Optical coherence tomography: Translation from 3D-printed vascular models of the anterior cerebral circulation to the first human images of implanted surface modified flow diverters

Abstract

Background

The new generation of flow diverters includes a surface modification with a synthetic biocompatible polymer, which makes the device more biocompatible and less thrombogenic. Optical coherence tomography (OCT) can be used to visualize perforators, stent wall apposition, and intra-stent thrombus. Unfortunately real world application of this technology has been limited because of the limited navigability of these devices in the intracranial vessels. In this report, we share our experience of using 3D-printed neurovascular anatomy models to simulate and test the navigability of a commercially available OCT system and to show the application of this device in a patient treated with the new generation of surface modified flow diverters.

Material and methods

Navigability of OCT catheters was tested in vitro using four different 3D-printed silicone replicas of the intracranial anterior circulation, after the implantation of surface modified devices. Intermediate catheters were used in different tortuous anatomies and positions. After this assessment, we describe the OCT image analysis of a Pipeline Shield for treating an unruptured posterior communicating artery (PCOM) aneurysm.

Results

Use of intermediate catheters in the 3D-printed replicas was associated with better navigation of the OCT catheters in favorable anatomies but did not help as much in unfavorable anatomies. OCT image analysis of a PCOM aneurysm treated with Pipeline Embolization Device Shield demonstrated areas of unsatisfactory apposition with no thrombus formation.

Conclusions

OCT improves the understanding of the flow diversion technology. The development of less thrombogenic devices, like the Pipeline Flex with Shield Technology, reinforces the need for intraluminal imaging for neurovascular application.

Introduction

The next frontier in cerebral aneurysm repair seems to be in surface modification of endovascular devices and sophisticated intravascular imaging. Flow diverter technology has emerged as a safe and effective endovascular technique for treating brain aneurysms worldwide. Despite the promising results, thromboembolic events are still one of the most common complications reported in multiple large series, with major ipsilateral ischemic strokes between 1.6 and 4.3%.^1–3 There is an ongoing effort to introduce new solutions to reduce the thromboembolic complications. The Pipeline Flex Shield is the new generation of Pipeline devices, which has been redesigned using surface modification to make it less thrombogenic.^4,5

Intravascular imaging with optical coherence tomography (OCT) has been widely used in cardiac angiography^6,7 and is known to provide the best available assessment of: (a) stent apposition to the vessel wall during implant; (b) acute thrombus formation on the implant; and (c) development of neointima.^8–13 Unfortunately, there has been a bit of disappointment after the initial enthusiasm for intravascular imaging using OCT, largely due to the limited navigability of these catheters. Neuroimaging with OCT is currently off-label.

In this article, we share our experience of using 3D-printed neurovascular anatomy models to simulate and test the navigability of a commercially available OCT system. After an initial assessment, we were able to document the implantation of a new generation flow diverter with Shield Technology using OCT. We describe the nuances and discuss the challenges faced in using this new technology for intravascular imaging.

Materials and methods

In vitro procedures

We first obtained approval from the institutional review board for using OCT in 3D-printed specimens. Four newer generation flow diverter stents (Pipeline Embolization Device with Shield Technology; Medtronic Neurovascular, Irvine, California, USA) were placed into four different 3D-printed silicone replicas of the intracranial anterior circulation (blood flow simulator model; Biomodex, France) under fluoroscopic guidance (Allura Xper FD20; Philips, Best, Netherlands). Elongation of cervical internal carotid artery (ICA) and siphon tortuosity was noted.

In order to test the distal navigability, the OCT system (Dragonfly OPTIS Imaging Catheter; St Jude Medical, Westford, Massachusetts, USA) was navigated over Synchro-2 0.014-in microwires (Stryker Neurovascular; Fremont, California, USA), through the stent as distal as possible, by using intermediate catheters (Navien 072, 115 cm; Medtronic Neurovascular, Irvine, California, USA) in different positions (Figure 1).

Figure 1.

Left supraclinoid saccular aneurysms showing digital subtraction angiography lateral view of type I carotid siphon.

A 4 × 20 mm Pipeline Embolization Device with Shield Technology was deployed in left internal carotid artery for treatment of a posterior communicating artery aneurysm (a to c). Optical coherence tomography (OCT) Dragonfly OPTIS Imaging Catheter was exchanged and navigated coaxially over the wire. As demonstrated in vitro, optical lens navigation (arrow) was not possible by placing the intermediate catheter proximal to the stent (d). The Navien 072 115 cm was navigated through the stent up to the aneurysm neck (e). For image acquisition, the OCT catheter was unsheathed by pulling the intermediate catheter (f).

Clinical case and OCT imaging

A patient is his sixties consulted because of recurrent headaches. Magnetic resonance (MR) angiography and subsequent digital subtraction angiography (DSA) demonstrated a 7-mm wide neck saccular PCOM aneurysm on the left side. Considering the location and size of the aneurysm in an elderly patient, endovascular treatment with a stent was recommended.

Tablet clopidogrel 75 mg/day and aspirin 100 mg/day were started one week prior to the intervention, and therapeutic effect was checked and dose adjusted by doing an antiagreggation test (Verify Now system, Accumetrics; San Diego, California, USA) before the intervention. OCT was used to assess opposition between stent struts and the vessel wall and to look for theoretical prothrombotic risks such as acute clot formation or stent malapposition. The procedure was done under general anesthesia in our angiolab (Allura Xper FD20; Philips, Best, Netherlands). Heparin was administered at the beginning of the procedure and intermittently during the procedure to maintain the activated clotting time between 250 and 290 seconds. We used a triaxial system for the delivery of Pipeline Embolization Device (PED) using Neuron Max 6Fr 90 cm (Penumbra, Inc.; Alameda, California, USA), a Navien 072 intermediate catheter (Medtronic Neurovascular; Irvine, California, USA) and a Marksman 027 catheter (Medtronic Neurovascular; Irvine, California, USA). A 4 × 20 mm Pipeline Embolization Device with Shield Technology was selected based on the maximum parent artery diameter on 3D DSA (Figure 1). We selected the Navien 072, 115 cm, to have enough length to unsheath the OCT catheter and adequate lumen for clearing the blood vessel with a contrast injection during intravascular imaging acquisition.

After deployment, the intermediate catheter Navien 072 was navigated beyond the carotid siphon through the stent to the aneurysm neck (Figure 1). The Dragonfly OCT catheter (Figure 2) was navigated through the Navien catheter and OCT imaging was performed using the high-resolution protocol on the ILUMIEN OPTIS system (St Jude Medical). Pullback OCT image series (540 frames covering 54 mm vessel length, 10 mm/s pullback speed) were taken along the parent artery and across the neck of the aneurysm. In order to remove blood from the artery during OCT imaging, iodine contrast (Visipaque 320; GE Healthcare, Marlborough, Massachusetts, USA) was injected manually every 4 seconds to clear the blood from the vessel (Figure 1).

Figure 2.

Dragonfly OPTIS Imaging Catheter.

(a and b) 2.7 F low-profile tip hydrophilic coating compatible with standard 0.014-in guidewires. Although catheter length is 135 cm, the additional connector length (*) should be considered if coaxial access is needed (a). The optical lens (white arrow) is the most fragile portion, where imaging acquisition starts in a pullback fashion for a 54-mm segment analysis. In our experience the critical point for intracranial navigation was the segment between the lens marker and the wire exit port (black arrow), where the catheter bent and stocked after applying tension (b). The more distal 27 mm of the catheter, above the optical lens (white arrow), is designed for navigation and microwire insertion (monorail) for support. Synchro-2 microwire (white dashed arrow) and the distal tip marker (black dashed arrow) are seen (b).

Offline OCT image analysis was performed with ILUMIEN OPTIS post-processing software. We selected three cross-sectional images per pullback (mid-stent at the level of the aneurysm neck, mid-stent at the paraclinoid curve, and proximal stent) for each of the time points of the study (Figure 3).

Figure 3.

Optical coherence tomography images.

Cross-sectional optical coherence tomography (OCT) images analysis after Pipeline Shield deployment in three different locations: mid-stent at the level of the aneurysm neck (a), mid-stent at paraclinoid curve (b), and proximal stent(c). Areas of device malapposition were identified close to the aneurysm neck (a) and paraclinoid segment (b) (white asterisk), and different areas of device opening were measured in these three locations. Complete wall apposition was visible at proximal segment. Note the aneurysm lumen (L).

Results

In vitro evaluation

Table 1 shows the evaluation of the navigability of the current commercially available OCT catheter Dragonfly regarding vessel anatomy¹⁴ and catheters used. We used four 3D-printed replicas for in vitro testing. Two of them had non-tortuous vascular anatomies (C-shaped cervical ICA and type I carotid siphon) and two had replicas of tortuous anatomies (S-shaped cervical ICA and type II carotid siphon). One PED was deployed per model with adequate neck coverage of the aneurysm located at supraclinoid, paraophthalmic and cavernous segments. All devices were distally anchored in M1 and deployed (Figure 4).

Table 1.

In vitro evaluation of optical coherence tomography (OCT) catheter navigation based on arterial anatomy and position of the intermediate catheter related to the braid.

Four 3D-printed replicas were used for in vitro testing. Two of them had non-tortuous vascular anatomies (C-shaped cervical internal carotid artery (ICA) and type I carotid siphon) and two had replicas of tortuous anatomies (S-shaped cervical ICA and type II carotid siphon). Intermediate catheters were used in proximal and distal locations.

DIC: distal intermediate catheter located at distal end of the Pipeline Shield; PIC: proximal intermediate catheter located at proximal end of the Pipeline Shield.

Navigability was tested as follows: 0: no navigation. 1: full navigation.

Figure 4.

3D-printed silicone replica of intracranial anterior circulation.

(a to e) C-shaped cervical internal carotid artery (ICA) and type I carotid siphon. (b) Fluoroscopy images: a 4 × 25 mm Pipeline Shield was deployed from middle cerebral artery (MCA) to cavernous segment covering two aneurysms located at paraclinoid/supraclinoid segments. (c) A Navien 072 115 cm was positioned at proximal end of the stent and optical coherence tomography (OCT) catheter was pushed over a Synchro-2 microwire. Optical lens navigation (white arrow) over paraophthalmic segment was not possible. (d) Distal intermediate catheter located at distal end of the Pipeline Shield. Optical lens navigation through supraclinoid segment and MCA were achieved. (e) Simulation of the image acquisition was based on unsheathing of the intermediate catheter leaving uncovered the pullback segment of analysis. (f to h) S-shaped cervical ICA and type II carotid siphon. (f) A 4.25 × 20 mm Pipeline Shield was deployed from right MCA to cavernous segment covering one aneurysm located at paraclinoid segment. (g and h) A Navien 072 115 cm was positioned at mid-stent (g) and stent distal end (h) and OCT catheter was pushed over a Synchro-2 microwire. Distal optical lens navigation (white arrow) was not possible despite the use of intermediate catheter. Note proximal (*) and distal markers (optical lens) of the pullback region under fluoroscopy. The distal end of the catheter above the optical lens, with a radiopaque tip marker (black dashed arrow), is designed for navigation and the monorail microwire insertion for support.

The distal 27 mm of the catheter, above the optical lens, is designed for navigation and microwire insertion (monorail) for support. The OCT intra-stent navigability was rated based on the optical lens position (which is where the pullback starts for imaging acquisition) (Figures 2 and 4).

Use of intermediate catheters was associated with better navigation of the OCT catheter, in favorable anatomies (C-shaped ICA and type I carotid siphon) as shown in Table 1, but did not help as much in unfavorable anatomies(S-shaped ICA and type II carotid siphon) (Figure 4).

Clinical evaluation

In the presented case, the coaxial system was needed because of cervical ICA and siphon elongation (Figure 1). The distal tip was navigated into the M1 segment, but the lens marker navigated up to the aneurysm neck because we did not navigate the Navien 072 115 cm more distally. As mentioned above, pullback was done from the aneurysm neck to the proximal end of the stent. OCT images showed that Pipeline Shield fully expanded to nominal at the paraclinoid and cavernous segments. We found discordance with DSA artery measurements, resulting in areas of malapposition despite nominal opening, especially at the paraclinoid portion (Figure 2). No thrombus was identified on the device surface.

Like previous reports,⁸ areas of malapposition and failure of the device to fully open were found close to the aneurysm neck (Figure 3), but this issue did not affect contrast stasis demonstrated on DSA. No further angioplasty was done and the patient was extubated with no complications. He was discharged home on post-operative day 3. The patient visited for one-month follow-up and was found to have an unremarkable recovery. He continued on a dual antiplatelet regimen (alternate-day clopidogrel 75 mg and daily aspirin 100 mg), to be maintained for the next three months, following which he would be changed to a mono drug regimen.

Discussion

Intravascular imaging can play a major role in improving the safety of endoluminal devices such as flow diverters. OCT technology has been around now for more than two decades. Large clinical series and prospective trials have shown that this technology is safe and effective.^1–3 OCT provides high-resolution cross-sectional in situ images from intact tissue, based on tissue reflectance of near-infrared or infrared light through a single fiber optic wire. The technique involves simultaneous rotation and withdrawal of the wire to obtain a series of cross-sectional images.^9,10

The spatial resolution yielded from OCT exceeds what is possible with routine imaging methods such as from catheter directed angiography or even non-invasive imaging techniques like computed tomography angiography and MR angiography. OCT has been widely used in cardiology to demonstrate high-resolution arterial wall analysis, intimal hyperplasia after stenting and plaque characterization. After selecting a proper stent, in conjunction with OCT, one can confirm optimal stent placement and stent wall apposition, as well as ensure full expansion of the stent, thereby reducing complications such as in-stent thrombosis.^6,7

The main limitation for the standardized use in the neurovascular field is the tortuous nature of intracranial circulation, especially at the carotid siphon and the cervical ICA, making the use of coaxial systems mandatory in most cases. Lopes et al.¹¹ demonstrated the utility of OCT for the study of intracranial artery perforators even in the middle cerebral artery segment, technically feasible with the use of intermediate catheters in a cadaver study. We corroborated the technical feasibility using in vitro testing and found that the tortuous anatomies are a predictive factor for failed navigation.

In our experience, the critical point for intracranial navigation was the segment between the lens marker and the wire exit port, where the catheter bent and stocked after applying tension (Figure 2). We also found in our in vitro testing that the intermediate catheters enable better trackability of the distal tip. But in more tortuous anatomies like S-shaped cervical ICA and type II carotid siphon, the lens marker could not be navigated until M1 despite the use of a distal intermediate catheter (Figure 4).

The main limitation of the silicone model used is that replicas were made from the distal cervical ICA, without considering proximal elongations like the common carotid artery and aorta requiring longer systems.

PED with Shield Technology is the third generation of the PED and is less thrombogenic and more biocompatible than its predecessors.⁴ Shield Technology is a surface modification where a synthetic phosphorylcholine (PC) polymer is covalently bonded to the strands that make up the Pipeline braid. PC has been used for years in biocompatible medical surfaces or stent coating in cardiology because it “resembles” the polar head of the phospholipids of the cell membrane, and hence has the ability to reduce protein adsorption and thrombin generation.^4,5

Both intra- and post-procedural intravascular imaging studies using the OCT have been done on the Pipeline Shield in animal models. Marosfoi et al.¹² showed that OCT is superior to DSA for identifying acute clot formation after PED Shield deployment, especially at branch origins. Similarly, King et al.⁸ demonstrated that OCT is also superior to DSA in identifying segments of malapposition, which can suggest a potential source for delayed or failed occlusion. At 21 days follow-up, Matsuda et al.¹³ reported a more homogeneous endothelialization pattern of PED Shield compared to classical PED in swine models, which could possibly be because of the biocompatibility.

In clinical practice, OCT is seen to be superior to angiography in measuring the arterial diameter and identifying areas of stent malapposition, which is demonstrated in our case. Although there have been studies⁸ showing that a malapposition scoring system applied to post-explant histological samples correlated well with occlusion rates, in the presented case we preferred to avoid angioplasty considering the significant contrast stagnation achieved after flow diverter deployment.

In our opinion, with the current technology available in the market, OCT catheter navigation is only feasible in posterior circulation and favorable anterior circulation arterial anatomies. Given et al.¹⁴ published an in vivo case of a wingspan valuation of OCT on posterior circulation and Patel et al.¹⁵ reported the use of OCT for a vertebral artery stenosis; both reports showed that posterior circulation segments are more amenable for navigation, not only because of less tortuosity but also because they need shorter catheters.

Access to supraclinoid ICAs may require coaxial systems with long intermediate catheters to overcome the tortuosity. One-way we could attempt to solve this is by increasing the length of the OCT catheters allowing complete unsheathing of the pullback segment. Also, reinforcing the segment between the lens and the wire exit port (Figure 2) would improve trackability beyond the carotid siphon.

Conclusion

OCT is a valuable tool for intravascular analysis which improves the understanding of the flow diversion technology for the treatment of intracranial aneurysms.

The development of less thrombogenic devices, such as the Pipeline Flex with Shield Technology, reinforces the need for intraluminal imaging for neurovascular application.

Despite that new intermediate catheters have improved the intracranial applications, current OCT technology needs some technical modifications especially to access the anterior circulation locations beyond the carotid siphon.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.