Clinical results of bioresorbable drug-eluting scaffolds in short and long coronary artery lesions using the PSP technique

Prior studies demonstrated a comparable safety and efficacy for the use of bioresorbable scaffolds Absorb in native and simple coronary lesions with a higher rate of thrombotic events like scaffold thrombosis in multiple randomized trials compared to metallic drug eluting stents [1–3]. These studies showed that without a consequent use of the renewed PSP-technique the scaffold thrombosis rate as well as clinical endpoints were higher compared to metallic drug-eluting stents [1, 4]. The PSP-technique was used in only 8% of lesions in the randomized Absorb III trial leading to a significant higher rate of target lesion failure compared to the everolimus-eluting stent [2]. Previous studies showed that even in very complex lesions like chronic total occlusions the implantation of Absorb BVS with the PSP-technique is associated with a low risk of thrombotic events and good clinical outcomes [5, 6]. From DES-trials it is known that patients with short coronary lesions have lower MACE-rates and a better outcome compared to DES in long lesions [7–10]. In addition, it is well known that BVS with Absorb is associated with a worse outcome compared to DES. Until now data including all comers stenosis with long coronary lesions treated with the bioresorbable scaffolds alone were scarce. Therefore we compared the outcome of long with short lesions treated with the Absorb BVS using the PSP technique with respect to clinical follow-up up to 36 months.

There were 190 patients treated with Absorb BVS for lesions < 20 mm length and 136 patients treated with Absorb BVS for lesions longer or equal than 20 mm by quantitative coronary angiography.

The clinical follow-up after 12 months was completed in 98.5% of available patients. The rate of patients was 95.7% after 24 months and 90.8% after 36 months. The rate of scaffold thrombosis did not differ between groups and was 0% after 12 months, 1% after 24 months and 1.7% after 36 months in short lesions compared to 1.4, 0.0 and 0.0% in long lesions. Kaplan-Meier estimates for scaffold thrombosis after 12 months were 0% for short lesions and 1.40% for long lesions (p = 0.05), 0.46% vs. 1.40% after 24 months (p = 0.2286) and 4.05% vs. 1.50% after 36 months (p = 0.7529). After 12 months two definite scaffold thrombosis (after ARC-criteria) with total incidence of 0.48% and after 24 months one definite (total incidence: 0.30%) very late scaffold thrombosis occurred. Three possible (total incidence 1.68%) occurred in the 36 months follow-up.

Target lesion revascularization was higher in long compared with short lesions after 12, 24 and 36 months (Tables 4, 5 and 6). Kaplan-Meier estimates for target lesion revascularization after 12 months were 1.46% for short and 7,69% for long lesions (p = 0.0012), 2.06% vs. 8.75% after 24 months (p = 0.0027) and 4.96% vs. 9.59% after 36 months of follow-up (p = 0.0109). Consequently the DOCE was numerically higher in patients with long lesions as compared to patients with short lesions. The difference between groups was 6.2, 4.4 and 5.1% after 12, 24 and 36 months favoring shorter lesions. Kaplan-Meier estimates for DOCE after 12 months were 2.63% for short lesions and 8.09% for long lesions (p = 0.0131), 5.51% vs. 11.35% (p = 0.0503) after 24 months and 8.00% vs. 18.00% (p = 0.0264) after 36 months of clinical follow-up (Fig. 1).

	Total	Lesion length < 20 mm	Lesion length ≥20 mm	P-Value
Number of lesions	417	274	143
Scaffold thrombosis, N (%)
acute	0 (0)	0 (0)	0 (0)	0.04
subacute	1 (0.2)	0 (0.0)	1 (0.7)
Late	1 (0.2)	0 (0.0)	1 (0.7)
TLR, N (%)	14 (3.4)	4 (1.5)	10 (7.0)	0.01
MI, N (%)	6 (1.4)	2 (0.7)	4 (2.8)	0.14
Number of patients	321	186	135
DoCE, N (%)	17 (5.3)	5 (2.7)	12 (8.9)	0.015
Cardiac mortality, N (%)	2 (0.6)	0 (0.0)	2 (1.5)	0.10

TLR ischemia-driven target lesion revascularization, MI myocardial infarction, not clearly related to a non-target vessel, DoCE device-oriented endpoints

	Total	Lesion length < 20 mm	Lesion length ≥20 mm	P-Value
Number of lesions	329	223	106
Scaffold thrombosis, N (%)
	1 (0.3)	1 (0.4)	0 (0.0)	0.38
TLR, N (%)	14 (4.3)	5 (2.3)	9 (8.7)	0.03
MI, N (%)	7 (2.1)	4 (1.8)	3 (2.8)	0.15
Number of patients	247	146	101
DoCE, N (%)	18 (7.3)	8 (5.5)	10 (9.9)	0.19
Cardiac mortality, N (%)	3 (1.2)	2 (1.4)	1 (1.0)	0.96

TLR ischemia-driven target lesion revascularization, MI myocardial infarction, not clearly related to a non-target vessel; DoCE device-oriented endpoints

	Total	Lesion length < 20 mm	Lesion length ≥ 20 mm	P-Value
Number of lesions	179	121	58
Scaffold thrombosis, N (%)	3 (1.1)	3 (1.7)	0 (0.0)	0.12
TLR, N (%)	11 (6.1)	5 (4.1)	6 (10.3)	0.08
MI, N (%)	8 (4.5)	5 (4.1)	3 (5.2)	0.21
Number of patients	139	84	55
DoCE, N (%)	16 (11.5)	8 (9.5)	8 (14.6)	0.37
Cardiac mortality, N (%)	5 (3.6)	3 (3.6)	2 (3.6)	0.95

TLR ischemia-driven target lesion revascularization, MI myocardial infarction, not clearly related to a non-target vessel; DoCE device-oriented endpoints

Using PSP-technique in 88% of lesions treated with Absorb BVS we were able to demonstrate that long lesions have a higher rate of TLR but not a higher risk for scaffold thrombosis up to 36 months of follow-up.

Kaplan-Meier estimates showed significant differences concerning DOCE after 12, 24 and 36 months. Kaplan-Meier-estimates for target lesion revascularization demonstrated significant lower rates in short lesions. For scaffold thrombosis no significant difference between short and long lesions concerning Kaplan-Meier estimates were observed.

Former studies especially the ABSORB-Trials showed good clinical results in selected lesions after 12 months follow-up [2, 15–17]. Regarding scaffold thrombosis rates an alarmingly trend towards higher rates of thrombotic events in patients treated with bioresorbable scaffolds could be seen. Considering scaffold thrombosis the randomized ABSORB-II trial with a follow-up period of three years showed a high scaffold thrombosis rate of 3% compared to 0% in metallic drug-eluting stents [1]. Causal for these findings missing predilation, wrong size selection and the lack of careful postdilation can be mentioned. Proper studies confirmed these findings in metallic drug eluting stents. Because of the strut thickness a careful implantation technique with predilation, proper sizing and postdilation is essential. As it is known from the one year analysis of ABSORB II the postdilation in this study was 61% [15].

The Absorb III study enrolled 2008 patients and randomized them 2:1 for treatment with scaffolds or drug-eluting stents [2]. After one year the scaffold thrombosis rate including probable or definite scaffold thrombosis was 1.5% in the bioresorbable scaffold group and 0.7% in the metallic drug-eluting-stent population without reaching significance [2]. Postdilation after scaffold implantation in this study was 65.5% [2]. The three-year data of the ABSORB-Trial in low and moderate complex lesions was associated with low and acceptable rates of major adverse clinical events with the proper PSP-technique. The ABSORB-Investigators could also demonstrate that the scaffold thrombosis rate was higher than with metallic DES [18].

The AIDA study enrolled 1845 patients and randomized them either in the bioresorbable scaffold group or the drug-eluting stent group [3]. They found no significant difference between the treatment with bioresorbable scaffolds or with metallic drug-eluting stents concerning target-vessel failure in a two year period (11.7% vs. 10.7%, p = 0.43) [3]. After 2 years of follow-up a higher rate of device thrombosis could be observed (3.5% vs. 0.9%, p = < 0.001) [3]. In this study predilation took place in 96.9% of treated patients and postdilation was used in 74% of patients [3].

In the GHOST-EU registry patients with exclusion criteria for randomized ABSORB trials were enrolled to represent a real-world population [19]. The scaffold thrombosis rate over 6 months was 2.1% [19]. A real-world population was studied by GABI-R as well. The GABI-R Investigators divided the population up into patients with scaffold thrombosis and patients without this thrombotic event to study predictors. In this international multicenter study 3137 patients were included between November 2013 and January 2016. Predilation took place in 95.6% of cases and postdilation was present in 77.3% of lesions. Differences between these two populations could be seen in a longer median scaffolded length (28 mm vs. 23 mm), a higher rate of scaffolded bifurcations (11.8% vs. 3.7%, p = < 0.0001) and more ostial lesions (3.9% vs. 0.8%, p = < 0.05).

Analyzing the patients with scaffold thrombosis, a longer scaffold length of 20.45 ± 5.66 mm (without scaffold thrombosis: 19.65 ± 6.23 mm; p = 0.27) and lower rates of postdilation (65.7% vs. 73.3%) were present [Differences between patients with and patients without scaffold thrombosis – Results of the German-Austrian ABSORB RegIstRy (GABI-R), J. Wöhrle and GABI-R Investigators, DGK 2017].

The idea that bioresorbable scaffolds could prove their benefits by regained vasomotion and complete resorption of the vessel-cage could only be studied at very long follow-up periods [19]. Therefore we planned to continue the follow-up time of this study until 10 years.

The 36 months data from the ABSORB II trial demonstrated a significant higher rate of adverse events including thrombotic ones using the bioresorbable vascular scaffold instead of metallic drug-eluting stents [1]. On top the study failed to reach the endpoint in late lumen loss [20]. In this study the PSP-technique was not routinely used. Meta-analysis summarizing important randomized studies showed a higher rate of scaffold thrombosis comparing to common metallic drug eluting stents [21–23]. Astonishingly the postdilation rate was between 52 and 66% [21–23]. Fernandes et al. analyzed a real world population with long coronary lesions after implantation of an everolimus-eluting stent [7]. The MACE and ST rates at 12, and 24-months follow-up were 2.1, 5.4% (MACE) and 0.7, 1.5% (ST) [7]. The scaffold thrombosis rate in our study was 1.4, 0.0% and for DOCE 8.9, 9.9%, after 12 and 24-month follow-up. Lesiak et al. compared a bioresorbable polymer sirolimus-eluting stent in patients with long coronary lesions with permanent polymer everolimus-eluting stent [8]. The target lesion revascularisation rate was 3.7% and the scaffold thrombosis rate was 1.2% in EES after 9 months of follow-up [8]. A recent study by Kang et al. demonstrated good results in very long coronary lesions (> 50 mm) with zotarolimus- and everolimus drug eluting stents [9]. Patra et al. showed 4% TLF after 12 months of follow up [10]. It is known that the bioresorbable scaffold has higher DOCE rates in a real world population. This comparison with results of the literature demonstrate that Absorb BVS is associated with a higher event rate compared with everolimus eluting stents [24].

With the introduction of the implantation-technique of Puricel and Gori et al. a significant reduction of scaffold thrombosis could be seen [4]. This implantation technique is composed of predilation with a non-compliant balloon up to the reference vessel diameter, then implantation of the scaffold of same size and careful high-pressure postdilation with a non-compliant balloon 0.5 mm larger than the implanted device – the nowadays so called PSP-technique [4]. It is to be assumed that careful and accurate implantation technique as well as the learning curve of interventionalists – as seen in GABI-R have a great impact on our outcome measures and especially thrombotic events [25]. The idea that bioresorbable scaffolds could prove their benefits by regained vasomotion and complete resorption of the vessel-cage could only be studied at very long follow-up periods [20]. Tanaka et al. showed a low scaffold thrombosis rate (1.2%) after a two year follow-up using the bioresorbable vascular scaffold in a population of 264 patients with 400 lesions with careful PSP-technique [26].

Sotomi et al. studied in his OCT-controlled population the most common reasons for scaffold thrombosis [27]. Malapposition (24%), incomplete lesion coverage (18%) and device underexpansion (12%) were reasons for early scaffold thrombosis [27]. Late ones fail at malapposition (35%), discontinuity (31%), peri-strut low intensity area (19%) and incomplete lesion coverage (12%) [27]. Also the device overlap seems to be a factor increasing the scaffold thrombosis rate [28]. The study group of Polimeni et al. enrolled 183 patients with ST-segment myocardial infarction in their study and could demonstrate that the scaffold thrombosis rate was reduced when an optimized implantation technique was used [29]. Many recent studies tried to find predictors for scaffold failure and figured out that the implantation technique is a very important factor in reducing cardiac events [28, 30–36]. The use in acute coronary syndroms studied by Anadol et al. seems to be safe as well [37]. But for long coronary lesions.

In conclusion the implantation of bioresorbable vascular scaffolds Absorb in long lesions compared to short ones is safe using the PSP-technique including accurate predilation, proper sizing and postdilation and shows an acceptable risk of thrombotic or adverse events.

In long lesions compared to short ones the bioresorbable scaffold Absorb has higher DoCE-rates but shows acceptable clinical follow-up results up to 36 months with the proper PSP-technique.