Optical Coherence Tomography Findings in Patients Presenting With In-Stent Restenosis: A Prospective Observational Study of Patterns of Neointimal Hyperplasia and Associated Risk Factors

Introduction Morphological features of neointimal tissue play a pivotal role in the pathophysiology of in-stent restenosis (ISR) after percutaneous coronary intervention (PCI). This study was designed to qualitatively and quantitatively assess neointimal characteristics of lesions using optical coherence tomography (OCT) in patients presenting with ISR. Methods This was a single-center, prospective, observational study performed at a tertiary-care center in India. Patients diagnosed with stable angina and acute coronary syndrome with post-procedural angiographically documented restenosis (>50%) were included. Results A total of 34 patients with ISR were studied. Neointimal hyperplasia was classified as (i) homogenous group (n = 18) and (ii) non-homogenous group (n = 16). Fourteen (77.8%) diabetics belonged to the homogenous group. Predominant plaque characteristics such as neoatherosclerosis, cholesterol crystals, and calcium were documented in 14 (77.8%), 12 (66.7%), and 11 (61.1%) patients in the homogenous group and 10 (62.5%), 10 (62.5%), and 9 (56.2%) patients in the non-homogenous group, respectively. Unexpanded stent struts were identified in 11 (61.1%) and 11 (68.8%) patients in the homogenous and non-homogenous groups, respectively. Mean strut thickness was 93.73 ± 31.03 µm and 83.54 ± 18.0 µm, ISR was 72.50 ± 15.93% and 65.37 ± 21.69%, the neointimal thickness was 588.06 ± 167.82 μm and 666.25 ± 218.05 μm, and neointimal hyperplasia was 54.54 ± 11.23% and 59.26 ± 8.86% in the homogenous and non-homogenous groups, respectively. Conclusion Neoatherosclerosis and stent underexpansion were predominantly observed in our study and only diabetes was found to be significantly associated with homogenous neointimal hyperplasia.


Introduction
In-stent restenosis (ISR) is triggered by suboptimal re-endothelization and exaggerated neointimal proliferation which can occur within months or years after stent implantation [1].The bare-metal stent era witnessed a 32-55% incidence of ISR whereas the introduction of second-generation drug-eluting stents prompted a decline in ISR incidence to 10-12.2% [2][3][4].Nonetheless, ISR still prevails to a lesser extent in routine clinical practice [5].Intracoronary optical coherence tomography (OCT) provides a higher axial resolution than conventional intravascular ultrasound.In a recently conducted multicentric randomized controlled trial by Ali ZA, et al. in 2023, it was observed that the minimum stent area after percutaneous coronary intervention (PCI) was 5.72 ± 2.04 mm 2 in the OCT group and 5.36 ± 1.87 mm 2 in the angiography group (p < 0.001).Target-vessel failure within two years occurred in 88 patients in the OCT group and in 99 patients in the angiography group (p = 0.45), OCT-related adverse events occurred in one patient in the OCT group and in two patients in the angiography group, and stent thrombosis within two years occurred in six patients (0.5%) in the OCT group and in 17 patients (1.4%) in the angiography group [6].It also proves useful for qualitative as well as quantitative evaluation of neointimal tissue [7].In earlier studies, it was hypothesized that the morphological characteristics underlying the ISR lesion, that is neoatherosclerotic or peri-stent calcium, would have an impact on the expansion of an implanted stent during revascularization.These morphologic characteristics can be evaluated using OCT providing highlights of the tissue covering the stent struts at follow-up [8].Pathological findings revealed that neointima within a stent consists of numerous tissue components including collagen, proteoglycan, smooth muscle, fibrin, and thrombus.This observation is suggestive of differential prognosis of stented lesions according to in-stent neointimal characteristics [9].Although previous studies have explored correlations between neointimal patterns, underlying stent type, and restenosis phase, the degree of intralesional neointimal characteristics has not been elucidated [5,10,11].Therefore, this study sought to qualitatively and quantitatively assess neointimal characteristics of lesions using OCT in patients presenting with ISR.Studying the patterns of ISR is very important and has clinical implications in planning the strategy of PCI after OCT imaging.For example, if the type of ISR is neoatherosclerosis with calcium (calcific neoatherosclerosis), cutting balloon/scoring balloon can be employed as the preferred modality of PCI, whereas if the type of ISR is severely calcified neoatherosclerosis, rotablation or intravascular lithotripsy can be employed as the preferred modality of PCI.
This article was previously presented as a poster at the ESC (European Society of Cardiology) ASIA 2022 with the Asian Pacific Society of Cardiology (APSC) and Asean Federation of Cardiology (AFC), at Raffles City Convention Centre, Singapore.

Study design and patient population
This was a single-center, prospective observational study conducted at a tertiary care center in India between 1 August 2020 and 30 December 2021.A total of 34 consecutive post-PCI patients, diagnosed with stable angina and acute coronary syndrome, who had undergone invasive coronary angiography followed by OCT, were enrolled.

OCT image acquisition
OCT images were acquired using a frequency-domain OCT system (C7-XR OCT Intravascular Imaging System; St. Jude Medical Inc., St. Paul, MN).OCT images were generated at 100 frames/s and the catheter was pulled back at a speed of 20 mm/s.A contrast medium was continuously flushed through a guiding catheter at a rate of 4-5 mL/s for 3-4 s.Continuous images were acquired and stored digitally for subsequent analysis [7].

OCT image analysis
The OCT examination was performed using off-line OCT (Light Lab Imaging Inc., Westford, MA) software.For qualitative analysis, the pattern of re-stenosed tissue structure in the cross-sectional images at every 1 mm interval was categorized into two types: homogeneous and non-homogenous patterns of neointimal hyperplasia (NIH), of which, non-homogenous was further divided into heterogenous and layered patterns.
Quantitative analysis of OCT images was performed at minimal lumen area (MLA) sites.The MLA and stent area were manually traced, and the mean neointimal thickness was automatically calculated.Stent and luminal cross-sectional area (CSA) were measured, and NIH CSA was then calculated as the stent CSA minus the luminal CSA.The percentage of NIH-CSA was calculated as NIH-CSA × 100/stent CSA.The NIH thickness, the distance between the endoluminal surface of the neointima and the strut, can be measured inside all the struts as a line perpendicular to the neointima and the strut.
Cross-sectional OCT images of in-stent segments were analyzed at every 1 mm of the stent body.The region of interest was defined as an in-stent segment and 5 mm proximal and 5 mm distal segments.An uncovered strut was defined as having an NIH thickness of 0 µm.Malapposition was defined as the separation of stent struts from the vessel wall with a strut-vessel lumen distance of >200 mm.

Definitions
ISR was defined as in-stent segments -5 mm proximal and distal segments.Briefly, homogeneous tissues have uniform optical properties without focal variation in the backscattering pattern, heterogeneous tissues have focally changing optical properties and various backscattering patterns, and layered tissues consist of concentric layers with different optical properties [9].Thrombus was defined as signal-rich, lowbackscattering protrusions (white thrombus), or high-backscattering protrusions (red thrombus) inside the lumen with signal-free shadowing.Neoatherosclerosis was defined as the presence of lipid-laden intima and/or calcification inside the stent.Thin-cap fibroatheroma-like neointima was defined as the presence of an area with signal attenuation and a diffuse border, along with <65 µm fibrous cap thickness at the thinnest part.The site of neoatherosclerosis was classified as the proximal section, the middle section, and the distal section of the stent body.Micro-vessel was defined as a small tubular or vesicular structure with a diameter of <200 µm [12] (Figure 1).

Statistical analysis
The Statistical Package for Social Sciences (SPSS; IBM Corp, Armonk, NY) program, version 21.0, was used for data analysis.Continuous variables were expressed as mean ± standard deviation and categorical variables as percentages.Data normality was checked using the Kolmogorov-Smirnov test.Independent ttest and Mann-Whitney test were used for normally and non-normally distributed data, respectively.Categorical variables were compared with the chi-square test or Fisher's exact test.A p-value of <0.05 was considered statistically significant.

TABLE 4: Optical coherence tomography-derived quantitative assessments of plaque characteristics among homogenous and non-homogenous groups
All data are expressed as mean ± standard deviation.

Discussion
We found the prevalence of homogenous NIH in 18 (52.9%)patients and non-homogenous in 16 (47.1%)patients.A similar prevalence was observed in a study by Kim et al. [9] who reported 207 (54.9%) patients with homogenous and 170 (45.1%) patients with non-homogenous NIH.A study by Xhepa et al. [10] demonstrated a significant difference between the homogenous and non-homogenous groups in diabetes mellitus (p = 0.026), although hypercholesterolemia has been an established risk factor for atherosclerosis [13].The lipid profile of patients in the present study showed that there was no significant difference between the homogenous and non-homogenous groups.Similarly, in our study, diabetes mellitus was more common in the homogenous group, 14 (77.8%),compared to the non-homogenous group, 5 (31.2%).
Neoatheroslecrosis progression occurs due to proteoglycan deposition which promotes the retention of lipoprotein and infiltration of inflammatory cells such as lymphocytes, macrophages, and giant cells.These cells may gradually replace smooth muscle cells, resulting in changes in the neointimal pattern [14].ISR if seen at the histological level can be attributed to the accumulation of lipid-laden foamy macrophages with the presence or absence of a necrotic core, or calcification within the nascent intima after stent placement.
The most common and the earliest lesion is foamy macrophage clusters, seen mostly in the luminal surface or in the peri-strut region.The above induces fibroatheroma formation.The core of necrosis mostly contains cell-free fragments with free cholesterol, with a mostly damaged extracellular matrix.Significant hemorrhage with fibrin accumulation could be seen in the necrotic core in neoatherosclerosis, originating after a fissure or rupture of the luminal surface or from leaky vasa vasorum developed into the adventitial layer near the stent struts.Also, foamy macrophages migrated into the neointimal layer can initiate the development of fibroatheroma with a thin cap, resulting in in-stent plaque rupture and thrombosis [15].In the present study, plaque characteristics such as neoatherosclerosis among the homogenous and nonhomogenous groups were reported in 14 (77.8%) and 10 (62.5%) patients, p = 0.329, respectively, which can be attributed to the mean prior PCI interval which is 5.79 years.The patient cohort in the present study is representative of patients vulnerable to the occurrence of late ISR.Our finding was in line with the previous study of Lee et al. [3] who identified 3 (14.3%)patients with early ISR and 36 (35.6%) patients with late ISR.
In contrast to our study, Yamamoto et al. [14] reported progression to neoatherosclerosis in 0.6% of patients with a homogeneous pattern, 5.6% of patients with a heterogeneous pattern, and 3.9% of patients with a layered pattern at the six-month follow-up.Microvessels among the homogenous and non-homogenous groups were reported in 9 (50%) and 8 (50%) patients (p = 1.000), respectively.In contrast, a previous study reported the presence of microvessels in 22 (24.2%) and 6 (25.0%) patients among homogenous and heterogenous groups, respectively [16].This difference in the prevalence of microvessels is due to the follow-up duration which was one year in the study of Shi et al. [16].Pathologically, infiltration and accumulation of macrophages are supposed to be a vital process in the development of susceptible plaques [16].In our study, macrophages were reported in 9 (50%) and 7 (43.8%)patients among the homogenous and non-homogenous groups, respectively.A previous study reported that mean macrophage detected by OCT is an independent morphological risk factor of neointimal heterogeneity [16].Calcium among the homogenous and non-homogenous groups was reported in 11 (61.1%) and 9 (56.2%)patients (p = 0.774), respectively.However, Shi et al. [16] found calcified plaque in 39 (42.9%) and 7 (29.2%)among the homogenous and heterogenous groups, respectively.
Stent underexpansion has long been recognized as a major mechanical factor triggering ISR.In the present study, stent underexpansion was found in 11 (61.1%) and 11 (68.8%)among the homogenous and nonhomogenous groups, respectively.Similarly, a previous study by Xhepa et al. [10]  [9] where minimal lumen CSA was 4.5 ± 1.6 mm 2 and 4.0 ± 1.9 mm 2 .MLD was reported as 1.38 ± 0.33 mm and 1.36 ± 0.42 mm, which coincides with the previous study of Jung et al. [11] where the mean lumen diameter was 1.1 ± 0.4 mm and 1.2 ± 0.4 mm.NIH in our study was 54.54 ± 11.23 and 59.26 ± 8.86.On the other hand, the previous findings of Habara et al. [18] reported NIH as 58.6 ± 10.3% and 55.6 ± 13.4% for very late ISR and early ISR.The length of the ISR lesions was 24.62 ± 4.89 mm and 24.63 ± 7.02 mm among the homogenous and non-homogenous groups, respectively.Chang et al. [19] observed that in the bare-metal stent group, the longer the lesion length, the higher the ISR rates.Similarly, stent length and lesion length have been reported as independent predictors of ISR in various drug-eluting stents such as sirolimus-eluting stents.However, there was no significant difference in the length of ISR lesions between the homogenous and non-homogenous groups in our study.Our study may have some limitations due to it being singlecentric, having a relatively smaller sample size, and there may be interobserver variations in the OCT findings.Hence, a larger multicentric study is required in the future for more generalization of the results in real-world practice.
A total of 34 patients with ISR were studied, in whom quantitative and qualitative plaque characteristics were analyzed using OCT.The mean age of the patients was 54.79 ± 9.98 years.The majority of the patients belonged to the 51-65 years age category.Homogenous NIH was observed in 18 (52.9%)patients while non- homogenous groups, respectively.In this study, the majorly affected vessel with ISR was the left anterior descending artery in 10 (55.6%) and 9 (56.2%)among the homogenous and non-homogenous groups, respectively.Baseline and angiographic characteristics among the homogeneous and non-homogeneous groups are demonstrated in Table1.Non-homogenous (n = 16)Total (n = 34) p-Value

TABLE 2 : Plaque and stent characteristics among homogenous and non-homogenous groups All
data are expressed as number (percentage).