Optical coherence tomography for elucidation of flow-diversion phenomena: The concept of endothelized mural thrombus behind reversible in-stent stenosis in flow-diverters

Abstract

Purpose

Flow-diverters have revolutionized the endovascular treatment of intracranial aneurysms, offering a durable solution to aneurysms with high recurrence rates after conventional stent-assisted coiling. Events that occur after treatment with flow-diversion, such as in-stent stenosis (ISS) are not well understood and require further assessment. After assessing an animal model with Optical Coherence Tomography (OCT), we propose a concept that could explain the mechanism causing reversible ISS after treatment of intracranial aneurysms with flow-diverters.

Methods

Six Pipeline Flex embolization devices (PED-Flex), six PED with Shield technology (PED-Shield), and four Solitaire AB devices were implanted in the carotid arteries (two stents per vessel) of four pigs. Intravascular optical coherence tomography (OCT) and digital subtraction angiography (DSA) images obtained on day 21 were compared to histological specimens.

Results

A case of ISS in a PED-Flex device was assessed with OCT imaging. Neointima with asymmetrical topography completely covering the PED struts was observed. Histological preparations of the stenotic area demonstrated thrombus on the surface of device struts, covered by neointima.

Conclusion

This study provides a plausible concept for reversible ISS in flow-diverters. Based on an observation of a previous experiment, we propose that similar cases of ISS are related to thrombus presence underneath endothelization, but further experiments focused on this phenomenon are needed. Optical Coherence Tomography will be useful tool when available for clinical use.

Introduction

Flow-diverters (FD) have revolutionized the endovascular treatment of intracranial aneurysms, offering a durable solution to aneurysms with high recurrence rates after conventional stent-assisted coiling, such as large/giant, wide necked and fusiform ones. ¹ Pipeline Embolization Device (PED) and PED-Flex are widely used first and second generation FDs shown to be effective also for small and unruptured saccular aneurysms of the anterior circulation.^2,3 PED with Shield technology (PED-Shield) involves the addition of a phosphorylcholine polymer covalently bound to the meshwork, providing a less thrombogenic surface. ⁴ Although these devices have been extensively used in the endovascular neurrgery practice, the events that occur after treatment with flow-diversion (e.g. delayed aneurysm rupture, distal intraparenchymal hemorrhage and in-stent stenosis) have not been completely understood and require further assessment. ⁵

Reversible in-stent stenosis (ISS) within an implanted FD is not a rare finding on follow-up digital subtraction angiography (DSA). Often minor with no clinical relevance, ISS is frequently considered to be a benign condition associated with flow-diversion rather than a true complication of FDs. ⁶ Further, previous studies have commented on its possible role in delayed ischemia and early aneurysm occlusion, without definitive conclusions.^5–8 The currently accepted hypothesis suggests that ISS is caused by either fibroelastic tissue formation or neointimal hyperplasia, similar to that of other intravascular devices.^2,6 However, the reversible and asymptomatic behavior of ISS in FDs suggests a different nature for this phenomenon.^6,7,9 Given the lack of clarity, investigation with more sophisticated imaging is warranted to elucidate the actual mechanism behind ISS and its implications. Optical coherence tomography (OCT) is a high-resolution imaging modality previously used in ophthalmology and cardiology. When applied to intravascular imaging in the field of neurosurgery, it can provide details that are unobtainable with standard vessel imaging (i.e. MRA, CTA, DSA). In the present study, we use an observation from one of our previous experiments with OCT to propose a new concept: an endothelized thrombus as a mechanism behind the asymptomatic reversible ISS observed in the follow-up imaging after treatment with PED.

Methods

We conducted an animal model experiment using PED Shield, PED-Flex, and Solitaire AB (Medtronic, Irvine, California, USA), and assessed the neointima development using OCT imaging. The results of this experiment were previously published. ¹⁰ The current study will focus on the ISS (>50%) observed during the experiment. The following methods (animal study design, procedural description, imaging protocol and tissue harvesting) have been previously described by Matsuda et al. ¹⁰

Animal study design

The local Institutional Animal Care and Use Committee approved the study protocol. In compliance with the Animal Welfare Act and the ‘Guide for the Care and Use of Laboratory Animals’ formulated by the Institute of Laboratory Animal Research, all animals received humane care. ¹¹

Procedural description

Six PED-Flex, six PED-Shield, and four Solitaire AB were implanted in carotid arteries (two stents per vessel) of four Yorkshire 25–35 kg pigs. The implanted devices sizes were 5 × 20 mm for PED’s and 4.0 × 20 mm for Solitaire AB. The animals were given 10 mg/kg of aspirin and clopidogrel daily. A mixture of telazol and xylazine weight-based was used to pre-anesthetize all animals. Animals were intubated once an adequate anesthetic status was achieved and were maintained using continuous inhalation of isoflurane (1–3%). The femoral artery was accessed percutaneously by a vascular access sheath (6 F). Heparin (3000–8000 U) was then administered through the inserted vascular sheath. Baseline imaging was acquired with DSA before implanting two stents in each of the carotid arteries. Intravascular images at day 21 post-procedure were obtained with both OCT and DSA.

OCT imaging protocol and analysis

Dragon fly Optis imaging system (LightLab Imaging, Inc., St. Jude Medical, St. Paul, MN, USA) was utilized to obtain OCT images. A motorized OCT catheter pull back at rate of 18.0 mm/s, while a simultaneous intra-arterial iodine contrast injection at rates of 4–6 ml/s for a total volume of 16–24 ml to clear the blood from the vessel where used to obtain all OCT images at 180 frames per second. Ilumien Optis post processing software was used for offline OCT image analysis. Three cross-sectional images were obtained per pullback: proximal stent, mid-stent, and distal stent. These were further matched with histological specimens of the same area.

Tissue harvesting and histology

Under anesthesia, all animals were euthanized after the angiogram (day 21). The carotid arteries were excised, and the stented segments were harvested. The distal edge of the stent was marked after histological processing. Three sections were selected and correlated to OCT imaging at an equal distance from the distal end of the stent. All histological specimens were post-fixed in 2.5% glutaraldehyde for at least 2 h. Slides created using the EXAKT system (EXAKT Technologies, Inc., Oklahoma City, OK, USA) were stained with Hematoxylin and Eosin (H&E). A total of 48 OCT images with histology matched cross-sections were analyzed.

Results

A case of ISS in a PED-Flex identified with DSA on day 21 was further assessed with OCT imaging (Figure 1(a) and (b)). Neointima with asymmetrical topography was completely covering the PED struts (Figure 1(c)). Further correlation with histology demonstrated thrombus on the device struts, covered by neointima (Figures 1(d) and 2).

Figure 1.

DSA showing in-stent stenosis in a PED-Flex (a). OCT imaging - topography of the stenotic area (b) correlates well with the digital subtraction angiography image, and complete neointimal coverage of the PED-Flex struts is seen (c). Histology section demonstrated an area of neointimal coverage of PED struts (black arrowhead) and two areas (arrow) of neointima covering thrombus and the PED struts (d).

Figure 2.

OCT imaging of a stenotic area of PED-Flex (a) and correlation with histology section (b). Amplification of the area of interest (c), showing an underlying thrombus covered by neointima (black arrows) over the stent struts (white arrowheads).

Discussion

In our experiment, we observed a match between the lumen topography on OCT imaging and the stenotic areas on DSA. When these images were correlated with their respective histological specimens, an endothelized thrombus was seen to be causing the lumen topography observed. This concept could explain the reversible ISS usually seen after treatment with PED, given that when thrombus resolves, the stenosis would disappear.

Two main mechanisms of stenosis after angioplasty and stenting have been described in interventional cardiology: vessel inflammatory response to injury (or negative remodeling), and neointimal hyperplasia.^12–14 Negative remodeling occurs when vascular overstretch triggers an inflammatory response that leads to tissue remodeling with subsequent loss of both inner and outer diameter of the artery, constraining the mesh of the device implanted.^12,13 Neointimal hyperplasia can occur after vascular overstretch (injury) or due to foreign body reaction to the metal struts, which trigger distinct levels and pathways of inflammation leading to different extents of neointimal growth within the stented segment.^12–14 Although negative remodeling can also occur when FDs are implanted (Figure 3), the reversible ISS after treatment with PED has distinctive angiographic features that suggest a different nature (Figure 4).

Figure 3.

Illustrative case of negative remodeling stenosis in a 20 s female patient who was found to have multiple intracranial aneurysms after imaging investigation for migraines and was treated with PED. Digital subtraction angiography lateral views: two aneurysms (arrowheads) in the ophthalmic segment of the left internal carotid artery, measuring 7.1 × 3.2 mm and 4 × 2.5 mm, respectively (a); PED mesh is well appositioned to the wall and along the aneurysms (b); three months follow-up revealed negative remodeling (c), with the distal third of the mesh (arrows) being constrained (d); follow-up 18 months after treatment (e) with some residual constraining of the distal third of the mesh (f).

Figure 4.

Illustrative case of in-stent stenosis (ISS) in a 50 s female patient treated with PED. Digital subtraction angiography (DSA) lateral view of the right internal carotid artery showing a superior hypophyseal aneurysm (arrowhead) and previous coiling of a middle cerebral artery aneurysm (a). The superior hypophyseal aneurysm was treated with a PED, with good apposition of the mesh (b). DSA at six months follow-up revealed ISS, fully expanded PED mesh (arrows) can be seen around stenotic lumen on lateral (c) and anterior (d) views. DSA at 20 months follow-up showing improvement of ISS, with mild stenosis remaining, on lateral (e) and anterior (f) views.

Chalouhi et al. ⁶ were the first to methodically assess ISS in PED, reporting the event in 15.8% of patients within a mean follow-up period of 6.7 months. In their series, all patients showing ISS were asymptomatic women, and only anterior circulation and no use of aspirin were found to be independent predictors of ISS. In following studies, the incidence of ISS after treatment with PED ranges from 0.3 to 18%,^3,5,6,15 depending on follow-up time and percentage of lumen stenosis considered. In general, lumen stenosis greater than 50% (>50%) is considered relevant, while lesser lumen narrowing is usually labeled as intimal hyperplasia.^5,8 The PUFS and PREMIER trials observed ISS > 50% in 1.1 and 2.9% of patients, respectively, at 1-year follow-up.^16,17 Despite this divergence in the reported rates, some common points about ISS are seen across the case series and trials. It is usually present on first follow-up, with the earliest image obtained at three months.^1,5,6,8,15 Almost all patients are asymptomatic and the majority of cases show complete resolution or at least some degree of involution within six months, even without additional intervention or changes in the antiplatelet therapy.^5–8 The rate of ISS greater than 50% in PED Shield does not differ from the range reported for other PEDs previous studies mentioned. The majority of ISS observed after PED-Shield were also asymptomatic, occurring at rates of 5.6% and 3.1% at six months and one-year follow-up, respectively. ¹⁸

A few authors speculated about a possible contributing role of ISS to aneurysm occlusion, after identifying early and complete aneurysm occlusion in the cases that developed ISS.^6,7 Moreover, studies of coronary stenting in the past found thrombus formation to be present in, and associated with, the neointima development.^19,20 The concept being presented here, an endothelized thrombus as underlying mechanism for reversible ISS after PED, would be in accordance with these findings and assumptions. However, we believe that further experiments with sophisticated imaging, such as OCT, focusing on ISS are needed for further understanding of this phenomenon.

OCT have been used to assess PED previously. Matsuda et al. ²¹ found that, despite being less thrombogenic under equal antiplatelet regimens, PED-Shield still had higher thrombus formation under Single Antiplatelet Therapy (SAPT) when compared to PED-Flex under Dual Antiplatelet Therapy (DAPT). Their study also showed that neointima had reduced coverage in the setting of DAPT in all devices tested, regardless of surface modification. ²¹ PED-Shield was found to develop earlier neointimal growth, but similar coverage with time, when compared to PED-Flex under equal antiplatelet regimens. ¹⁰ Conversely, another study observed less neointimal coverage in PED-Shield when compared to PED-Flex, when both devices were under DAPT. ²² These findings would not be acquired with standard vessel imaging. Therefore, it is expected that OCT will help to elucidate many events associated with flow diversion when it becomes available for clinical use in neurosurgery.

Limitations

The results of our study should be interpreted in light of the limitations of porcine models. Not all hemodynamic forces were involved, given that all treated vessels were straight without aneurysms or branches. The initial endothelial coverage can be damaged by intravascular manipulation before or after device implantation, which could result in the misinterpretation of the degree of final tissue coverage. Additionally, in animal models, full de-endothelialization of the vessels can be caused by angioplasty.^23,24 This study is further limited by thrombosis observed in six PED Flex and two PED Shield at day 7, which can alter the process of neointimal formation.

Conclusion

We observed the presence of ISS after FD placement in an animal model using OCT assessment. After correlation with DSA and histology, we proposed a new concept in which cases of reversible ISS observed after treatment with FD may be related to thrombus presence underneath endothelization, explaining both the occurrence of stenosis and its reversibility over time. But further studies are certainly needed. With the upcoming advent of clinical use of OCT, the different mechanisms of ISS and other flow-diverter phenomena will likely be elucidated.