The first bioresorbable vascular scaffold (BVS), the Absorb BVS™ (Abbott Vascular, Santa Clara, CA, USA) made of poly-l-lactic acid came to the market in 2011 raising great expectations of another “revolution” in interventional cardiology. BVS was designed to mimic the acute results of drug-eluting metallic stents (DES) while avoiding the long-term penalties of stent thrombosis by maintaining the natural architecture of the vessel through temporary mechanical support with subsequent scaffold resorption. It was hypothesized that scaffold resorption would restore vasomotion, allowing luminal expansion and constriction as physiologically appropriate, and lessen subsequent plaque accretion. Empiric studies have not fulfilled these hopes. Rather, the Absorb BVS was associated with a higher risk of myocardial infarction and target lesion failure compared with 2nd generation DES [1]. An individual-patient-data pooled meta-analysis of the 4 randomized ABSORB trials in which patients were treated with everolimus-eluting Absorb BVS or cobalt-chromium everolimus-DES showed at 3 years a higher rate of target lesion failure (11.7% versus 8.1%; risk ratio [RR], 1.38; 95% confidence interval [CI], 1.10–1.73; P = 0.006), driven by greater target vessel myocardial infarction (7.8% versus 4.2%; RR, 1.72; P = 0.0006) and ischemia-driven target lesion revascularization (6.6% versus 4.4%; RR, 1.44; p = 0.02), with comparable cardiac mortality of 1.1%. Device thrombosis rates were also higher with BVS (2.4% versus 0.6%; RR, 3.71; P = 0.001) [2].

A subsequent BVS approach is the magnesium-based metallic Magmaris™ BVS (Biotronik AG, Bülach, Switzerland). In principle, it offers greater radial force and complete resorption in months vs years for Absorb BVS [2]. In this issue of CDT, Tousek et al. present the 12-month results of the PRAGUE-22 study which investigated late lumen loss in acute coronary syndrome patients receiving Magmaris sirolimus-eluting scaffold or permanent Xience DES [3]. The study was a small but carefully performed two-center randomized study with 25 patients in each group. Baseline and procedural characteristics were similar between groups except for higher rates of pre- and post-dilatation in the Magmaris group. This procedural difference was necessitated by vessel preparation needed for the Magmaris stent which included predilation using a balloon of diameter ≤ 0.5 mm than that of the vessel and post-dilation using a noncompliant balloon inflated to > 16 atm. Angiographic late lumen loss at 12 months was greater in the Magmaris group than in the Xience group (0.54 ± 0.70 vs. 0.11 ± 0.37 mm; p = 0.029). Similarly, OCT-measured late lumen diameter loss in the Magmaris group was also significantly higher than that in the Xience group (0.59 ± 0.37 vs 0.22 ± 0.20 mm; p = 0.01). During follow-up, three clinical events occurred in the Magmaris group requiring re-intervention. All these cases presented as unstable angina, two attributed to early strut collapse and one due to neointimal proliferation. One MI occurred in the Magmaris group due to scaffold thrombosis in the setting of DAPT discontinuation. One target lesion revascularization at 5 months occurred in the Xience group due to edge restenosis. The authors concluded that a greater degree of late lumen loss occurs at 12 months with the use of a magnesium-based bioresorbable stent in patients with acute coronary syndrome compared to a permanent, everolimus eluting metallic stent. The small size of the study precluded conclusion regarding differences in clinical outcomes.

The current study is in line with previous (somewhat disappointing) results of Magmaris BVS in humans. Previous single-arm experience has suggested that Magmaris is relatively safe [2]. However, in STEMI patients, the randomized MAGSTEMI trial showed that despite improved vasomotion at 1 year in the treated segment in the BVS group, late lumen loss was higher (vs DES) in the Magmaris group (in-segment: 0.39 ± 0.49 mm vs 0.02 ± 0.27 mm, P < 0.001; in-device: 0.61 ± 0.55 mm vs 0.06 ± 0.21 mm; P < 0.001) as was angiographic restenosis rate (in-segment 0% vs 21.5%, p < 0.001) [4]. At 3 years in the single-arm BIOSOLVE studies, the primary mode for target-lesion failure in 189 patients was target-lesion revascularization (3.4%) and cardiac death (2.3%) but on a positive note, there were no definite or probable scaffold thromboses [5].

Based on previous studies and the current study by Tousek et al., it is fair to ask whether the concept of BVS is wrong or whether there are technical issues that ultimately can be overcome to improve long-term outcomes. The European Society of Cardiology has acknowledged the current concerns about BVS by providing a Class IIIC recommendation, recommending their use only as part of a well-designed clinical trial. Further, for those patients who already received a BVS while they were commercially available, it is recommended that DAPT should be continued for 3 years or longer. Despite the rather discouraging results to date, there are reasons to believe that BVS has a brighter future based on some insights gleaned from bench testing and empiric results in clinical studies.

Implantation technique for BVS has, by and large, followed the DES approach. There may be better methods for BVS deployment, with each material used having unique implantation requirements. Blachutzik et al. systematically evaluated the mechanical effect of non-compliant balloon post-dilatation on Magmaris BVS as measured with OCT. Post-dilatation reduced malapposition without causing strut fracture and led to a larger scaffold diameter [6]. Bench testing by Barkholt et al. also demonstrated that the Magmaris stent could be post-dilated to a larger diameter without fracture and also observed that recoil was greater at 120 min after post-dilatation than metallic stents, Absorb BVS and a novolimus-eluting PLLA-based scaffold [7].

Relative to current DES, BVS strut thickness is much larger (60–80 μm for DES vs 140–150 μm for Magmaris and Absorb). Previous studies in the “ancient days” of bare metal stents demonstrated conclusively that simply decreasing stent strut thickness was associated with a decrease in long-term adverse cardiac events [8]. As such, decreasing BVS strut thickness while maintaining tensile strength is one potential way to improve outcomes. Second, the BVS degradation profile may need to be optimized. Dissolution times need to follow the “Goldliocks” principle: not too short so as to be an effective temporary scaffold (which may not be the case at present), and not too long, so as to not create its own issues by being present when no longer needed. Further, the method of dissolution may need to be improved so that BVS remnants do not become part of the problem by being the nidus for thrombus formation. Further improvement in biocompatibility (already apparently acceptable) may allow for prompt, complete and predictable neointimal strut encapsulation and dissolution. The impact of the clinical situation, i.e., acute coronary syndrome vs stable disease, in regard to strut resorption is another avenue of research.

We hypothesize that BVS will ultimately improve long-term results when technical issues are overcome. Our optimism is based on major developments in interventional cardiology that seemed fruitless for a long period. For those with a long memory, there were innumerable failures in preventing restenosis before DES. Thus, we hope that developers keep persevering, bearing in mind that, “It’s darkest before the dawn”.