one of the most intuitive ways of minimizing the risk of ST and excessive neointimal proliferation.4 Therefore, it comes as no surprise that soon after the discovery of endothelial progenitor cells (EPCs),5 a novel coronary stent, the Genous™ EPC capture stent (OrbusNeich, Florida, USA), was developed which had a proprietary coating that contained anti-human CD34 antibodies (Figure 1). These antibodies were able to capture circulating EPCs, such that the EPC capture stent was considered a potential solution to many of the problems of coronary stenting. Schematic diagram illustrating the underlying principal behind the EPC capture stent. The CD-34 antigen on the EPC binds to the anti CD-34 antibody on the stent's surface to promote rapid endothelialisation. Six years on, and Beijk et al.6 report the 1-year outcomes of the TRIAS HR study, the first randomized study comparing the Genous™ EPC stent with a conventional DES (the Taxus Liberté stent). This study adds to the paucity of data already available on the EPC stent, and disappointingly reaffirms the disparity between technological promise and real-life clinical results. Prior to the completion of 1-year follow-up this single-centre study was transformed into a multicentre study whose enrolment was subsequently discontinued prematurely on 19 February 2009. This followed the concerns of the data and safety monitoring board who felt that the non-inferiority primary endpoint would not be reached.7 Beijk et al. report a mean late luminal loss of 1.14 ± 0.64 mm at 6–12 months in the 53 patients treated with the EPC stent who returned for angiographic follow-up. This follow-up was not mandated by protocol, and therefore the late loss is potentially subject to a selection bias. Nevertheless, the results are consistent with earlier single arm assessments of the EPC stent in the HEALING First-in-Man (FIM) study, the HEALING II study, and the HEALING IIb study that all assessed the stent's performance in simple patients and reported 6-month mean late luminal loss of 0.63 ± 0.52, 0.78 ± 0.39, and 0.76 ± 0.50 mm, respectively.8–10 Similarly, Miglionico et al. also reported a late lumen loss of 0.88 ± 0.62 mm amongst complex patient groups enrolled in a prospective registry who were treated with the EPC stent.11 Disappointingly, these figures are inferior to other new generation coronary stents such as DES with biocompatible polymers (e.g. Endeavor Resolute™), DES with biodegradable polymers (e.g. BioMatrix™ biolimus-eluting stent), DES which are polymer free (e.g. VESTAsync sirolimus-eluting stent), stents which are fully biodegradable (the Bioabsorbable Vascular Solutions everolimus-eluting stent), and bare metal stents (BMS) with novel coatings (e.g. Titan-2 stent) which have reported late luminal loss values of 0.12 ± 0.26, 0.13 ± 0.46, 0.36 ± 0.23, 0.44 ± 0.35, and 0.55 ± 0.63 mm, respectively.12–16 The late loss results with the EPC stent appear to be more synonymous with the use of conventional BMS, and are certainly the first pointers that the idealistic model that rapid stent endotheliazation results in improved and faster healing, and thus less neointima, does not always ring true.4 Post hoc analysis of the HEALING II study indicated that the reduction in late loss in some patients could correlate with EPC levels. Therefore, in the TRIAS HR study, at least 1 week of mandatory statin therapy, which can increase EPC numbers, was required prior to the index percutaneous coronary intervention (PCI).9 Unfortunately, no measure of the number or functionality of EPCs was performed. Nevertheless, the late loss values suggest either a lack of beneficial effect from statin therapy, the presence of other confounding factors affecting EPC levels, or the possible recruitment of dysfunctional EPCs. In the HEALING IIB study, 2 weeks of statin therapy led to a 459% relative increase in the number of EPCs; however, this corresponded to only a 12% increase in circulating CD34+ cells, and unsurprisingly no benefit in terms of outcome.10 Digesting all these results, it is perhaps not surprising that the overall rates of target vessel failure were worse in patients treated with the Genous stent compared with DES controls. Some may argue that we are being too quick to judge the stent's performance, particularly in light of long-term follow-up data from the HEALING II study, which demonstrated a reduction of 16.9% in late lumen loss between 6 and 18 months follow-up.9 In addition, the preliminary results of 2-year follow-up from the TRIAS HR study also demonstrated a higher absolute increase in target lesion revascularization between 1 and 2 years in those treated with a DES compared with the EPC stent.17 This may reflect regression in those patients treated with the EPC stent—however, no formal angiographic follow-up is available—and/or ‘late-catch’ in those treated with a DES, which is a well documented phenomenon.18,19 In contrast to clinical efficacy, the EPC stent appears to live up to its promise with respect to ST, with no observed ST events reported in published studies, despite the use of only 1 month of dual antiplatelet therapy.8,9,11 Unfortunately, preliminary data from the HEALING IIB study in patients with simple lesions, and the single-centre GENIUS-STEMI trial in patients with ST-elevation myocardial infarction (STEMI), have both reignited doubts with respect to the potential benefits of the EPC capture stent technology. The rate of ST at 6-months follow-up in the HEALING IIB study was 3%, whilst in the GENIUS-STEMI study a rate of 6.0% was reported in STEMI patients treated with the EPC stent, compared with 0.0% in controls treated with a BMS.10,20 Whether EPC cells are still captured, or even functional in the thrombotic milieu of an STEMI remain questions to be answered. The clinical results seen with the Genous™ stent thus far raise questions over the initial proof of concept. Were we blinded by the potential of this new technology, with the promise of improved outcomes and lower ST, at a time when conventional DES were being accused of ‘killing patients’? Let us briefly revisit aspects of this technology, which may suggest that these clinical results were perhaps expected, rather than a ‘great’ surprise. The fundamental principle of this technology has been portrayed as a simple concept; however, the underlying cellular processes are far from simple, and, as yet, are still not completely understood. This lack of complete understanding is certainly one factor that has contributed to the discrepancies between pre-clinical and clinical results.21 Another important issue is the difficulty in accurately identifying EPCs, such that similar subsets of EPCs have been defined differently in both in vivo and in vitro studies. Further complicating this area of research is the difficulty in accelerating the already rapid process of healing seen after implantation of a standard BMS in non-atherosclerotic rabbit and porcine animal models. Ideally, the captured CD34+ EPCs will either differentiate into mature endothelial cells or, through the actions of paracrine factors, stimulate growth of surrounding endothelium. Unfortunately, however, the CD34+ markers used to phenotype EPCs are not only non-specific, but are also shared by both haematopoietic stem cells and mature endothelial cells. Therefore, it is possible for the EPC capture stent to sequester other bone marrow cell lines such as immune complement, or smooth muscle progenitor cells, which in turn can lead to neointimal proliferation.22 Similarly, haematopoietic CD34+ stem cells can potentially differentiate into vascular cells that can be involved in the formation of atherosclerotic plaque.23 Moreover, this unavoidable recruitment of non-EPCs to the stent's surface can ultimately interfere with the proliferation, differentiation, and function of sequestered EPCs, potentially inducing more intimal hyperplasia. The observation of an ‘endothelial covering’ over the EPC capture stent is not synonymous with a normal functioning endothelium. The vascular endothelium plays such an essential role in controlling neointimal hyperplasia that strategies to accelerate the endothelialization of coronary stents can, in theory, reduce excessive neointimal formation.24 This theory stems from the observations of the patterns of vascular injury induced by balloon angioplasty, and from subsequent stent studies.4,25 Unfortunately, data from pre-clinical studies and the clinical studies presented here suggest that the relationship between the rate of endothelialization and neointimal formation is somewhat more complicated. Moreover, the process of neointimal formation after vessel injury is also complex and remains incompletely understood.26 The accelerated endothelialization does offer the advantage of covering exposed stent struts, reducing the potential risk of ST. This has been demonstrated in the initial clinical studies; however, the results from the HEALING IIB and GENIUS-STEMI trial indicate that this relationship is not mutually exclusive. The addition of an antiproliferative drug coating onto BMS was seen as the solution to the problem of restenosis. Likewise, the Genous™ stent has been combined with an abluminal biodegradable polymer that elutes a low dose of sirolimus. Although conceptually this appears to be an unlikely combination, with the attraction of EPCs on one hand and the sirolimus-mediated cell cycle arrest on the other, early pre-clinical data indicate that the addition of anti-CD34 antibodies onto the surface of a sirolimus-eluting stent (SES) does in fact facilitate endothelialization. The use of an abluminal coating has reduced the concentration of sirolimus compared with a standard SES without significantly affecting the suppression of neointimal hyperplasia. In the porcine model, scanning electron microscopy demonstrates a significantly greater rate of endothelialization in stents combining EPC capture with either half or quarter dose sirolimus, when compared with the standard SES. Further data from histology and optical coherence tomography at 28 days follow-up in the same model also indicate that these EPC-SES stents induce less neointimal hyperplasia, and have lower inflammation scores when compared with the standard SES and Genous™ EPC stent. Overall, the EPC-SES stent offers the potential to improve vascular healing whilst still maintaining effective control over neointima proliferation. The REMEDEE FIM study which commenced on 1 December 2009 aims to randomize 180 patients to treatment with either the EPC-SES Stent or the Taxus Liberté paclitaxel-eluting stent, with a primary endpoint of late loss at 9 months follow-up.27 EPC capture stent technology remains an attractive concept, and may yet find its niche in interventional practice. Presently, however, there is incomplete understanding of the biological processes involved, and, as such, the technology may have been adopted ‘before its time’. Conflict of interest: none declared.