Current Status of Bioresorbable Scaffolds in the Treatment of Coronary Artery Disease

Material Composition and Properties

The optimal BRS should ensure adequate short- to mid-term scaffolding of the previously stenosed vessel to avoid recoil and completely dissolve afterward to prevent side effects. Thus, temporarily sufficient radial support is needed, with struts as thin as possible. The design should warrant deliverability and straightforward handling, flexibility in different anatomic circumstances, and integrity during resorption. The optimal duration until full resorption is not yet defined. To achieve these goals, a considerable variety of materials and designs are under investigation.

Furthermore, the use of drug elution is inconsistent. Several different substances have been applied, and some BRS are noneluting. The current trend is toward broader use of drug elution, with mTOR (mammalian target of rapamycin) inhibitors as the most frequently used antiproliferative drugs in DES.

The Central Illustration provides an overview of different designs and characteristics of existing BRS, with representative images in Figures 1 and 2⇓⇓.

• Download figure
• Open in new tab
• Download powerpoint

• Download figure
• Open in new tab
• Download powerpoint

Central Illustration

Results for Current Existing Bioresorbable Scaffolds

Various bioresorbable scaffolds (BRS) are under investigation, and an overview of the different compositions, designs, current status, and results are presented. CE = Conformité Européenne; DES = drug-eluting stent; CSA = cross-sectional area; ID = ischemia driven; FiM = first-in-man; MLD = minimal lumen diameter; LLL = late lumen loss; PDLA = poly-d-lactic acid; PLLA = poly-l-lactic acid.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Representative Images of Various Existing Scaffolds

Different materials are currently used for bioresorbable scaffold (BRS). The Igaki-Tamai stent (Kyoto Medical Planning Co., Ltd., Kyoto, Japan) (A), the ABSORB Bioresorbable Vascular Scaffold (Abbott Vascular, Santa Clara, California) (B), and the DESolve bioresorbable scaffold (Elixir Medical Corporation, Sunnyvale, California) (C) are all manufactured from poly-l-lactic acid. The DREAMS magnesium alloy (Biotronik, Berlin, Germany) (D) is a metal bioresorbable vascular scaffold. The ReZolve 2 BRS (Reva Medical Inc., San Diego, California) (E) is produced on a desaminotyrosine polycarbonate basis, and the Ideal BioStent (Xenogenics Corp., Canton, Massachusetts) (F) is composed of a salicylate polymer and linker.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Angiographic and Optical Coherence Tomography Imaging of the DESolve Scaffold, Absorb BVS, and ReZolve 2 BRS

Angiographic illustration of pre-implantation vessels (A–C). No residual stenoses were found via angiography after implantation of a DESolve scaffold (Elixir Medical Corporation, Sunnyvale, California) (D), Absorb BVS (bioresorbable vascular scaffold) (Abbott Vascular, Santa Clara, California) (E), and ReZolve 2 BRS (Reva Medical Inc.) (F). Post-implantation optical coherence tomography imaging confirms adequate scaffold positions with no evidence of strut malapposition (G to I).

Poly-l-lactic acid

Different materials are used for manufacturing BRS, with poly-l-lactic acid (PLLA) being the most commonly used. For most existing PLLA-based devices, strut thickness is 150 μm. A BRS currently being developed has the thinnest struts (100 μm) of all BRS, irrespective of composition. According to the manufacturer, a PLLA-based scaffold has radial strength comparable to that of current drug-eluting metal stents. Directly after implantation, radial strength is ∼1,200 mm Hg, and the observed radial force can still be as great as 800 mm Hg after 1 year. Degradation by hydrolysis of interlactic bonds of the long PLLA chains results in particles that macrophages can phagocytose. The end product is lactic acid, metabolized via pyruvate into carbon dioxide and water through the Krebs cycle (15), with complete degradation achieved in 1 to 3 years (Central Illustration). Figure 3 shows degradation over time of the BRS compared with the Xience DES (Abbott Vascular, Santa Clara, California). PLLA-based devices ensure radial support for ∼6 months.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Degradation and Late Lumen Enlargement of the Absorb BVS Between 1 and 42 Months After Implantation and Comparison With an Everolimus-Eluting Stent

Porcine coronary artery model photomicrographs demonstrating continuous degradation of the poly-l-lactic acid–based Absorb BVS (bioresorbable vascular scaffold) (Abbott Vascular, Santa Clara, California), ending in almost complete scaffold strut replacement by a provisional cellular matrix after 42 months. Late lumen enlargement can be observed compared with an everolimus-eluting stent. (Image provided by Abbott Vascular, Santa Clara, California.)

Other Materials

A tyrosine polycarbonate–based BRS providing up to 6 months of radial support is also under investigation. Resorption takes between 24 and 36 months. Final products of degradation, which starts with hydrolysis and ends with the Krebs cycle, are ethanol, water, and carbon dioxide (19).

A BRS made of polylactic anhydride containing 2 salicylic acid molecules linked to 1 sebacic acid molecule was developed to provide mechanical support. Degradation into salicylate, water, and carbon dioxide is complete within ∼15 months (19).

Potential Benefits

BRS achieve successful acute revascularization of coronary artery lesions and show reasonably low rates of TLR and major adverse cardiac events (MACE) during early follow-up (Central Illustration). Multiple imaging analyses reveal beneficial plaque stabilization and sealing caused by BRS-induced remodeling (20), although the clinical impact needs further assessment. Due to BRS degradation, no foreign body remains in the vessel long term. Thus, late and very late stent thrombosis risks are potentially reduced or eliminated, depending on resorption duration. Total stent length is a well-known stent thrombosis risk factor. Because BRS dissolve, this risk may be reduced, especially in long or complex lesions and diffuse disease, when several would be implanted simultaneously. Additionally, the permanent complete side-branch occlusion risk may decrease. Because struts degrade, long-term uncovered stent struts are unlikely to factor in stent thrombosis. Incomplete endothelialization was observed for DES as long as 40 months after implantation (7). There are also reduced neointimal tissue growth and neoatherosclerosis and chronic inflammation risks as reactions to a permanent metal implant, all well-known late and very late stent thrombosis triggers in DES. Because the BRS coating is degradable, not durable, another stent thrombosis stimulus is absent. Intravascular ultrasound (IVUS) and optical coherence tomography (OCT) examinations display late lumen enlargement in numerous patients with the Absorb bioresorbable vascular scaffold (BVS) and DESolve BRS (Elixir Medical Corporation, Sunnyvale, California) (15,21–29).

The initial mechanical flexibility of some BRS may maintain original vessel geometry better than rigid metal stents; this would reduce their influence on biomechanical properties and blood flow. Furthermore, minor malapposition can be resolved by BRS self-correction, and its degradation avoids long-term malapposition (29). Incomplete stent apposition, as observed with DES after thrombus resolution, is also unlikely. Because there is no long-term vessel caging, abnormal shear stress may be reduced, as revealed by restored vasomotion (30). In contrast, paradoxical vasoconstriction was observed after DES implantation, most likely due to impaired endothelial function (31). BRS are better suited than metal stents for noninvasive imaging, such as coronary computed tomography and magnetic resonance imaging, because they do not cause artifacts, and follow-up is possible with these modalities. Moreover, BRS implantation might allow surgeons to carry out anastomosis of coronary artery bypass grafts at distal segments, and in patients who might require multiple interventions, there will be no interference with previously implanted DES because side branches can sometimes be especially difficult to recross.

The First Bioresorbable Scaffold

The Igaki-Tamai stent (Kyoto Medical Planning Co., Ltd., Kyoto, Japan) was the first BRS used in humans. This PLLA-based BRS is self-expandable when heated; consequently, contrast dye at 80°C is used for balloon inflation. Expansion continues at body temperature until dilation and vessel wall resistance reach equilibrium. In 2000, Tamai et al. (14) reported initial results from 15 patients in whom 25 stents were successfully implanted. Long-term data with >10 years of follow-up for 50 patients treated with 84 Igaki-Tamai biodegradable stents are available (32). Interestingly, during the first 6 months, minimal lumen diameter decreased and then constantly increased to 2.22 ± 0.56 mm at the 3-year follow-up. IVUS analysis showed almost constant stent cross-sectional area after 1, 2, and 3 years, whereas minimal lumen cross-sectional area decreased from 5.44 mm² immediately after the procedure to 3.64 mm² after 6 months, then increased to 5.18 mm² after 3 years. During the follow-up period, a total of 14 TLRs, 1 acute scaffold thrombosis and 1 very late scaffold thrombosis, 1 lesion-related myocardial infarction, and 1 cardiac death were noted. Accordingly, cumulative TLR rates per patient were 16% after 1 and 3 years, 18% after 5 years, and 28% after 10 years (32).

Despite these promising results, development of the Igaki-Tamai biodegradable stent was discontinued due to 2 limitations: first, implantation requires an 8-French guiding catheter, and second, the heated contrast dye may cause vessel wall injury. A new version of the device is currently undergoing pre-clinical evaluation.

Bioresorbable Scaffolds Currently Available in Clinical Practice

Several BRS types are currently in development, but only 2 have the Conformité Européenne (CE) mark for use in coronary artery disease: the Absorb bioresorbable vascular scaffold (BVS) (Abbott Vascular) and the DESolve scaffold (Elixir Medical Corporation). The Igaki-Tamai stent (Kyoto Medical Planning Co., Ltd.) also has the CE mark, but only for peripheral use. No devices have U.S. Food and Drug Administration approval.

The most widely investigated bioresorbable scaffold to date is the BVS. It is made of PLLA, its strut thickness is 150 μm, and it elutes a 1:1 mixture of poly-d,l-lactic acid and the antiproliferative drug, everolimus. Full hydrolytic degradation takes as long as 3 years (21).

The first-generation BVS was investigated in the ABSORB Cohort A, a first-in-humans trial. This prospective, open-label study included 30 patients with a single de novo coronary artery lesion. Device success was achieved in 94% of the patients (15). During 5 years of follow-up, the MACE rate was 3.4%, owing to 1 non–Q-wave myocardial infarction in a previously untreated vessel (22). Furthermore, restoration of vasomotion was demonstrated by OCT (23). A decrease in stent minimal lumen diameter during 2 years of follow-up was also noted (23). Additional evaluation showed BVS shrinkage due to acute recoil in quantitative coronary angiography and late recoil in IVUS (24,25).

Taking these findings into account, a new version, BVS 1.1, was developed. The struts were redesigned, leading to increased radial strength, longer radial support, and optimized drug transfer. BVS strut thickness was unchanged, and room-temperature BVS storage was now possible (26). The ABSORB Cohort B study evaluated the second-generation BVS. During 2 years of follow-up, the overall MACE rate was 9.0% (27). Late lumen loss in quantitative coronary angiography was substantially lower than in Cohort A. IVUS demonstrated a significant minimal lumen area decrease at the 6-month follow-up, which remained almost unchanged at the 2-year follow-up. Remarkably, the mean lumen area also decreased initially, but then increased significantly 2 years after the index procedure. OCT showed a constantly increasing scaffold area (27). In contrast, the mean scaffold area did not change significantly between the post-procedural and 12-month OCT and IVUS examinations in the Cohort B2 study (28). Nevertheless, there was no evidence of late or very late scaffold recoil. Vasomotion was tested by application of either acetylcholine or methylergonovine and subsequent lumen measurements, which revealed restoration of pharmacologically induced vasomotion 12 months after the procedure (28).

ABSORB EXTEND is another international multicenter study, including those with long lesions and small vessels (33). Preliminary results published with 24 months of clinical follow-up of 250 patients demonstrated a MACE rate of 7.3%, an ischemia-driven TLR rate of 4.0%, and a stent thrombosis rate of 0.8% (according to the Academic Research Consortium possible/definitive definition) (34). A randomized, controlled trial comparing BVS and DES is currently ongoing.

The other CE-approved, commercially available (in Europe) BRS is the DESolve scaffold (Elixir Medical Corporation), manufactured from PLLA and eluting a mixture of anti-inflammatory myolimus and PLLA. Strut thickness is 150 μm, and >95% of the device is resorbed after ∼1 year. Its advantage, compared with other BRS, is a wider range of expansion, with consequently reduced strut fracture risk, and self-correction of minor malapposition. During the 12-month follow-up period of the multicenter DESolve first in humans trial, 1 TLR occurred between 30 days and 6 months; there were 2 other MACE events but no evidence of scaffold thrombosis.

The DESolve Nx study is currently investigating a DESolve scaffold refinement, including elution of antiproliferative novolimus and a larger device size spectrum. The MACE rate at 1-year follow-up was 5.7% for 126 enrolled patients. Additionally, the lumen area, assessed by OCT, IVUS, and computed tomography, was almost constant during follow-up (35). Further evaluation in the DESolve NX II pivotal trial is ongoing. A refined version, with 100-μm strut thickness, will soon be available commercially.

In addition to the devices carrying the CE mark, other BRS concepts are under clinical and pre-clinical evaluation. The Central Illustration provides an overview of different designs and characteristics of existing BRS, with representative images in Figures 1 and 2.

Off-Label Use and Special Considerations

Preliminary results from registries investigating patients with acute coronary syndrome and ST-segment elevation myocardial infarction are available and show reasonable short-term outcomes and results comparable to those of patients treated with DES (36,37). Data comparing BRS use in diabetic with nondiabetic patients and diabetic patients who underwent DES implantation found no significant differences regarding the composite endpoint of cardiac death, target vessel myocardial infarction, and TLR at 1-year follow-up (38). However, diabetic patients have an increased repeat revascularization risk due to diffuse diseased vessels and a higher in-stent restenosis and stent thrombosis risk (3). Hence, from a long-term perspective, diabetic patients might benefit from BRS rather than DES.

Although BRS use for left main coronary artery disease was described (39), because the left main coronary artery usually has a larger vessel diameter than other coronary arteries and BRS are only available in limited sizes and have strict expansion ranges, its application should be limited. In contrast, BRS use is more feasible in small vessels and with reasonable results (40).

BRS seem attractive for treatment of in-stent restenosis (41). Compared with drug-eluting balloons, they provide short-term scaffolding, ensure drug delivery, and do not affect vessel diameter with another metal layer, as with DES. Indeed, BRS strut thickness must be considered in deciding whether to use BRS or another option.

BRS implantation for treatment of chronic total occlusions is feasible (42) because stent length is a predictor of in-stent restenosis, and stent thrombosis can be avoided by device dissolution. If multiple BRS are required, special care should be taken to minimize overlap due to the strut thickness of currently available BRS. In addition, the feasibility of subintimal tracking is thus far unclear.

Successful BRS use in bifurcations was also described (43,44). When treating bifurcations, simultaneous post-dilation of the main and side branches (“kissing”) is often required to resolve stent distortion. Due to scaffold fracture risk, post-dilation in BRS-treated patients should be performed with a “mini-kissing” post-dilation technique with minimal balloon overlap (44).

Although challenging, BRS implantation is also feasible for ostial lesions (45). Proper lesion sizing and BRS positioning are essential to avoid malapposed scaffold struts that protrude from the ostium because subsequent post-dilation may cause scaffold fracture.

BRS implantation in vein grafts was also performed (46), but vessel dimensions and a tapered shape often limit their use.

These findings are derived from registry data and case reports, all performed with the Absorb BVS. All cases were highly selected, and procedures were performed in high-volume centers experienced in PCIs with BRS. General BRS use restrictions must always be considered, including the presence of heavy calcification, severe tortuosity, and the BRS expansion ranges. Thus, BRS implantation should currently be reserved for patients with simple lesions and, until more evidence is available, should always be performed according to the manufacturer’s recommendation.

Current Limitations

Some possible benefits of BRS are hypothetical or only demonstrated in animal testing or small human cohorts, for example, the impact of restoring vasomotion. Most data derive from small, nonrandomized studies investigating patients with stable coronary artery disease and de novo lesions. Hence, relatively few patients were treated and experience in other anatomic settings (e.g., bifurcation lesions, long lesions, small vessels) or clinical presentations (e.g., acute coronary syndrome) is limited. Thus, in complex lesions, intravascular imaging should support BRS implantation.

There are practical concerns regarding strut thickness, which is greater than in conventional stents and may lead to vessel injury, nonlaminar flow, platelet deposition, and poor deliverability. Thus, mechanical considerations seem more challenging, especially when calcification or tortuosity are present. Regardless of lesion anatomy, pre-dilation is mandatory; direct stenting is not possible. Due to pre-dilation, a longer balloon inflation time, and post-dilation, if necessary, more induced myocardial ischemia might be associated with BRS implantation than DES. Due to the lack of radial strength of some BRS and poor deliverability in complex lesions, prolonged, extensive, and time-consuming pre-dilation is required, increasing the risk of pre-dilation dissections that might require longer and/or overlapping stents. There is an apparently increased scaffold fracture risk with overdilation; thus, significant BRS upsizing is impossible. The relatively long dilation time during deployment may also need to be reduced. Only limited scaffold sizes are currently available, and special facilities are needed for storage of some. Due to these technical particularities, the total cost and duration of PCI with a BRS may be higher than with a conventional DES.

Thus far, the optimal duration of dual-antiplatelet therapy in conjunction with BRS application is unclear. Although with some devices, resorption is achieved after a relatively short time and a reduction seems prudent, reduced shear stress from the thick struts of current BRS may cause platelet activation.

Multiple processes, including release of active drugs and polymer resorption, could adversely affect metal DES. This is illustrated by the impact of acquired malapposition, in the sense of defective healing, compared with the lack of stent strut apposition at the end of the procedure. Thus, defective healing and late adverse reactions may not be completely avoided with BRS.

Current BRS limitations will likely be resolved in the future. Although their advantages already outnumber their disadvantages, large, randomized, controlled trials are still needed.