Summary
Following the development of stents, then drug-eluting stents (DES), bioresorbable scaffolds are proposed as a third evolution in coronary angioplasty, aiming to reduce the incidence of restenosis and stent thrombosis and to restore vascular physiology. At least 16 such devices are currently under development, but published clinical data were available for only three of them in September 2014. The first device is Abbott’s BVS ® , a poly-L-lactic acid (PLLA)-based everolimus-eluting device, which has been tested in a registry and two non-randomized trials. Clinical results seem close to what is expected from a modern DES, but possibly with more post-procedural side-effects. Two randomized trials versus DES are underway. This device is already marketed in many European countries. The second device is Elixir’s DESolve ® , a PLLA-based novolimus-eluting device, which has been evaluated in two single-arm trials. Results are not widely different from those expected from a DES. The third device is Biotronik’s DREAMS ® , a metallic magnesium-based paclitaxel-eluting device, which has been assessed in an encouraging single-arm trial; its second version is currently undergoing evaluation in a single-arm trial. The available results suggest that the technological and clinical development of bioresorbable scaffolds is not yet complete: their possible clinical benefits are still unclear compared with third-generation DES; the impact of arterial physiology restoration has to be assessed over the long term; and their cost-effectiveness has to be established. From the perspective of a health technology assessment, there is no compelling reason to hasten the clinical use of these devices before the results of ongoing randomized controlled trials become available.
Résumé
Après le développement des stents, puis des stents actifs, les stents actifs entièrement résorbables constituent la troisième évolution de l’angioplastie coronaire. Ces dispositifs visent à réduire l’incidence des resténoses et des thromboses et à restaurer la physiologie vasculaire. Au moins seize dispositifs sont en développement mais seulement trois ont fait l’objet de publications jusqu’en septembre 2014. Le premier (ABBOTT–BVS ® ) est en polymère d’acide L-lactique (PLLA) délivrant de l’évérolimus. Il a été évalué par un registre et deux essais non comparatifs. Les résultats cliniques semblent proches de ceux d’un stent actif mais avec peut-être plus d’événements indésirables. Deux essais comparatifs randomisés versus stents actifs sont en cours. Ce stent est commercialisé dans la plupart des pays européens. Le deuxième (Elixir–DESolve ® ), (PLLA, novolimus), a été étudié par deux suivis de cohorte. Les résultats ne semblent pas différents de ceux obtenus avec les stents actifs. Le troisième (Biotronik–DREAMS ® ), (alliage magnésium-terres rares, paclitaxel), a fait l’objet d’un suivi de cohorte encourageant ; l’industriel étudie une deuxième version dans le cadre d’une cohorte. Les résultats disponibles suggèrent que le développement technique et clinique des stents actifs entièrement résorbables n’est pas finalisé. Les bénéfices cliniques semblent comparables à ceux des stents actifs de dernière génération et ceux imputables à la récupération de la vasomotricité artérielle ne sont pas encore démontrés à long terme. Leur rapport coût-efficacité n’est pas encore connu. En termes d’évaluation des technologies de santé, il n’existe pas d’argument convainquant pour en recommander l’utilisation en pratique clinique avant le résultat des études contrôlées en cours.
Background
Restenosis is an important limitation of myocardial revascularization after routine balloon coronary angioplasty . An ‘elastic recoil’ phenomenon was mainly involved in the mechanism of restenosis (along with constrictive remodelling and neointimal proliferation), leading to the development of bare-metal stents (BMS) .
While the widespread use of BMS enabled a reduction in early restenosis and in the incidence of arterial wall dissection, an increase in the incidence of late restenosis was noted and attributed primarily to neointimal proliferation. These observations led to the development of drug-eluting stents (DES) , whose diffusion of antiproliferative agents was accompanied by a reduction in post-stenting reintervention rates . However, their impact on clinical events occurring after revascularization remains poorly evaluated.
The attention of interventional cardiologists has now turned to very late restenosis and thrombosis. These events may be induced by a long-term effect of the polymer bonding the stent itself to the drug to be delivered into the arterial wall. The first proposed solution to this issue involved bioresorbable polymer stents (the metal frame of which stays in the artery), which remain in development . Moreover, persistent late acquired malapposition with durable metal stents may be associated with chronic inflammation, neoatherosclerosis, very late lumen loss and stent thrombosis , contrasting with the fact that a stent has no mechanical function after a few months. It has also been argued that the presence of the stent suppresses arterial wall motility and vasodilation ability. This line of reasoning has led some manufacturers to start developing fully bioresorbable devices (usually called ‘scaffolds’ rather than ‘stents’) for coronary artery stenting. Some feasibility studies are available, but large-scale trials are yet to be published.
The present paper reviews the available evidence on fully bioresorbable coronary scaffolds, with a heath technology assessment perspective, setting aside the development of resorbable polymer stents.
Methods
The Comité d’Évaluation et de Diffusion des Innovations Technologiques (CEDIT) undertook an early assessment of this emerging health technology for the Assistance Publique–Hôpitaux de Paris (AP–HP). A systematic search for relevant literature, using subsets of the set (‘coronary’, ‘scaffold’, ‘resorbable’ ‘stent’), up to September 2014, was conducted using the MEDLINE and EMBASE databases, completed by a manual search of the references of retrieved papers. References pertaining to ‘resorbable polymer stents’ only were rejected.
This search aimed for the exhaustive retrieval of formally published clinical and economic results relating to fully bioresorbable scaffolds currently on the market or under development in humans, while giving enough technical and organizational information to allow for decision making at a hospital management level. We limited our review to devices for which we retrieved at least one formally published human clinical use and that are still available or under further development.
As is customary for CEDIT assessment reports, the information was organized under four headings: technical, clinical, economic and organizational. The scarcity of available information precluded formal aggregation (meta-analysis) of the results and led us to quote some indirect information, such as press releases.
Results
The available background papers summarize a large quantity of information on the current development of these devices. According to these papers, various manufacturers have started the development of at least 16 different devices to date.
Technical aspects
The technical characteristics of fully resorbable scaffolds with published human use and currently under development are listed in Table 1 ; these devices are pictured in Fig. 1 .
| BVS 1.1 | DESolve 1.0 | DREAMS 1G | |
|---|---|---|---|
| Manufacturer | Abbott | Elixir | Biotronik |
| Availability | CE marked | CE marked | |
| Material | PLLA | PLLA | Magnesium rare-earth alloy |
| Eluted drug | Everolimus | Novolimus | Paclitaxel a |
| Design | In-phase hoops with straight links | In-phase hoops with straight links | 6-Crown |
| Strut thickness (μm) | 156 | 150 | 125 |
| Radial support duration | 6 months | – | 3–6 months |
| Time to resorption | 2–3 years | 2–3 years | 1 year |
A fully bioresorbable device for coronary artery stenting was attempted early in the 1990s ; however, the initial success of DES targeting the same clinical problem caused these initial efforts to be discarded.
All currently developed fully bioresorbable scaffolds but one are drug-eluting devices; therefore, current research concerning the associated drugs (sirolimus or paclitaxel families) is relevant to these devices. In contrast, their mechanical properties strongly depend on their base material and are, by definition, unstable, as the device is bound to disappear eventually. The question to be answered is whether the mechanical function of a scaffold is sufficient for its clinical purposes, in terms of strength of the acute stent recoil and duration.
The devices for which we retrieved results regarding human clinical use are based on two classes of biomaterials: poly-L-latic acid (PLLA) polymers and magnesium rare-earth alloys.
PLLA polymers
At least four manufacturers have attempted to create a bioresorbable scaffold based on this material.
In 2000, Kyoto Medical Planning (Kyoto, Japan) published the results of the first clinical trial assessing their Igaki-Tamai ® biodegradable coronary stent. This first-generation device was thermoplastic, its deployment needing the injection of contrast medium heated to 80 °C; furthermore, this deployment used a large-calibre guide. These drawbacks and the initial results did not lead the manufacturer to undertake further development at the time of the publication of the first DES clinical results. One should note, however, the recent publication of the 10-year follow-up of this cohort . A second iteration of this device is said to be in development.
Abbott (Chicago, IL, USA) has also developed a PLLA device (the ABSORB Bioresorbable Vascular Scaffold [BVS ® ]). This device is the oldest of those presently aimed at the market and, therefore, the best documented. Some information is available about its pharmacological and resorption dynamics : the everolimus elution is maximal during the first weeks and null after 3 months; the scaffold provides mechanical support for 3 months, but its strength is then lost rapidly; the structure is lost 6 months after implantation, but scaffold elements are visible up to 2 years after implantation.
The manufacturer Elixir (Sunnyvale, CA, USA) has undertaken a cohort study of the DESolve ® bioresorbable coronary scaffold system (16 patients); a larger single-arm trial (DESolve Nx) has been completed , but the first results, presented orally at EuroPCR 2013 , have not yet been published.
The French start-up company Arterial Remodeling Technologies (Noisy le Roi, France) has undertaken a first cohort study (ARTDIVA, 30 patients, five centres) assessing a PLLA-based non-drug-eluting resorbable scaffold; the first 30-day follow-up results were presented at TCT 2013 , but await formal publication.
Magnesium rare-earth alloys
The manufacturer Biotronik (Berlin, Germany) has undertaken trials evaluating such a device, called the Drug Eluting Absorbable Metal Scaffold (DREAMS ® ) , after a first iteration produced disappointing clinical results . Detailed information about the mechanical and the pharmacological properties of this paclitaxel-eluting device does not seem to have been published; according to Patel and Banning , the mechanical strength of the metallic alloy would allow for less beam section inflation (150%) than for a PLLA-based device (240%), to achieve the same strength as a chrome-cobalt device.
Other devices
Other manufacturers have announced their intention to work on similar devices. Among them are Reva Medical (San Diego, CA, USA) (tyrosine polycarboxylate-based device), whose first clinical trial (RESTORE) results were presented at TCT 2012; a second pivotal trial (RESTORE II) has been initiated . A first iteration of the IDEAL (modified PLLA-based) device led to disappointing clinical results; a second iteration was in the preclinical evaluation stage in 2013. Other devices have not yet reached the stage of human evaluation .
Clinical results
The main inclusion criteria of the published trials are listed in Table 2 , the patient characteristics are described in Table 3 and the lesion characteristics in Table 4 . Published clinical and angiographical results are summarized in Table 5 .
| Study | Age (years) | Number of lesions | Artery | Length (mm) | Diameter (mm) | TIMI | Stenosis (%) | AMI | LVEF (%) | Diff a | Other |
|---|---|---|---|---|---|---|---|---|---|---|---|
| ABSORB A | > 18 | One de novo | Native | ≤ 8 | 3 | > 1 | 50 ≤ s < 100 | No | > 30 | No | |
| ABSORB B | > 18 | One or two de novo (different vessels) | Native | ≤ 4 | ≤ 3 | > 1 | 50 < s < 100 | No | > 30 | No | |
| ABSORB Extend | > 18 | One or two de novo (different vessels) | Native | ≤ 28 | 2 ≤ d ≤ 3.8 | ≥ 1 | < 100 | No | – | No | In-stent restenosis or thrombus |
| ABSORB II | 18 < a ≤ 85 | One or two de novo (different vessels) | – | ≤ 48 | 2.25 ≤ d ≤ 3.8 | ≥ 1 | 50 < s < 100 | No | > 30 | No | No recent PCI, bypass lesion |
| Prague 19 | – | – | – | ≤ 24 | 2.3 ≤ d ≤ 3.7 | – | – | Yes b | – | – | In-stent restenosis or thrombus |
| BVS STEMI | 18 < a ≤ 85 | – | – | – | 2.0 ≤ d ≤ 3.8 | – | – | Yes b | – | – | Previous CABG, thrombus in stent |
| DESolve First-in-Man | – | One or two de novo | Native | ≤ 10 | ≤ 10 | – | ≤ 80 | No | > 30 | – | Recent myocardial infarction, restenosis, calcifications |
| BIOSOLVE-1 | – | One or two de novo (different vessels) | – | ≤ 12 | 3.0 ≤ d ≤ 3.5 | – | 50 ≤ s ≤ 99 | No | – | No |
a Most studies excluded difficult lesions (calcified, ostial or furcation lesions, angulations, thrombus).
| Study | ABSORB A ( n = 30) | ASBORB B Small vessels a ( n = 41) | ABSORB B Large vessels ( n = 60) | ABSORB Extend (first 512 patients) ( n = 512) | Prague 19 ( n = 40) | BVS STEMI ( n = 49) | ABSORB II (BVS group) ( n = 335) | DESolve First-in-Man ( n = 16) | BIOSOLVE-1 ( n = 46) |
|---|---|---|---|---|---|---|---|---|---|
| Age (years) | 62 ± 9 | 62.4 ± 9.4 | 62.2 ± 8.7 | 62 ± 11 | 58.9 ± 10.9 | 58.9 ± 10.5 | 61.5 ± 10.0 | 69.3 ± 8.4 | 65.3 ± 9.7 |
| Male sex | 18 (60) | 26 (63) | 47 (78) | 74% | 31 (78) | 38 (78) | 253 (76) | 10 (63) | 34 (74) |
| Smoker b | 6 (20) | 9 (22) | 8 (14) | 23% | 25 (62) | 27 (69) | 79 (24) | 11 (69) | 17 (37) |
| Diabetes b | 1 (3) | 9 (22) | 8 (13) | 26% | 1 (3) | 4 (8) | 80 (24) | 1 (6) | 7 (15) |
| Hypertension b | 18 (60) | 24 (58) | 38 (64) | 65% | – | 19 (39) | 231 (69) | 10 (63) | 40 (87) |
| Hyperlipidaemia b | 19 (63) | 35 (85) | 44 (73) | 67% | – | 11 (22) | 252 (75) | 11 (69) | 41 (89) |
| Previous target vessel intervention | 3 (10) | – | – | 5% | – | – | 14/120 (12) | – | – |
| Previous PCTA | – | 6 (15) | 15 (25) | – | 1 (3) | 0 (0) | – | – | 27 (59) |
| Previous CABG | – | 1 (2) | 2 (3) | – | 0 | 0 | – | – | |
| Previous myocardial infarction | 1 (3) | 7 (18) | 18 (30) | 16% | 1 (3) | 1 (2) | 93 (28) | 4 (25) | 15 (33) |
| Functional status | |||||||||
| Silent ischaemia | 1 (3) | – | – | 5% | 0 | – | 42 (13) | 5 (31) | 1 (2) |
| Stable angina | 21 (70) | – | – | 64% | 0 | – | 214 (64) | 11 (69) | 43 (93) |
| Unstable angina | 8 (27) | – | – | 31% | 0 | – | 68 (20) | 0 | 2 (4) |
| Myocardial infarction | 0 | 0 | 0 | 0 | 40 (100) | – | 0 | 0 | 0 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
