Author + information
- Received December 22, 2015
- Accepted January 15, 2016
- Published online April 25, 2016.
- Shimpei Nakatani, MDa,
- Yuki Ishibashi, MD, PhDa,
- Yohei Sotomi, MDb,
- Laura Perkins, DVM, PhDc,
- Jeroen Eggermont, PhDd,
- Maik J. Grundeken, MDb,
- Jouke Dijkstra, PhDd,
- Richard Rapoza, PhDc,
- Renu Virmani, MDe,
- Patrick W. Serruys, MD, PhDf and
- Yoshinobu Onuma, MD, PhDa,∗ ()
- aThoraxCenter, Erasmus Medical Center, Rotterdam, the Netherlands
- bAcademic Medical Center, University of Amsterdam, Amsterdam, the Netherlands
- cAbbott Vascular, Santa Clara, California
- dLeiden University Medical Center, Leiden, the Netherlands
- eCVPath Institute, Gaithersburg, Maryland
- fInternational Centre for Circulatory Health, National Heart and Lung Institute, Imperial College London, London, United Kingdom
- ↵∗Reprint requests and correspondence:
Dr. Yoshinobu Onuma, Thoraxcenter, Ba-583, ’s Gravendijkwal 230, 3015 CE Rotterdam, the Netherlands.
Objectives The aim of the present study was to investigate the relationship between the integration process and luminal enlargement with the support of light intensity (LI) analysis on optical coherence tomography (OCT), echogenicity analysis on intravascular ultrasound, and histology up to 4 years in a porcine model.
Background In pre-clinical and clinical studies, late luminal enlargement has been demonstrated at long-term follow-up after everolimus-eluting poly-l-lactic acid coronary scaffold implantation. However, the time relationship and the mechanistic association with the integration process are still unclear.
Methods Seventy-three nonatherosclerotic swine that received 112 Absorb scaffolds were evaluated in vivo by OCT, intravascular ultrasound, and post-mortem histomorphometry at 3, 6, 12, 18, 24, 30, 36, 42, and 48 months.
Results The normalized LI, which is the signal densitometry on OCT of a polymeric strut core normalized by the vicinal neointima, was able to differentiate the degree of connective tissue infiltration inside the strut cores. Luminal enlargement was a biphasic process at 6 to 18 months and at 30 to 42 months. The latter phase occurred with vessel wall thinning and coincided with the advance integration process demonstrated by the steep change in normalized LI (0.26 [interquartile range (IQR): 0.20 to 0.32] at 30 months versus 0.68 [IQR: 0.58 to 0.83] at 42 months, p < 0.001).
Conclusions In this pre-clinical model, late luminal enlargement relates to strut integration into the arterial wall. Quantitative LI analysis on OCT could be used as a surrogate method for monitoring the integration process of poly-l-lactic acid scaffolds, which could provide insight and understanding on the imaging-related characteristics of the bioresorption process of polylactide scaffolds in human.
- biodegradable polymer
- bioresorbable scaffold
- coronary intervention
- intravascular ultrasound
- optical coherence tomography
As an alternative approach to metal drug-eluting stents, fully bioresorbable polymeric drug-eluting scaffolds provide transient vessel support with drug-delivery capability. As the scaffold begins to resorb, the vessel is no longer caged, and therefore luminal area as well as vessel area could increase simultaneously without creating evagination (1–5). The everolimus-eluting scaffold (Absorb; Abbott Vascular, Santa Clara, California) consists of a semicrystalline poly-l-lactic acid (PLLA) backbone coated by a thin amorphous layer of poly-d,l-lactic acid containing the antiproliferative agent everolimus. After implantation, the polylactide strut progressively degrades by hydrolysis, and its molecular weight starts to decrease from its initial molecular weight of around 100 kDa (molecular weight loss) (6). The PLLA molecules remain at the implanted site until the polymeric chains become small enough to diffuse from the site into the surrounding tissue (mass loss). As small oligomers or monomers gradually leave the site, there is progressive replacement by a provisional matrix initially composed of a milieu of extracellular matrix components. This initially acellular provisional matrix is gradually cellularized with connective tissues, and the struts and footprints eventually become fully integrated into the surrounding neointimal tissue of the vessel wall (6,7).
It is well-established that the scaffolding efficacy of the device is related to the timing of molecular weight reduction and the loss of mechanical integrity (8). However, at a late phase, it is still unclear whether the integration of strut footprints is associated with the late luminal enlargement. In the pre-clinical assessment of fully bioresorbable scaffolds, it is therefore important to assess the processes of molecular weight loss and integration in vivo. In humans, intravascular imaging has been used in vivo as a surrogate marker to understand the bioresorption and integration process, but the correlation between the surrogate assessment and the true bioresorption process needs to be established.
On intravascular ultrasound (IVUS), quantitative echogenicity has been demonstrated to correlate with the molecular weight of PLLA (9). On optical coherence tomography (OCT), the visual categorizations of strut appearance have previously been demonstrated to correlate with the integration process (10). However, this visual categorization was limited by its moderate reproducibility (k = 0.58). Recently, log-transformed optical coherence tomographic signal measurement (light intensity analysis) of strut cores was introduced as a feasible and reproducible method to assess the degree of strut integration after scaffold implantation (11). In humans, the median intensity value of strut cores increased significantly at 24 months and kept increasing up to 36 months, and most of pre-existing struts were indiscernible at 60 months on OCT (Figure 1). It was hypothesized that light reflectivity is correlated with connective tissue infiltration of the strut cores. However, this hypothetical correlation between light intensity and histological changes has so far not been demonstrated with strut histology-matched light intensity.
The aim of this study was to demonstrate the relationship between light intensity and histological changes with regard to strut integration at 3, 6, 12, 18, 24, 30, 36, 42, and 48 months in a porcine coronary artery model. In addition, IVUS echogenicity analysis was also assessed to monitor early changes in molecular weight.
The present study was conducted from 2009 to 2013. Eight nonatherosclerotic juvenile domestic crossbred farm swine and 65 Yucatan mini-swine underwent Absorb scaffold implantation with a targeted balloon-to-artery ratio of 1.0:1.1. Each animal received a single everolimus-eluting scaffold (Absorb; 3.0 × 18 mm for 1, 3, and 6 months and 3.0 × 12 mm for 12 to 48 months) in 1 or 2 main coronary arteries. The Absorb scaffold used in the present study is the same as the device used in cohort B of the ABSORB clinical trials. Seventy-three pigs with 112 Absorb scaffolds implanted were examined by OCT at baseline and anesthetized at the designated endpoints with the optical coherence tomographic and IVUS examinations at 3 months (n = 10 Absorb scaffolds in 8 farm swine), as well as 6 (n = 10 Absorb scaffolds in 8 Yucatan mini-swine), 12 (n = 11 in 7), 18 (n = 12 in 7), 24 (n = 12 in 7), 30 (n = 12 in 8), 36 (n = 12 in 8), 42 (n = 13 in 8), and 48 (n = 20 in 12) months (Figure 2).
Optical coherence tomographic acquisition was executed by the frequency-domain optical coherence tomographic imaging system (C7 Dragonfly or Dragonfly Duo, St. Jude Medical, St. Paul, Minnesota), with the exception of a few early investigations performed with the time-domain (TD) OCT imaging system in 2 animals. IVUS runs were acquired with 40-MHz mechanical systems, using Galaxy version 2.02 or iLab (Boston Scientific, Natick, Massachusetts). After performing intravascular imaging studies (OCT and IVUS), animals were humanely euthanized. Hearts were excised and pressure perfused with 0.9% saline solution, followed by pressure perfusion fixation with 10% neutral buffered formalin overnight in preparation for histology. Embedded arteries were divided into a minimum of 3 blocks representing the proximal, medial, and distal regions of the scaffold. Duplicate 4- to 6-mm sections from each of the 3 blocks were collected, and each was stained with Movat’s pentachrome and hematoxylin and eosin for evaluation by light microscopy.
Experimental studies received protocol approval from the Institutional Animal Care and Use Committee and were conducted in accordance with American Heart Association guidelines for pre-clinical research and the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, 1996).
Histological categorization of strut and strut footprint
On histology, the integration grade of each strut was semiquantitatively classified into 6 groups according to the connective tissue composition within a strut core in Movat’s pentachrome–stained sections: 0) acellular, no connective tissue composition; 1) hypointegration, 1% to 10% connective tissue composition; 2) low integration, 11% to 25% connective tissue composition; 3) moderate integration, 26% to 50% connective tissue composition; 4) moderate to high integration, 51% to 75% connective tissue composition; and 5) high integration, >75% connective tissue composition. Figure 3 shows examples of strut appearance on the basis of histological classification of integration (10).
Optical coherence tomographic quantitative measurements
Cross sections with unacceptable image quality for measurements due to suboptimal flushing (4 scaffolds at 48 months) were excluded from the quantitative analysis and light intensity analysis. Quantitative measurements (scaffold area, incomplete scaffold apposition area, and neointimal area) were performed in the scaffolded segment and periscaffolded segments (within 5 mm proximal and distal to the stent edge) at 1-mm intervals according to previously published methods (12). As the porcine coronary artery grows, the luminal area in the periscaffolded segment (reference luminal area) is enlarged (2,13). To compare the time-dependent changes, the normalized scaffold and luminal area was calculated as the ratio of the scaffold and lumen to the reference luminal area (2).
Optical coherence tomographic light intensity analysis of strut cores
Because the light intensity values vary between TD and frequency-domain OCT, images obtained by TD OCT (1 scaffold at 3 months and 2 at 12 months) were excluded from the light intensity analysis. Pull-backs with high intraluminal signal intensity with shadows due to suboptimal flushing were excluded from the light intensity analysis (1 scaffold at 6 months, 3 at 36 months, and 4 at 48 months).
Light intensity analysis of the strut cores was performed using dedicated software (QCU-CMS version 4.69 [research version]; Leiden, the Netherlands). Raw images with 16 bits in original polar format were used to ensure that interpolation, dynamic range compression, or other image processing did not alter the signal and bias the analysis. The contours of strut cores were delineated manually with “box-shape” by visual inspection in Cartesian images (Online Figure 1) (11). According to the strut contours, the median light intensity values were computed with the subtraction of 2 pixels inside of the manual strut core contour by the software automatically, as described previously (11).
To minimize bias in light intensity measurement caused by the variation in optical signal due to eccentric location of the optical coherence tomographic catheter or the uneven distribution of neointima on top of the strut, the light intensity values of strut cores were normalized by the median light intensity value of the interstrut neointima in the vicinity (bilaterally 22.5° wide originating from the strut center) of each strut (referred to as the normalized light intensity of strut cores) (Online Figure 1).
The light intensity analysis was performed in cross sections at intervals of 1 mm and additionally in the OCT-imaged struts matched with histology. Figure 3 shows examples of strut appearance on the basis of normalized light intensity of strut cores on OCT.
Matching of struts on histology and OCT
To correlate strut cores and histology at a strut level, 1 observer (Y.O.) aware of the histological image selected the matched OCT-histology cross sections at each time point using landmarks such as side branches, metallic radiopaque markers, or the appearance of neointima and media (Online Appendix). In the selected cross sections of OCT and histology, individual struts were further matched by identifying struts with similar angular orientation and peculiar appearance of struts (strut-level matching) (Online Figure 2). To ensure accurate identification of the strut areas by OCT and histology, especially at later time points (36, 42, and 48 months), when struts are more integrated and thus more difficult to identify, strut contours drawn on optical coherence tomographic analysis were superimposed on histology to match the strut core regions and to calculate the percentage of infiltration by connective tissue normalized for the strut footprint area.
These matched strut images were sent to the independent optical coherence tomographic analyst (S.N.) and the pathologist (L.P.). The normalized light intensity of the OCT-imaged strut and the histological categorization of integration of this strut into the surrounding neointimal tissue was blinded to the observer who had performed the matching.
Echogenicity analysis in IVUS grayscale
IVUS quantitative analysis (vessel, scaffold, and luminal area) was performed in 0.5-mm intervals in the scaffolded segment and periscaffolded segments (defined by a length 5 mm proximal and distal to the stent edge). The normalized vessel area was calculated as the ratio of the vessel to the reference vessel area, and the normalized scaffold and luminal area was calculated as the ratio to the reference luminal area.
According to the contours of lumen and vessel, the dedicated software (QCU-CMS version 4.69 [research version]) calculates the areas of 5 tissue types categorized by using the median brightness of the adventitia as a reference in the lumen-vessel compartments automatically: 1) hypoechogenic, 2) hyperechogenic, 3) calcified, 4) upper echogenic, and 5) unknown (9).
All statistical analyses were performed using the SPSS version 23.0 (SPSS, Chicago, Illinois). Normality of distributions was tested with the Kolmogorov-Smirnov statistic. Continuous variables are presented as mean ± SD or median (interquartile range [IQR]), as indicated in the tables. Generalized estimating equations modeling was performed to take into an account the clustered nature of >1 scaffold analyzed from the same pig, which might result in unknown correlations among measurements within these scaffold clusters. Categorical variables are presented as absolute values and percentages. Relations between histological categories and normalized light intensity of strut cores were analyzed by Spearman’s rank-order correlation. Bayesian analysis of normalized light intensity included estimation of area under the receiver-operating characteristic, curve with the optimal cutoff value for the detection of the onset of “strut integration” into the neointimal surrounding and the shift to “moderate to high filtration into the strut core,” with associated sensitivity and specificity. A modification of the classification of Swets (13) was used to classify diagnostic efficiency of normalized light intensity according to the values of the area under the curve as low (<0.70), moderate (0.70 to 0.90), or high (>0.90).
In total, 336 histological cross sections and 112 optical coherence tomographic pull-backs were available (Figure 2). After excluding 4 optical coherence tomographic pull-backs because of impaired image quality, 108 pull-backs were analyzed for quantitative measurements. Of 108 pull-backs, 7 were excluded from the light intensity analysis for technical reasons (3 because of TD OCT and 4 because of inappropriate image quality for light intensity analysis), resulting in 101 OCT pull-backs available for light intensity analysis. After excluding 33 histological cross sections because of lack of analyzable OCT pull-backs, 2,979 struts were identified in 303 histological cross sections. Using sectorial approximate locations and landmarks such as metallic radiopaque markers, side branches, and neointimal formation, a total of 2,455 struts were matched between OCT and histology.
Light intensity analysis over time
The light intensity analysis in 2,455 matched struts over time is summarized in Table 1 and Figure 4A. There was no significant difference in the normalized light intensity of strut cores until 18 months after scaffold implantation: 0.14 (IQR: 0.11 to 0.22) at 3 months, 0.12 (IQR: 0.08 to 0.17) at 6 months, 0.15 (IQR: 0.10 to 0.21) at 12 months, and 0.15 (IQR: 0.11 to 0.19) at 18 months. The normalized light intensity increased gradually between 18 and 30 months (0.19 [IQR: 0.14 to 0.26] at 24 months; p = 0.001 vs. 18 months; 0.26 [IQR: 0.20 to 0.32] at 30 months; p < 0.001 vs. 24 months). After 30 months, the normalized light intensity surged until 42 months: 0.48 (IQR: 0.37 to 0.62) at 36 months (p < 0.001 vs. 30 months) and 0.68 (IQR: 0.58 to 0.83) at 42 months (p < 0.001 vs. 36 months) and was close to 1.0 at 48 months (0.76 [IQR: 0.61 to 0.90]; p = 1.00 vs. 42 months), suggesting that the strut cores were completely integrated into the surrounding tissue.
The categorization of histological integration of 2,455 matched struts over time is summarized in Table 1 and Figure 4B. From 3 to 12 months, the struts were completely separated from the lumen by a thin, fibromuscular neointima and had well-defined, squared appearances. Most of the struts were classified as acellular (96.5% at 3 months, 90.6% at 6 months, and 88.6% at 12 months). From 18 to 30 months, the strut footprints maintained their discrete borders but began to appear blue with Movat’s pentachrome. The percentage of acellular struts decreased gradually (82.6% at 18 months, 79.3% at 24 months, and 75.5% at 30 months). Six months later (at 36 months), the percentage of acellular struts has abruptly decreased to 13.6%, and the strut footprints were generally colonized by connective tissue. These strut footprints further progressed up to the point at which the strut footprint was only poorly discernible. Twenty-nine percent of matched struts were classified as grade ≥3 (>25% of cellularization and connective tissue composition). After 42 months, there was no strut without integration, and the rate of highly integrated struts increased gradually up to 48 months. At 48 months, 61.1% of matched struts showed >50% of connective tissue composition.
Comparison between optical coherence tomographic light intensity analysis and histological classification
The correlation between the normalized light intensity on OCT and histological categorization of 2,455 matched struts is summarized in Table 1 and Figure 4C. There were significant differences among the histological categories with respect to normalized light intensity: 0.16 (IQR: 0.11 to 0.23) in category 0 (acellular), 0.25 (IQR: 0.17 to 0.38) in category 1 (1% to 10%), 0.43 (IQR: 0.32 to 0.61) in category 2 (11% to 25%), 0.65 (IQR: 0.54 to 0.77) in category 3 (26% to 50%), 0.72 (IQR: 0.58 to 0.88) in category 4 (51% to 75%), and 0.80 (IQR: 0.67 to 0.96) in category 5 (>75%) (p < 0.01). In paired comparison, normalized light intensity was significantly different between categories 0 and 3, whereas there were no significant differences between categories 3 and 5 (Table 1, Figure 4C). In Spearman rank-order correlation analysis, there was a significant positive correlation between histological category of integration and normalized light intensity of strut cores (r = 0.791; p < 0.01).
Receiver-operating characteristic curves demonstrated that the diagnostic efficiency of the normalized light intensity of strut cores for the detection of the onset of the “strut integration” into the surrounding neointima (histological category ≥1, area under the curve 0.924, cutoff value 0.326) and for the detection of the shift to “moderate to high filtration into the strut core” (histological category ≥3, area under the curve 0.965, cutoff value 0.413) were excellent (Online Figure 3).
Quantitative measurements on OCT
The optical coherence tomographic quantitative measurements and light intensity analysis of strut cores performed in 1-mm interval cross sections are summarized in Table 2 and Online Figure 4. At baseline, there were no significant difference in scaffold and flow area among all time-point groups (scaffold area, p = 0.113; flow area, p = 0.124). Compared with 3 months (0.54 ± 0.13), the normalized luminal area was significantly larger after 12 months (p < 0.05). The neointimal thickness on top of struts decreased significantly at 48 compared with 3 months (218 ± 73 μm at 3 months vs. 140 ± 19 μm at 48 months, p < 0.001).
The light intensity analysis of strut core in 1-mm intervals was in line with the matched struts, showing a surge between 30 and 42 months.
Quantitative measurements and echogenicity changes on IVUS
The IVUS quantitative measurements and echogenicity analysis on the basis of 0.5-mm-interval cross sections are summarized in Table 3. The area of vessel wall (including media and neointima) did not change over time, whereas the thickness of vessel wall significantly decreased at 48 months (0.33 ± 0.06 mm) compared with 3 months (0.41 ± 0.05 mm) (p < 0.001).
The hyperechogenicity plus upper echogenicity area, a surrogate parameter for the molecular weight of PLLA (9), significantly decreased during the first 24 months (0.42 [IQR: 0.29 to 0.56] at 24 months vs. 1.10 [IQR: 1.07 to 1.48] at 3 months, p < 0.001).
The main findings of the present analysis are as follows (Figures 5 and 6A). First, following polymeric scaffold implantation, the normalized light intensity of strut cores did not change between 3 and 18 months; thereafter it gradually increased from 18 to 30 months, but the change in light intensity significantly surged between 30 and 42 months. Second, the histological evaluation showed that the frequency of acellular strut cores (absence of tissue infiltration) remained the same up to 12 months and then gradually decreased up to 30 months and began to dramatically change between 30 and 42 months, virtually disappearing at 48 months. Third, in matched struts for OCT and histology, the normalized light intensity was able to differentiate the degree and intensity of connective tissue infiltration inside the strut cores. Using a cutoff value of 0.413, the normalized light intensity could detect the advanced process of integration of the strut cores (histological grade ≥3) with an accuracy of 0.922. Fourth, the optical coherence tomographic quantitative measurements demonstrated that the scaffold and luminal area normalized to the reference luminal area increased between 30 and 36 months. The neointimal thickness on top of the struts and the black core strut thickness (if visible) decreased concurrently between 36 and 42 months. Fifth, the IVUS quantitative measurements were in line with the results of optical coherence tomographic measurements but were not as discrete as those of OCT. The area of neointima plus media in absolute value did not show a significant change over time, whereas there was a thinning of the vessel wall (thickness between lumen and vessel) due to the late luminal enlargement. Finally, the upper echogenicity plus hyperechogenicity area steadily decreased during the first 24 months, reflecting the early loss of molecular weight.
Luminal enlargement in the first 2 years: Loss of mechanical integrity and adaptive expansion to the inherent arterial growth of the model
One of the most important findings of the present study is that there are 2 phases of vessel and luminal growth in 4 years after implantation of the Absorb bioresorbable scaffold. On IVUS and OCT, the first significant luminal and scaffold enlargement occurs between 6 to 18 months, whereas the second occurs between 30 and 42 months (Figure 6A). Of note, these luminal gains were not observed following permanent metallic stent implantation, because of the permanent mechanical caging (2,14).
The first growth of the scaffold and luminal area seems to be related to the loss of mechanical integrity and the natural growth of coronary artery. The Absorb device loses its mechanical strength 6 months after implantation (15,16). On OCT, at 3 months, luminal and scaffold area is normalized by the reference lumen smaller than the reference area (0.54), but after 6 months, and coinciding with the loss of mechanical integrity, the scaffold and lumen started to follow the enlargement of reference area (normalized lumen area at 12 to 30 months: 0.72 to 0.77). The current animal study was performed in the Yucatan pig model, which is known to have inherent coronary arterial growth over time (2,17).
Second-phase luminal enlargement: Integration of strut core
The second phase of vessel remodeling, which occurred between 30 and 42 months, is presumably due to the integration process. During this period, the normalized luminal area further increased from 0.77 to 0.93. On histology, rapid integration of the struts in the surrounding neointima was observed. The frequency of moderately to highly integrated struts (histological grade ≥3) increased from 0.7% to 94%, which was clearly illustrated by the surge of light intensity inside the strut core void on OCT.
During the infiltration of connective tissue into the provisional matrix, the thickness of the neointima on top of struts decreased. It is still unclear whether maturation from a provisional matrix to collagen-rich connective tissue may further influence mechanotransduction by improving the transmission of mechanical signal (18). The stimulation of initial smooth muscle cells can induce matrix metalloproteinase release, which plays a key role in matrix deposition and reorganization by collagen type I deposition that leads to negative arterial remodeling and potential neointimal shrinkage (18–21).
The present analysis confirms that 1) OCT light intensity correlated with the integration process and that 2) IVUS echogenicity can detect early changes in molecular weight. By using IVUS echogenicity, molecular weight loss could be monitored, and the timing of loss of mechanical integrity could be predicted. Without any reference to gel permeate chromatography in vivo, it would be clinically useful to judge by echogenicity the degree of biodegradation related to the loss of mechanical strength and subsequent restoration of mechanotransduction.
Optical coherence tomographic light intensity is more sensitive for the integration process, which is associated with the thinning of the vessel wall as well as very late luminal enlargement. By using the classification of strut (using a proposed cutoff of 0.413 for infiltration greater than 25%), clinicians can assess the stage of integration that heralds late luminal enlargement.
When the quantitative light intensity analysis is applied to the human data obtained from ABSORB cohort B (Figure 6B), the light intensity surge had not yet been detected at 36 months (the average of normalized light intensity was 0.22 in 161 sequentially matched struts). This suggests that the integration process in humans is somewhat slower than that in animals and that either late luminal enlargement could commence later than 3 years or that the underlying plaque fails to allow positive remodeling because it is rich in type I collagen.
The influence of the underlying plaque on the integration process remains unclear. In previous clinical studies, the Absorb scaffold was associated with a decrease of the plaque area on IVUS in the long term when compared with post-implantation (22). It was questioned whether this is due to the disappearance of struts or the real reduction of atherosclerotic plaque. In the present animal study, the area of neointima plus media remained unchanged throughout the 4-year follow-up period, whereas vessel wall thickness was reduced as a result of the very late luminal expansion or conversion of type III collagen to type I with cross-linking of collagen (Figure 6A). This suggests that the integration of the struts does not significantly affect the area of neointima plus media but induces the expansion of the lumen and the vessel, which could be mediated by the mechanotransduction and the restoration of shear stress (1).
From visual categorization to quantitative measurement of light intensity
In the present analysis, we used quantitative assessment of light intensity, which is more reproducible than visual assessment. The method has been applied to optical coherence tomographic analysis of struts in humans. The intraclass correlation coefficient for interobserver variability was as high as 0.91 (11).
The present OCT analysis using the second iteration of the Absorb device showed that using a cutoff of 0.413, the measured light intensity can differentiate moderately to highly integrated struts from low-integrated struts with an accuracy of 0.922. This integration progressively occurs between 30 and 42 months after implantation, which seems to be the critical timing for maximal vessel remodeling.
The other possible optical coherence tomographic methods to evaluate the bioresorption process include the refractive index or dispersion to differentiate the provisional matrix and polymer, or birefringence analysis using polarization-sensitive OCT (23). In terms of quantitative measurement, OCT is presumably more close to the in vivo dimension than histomorphometric measurement, because of the absence of vessel shrinkage from histological preparation (24).
On histology, we used a semiquantitative scale to classify the degree of integration. Theoretically, the infiltration rate of connective tissues into the strut footprint could also be quantifiable on histology; however, because of tissue shrinkage due to formalin fixation and dehydration with tissue processing, it is unclear which segment of the vessel may be most affected. It is likely to be proteoglycans, which are water rich, and dehydration affects water-rich areas more than other regions. It was also possible that the strut footprint could deform during the histological processing; after 4 years, the original area occupied by the polymeric strut is poorly discernible by light microscopy, which made quantitative measurement on histology difficult. The use of time-of-flight secondary iron mass spectroscopic analysis might facilitate future pre-clinical studies to quantify the degrees of degradation and integration (25).
The present study was performed in a healthy porcine model without atherosclerosis, so the generalizability of the concepts to human is therefore limited. The methodology for normalization is not necessarily applicable to humans, because homogeneity of neointima is different between human and porcine models. The expected complex relationships among plaque burden, mechanotransduction, luminal enlargement and vascular remodeling, and natural coronary artery growth are not yet elucidated in the current models, although the initial molecular weight loss due to depolymerization is not different between human and pigs, because the process is purely chemical via hydrolysis.
From a regulatory perspective, it is mandatory to investigate the process of bioresorption of each fully bioresorbable scaffold, because the rate can vary according to the manufacturing process. The current assessment for monitoring bioresorption and integration process by echogenicity on IVUS and light intensity on OCT could be applied only to scaffolds made of PLLA, with similar molecular weights and similar manufacturing processes, but it could not be applied to other scaffolds made of different materials (e.g., magnesium), with different molecular weights (e.g., Igaki-Tamai) or different manufacturing processes (e.g., Mirage).
In this pre-clinical model, luminal enlargement is a biphasic process in which the latter phase likely relates, at least in part, to strut integration of the Absorb scaffold into the arterial wall. The quantitative light intensity analysis of strut cores on OCT could be used as a surrogate method for monitoring matrix infiltration and integration of collagen-rich connective tissue within the polymeric struts that coincide with the time of late luminal enlargement. These intravascular methods may provide insight and understanding of the imaging-related characteristics of the bioresorption process of various polylactide scaffolds in human.
WHAT IS KNOWN? Visual categorizations of strut appearance of bioresorbable scaffolds on OCT have previously been demonstrated to correlate with the integration process with limited reproducibility. Light intensity analysis of strut cores has been introduced as a feasible and reproducible method for a quantitative assessment of strut integration after scaffold implantation.
WHAT IS NEW? In a porcine pre-clinical model, IVUS echogenicity analysis and light intensity analysis on OCT were well correlated with the depolymerization process of the strut and the integration process after the complete bioresorption, respectively. The late luminal enlargement observed between 3 and 4 years seems to be related to strut integration.
WHAT IS NEXT? Further studies are needed to evaluate the generalizability of the concepts to human and to elucidate the expected complex relationships among plaque burden, mechanotransduction, luminal enlargement and vascular remodeling, and natural coronary artery growth.
For additional methods and supplemental figures, please see the online version of this article.
This study was funded by Abbott Vascular. Drs. Perkins and Rapoza are full-time employees of Abbott Vascular. Drs. Serruys and Onuma are members of the advisory board of Abbott Vascular. Dr Virmani receives research support from Abbott Vascular, BioSensors International, Biotronik, Boston Scientific, Medtronic, MicroPort Medical, OrbusNeich Medical, SINO Medical Technology, and Terumo Corporation; has speaking engagements with Merck; receives honoraria from Abbott Vascular, Boston Scientific, Lutonix, Medtronic, and Terumo Corporation; and is a consultant for 480 Biomedical, Abbott Vascular, Medtronic, and W.L. Gore. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- interquartile range
- intravascular ultrasound
- optical coherence tomography
- Received December 22, 2015.
- Accepted January 15, 2016.
- American College of Cardiology Foundation
- Serruys P.W.,
- Garcia-Garcia H.M.,
- Onuma Y.
- Lane J.P.,
- Perkins L.E.,
- Sheehy A.J.,
- et al.
- Onuma Y.,
- Dudek D.,
- Thuesen L.,
- et al.
- Radu M.D.,
- Raber L.,
- Kalesan B.,
- et al.
- Otsuka F.,
- Pacheco E.,
- Perkins L.E.,
- et al.
- Onuma Y.,
- Serruys P.W.,
- Perkins L.E.,
- et al.
- Nakatani S.,
- Onuma Y.,
- Ishibashi Y.,
- et al.
- Serruys P.W.,
- Onuma Y.,
- Ormiston J.A.,
- et al.
- Swets J.A.
- Strandberg E.,
- Zeltinger J.,
- Schulz D.G.,
- Kaluza G.L.
- Onuma Y.,
- Serruys P.W.
- Chien S.
- Deguchi J.O.,
- Aikawa E.,
- Libby P.,
- et al.
- Galis Z.S.,
- Khatri J.J.
- Nagase H.,
- Visse R.,
- Murphy G.
- Sarno G.,
- Bruining N.,
- Onuma Y.,
- et al.
- Tearney G.J.,
- Bouma B.E.
- Gogas B.D.,
- Radu M.,
- Onuma Y.,
- et al.
- Simsek C.,
- Karanasos A.,
- Magro M.,
- et al.