https://orcid.org/0000-0001-5100-0471
Abstract
The aim of the study is to assess the impact of the baseline plaque composition on the DREAMS 3G luminal late loss and to compare the serial plaque changes between baseline and 6 and 12 months (M) follow-up.
A total of 116 patients were enrolled in the BIOMAG-I trial. Patients were imaged with optical coherence tomography (OCT) pre- and post-DREAMS 3G implantation and at 6 and 12 M. OCTPlus software uses artificial intelligence to assess composition (i.e. lipid, calcium, and fibrous tissue) of the plaque. The differences between the OCT-derived minimum lumen area (MLA) post-percutaneous coronary intervention and 12 M were grouped into three terciles. Patients with larger MLA differences at 12 M (P = 0.0003) had significantly larger content of fibrous tissue at baseline. There was a reduction of 24.8% and 20.9% in lipid area, both P < 0.001, between the pre-DREAMS 3G OCT and the 6 and 12 M follow-up. Conversely, the fibrous tissue increased by 48.4% and 36.0% at 6 and 12 M follow-up, both P < 0.001.
The larger the fibrous tissue in the lesion at baseline, the larger the luminal loss seen at 6 and 12 M. Following the implantation of DREAMS 3G, favourable healing of the vessel coronary wall occurs as shown by a decrease in the lipid area and an increase in fibrous tissue.
Serial changes that take place in the vessel wall after the implantation of DREAMS 3G. Device main characteristics are outlined in the central rectangle. At the top row, a calcified (white) and lipid (yellow) plaque was imaged serially, while at the bottom row, a plaque with a larger fibrotic (green) content is depicted. At 12 months (M), a greater lumen reduction (compare lumen size between post-DREAMS 3G and 12 M) was observed in the latter compared with the top example. In addition, irrespective of the type of plaque before the implantation of DREAMS 3G, at 12 M, a reduction lipid and an increase in fibrous tissue (especially the fibrous cap overlying the lipid tissue—box–whisker plot) was observed. Please note that all the blue circles superimposed to the optical coherence tomography catheter are of the same size.
Introduction
Bioabsorbable stents were devised to prevent long-term stent-related adverse events (e.g. neoatherosclerosis and late thrombosis) compared with metallic and permanent stents. A promising second-generation drug-eluting absorbable metallic scaffold (Magmaris, Biotronik, Bülach, Switzerland) had a better in-scaffold late lumen loss (LLL) at 12 months (M) (0.39 ± 0.27 vs. 0.52 ± 0.39 mm with the first-generation device) and no scaffold thrombosis up to 3 years.1 Yet, the best iteration so far is the bioresorbable DREAMS 3G, made with a new magnesium alloy, which has further improved the stent’s mechanics compared with its predecessors. The safety and performance of this stent was evaluated in the BIOMAG-I trial, a prospective, multicentre, non-randomized, first-in-human study of 116 patients.2–4 At 6 and 12 M, in-scaffold LLL was 0.21 ± 0.31 and 0.24 ± 0.36 mm, respectively. After this period, its resorption and integration into the coronary wall process is completed at 12 M. It is, therefore, important to evaluate the serial compositional changes in the vessel wall and determine whether luminal loss over time is related to the composition of the de novo vessel wall at baseline.
Optical coherence tomography (OCT) has a unique resolution to permit coronary plaque content characterization.5 Newer artificial intelligence (AI)-powered software, such as OCTPlus (Pulse Medical, Shanghai, China), enables automatic analysis of plaque components.6
In the current study, we present the results of temporal changes in the atherosclerotic plaque characteristics after implantation of DREAMS 3G scaffold through one year.
Methods
Study population
This is a post hoc analysis of the patients enrolled in the BIOMAG-I trial (NCT04157153), which is a prospective, multicentre, non-randomized, first-in-human study of 116 patients evaluating the safety and performance of the DREAMS 3G (sirolimus-eluting resorbable coronary magnesium scaffold system) in patients with de novo coronary lesions. The study was conducted at 14 centres across Europe. Inclusion and exclusion criteria have been described elsewhere.3 Briefly, patients enrolled had symptomatic coronary artery disease with stable or unstable angina, documented silent ischaemia, or non–ST-elevation myocardial infarction (NSTEMI). A maximum of two de novo single lesions in two separate coronary arteries was permitted. Patients with left main disease, ST-elevation myocardial infarction, unsuccessful predilatation, or ostial lesions were excluded. The study was approved by the ethics committees of all participating institutions.
DREAMS 3G device
The DREAMS 3G is a scaffold system with a backbone composed of a resorbable magnesium alloy that is completely coated with bioresorbable poly-L-lactide acid, which incorporates sirolimus as the antiproliferative drug. It contains two permanent X-ray markers made from tantalum on its distal and proximal ends. Magnesium degrades to amorphous calcium phosphate via magnesium hydroxide and magnesium phosphate, and resorption is completed within 12 M. The stent was implanted in accordance with the ‘4P’ strategy.7
Quantitative coronary angiography analysis
Angiograms were recorded in two orthogonal views after intracoronary injection of nitroglycerine (200 μg), with matching projections taken before and after the procedure.
All QCA data of patients included in this report were analysed by an independent core laboratory that was blinded to the clinical outcomes (MedStar Cardiovascular Research Network, Washington, DC, USA). Late loss and/or late luminal gain are defined as the difference between minimum lumen diameter (MLD) at post-procedure minus MLD at follow-up. For lumen diameter reduction, this is a positive number; for late increase in lumen size, this is a negative number.
OCT acquisition and analysis
The BIOMAG-I trial protocol required angiographic and OCT evaluations at pre- and post-procedure, as well as at 6 and 12 M. The OCT data were analysed using an automated software, OCTPlus, at an independent core laboratory (MedStar Cardiovascular Research Network). The software uses AI to measure the luminal dimensions [i.e. minimum lumen area (MLA)] and characterize plaque into cholesterol crystal, fibrous tissue, calcium, lipid, and macrophage content (Figure 1) in each of the frames of the treated segment. Each of the five tissue types is further analysed for various parameters [area (mm2), span angle (°), and thickness (mm)].8
OCT matching frames (top row) are shown from before and after the implantation of the device (pre- and post-DREAMS 3G), at 6 and 12 M for two different patients; the top one experienced late lumen enlargement and the bottom LLL. The corresponding frames analysed by the OCTPlus software are shown in the relevant bottom rows. These frames depict the colour coding used for the different tissue types. The frame after the implantation is not presented because the struts produce an artefact that it is misinterpreted by the software. Of note, at the top example, the lesion (top left frame) was severely calcified and further modified with the predilation and device implantation. On the top row, the frames show resorption and full integration of the device into the vessel wall at 6 and 12 M. On the same example bottom row, please note the new layer of fibrous tissue (green) overlying the calcium and other constituents of the plaque. The second example showing LLL (compare lumen size between post-DREAMS 3G and 12 M) had a more fibrotic plaque before the intervention. Please note that all the blue circles superimposed to the OCT catheter are of the same size.
Study objectives
We aimed to compare the baseline (before implantation with the DREAMS 3G) coronary plaque composition in relation to the luminal late loss by angiography and by looking at the differences in MLA changes by OCT and compare the serial coronary plaque changes between baseline (before implantation with the DREAMS 3G) and 6 and 12 M follow-up.
Statistical analysis
Categorical variables are presented as a number (percentage), with comparison using χ2 or Fisher’s exact test (for an expected cell value <5). Continuous variables were presented as mean ± standard deviation (SD). The means of the three groups were then compared using Welch’s test within one-way analysis of variance. The absolute difference [6 and 12 M minus pre-percutaneous coronary intervention (PCI)] and relative difference {e.g. [(12 M − Pre)/Pre] × 100%} were calculated, and P value was reported. Cumulative curves depicting the angiographic LLL and difference in MLA by OCT were created. Complete case analyses on serial changes in plaque characteristics were performed by means of one-sample t-test to evaluate the changes between two timepoints.
Furthermore, we also explore the univariable test statistics for fibrous tissue in % and demographic parameters captured in electronic case report form (eCRF).
JupyterLab version 3.4.4 based on Python version 3.9.13 and SAS 9.4 were used for data analysis. A P value < 0.05 was considered statistically significant. No adjustment of alpha for multiple testing was performed.
Results
Study population
A total of 116 patients were enrolled in the BIOMAG-I trial. Of these patients, 110 had OCT data pre-DREAMS 3G implantation, 113 had OCT data post-DREAMS 3G implantation, 104 had OCT data at 6 M follow-up, and 96 had OCT data at 12 M as shown in Supplementary data online, Figure S1. The mean age was 61.0 ± 9.0 years, and most participants were male (77.6%). Clinical presentation included stable angina (42.2%) and silent ischaemia (24.1%), while 20.7% had NSTEMI. Other baseline demographics are outlined in Supplementary data online, Table S1. Regarding the lesion characteristics, the following should be noted: the majority was in the left anterior descending artery and were type B2 according to the American College of Cardiology/American Heart Association classification.
Coronary plaque composition and luminal loss
At 6 M (see Supplementary data online, Table S1A), the patients were split into three groups by angiographic LLL terciles. There were no statistically significant differences in all tissue types across LLL groups. At 12 M (Table 1), however, the fibrous tissue as measured before implantation of DREAMS 3G at baseline was different with 2.39 ± 1.05 vs. 2.30 ± 0.87 and 2.95 ± 0.97 mm2 for the lower, middle, and high LLL groups, respectively, with a significant P value (0.0172). Furthermore, the correlation with a spearman coefficient of 0.260 (P = 0.011) was detected.
OCT-derived tissue types as measured before implantation of DREAMS 3G at 12 M LLL terciles
| Angiographic LLL between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤0.09 mm) | Tercile 2 (0.09–0.27 mm) | Tercile 3 (>0.27 mm) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.39 ± 1.05 | 33 | 2.30 ± 0.87 | 31 | 2.95 ± 0.97 | 0.0172 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 33 | 0.01 ± 0.01 | 31 | 0.01 ± 0.01 | 0.5954 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 33 | 0.03 ± 0.01 | 31 | 0.03 ± 0.01 | 0.7073 |
| Lipid, mm2 | 31 | 1.18 ± 0.52 | 33 | 1.02 ± 0.38 | 31 | 1.03 ± 0.50 | 0.3623 |
| Calcification, mm2 | 30 | 0.30 ± 0.24 | 33 | 0.29 ± 0.18 | 31 | 0.25 ± 0.19 | 0.5327 |
| Angiographic LLL between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤0.09 mm) | Tercile 2 (0.09–0.27 mm) | Tercile 3 (>0.27 mm) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.39 ± 1.05 | 33 | 2.30 ± 0.87 | 31 | 2.95 ± 0.97 | 0.0172 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 33 | 0.01 ± 0.01 | 31 | 0.01 ± 0.01 | 0.5954 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 33 | 0.03 ± 0.01 | 31 | 0.03 ± 0.01 | 0.7073 |
| Lipid, mm2 | 31 | 1.18 ± 0.52 | 33 | 1.02 ± 0.38 | 31 | 1.03 ± 0.50 | 0.3623 |
| Calcification, mm2 | 30 | 0.30 ± 0.24 | 33 | 0.29 ± 0.18 | 31 | 0.25 ± 0.19 | 0.5327 |
OCT-derived tissue types as measured before implantation of DREAMS 3G at 12 M LLL terciles
| Angiographic LLL between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤0.09 mm) | Tercile 2 (0.09–0.27 mm) | Tercile 3 (>0.27 mm) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.39 ± 1.05 | 33 | 2.30 ± 0.87 | 31 | 2.95 ± 0.97 | 0.0172 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 33 | 0.01 ± 0.01 | 31 | 0.01 ± 0.01 | 0.5954 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 33 | 0.03 ± 0.01 | 31 | 0.03 ± 0.01 | 0.7073 |
| Lipid, mm2 | 31 | 1.18 ± 0.52 | 33 | 1.02 ± 0.38 | 31 | 1.03 ± 0.50 | 0.3623 |
| Calcification, mm2 | 30 | 0.30 ± 0.24 | 33 | 0.29 ± 0.18 | 31 | 0.25 ± 0.19 | 0.5327 |
| Angiographic LLL between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤0.09 mm) | Tercile 2 (0.09–0.27 mm) | Tercile 3 (>0.27 mm) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.39 ± 1.05 | 33 | 2.30 ± 0.87 | 31 | 2.95 ± 0.97 | 0.0172 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 33 | 0.01 ± 0.01 | 31 | 0.01 ± 0.01 | 0.5954 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 33 | 0.03 ± 0.01 | 31 | 0.03 ± 0.01 | 0.7073 |
| Lipid, mm2 | 31 | 1.18 ± 0.52 | 33 | 1.02 ± 0.38 | 31 | 1.03 ± 0.50 | 0.3623 |
| Calcification, mm2 | 30 | 0.30 ± 0.24 | 33 | 0.29 ± 0.18 | 31 | 0.25 ± 0.19 | 0.5327 |
In Supplementary data online, Table S1B, the differences between OCT-derived MLAs post-PCI and 6 M were grouped into three categories. The mean fibrous tissue at baseline was significantly different in the MLA terciles (P = 0.0057). Likewise, in Table 2, the differences between MLAs post-PCI and 12 M were also grouped into three categories. Larger MLA differences in tercile 3 (>3.07 mm2, larger luminal loss) were observed in patients who had significantly larger content of fibrous tissue at baseline compared with the smallest group (P < 0.05). Nevertheless, the mean fibrous tissue differs significantly between the terciles (P = 0.0003). The Spearman rank correlation coefficient between fibrous tissue and difference in MLA post-PCI and 12 M was determined as 0.390 (P = 0.0001). Of note, using the 12 M OCT terciles, a comparison of the scaffold expansion (right after PCI) and plaque burden (before intervention) across terciles resulted in no significant differences.
Three-dimensional OCT-derived MLA at 12 M differences
| OCT MLA difference between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤1.86 mm2) | Tercile 2 (1.86–3.07 mm2) | Tercile 3 (>3.07 mm2) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.10 ± 0.86 | 29 | 2.58 ± 1.05 | 30 | 3.04 ± 0.84 | 0.0003 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 29 | 0.01 ± 0.01 | 30 | 0.01 ± 0.01 | 0.9564 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 29 | 0.03 ± 0.01 | 30 | 0.03 ± 0.01 | 0.3515 |
| Lipid, mm2 | 31 | 1.13 ± 0.60 | 29 | 0.98 ± 0.36 | 30 | 1.14 ± 0.44 | 0.2678 |
| Calcification, mm2 | 30 | 0.31 ± 0.20 | 29 | 0.24 ± 0.19 | 30 | 0.30 ± 0.23 | 0.3723 |
| OCT MLA difference between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤1.86 mm2) | Tercile 2 (1.86–3.07 mm2) | Tercile 3 (>3.07 mm2) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.10 ± 0.86 | 29 | 2.58 ± 1.05 | 30 | 3.04 ± 0.84 | 0.0003 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 29 | 0.01 ± 0.01 | 30 | 0.01 ± 0.01 | 0.9564 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 29 | 0.03 ± 0.01 | 30 | 0.03 ± 0.01 | 0.3515 |
| Lipid, mm2 | 31 | 1.13 ± 0.60 | 29 | 0.98 ± 0.36 | 30 | 1.14 ± 0.44 | 0.2678 |
| Calcification, mm2 | 30 | 0.31 ± 0.20 | 29 | 0.24 ± 0.19 | 30 | 0.30 ± 0.23 | 0.3723 |
Three-dimensional OCT-derived MLA at 12 M differences
| OCT MLA difference between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤1.86 mm2) | Tercile 2 (1.86–3.07 mm2) | Tercile 3 (>3.07 mm2) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.10 ± 0.86 | 29 | 2.58 ± 1.05 | 30 | 3.04 ± 0.84 | 0.0003 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 29 | 0.01 ± 0.01 | 30 | 0.01 ± 0.01 | 0.9564 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 29 | 0.03 ± 0.01 | 30 | 0.03 ± 0.01 | 0.3515 |
| Lipid, mm2 | 31 | 1.13 ± 0.60 | 29 | 0.98 ± 0.36 | 30 | 1.14 ± 0.44 | 0.2678 |
| Calcification, mm2 | 30 | 0.31 ± 0.20 | 29 | 0.24 ± 0.19 | 30 | 0.30 ± 0.23 | 0.3723 |
| OCT MLA difference between post-PCI and 12 M | |||||||
|---|---|---|---|---|---|---|---|
| Tercile 1 (≤1.86 mm2) | Tercile 2 (1.86–3.07 mm2) | Tercile 3 (>3.07 mm2) | P value | ||||
| n | Mean ± SD | n | Mean ± SD | n | Mean ± SD | Welch test | |
| Fibrous tissue, mm2 | 31 | 2.10 ± 0.86 | 29 | 2.58 ± 1.05 | 30 | 3.04 ± 0.84 | 0.0003 |
| Cholesterol crystal, mm2 | 31 | 0.01 ± 0.00 | 29 | 0.01 ± 0.01 | 30 | 0.01 ± 0.01 | 0.9564 |
| Macrophage, mm2 | 31 | 0.02 ± 0.01 | 29 | 0.03 ± 0.01 | 30 | 0.03 ± 0.01 | 0.3515 |
| Lipid, mm2 | 31 | 1.13 ± 0.60 | 29 | 0.98 ± 0.36 | 30 | 1.14 ± 0.44 | 0.2678 |
| Calcification, mm2 | 30 | 0.31 ± 0.20 | 29 | 0.24 ± 0.19 | 30 | 0.30 ± 0.23 | 0.3723 |
The cumulative curves (Figure 2A and B) show the OCT-derived MLA differences between post-PCI and 6 M and between post-PCI and 12 M, respectively. In both panels, the patients who experienced larger luminal loss (larger difference in MLAs = tercile 3) had larger percentages of fibrous tissue (green colour in the pie charts) at baseline.
Cumulative curves of the OCT-derived MLA (mm2) differences between post-PCI and 6 M (A) and between post-PCI and 12 M (B). Pie charts show the percentage of each tissue type as assessed at baseline before the implantation of the DREAMS 3G device in each of the terciles. In both panels, the patients who experience larger luminal loss (larger difference in MLAs = the tercile 3) had a larger percentage of fibrous tissue at baseline.
In Supplementary data online, Table S1C and D, the univariable analysis of percentage of fibrous tissue and patient’s demographics showed in all analyses that the fibrous tissue remains highly significant and none of the demographic factors comes close to the significance level.
Serial changes in coronary plaque composition in response to DREAMS 3G implantation
There was a reduction of 24.8% and 20.9% in lipid area, which were statistically significant (both P values < 0.001), between the OCT taken before the DREAMS 3G implantation (i.e. pre-PCI) and the 6 and 12 M follow-up as shown in Supplementary data online, Tables S3–S5. Conversely, the fibrous tissue increased by 48.4% and 36.0% between the OCT taken pre-PCI and 6 and 12 M follow-up. These were also statistically significant (both P values < 0.001) as seen in Supplementary data online, Tables S3 and S4.
In Table 3, a total of 83 patients had truly serial OCT imaging, and therefore, changes in plaque characteristics within the same patient could be evaluated by OCT before DREAMS 3G implantation (pre) and 6 and 12 M follow-up. Similar to the relative changes mentioned above, in this group of patients, the absolute differences showed a significant decrease in lipid area at 6 and 12 M, both P values < 0.001. There was also a significant increase in fibrous tissue area at 6 and 12 M, both P values < 0.001. Of note, the fibrous cap thickness overlying the lipid content increases significantly by 25.13% and 14.8% between pre-PCI and 6 and 12 M (see also Graphical Abstract).
Serial changes in plaque characteristics between before DREAMS 3G implantation (pre) and 6 and 12 M follow-up (n = 83)
| Pre | 6 M FU | 12 M FU | Absolute difference (6 M − Pre) | P value (6 M − Pre) | Absolute difference (12 M − Pre) | P value (12 M − Pre) | Absolute difference (12 − 6 M) | P value (12 − 6 M) | |
|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | Mean ± SD | Mean ± SD | Mean ± SD | t-test | Mean ± SD | t-test | Mean ± SD | t-test | |
| Fibrous tissue | |||||||||
| Area (mm2) | 2.54 ± 0.97 | 3.81 ± 1.22 | 3.45 ± 1.14 | 1.27 ± 1.00 | <0.0001 | 0.91 ± 0.96 | <0.0001 | 0.36 ± 0.98 | 0.0012 |
| Fibrous cap thickness (microns) | 291.21 ± 75.23 | 364.39 ± 81.99 | 334.32 ± 63.05 | 73.18 ± 6.76 | <0.0001 | 43.11 ± 12.18 | <0.0001 | 30.07 ± 18.94 | 0.0081 |
| Cholesterol crystal | |||||||||
| Area (mm2) | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.00 ± 0.01 | 0.4767 | 0.00 ± 0.01 | 0.8875 | 0.00 ± 0.01 | 0.3899 |
| Macrophage | |||||||||
| Area (mm2) | 0.03 ± 0.01 | 0.02 ± 0.01 | 0.02 ± 0.01 | 0.00 ± 0.01 | 0.3980 | 0.00 ± 0.01 | 0.6996 | 0.00 ± 0.01 | 0.6272 |
| Lipid | |||||||||
| Area (mm2) | 1.10 ± 0.47 | 0.83 ± 0.42 | 0.87 ± 0.39 | 0.27 ± 0.55 | <0.0001 | 0.24 ± 0.55 | 0.0002 | 0.04 ± 0.39 | 0.4128 |
| Calcification | |||||||||
| Area (mm2) | 0.29 ± 0.21 | 0.26 ± 0.19 | 0.28 ± 0.20 | 0.03 ± 0.20 | 0.2252 | 0.01 ± 0.20 | 0.6755 | 0.02 ± 0.17 | 0.3555 |
| Pre | 6 M FU | 12 M FU | Absolute difference (6 M − Pre) | P value (6 M − Pre) | Absolute difference (12 M − Pre) | P value (12 M − Pre) | Absolute difference (12 − 6 M) | P value (12 − 6 M) | |
|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | Mean ± SD | Mean ± SD | Mean ± SD | t-test | Mean ± SD | t-test | Mean ± SD | t-test | |
| Fibrous tissue | |||||||||
| Area (mm2) | 2.54 ± 0.97 | 3.81 ± 1.22 | 3.45 ± 1.14 | 1.27 ± 1.00 | <0.0001 | 0.91 ± 0.96 | <0.0001 | 0.36 ± 0.98 | 0.0012 |
| Fibrous cap thickness (microns) | 291.21 ± 75.23 | 364.39 ± 81.99 | 334.32 ± 63.05 | 73.18 ± 6.76 | <0.0001 | 43.11 ± 12.18 | <0.0001 | 30.07 ± 18.94 | 0.0081 |
| Cholesterol crystal | |||||||||
| Area (mm2) | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.00 ± 0.01 | 0.4767 | 0.00 ± 0.01 | 0.8875 | 0.00 ± 0.01 | 0.3899 |
| Macrophage | |||||||||
| Area (mm2) | 0.03 ± 0.01 | 0.02 ± 0.01 | 0.02 ± 0.01 | 0.00 ± 0.01 | 0.3980 | 0.00 ± 0.01 | 0.6996 | 0.00 ± 0.01 | 0.6272 |
| Lipid | |||||||||
| Area (mm2) | 1.10 ± 0.47 | 0.83 ± 0.42 | 0.87 ± 0.39 | 0.27 ± 0.55 | <0.0001 | 0.24 ± 0.55 | 0.0002 | 0.04 ± 0.39 | 0.4128 |
| Calcification | |||||||||
| Area (mm2) | 0.29 ± 0.21 | 0.26 ± 0.19 | 0.28 ± 0.20 | 0.03 ± 0.20 | 0.2252 | 0.01 ± 0.20 | 0.6755 | 0.02 ± 0.17 | 0.3555 |
Serial changes in plaque characteristics between before DREAMS 3G implantation (pre) and 6 and 12 M follow-up (n = 83)
| Pre | 6 M FU | 12 M FU | Absolute difference (6 M − Pre) | P value (6 M − Pre) | Absolute difference (12 M − Pre) | P value (12 M − Pre) | Absolute difference (12 − 6 M) | P value (12 − 6 M) | |
|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | Mean ± SD | Mean ± SD | Mean ± SD | t-test | Mean ± SD | t-test | Mean ± SD | t-test | |
| Fibrous tissue | |||||||||
| Area (mm2) | 2.54 ± 0.97 | 3.81 ± 1.22 | 3.45 ± 1.14 | 1.27 ± 1.00 | <0.0001 | 0.91 ± 0.96 | <0.0001 | 0.36 ± 0.98 | 0.0012 |
| Fibrous cap thickness (microns) | 291.21 ± 75.23 | 364.39 ± 81.99 | 334.32 ± 63.05 | 73.18 ± 6.76 | <0.0001 | 43.11 ± 12.18 | <0.0001 | 30.07 ± 18.94 | 0.0081 |
| Cholesterol crystal | |||||||||
| Area (mm2) | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.00 ± 0.01 | 0.4767 | 0.00 ± 0.01 | 0.8875 | 0.00 ± 0.01 | 0.3899 |
| Macrophage | |||||||||
| Area (mm2) | 0.03 ± 0.01 | 0.02 ± 0.01 | 0.02 ± 0.01 | 0.00 ± 0.01 | 0.3980 | 0.00 ± 0.01 | 0.6996 | 0.00 ± 0.01 | 0.6272 |
| Lipid | |||||||||
| Area (mm2) | 1.10 ± 0.47 | 0.83 ± 0.42 | 0.87 ± 0.39 | 0.27 ± 0.55 | <0.0001 | 0.24 ± 0.55 | 0.0002 | 0.04 ± 0.39 | 0.4128 |
| Calcification | |||||||||
| Area (mm2) | 0.29 ± 0.21 | 0.26 ± 0.19 | 0.28 ± 0.20 | 0.03 ± 0.20 | 0.2252 | 0.01 ± 0.20 | 0.6755 | 0.02 ± 0.17 | 0.3555 |
| Pre | 6 M FU | 12 M FU | Absolute difference (6 M − Pre) | P value (6 M − Pre) | Absolute difference (12 M − Pre) | P value (12 M − Pre) | Absolute difference (12 − 6 M) | P value (12 − 6 M) | |
|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | Mean ± SD | Mean ± SD | Mean ± SD | t-test | Mean ± SD | t-test | Mean ± SD | t-test | |
| Fibrous tissue | |||||||||
| Area (mm2) | 2.54 ± 0.97 | 3.81 ± 1.22 | 3.45 ± 1.14 | 1.27 ± 1.00 | <0.0001 | 0.91 ± 0.96 | <0.0001 | 0.36 ± 0.98 | 0.0012 |
| Fibrous cap thickness (microns) | 291.21 ± 75.23 | 364.39 ± 81.99 | 334.32 ± 63.05 | 73.18 ± 6.76 | <0.0001 | 43.11 ± 12.18 | <0.0001 | 30.07 ± 18.94 | 0.0081 |
| Cholesterol crystal | |||||||||
| Area (mm2) | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.00 ± 0.01 | 0.4767 | 0.00 ± 0.01 | 0.8875 | 0.00 ± 0.01 | 0.3899 |
| Macrophage | |||||||||
| Area (mm2) | 0.03 ± 0.01 | 0.02 ± 0.01 | 0.02 ± 0.01 | 0.00 ± 0.01 | 0.3980 | 0.00 ± 0.01 | 0.6996 | 0.00 ± 0.01 | 0.6272 |
| Lipid | |||||||||
| Area (mm2) | 1.10 ± 0.47 | 0.83 ± 0.42 | 0.87 ± 0.39 | 0.27 ± 0.55 | <0.0001 | 0.24 ± 0.55 | 0.0002 | 0.04 ± 0.39 | 0.4128 |
| Calcification | |||||||||
| Area (mm2) | 0.29 ± 0.21 | 0.26 ± 0.19 | 0.28 ± 0.20 | 0.03 ± 0.20 | 0.2252 | 0.01 ± 0.20 | 0.6755 | 0.02 ± 0.17 | 0.3555 |
Discussion
The main results can be summarized as follows: (i) the larger the fibrous tissue in the lesion that was treated with DREAMS 3G, the larger the luminal loss seen at 6 and 12 M, albeit the correlations were low. In particular, at 12 M, the results are consistent irrespective of whether the luminal loss is evaluated by the conventional angiographic LLL or by OCT-derived MLA. (ii) Following the implantation of DREAMS 3G, there was favourable healing of the vessel coronary wall as shown by a decrease in the lipid area that was present in the OCT before the implantation of the device in comparison with the OCT evaluations at 6 and 12 M. This was accompanied by an increase in fibrous tissue. Both changes indicate a positive impact of DREAMS 3G on plaque stabilization.
In this report, the observed changes in luminal dimensions between post-DREAMS 3G OCT and 6 and 12 M follow-up likely relates to the remodelling process that occurs within the vessel wall during the resorption and integration process of the device in the context of proper scaffold expansion after PCI in the three OCT groups at 12 M. This remodelling process can lead to either lumen loss (i.e. size reduction) or luminal gain (i.e. size increase).2 The lumen loss can be attributed to either neointimal formation or to late recoil. These two processes have been studied using the previous iteration of the device (i.e. Magmaris BIOSOLVE II study NCT01960504).9 Unlike this report, where an OCT AI analysis was used, the data analysis conducted in the Magmaris BIOSOLVE II study used visual assessment by expert readers for the tissue type characterization. It was concluded that ‘the extent of late scaffold recoil was dependent on the underlying plaque morphology and was the highest among fibrotic lesions.’ Even though there is an obvious methodological difference (human visual assessment vs. computer-aided AI analysis), in this report, which studied the most recent iteration of the device, we had similar conclusions that the larger the fibrous tissue at baseline, the larger the luminal loss at follow-up. It has been shown previously that fibrous tissue contributes to lumen loss. O’Brien et al.10 described that active transforming growth factor-beta (TGF-beta) plays a role in restenosis. Restenotic coronary lesions showed higher levels of immunodetectable beta ig-h3 protein, which is a TGF-beta–inducible gene h3, especially in areas of dense fibrous connective tissue. One must acknowledge that the most abundant tissue in coronary plaques is fibrous tissue, and therefore, some degree of restenosis is unavoidable. This information is nevertheless relevant when considering that future drug developments can target this pathway.
Devices such as DREAMS 3G do not only induce the remodelling process but also trigger the formation of healing tissue, which is also mostly fibrotic tissue. This fibrotic tissue is mostly built up as the most inner layer of the vessel wall, forming a new ‘intima,’ called neointima.2 In this report, we documented an increase in fibrous cap thickness, overlying the lipid tissue, from pre-PCI to 6 and 12 M, so this neointima is then ‘capping’ the underlying plaque. Brugaletta et al.11 showed that after implantation of the ABSORB bioresorbable vascular scaffold, as evaluated by visual assessment using OCT, a new thick inner fibrous cap was formed. They concluded that this new characteristic may contribute to plaque stability. As a corollary, in the PROSPECT ABSORB study,12 the concept of treating non–flow-limiting lesions with large plaque burden using ABSORB was tested. Results showed that the follow-up MLA was substantially enlarged and that the amount of lipid, as assessed by near-infrared spectroscopy (NIRS), was lower in the group treated with the device (maxLCBI4mm 62.0 vs. 268.8, P < 0.0001) compared with only medical treatment. Furthermore, these results were associated with favourable long-term clinical outcomes. In this report, we used AI methods for tissue characterization and found that there was significant increase of fibrous tissue, as seen in Figure 1 (note inner green layer overlying calcium) from baseline to 6 and 12 M. This is in accordance with the concept of plaque stabilization. More importantly, in our report, we also noted a significant decrease in lipid content, which is likely in part related to the systemic medical treatment that patients received, but it could also be partly attributed to the effects of the limus drug. These latter findings are in line with the PROSPECT ABSORB study results.
Whether bioresorbable scaffolds or stents can be used pre-emptively to treat non–flow-limiting lesions is being currently tested in the following studies: PREVENT (NCT02316886), VULNERABLE (NCT05599061), INTER-CLIMA (NCT05027984), and COMBINE-INTERVENE (NCT05333068). The intense ongoing research suggests that there was already enough evidence of positive plaque changes induced by bioresorbable scaffolds to plan those studies that are now again confirmed by this study.
Limitations
There are several limitations to be noted: (i) this was a single-arm study without a comparator; (ii) OCT, due to its limited penetration, can only assess the most superficial layers of the coronary vessel wall; (iii) the OCTPlus software, thus, has limited the analysis of deeper parts of the plaque, and in the original validation report, the characterization of macrophages was suboptimal; (iv) newer OCT technology combined with NIRS assessment is being introduced with improved lipid detection and, thus, should be preferred over OCT standalone catheters; and (v) still small sample size to provide robust evidence with the analyses.
Conclusion
The larger the fibrous tissue in the lesion, the larger the luminal loss seen at 6 and 12 M; following the implantation of DREAMS 3G, there is a favourable healing of the vessel coronary wall as shown by a decrease in the lipid area and by an increase in fibrous tissue.
Supplementary data
Supplementary data are available at European Heart Journal - Cardiovascular Imaging online.
Funding
The BIOMAG-I trial was sponsored by Biotronik AG.
Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.
References
Author notes
Conflict of interest: H.M.G.-G., R.W., G.D.M., M.G., S.B., A.K., and M.B.L. were core laboratory members, and the remaining authors were investigators of the trial. H.M.G.-G. has grants or contracts from Medtronic, Biotronik, Abbott, Neovasc, Corflow, Alucentbio, Philips, and Chiesi (paid to institution), received consulting fees from Boston Scientific and ACIST, and participates in DSMB/advisory board of the VIVID study. R.W. has grants or contracts from Amgen, Biotronik, Boston Scientific, Medtronic, and Philips IGT; received consulting fees from Abbott Vascular, Biotronik, Boston Scientific, Cordis, Medtronic, Philips IGT, Pi-Cardia Ltd, Swiss Interventional Systems/SIS Medical AG, Transmural Systems Inc, and Venous MedTech; received honoraria from AstraZeneca; participates in DSMB/advisory boards of Abbott Vascular, Boston Scientific, Medtronic, Philips IGT, and Pi-Cardia Ltd; and is an investor in MedAlliance and Transmural Systems Inc. J.T. reports grants and contracts from Abbott paid to his institution, speaker honoraria and support for attending meetings from Biotronik, and is an associate editor of Cardiovascular Biologics and Regenerative Medicine and Frontiers in Cardiovascular Medicine. J.E. reports personal fees/speaker honoraria from Abbott, Boston Scientific, Philips, and Shockwave, patents from Shared, and participation in advisory boards of Abbott and Phillips. The institution of J.F.I. receives grants or contracts from Terumo Corp, Biosensors, Concept Medical, Biotronik, Abbott Vascular, and Philips Volcano. J.F.I. reports consulting fees from Biotronik, Medtronic, Cordis, Terumo Corp., and ReCor Medical; speaker fees/honoraria from Terumo Corp, Biosensors, Medalliance, OrbusNeich, Concept Medical, Bristol Myers Squibb/Pfizer, Novartis, Cordis, AstraZeneca, and Philips Volcano; and support to attend meetings from Biotronik and Amgen. The institution of J.B. receives grants or contracts from Shockwave IVLS. J.B. receives consulting fees from Biotronik AG and Boston Scientific and speaker fees/honoraria from Biotronik AG, Boston Scientific, and Abbott Vascular, participates in the DSMB of Boston Scientific, and has a leadership or fiduciary role for Biotronik. G.G.T. reports consulting fees from Biotronik, Medtronic, Abbott, and Terumo and honoraria from Biotronik, Medtronic, Abbott, and Terumo. M.J. reports grant support from Boston Scientific, Cardiac Dimensions, Edwards Lifesciences, and Infraredx; consulting fees from AlchiMedics SAS, Biotronik, TriCares, Veryan, and Shockwave; speaker fees/honoraria from Abbott Vascular, Biotronik, Boston Scientific, Edwards Lifesciences, Cardiac Dimensions, AstraZeneca, Recor Medical, and Shockwave; travel support from SIS Medical, Edwards Lifesciences, Boston Scientific, and Cardiac Dimensions; and participation in Steering Committees of Biotronik and Edwards Lifesciences. R.T. reports lecture fees from Biotronik. M.W. reports speaker honoraria and conference attendance support from Biotronik. G.O. reports lecturer honoraria from Abbott Vascular, Biotronik, and Cordis and is a DSMB member of the SCIENCE trial and a CEC-member of the BIOFREEDOM STEMI trial. M.H. reports grants/contracts from Biotronik, Cardiac Dimensions, OrbusNeich, and Philips; consulting fees from Biotronik, Cardiac Dimensions, Shockwave Medical, and OrbusNeich; honoraria/speaker fees from Biotronik, Cardiac Dimensions, Shockwave Medical, OrbusNeich, and Philips; and support to attend meetings/travel support from Biotronik, is a steering committee member of the BIOSOLVE and BIOMAG trials, and is a past president of EAPCI. All other authors have no conflict of interest to declare.
