The ABSORB cohort A and B experiences provided objective characterization of BRS resorption and coronary healing process in humans.10)11) Serial intravascular imaging studies showed that strut resorption and vascular healing after BRS implantation were associated with late lumen enlargement.2) The formation of a neointima layer after BRS resorption suggested to seal or “cap” the necrotic core plaque, and to prevent plaque rupture in the future.3) In addition, the restoration of vasomotor response to stimuli was demonstrated by 5 years.4) Early promising results called for clinical comparison studies with standard metallic DES.

ABSORB II is the first randomized controlled trial to compare Absorb BRS with everolimus-eluting stent (EES) in 501 patients.1) The primary endpoint was angiographic vasomotor reactivity after nitrate injection and angiographic late luminal loss. At 1 year, first new or worsening angina was lower with BRS, although clinical outcomes were similar between groups. However, at 3-year follow-up, the vasomotor reactivity, angina status, and exercise capacity were not different. In addition, the late luminal loss was larger in the BRS group. The rate of a device-oriented composite endpoint (DOCE) was significantly higher in the BRS group, mainly driven by target vessel MI. Definite or probable device thrombosis was also significantly higher in the BRS group (Figure 2).5)

ABSORB III is the first large-scale, multicenter, randomized trial for US FDA regulatory approval.7) This study enrolled 2,008 patients with relatively simple coronary lesions. This study demonstrated that the BRS group was non-inferior to the EES group in the respective to target lesion failure (TLF) at 1-year. However, the 2-year results presented at American College of the Cardiology (ACC) 2017 that the rates of TLF were significantly higher in the BRS group due to the increased risk of target vessel MI in the BRS group. However, ABSORB China and ABSORB Japan at EuroPCR 2017 showed that BRS had comparable safety and efficacy to EESs at 3 years.

Recently, the Amsterdam Investigator-initiateD Absorb Strategy (AIDA) study reported early because of the higher incidence of device thrombosis in the BRS group.6) This study enrolled 1,845 patients with more complex lesion subset than ABSORB III trial in the context of routine clinical practice: acute coronary syndrome (ACS) was 54% and small vessel disease was about 20%. In addition, postdilatation (74%) was still underused. Although target vessel failure (TVF) was not significantly different, device thrombosis, and subsequently target vessel myocadial infarction were significantly higher in the BRS group. Table 1 summarized currently available randomized trials.

Update meta-analysis of 2-year outcomes from 7 randomized trials showed that BRS had higher 2-year risk of the DOCE than EES.12) This difference was mainly derived from target vessel MI and ischemic-driven target lesion revascularization (TLR). In addition, device thrombosis was significantly higher in the BRS group. However, cardiac mortality was not different between groups (Figure 3).

The Gauging coronary Healing with bioresorbable Scaffolding plaTforms in EUrope (GHOST-EU) registry evaluated the performance of the Absorb BRS in a real-world practice from 10 European heart centers.13) The incidence of TLF was 2.2% at 30 days and 4.4% at 6 months. However, definite or probable scaffold thrombosis was 1.5% at 30 days and 2.1% at 6 months. ABSORB expand registry reported that 12-month clinical outcomes in the first 512 patients.14) At one year, the ischemia-driven TVF was 4.9%. The cumulative rate of definite and probable scaffold thrombosis for this population was 0.8%. A Prospective, Randomized Trial of Bioresorbable Vascular Scaffold Versus Everolimus Eluting Stent in Patients Undergoing Coronary Stenting for Myocardial Infarction (ISAR-ABSORB MI) registry enrolled more complex population with diabetes (31.5%), ACS (39.0%), and bifurcation (13.1%).15) At 2 years, the composite of death, MI, or TLR occurred in 21.6%. Definite scaffold thrombosis occurred in 3.8%. This study showed the higher event rates than expected, which raised concerns about the daily use of BRS. The Registro Absorb Italiano (RAI) registry enrolled 1,505 patients (22.4% diabetes, 59.6% ACS) from 23 Italian heart centers16). All lesions were predilated and 96.8% lesions were post-dilated after BRS implantation. At 30 days, TLR occurred in 0.6% and definite or probable scaffold thrombosis occurred in 0.8%. This registry suggested that when accurate BRS implantation technique was used, an unrestricted BRS use would be safe and effective.17)18)19)20)21) In addition, BRS was evaluated in the complex patients and lesions subset including diabetes,22) ACS,23)24)25) MI,26)27)28)29)30)31)32)33) small vessel,34) and in-stent restenosis.35)36)37) However, the interpretation of such studies should be careful considering the biased nature of registry studies. Table 2 summarized currently available registry studies.

As current BRSs resolve 2–3 years after implantation, improvement in outcome comparing with metallic DESs is expected in long-term over those periods. Therefore, ideal BRS candidate is a relatively young with a good life expectancy (>5 years). On the other hand, the use of BRS in patients with limited life expectancy with multiple comorbidities, and at high bleeding risk was not supported.38)

BRS can be implanted in non-complex lesions including de novo lesions with diameter of 2.25–4.0 mm on on-site quantitative coronary angiography (QCA), relative short lesions, and stable angina presentation. The culprit lesion of ST segment elevated MI, bifurcation lesion treating with 2 scaffolds, and aorto-ostial lesion, extreme tortuous vessel and small vessel (<2.25 mm) were less favorable lesions for BRS implantation. Table 5 summarized the BRS favorable patient and lesion characteristics.38)

Optimal sizing and BRS implantation technique is of paramount importance for achieving favorable long-term outcomes. Operators should understand that thick scaffold struts and their plastic nature results in a lower lumen gain and a smaller post-implanted minimal lumen diameter compared with conventional metallic DES. Importantly, report that early scaffold thrombosis occurs at a time when most patients received DAPT suggested that scaffold thrombosis would be related to procedure related factors.5) In addition, in randomized trials and registry studies, the optimal techniques for BRS implantation was underused. The systemic introduction of a BRS-specific protocol which has come to be known as preparation, sizing, and postdilatation (PSP) are associated with an up to 70% decrease in scaffold thrombosis to rates similar to those reported in metallic DESs (Figure 4).18) We suggest the “effective” PSP methods for BRS implantation (Table 6, Figure 5).38)39)

The lesions should be prepared using adequate predilatation with semi or non-compliant balloon (1:1 vessel to balloon ratio). At the same time, operator should be cautious to avoid severe dissection over the BRS covered zone. Particularly for calcified lesions, aggressive lesion preparation is mandatory due to the relatively lack of sufficient radial force in BRS. Scoring/cutting balloon or rotational atherectomy can be used with lower threshold.40) If predilatation balloons cannot fully expand lesions, BRS implantation should be avoided.

Intracoronary imaging including intravascular ultrasound or optical coherent tomography is a useful tool to assess pre-intervention lesion characteristics and optimize stent implantation.41) However, even in clinical trial setting, intracoronary imaging was significantly underused.7)42)43) In addition, angiography guidance with visual estimation may be inaccurate.44)

Instead, online QCA with automatic calibration offers reliable assessment of vessel sizing without additional cost.45)46) We suggest that BRS sizing relies on proximal and distal maximal lumen diameter (Dmax) at the level of intended BRS implantation zone after nitroglycerin administration.47) Except in case of extreme vessel tapering, the scaffold selection should match the proximal Dmax in proximal device landing zone. Considering Dmax from online QCA is about 0.5 mm smaller than vessel diameter from intracoronary imaging,48) this sizing method with additional bigger non-compliant balloon inflation can well-negotiate the risk between underexpansion of proximal edge and dissection of distal edge. In addition, this strategy can limit excessive scaffold/artery ratio, and may decrease thrombogenicity, and neointimal thickness.49)50)

BRS should cover normal looking segment at either edge of lesion. Scaffold deployment should be performed slowly with long-duration. In general, high-pressure inflation with delivery balloon was not recommended. When implanting multiple BRS, minimal overlapping techniques to minimize scaffold thickness are suggested such as “marker-to marker” (up to 1 mm of overlap) or “scaffold-to-scaffold” (no overlap) technique.8)

Postdilitation is also very important during BRS implantation. In previous registry, all acute or subacute BRS thrombosis occurred in severe underexpanded scaffold.51) Postdilatation uses a non-compliant balloon with 0.5 mm bigger size than scaffold device with high inflation pressure (16–25 atm). Recent data showed that higher balloon to artery ratio (>1.2, or balloon to device ratio >1.15) was associated with expansive vessel wall remodeling.52) Target pressure should be maintained for at least 30 seconds, because a significant larger lumen diameter is obtained with a longer inflation time.53)