Article Text
PDF
Abstract
Background High-resolution vessel wall MRI (VWI) is increasingly used to characterize intramural disorders of the intracranial vasculature unseen by conventional arteriography.
Objective To evaluate the use of VWI for surveillance of flow diverter (FD) treated aneurysms.
Materials and methods Retrospective study of 28 aneurysms (in 21 patients) treated with a FD (mean 57 years; 14 female). All examinations included VWI and a contemporaneously obtained digital subtraction angiogram. Multiplanar pre- and post-gadolinium 3D, variable flip-angle T1 black-blood VWI was obtained using delay alternating nutation for tailored excitation (DANTE) at 3T. 3D time-of-flight MR angiography (MRA) was also carried out. Images were assessed for in-stent stenosis, aneurysm occlusion, presence and pattern/distribution of aneurysmal or parent vessel gadolinium enhancement.
Results The VWI-MRI was performed on average at 361±259 days after the intervention. Follow-up DSA was performed at 338±254 days postintervention. Good or excellent black-blood angiographic quality was recorded in 22/28 (79%) pre-contrast and 21/28 (75%) post-contrast VWI, with no cases excluded for image quality. Aneurysm enhancement was noted in 24/28 (85.7%) aneurysms, including in 79% of angiographically occluded aneurysms and 100% of angiographically non-occluded aneurysms. Enhancement of the stented parent-vessel wall occurred significantly more often when aneurysm enhancement was present (92% vs 33%, p=0.049).
Conclusion Advanced VWI produces excellent depiction of FD-treated aneurysms, with robust evaluation of the parent vessel and aneurysm wall to an extent not achievable with conventional MRI/MRA. Gadolinium enhancement may, however, continue even after enduring catheter angiographic occlusion, confounding interpretation, and requiring cognizance of this potentially prolonged effect in such patients.
- aneurysm
- flow diverter
- vessel wall
- MRI
Data availability statement
Data are available upon reasonable request. as above.
Statistics from Altmetric.com
Introduction
High-resolution vessel wall MRI (VWI) is increasingly used to characterize intramural disorders of the intracranial vasculature, such as atherosclerotic plaques, vasculitis, and aneurysms.1 Specifically in the case of aneurysms, emerging interest in VWI as a biomarker of aneurysm wall integrity has supported its use to identify aneurysms at high risk of rupture, to surveil aneurysms, or to remove uncertainty about the source of hemorrhage and to identify culprit aneurysms preoperatively.2 3 VWI is particularly well suited to longitudinal characterization in light of its non-invasive nature, potentially obviating the need for invasive catheter angiographic monitoring; however, past studies have focused primarily on the characterization of treatment-naïve, unruptured, and ruptured aneurysms while its usefulness for treated aneurysms remains mostly unexplored or restricted to limited series.4–7 We recently have implemented a custom-optimized black-blood VWI for characterization of treated aneurysms, with particular attention paid to the imaging evolution of aneurysms following flow diversion, now widely considered as the first-line treatment for aneurysms of varying subtypes.
Recent advancements in VWI have enabled ever-increasing spatial resolution and continued improvements in high-fidelity signal suppression from flowing blood, known to confound discrimination between the vessel lumen and wall due to partial volume averaging or flow effects, respectively.8 Existing flow suppression strategies, including double inversion recovery,9 motion-sensitized driven equilibrium, and variable flip angle 3D turbo spin echo sequences,10 particularly at conventionally attainable spatial resolution, might therefore be better suited to the evaluation of larger-caliber extracranial vasculature. The intracranial circulation, however, due to its smaller vessel caliber and paucity of a vasa vasorum, demands high performance VWI optimized both to the cerebrospinal fluid and flow suppression, and ultra-high (≥500 μm isotropic) spatial resolution.11 A promising contemporary strategy includes the addition of delay alternating with nutation for tailored excitation (DANTE) pulse trains to achieve uniform flow suppression while preserving signal from static spins.12 DANTE-prepared sequences have prove valuable in other applications, including for spatial tagging and more recently for robust flow suppression in black-blood arteriography.13 14
We have evaluated use of state-of-the-art 400–500 μm isotropic whole brain DANTE-prepared 3D variable flip-angle turbo spin echo VWI sequences accelerated with controlled aliasing in Parallel imaging (CAIPIRINHA) undersampling in a consecutive series of adult patients undergoing endovascular flow diversion of intracranial aneurysms in routine clinical practice. Owing to its unprecedented resolution and performance following flow diversion, we have observed a number of previously unfamiliar and sometimes paradoxical imaging patterns in such patients, including those indicating potential aneurysm recanalization despite enduring angiographic occlusion. We thus felt that a formalized introduction to this spectrum of findings was timely, in light of the widespread use of flow diversion and growing interest in the use of VWI for treated and untreated intracranial aneurysms.
Materials and methods
This is a retrospective, institutional review board-approved, Health Insurance Portability and Accountability Act (HIPAA)-compliant study. Inclusion criteria were: patients with intracranial unruptured aneurysms treated with a Pipeline embolization device (PED; Medtronic Neurovascular, Irvine, California, USA) for flow diversion, in whom post-treatment MRI, including ultra-high resolution black-blood VWI, was obtained before and after intravenous gadolinium contrast as part of routine clinical practice in our institution. Exclusion criteria included incomplete MRI protocols, unsuccessful intravenous gadolinium administration, inability to image with the high order (64-channel) head array coil to permit suitable acceleration and signal-to-noise ratio, or the absence of follow-up catheter angiographic evaluation.
Procedures
The interventions were performed by one of five neurointerventionalists. All procedures were performed under general anesthesia on a biplane Siemens machine. A standard guide sheath, standard distal access catheter and 0.027 microcatheter were used to deploy the PED. The decision to implant multiple PEDs was made at the discretion of the treating physician towards creation of a stable endoluminal construct.15 Follow-up catheter angiography according to routine practice in our institution was performed by the treating neurointerventionalist between 6 and 12 months post-treatment using the same biplane machine.
MRI protocol
Multiplanar, non-selective, pre- and post-gadolinium 3D, variable flip-angle T1 turbo spin echo VWI was obtained with flow suppression using DANTE in the sagittal plane on a clinical 3T whole body scanner (PRISMAfit, Siemens Healthineers). Frequency selective chemical fat saturation was obtained in select cases following gadolinium administration (gadobutrol (Gadavist); Bayer Schering Pharma, 1 mmol/kg IV infusion). 3D slabs were obtained at between 400 and 500 μm whole brain isotropic resolution, varying depending on the patient's head size to ensure whole brain coverage and mitigate aliasing within the slab. Unless modifications were required to allow for greater slab coverage, acquisition parameters optimized specifically for this protocol were: repetition time/echo time 1000/17 ms, 0.4 mm isotropic 3D variable flip-angle T1 turbo spin echo, single signal average, matrix 4362, CAIPI acceleration factor 2×1 with an integrated calibration scan using a 64-channel head receive array coil and total acquisition time ~11–13 min depending on total slices, and repetition time to accommodate brain coverage. 3D time-of-flight. MR angiography without contrast was also carried out in all cases.
Imaging review
DSA images were reviewed for the following: aneurysm size (largest diameter); number of PEDs used; aneurysm occlusion (yes/no); in-stent stenosis, defined as visible thickening of the intima within the stent construct (yes/no); flow diverter used (size and number). VWI were reviewed in consensus by two subspecialty certified neuroradiologists specializing in cerebrovascular imaging and blinded to catheter angiographic results, and were evaluated qualitatively for overall VWI black-blood angiographic quality, spanning limited, acceptable, good, and excellent visual ratings for both pre- and post-contrast VWI. Studies were examined for any artifacts corrupting evaluation of the treated aneurysm, parent vessel, or remainder of the image (eg, phase artifacts, free induction decay artifacts, motion, and susceptibility artifacts arising from the endovascular construct or directly from the skull base or paranasal sinuses. All aneurysms were classified with respect to location/vessel of origin, and morphologically as saccular, sidewall, or fusiform, and any variability in either performance or in imaging features relating to the preceding was noted. VWI scans were further assessed for binary determination of in-stent stenosis; intra-aneurysmal gadolinium enhancement; gadolinium enhancement of the aneurysm wall; and parent-vessel wall enhancement. Aneurysm and parent-vessel enhancement were defined by comparing the same plane pre- and post-contrast VWI sequence side by side.
Statistical analysis
Categorical variables were compared with Fisher’s exact test and reported as number and percent of occurrence. For continuous variables, normality was tested using visual inspection of histograms and the Shapiro-Wilk test. Normally distributed variables are reported as mean and SD. When continuous variables were not normally distributed, non-parametric tests were used, including the Mann-Whitney, Kruskal–Wallis, and Wilcoxon signed-rank tests, when appropriate, and reported as median and IQR. Non-parametric areas under the receiver operating characteristic (AUROC) curves were calculated to determine how accurately the time from PED placement to both VWI and follow-up angiography distinguish between the presence or absence of aneurysm enhancement on VWI, with interpretation of the area under the curve as previously reported.16 Statistical analysis was performed using Stata 12.1 (StataCorp, College Station, Texas, USA). A two-tailed p value of less than 0.05 was considered statistically significant.
Results
We identified 28 aneurysms in 21 patients (mean age 57±17.6 years; 14 female, 7 male) who met the inclusion criteria and who underwent MRI between January 2018 and 2020. The aneurysms were distributed as follows: 12 paraophthalmic, four posterior communicating, three anterior communicating, three intradural vertebral, two basilar, and four in other locations.17 Mean aneurysm size was 10.6±9.9 mm. On average, the aneurysms were treated with two PEDs (range 1–6). Six aneurysms were also coiled. VWI-MRI was performed on average at 361±259 days following intervention. Follow-up DSA was performed at 338±254 days post-intervention.
Across the cohort, black-blood angiographic quality was preserved, with no cases identified as limited in quality, and overall good or excellent angiographic quality as obtained in 22/28 (79%) pre-contrast and 21/28 (75%) post-contrast VWI. Aneurysm enhancement on VWI was noted in 24/28 (85.7%) aneurysms, including in 79% of angiographically occluded aneurysms and 100% of angiographically non-occluded aneurysms (figures 1 and 2). Aneurysm enhancement, irrespective of the observed pattern, was not associated with angiographic complete aneurysm occlusion (p=0.273; table 1). During our initial experience with the technique this finding was misinterpreted commonly as reticular intrasaccular residua or, when in the wall along the aneurysm dome, as flow undermining the wall and interposed between partial occlusion (see Discussion below). By comparison, all angiographically non-occluded aneurysms on DSA exhibited VWI enhancement (table 2), including within the aneurysm sac and the aneurysm wall, and in some cases in residual, new, or enlarging lobes of the aneurysm or along the neck. AUROC curve values for the time from PED placement to VWI, and PED placement to follow-up angiogram are 0.72 (95% CI 0.52 to 0.93) and 0.74 (95% CI 0.51 to 0.97), respectively, representing fair discrimination of aneurysm enhancement based on the timing of VWI from PED placement. No significant association between aneurysm enhancement and complete occlusion on angiography was noted, although occlusion rates were lower in the wall enhancement group (table 3).
DSA (A) and time-of-flight MR angiography (MRA) (B) demonstrate a saccular aneurysm arising from the left posterior cerebral artery (circle). Follow-up DSA (C) after flow-diverting stent placement demonstrates complete occlusion of the aneurysm. Follow-up MRA demonstrates questionable flow-related enhancement, which is not easily discriminated from shine-through effects from T1 shortening arising from organizing the clot in the excluded aneurysm sac (D). Pre- (E) and post-contrast (F) vessel wall imaging performed at 1-year follow-up demonstrates preserved high-resolution black-blood angiography even across the stented segment in the posterior cerebral artery (dashed arrow). Organizing the clot within the angiographically occluded aneurysm sac (circle) is noted, however, without flow voids in the aneurysm sac itself (case courtesy of Maksim Shapiro, MD).
Adult patient after coil embolization and flow diverting stent treatment of a right paraophthalmic internal carotid aneurysm. DSA (A) demonstrates residual filling of the aneurysm (arrow). The flow-related signal is lost in the region of the aneurysm on time-of-flight MR angiography (B) due to an artifact related to the stent and coil mass. Pre- (C) and post-contrast (D) vessel wall imaging at 1-year follow-up demonstrate a residual flow void in the region of the aneurysm without enhancement, representing residual filling of the aneurysm sac adjacent to the organizing clot within the aneurysm sac (circle).
Associations with angiographically determined complete aneurysm occlusion
Aneurysm occlusion versus enhancement on VWI
Associations with aneurysm enhancement
Focal mural enhancement of the parent vessel was commonly observed (24/27, 88.9%) conforming to the precise location of the proximal and distal landing of the flow diverter (FD) stent (figures 3 and 4), irrespective of either residual aneurysm flow or in-stent stenosis. Stent wall enhancement was significantly higher when aneurysm enhancement was present (92% vs 33%, p=0.049) but not independently associated with either aneurysm residual or in-stent stenosis.
Axial T2-weighted image (A) and DSA (B) demonstrate a large fusiform basilar aneurysm. After coil and flow diverting stent embolization, 3 years follow-up DSA (C) and time-of-flight MR angiography (D) show no evidence of residual aneurysm, with images obscured on both techniques by the large metallic coils mass. Sagittal pre- (E) and post-contrast (F) vessel wall imaging at 3-years' follow-up demonstrates enhancement of the basilar artery wall as well as the of the aneurysm sac (case courtesy of Peter Kim Nelson, MD).
3D DSA (A) and pretreatment lateral DSA image of a posterior communicating artery aneurysm treated with flow diverter. Pre- (C) and post-contrast (D) vessel wall imaging at 1-year follow-up demonstrating resolution of the aneurysm (yellow arrow) without evidence of enhancement. DSA performed a few days later (E and F) demonstrating complete aneurysm occlusion.
Discussion
This study introduces the use of ultra-high resolution black-blood VWI following endovascular flow diversion of intracranial aneurysms, underscoring the frequency with which certainly unexpected and even paradoxical findings may be encountered in real-world practice, probably owing to the unprecedented spatial resolution, flow suppression, gadolinium sensitivity, and artifact immunity of the modern armament of black-blood techniques. The findings highlight the importance of familiarity with the normal imaging appearance and evolution of treated aneurysms in the era of ultra-high resolution imaging of the vessel lumen and wall. Enhancement within and around treated aneurysms on such techniques may represent a potential biomarker of healing, largely decoupled from angiographic occlusion itself.
Although the exact nature of such enhancement will remain a matter of some conjecture, based on its consistency we propose that it reflects both the unprecedented spatial resolution and remarkable sensitivity of MRI to small concentrations of modern high-relaxivity gadolinium chelates, the latter exceeding the sensitivity of X-ray based techniques to iodine by several orders of magnitude.18 19 Notably, five aneurysms in our series were morphologically fusiform, previously reported to enhance more strongly, which may partly explain the results.20
We suggest a heightened attention to vascular, aneurysmal, and perianeurysmal enhancement, but with a measure of circumspection when encountering such findings as a predictor of residual or recurrent disease. Because of the presence of enhancement even with angiographic occlusion, and the highly unlikely occurrence of a true recurrence following successful FD embolization, we are uncertain as to what, if any, biological hemodynamic influence the flow subserving these enhancing areas might have without further longitudinal monitoring or histologic correlation.21 22
Beginning with the Pipeline for Uncoilable or Failed Aneurysms (PUFS) trial, which established safety and efficacy of the Pipeline embolization device,15 flow diversion has proved a changing strategy in the treatment of intracranial aneurysms. The flow diversion established by the PED and other similar devices acts in two ways, first by altering the flow into the aneurysm, hence flow diversion, and second, by providing a scaffold for new endothelialization which serves to exclude the aneurysm from the circulation while reconstructing the parent vessel.
We reiterate that the enhancement patterns described herein arise from presently indeterminate sources, but considering their nearly ubiquitous nature, and their discordance with angiographic occlusion, we believe they might represent a combination of healing effects/endothelialization within and around the aneurysm sac, and potentially the formation of nascent vascular channels, vascularized tissues, or less probably, vasa vasorum recruitment associated with the thrombosed component of the aneurysm, as have been reported in past histological studies.23–25 The timeline for development of these findings following treatment, or to what extent they may have existed prior to treatment, are not conclusively assessed in our cohort owing to the absence of consistent pretreatment VWI, as well as the variable post-treatment durations before VWI is carried out. Based on the existing body of literature examining the vessel wall of untreated aneurysms, we expect that such enhancement could have existed in some of our patients even before treatment, including from inflammatory (eg, immunoglobulin, complement, and lymphocyte-mediated) pathways, myeloperoxidase deposition, or pathologic vasa vasorum formation, which have been reported to predict near-rupture or perhaps even delayed rupture risk in some patients.26–29
In this study, none of the enhancing thrombosed aneurysms ruptured and most were confirmed as occluded on follow-up DSA. The highly effective flow suppression provided by DANTE coupled with variable flip-angle 3D turbo spin echo makes the presence of insufficient flow suppression or artifacts from the metallic construct unlikely despite the well-recognized vulnerability of historic VWI techniques. Because of the persistence of insufficient flow suppression, even in treated aneurysms without residual flow by DSA,30 we would consider possible explanations to be the ongoing cascade of mural inflammation, organizing intra-aneurysmal thrombosis, neointimal angiogenesis, and vasa vasorum recruitment within the adventitia23
Prior studies examining the vessel wall following endovascular coiling reported aneurysmal enhancement as relating to the healing process, although prior studies might have been technically limited by artifact arising from the coil mass.4 Similar findings to ours were reported in a small series of three aneurysms undergoing flow diversion by Guan et al using similar imaging strategies, although at one-eighth the spatial resolution used in our study.7 Several other studies reporting VWI for stented aneurysms,5 did not include FDs, and the difference may be non-trivial considering the unique metallic alloy and mesh framework of PED, which pose unique challenges to imaging and potentially unique reactive and reparative features histologically. Further, the mechanisms facilitating aneurysm thrombosis differ between flow diverters and other stents, with important implications for the time course and nature of healing and aneurysm thrombosis.
Of particular future interest is whether similar findings can be expected with new-generation flow diverters engineered with antithrombotic surface modifications, such as PED Shield, which was recently introduced in the US market and which we expect may find routine use.
Several limitations to our series merit attention, including the relatively small size of the sample cohort and retrospective nature of our review, which could introduce bias.
The enhancement, which might represent a potential biomarker of healing, was evaluated subjectively and as such it is limited in its comparability with other studies and its reproducibility. Accordingly, previously reported approaches to normalization and internal signal references may be beneficial additions to the subjective, qualitative approach described herein.31 32 Moreover, the enhancement might be related to other phenomena, including thickening of the arterial wall, associated vasculopathy, thrombosed aneurysms, or artifactual sources.20 33 34
Our series is one of the first, and the largest such experience with VWI following flow diversion, and to our knowledge the first using advanced flow suppression and acceleration techniques at such resolutions. We also acknowledge that any conclusions about the underlying biology may be conjectural in the absence of histologic confirmation; however, we believe in large part that the experience from mural arteriography and histology in untreated aneurysms offers important lessons and a framework from which to build ongoing paradigms of non-invasive MR monitoring following treatment. The follow-up VWI and the DSA were not performed on the same day and the findings may not reflect identical hemodynamic or pathologic conditions, although we believe the studies to be reasonably well-aligned, having been separated, on average, by approximately 24 days.
To conclude, high performance VWI can be obtained even following commonly prescribed FD treatment of intracranial aneurysms. The spatial and contrast resolution and sensitivity to intravenous gadolinium requires careful interpretation and cognizance of normal imaging patterns to avoid false imputation of aneurysm residua or recurrence. Further study in larger longitudinal cohorts is necessary to better formulate theories of normal evolution and to identify potential biomarkers of treatment response.
Supplemental material
Data availability statement
Data are available upon reasonable request. as above.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by NYU Grossman School of Medicine’s institutional review board. IRB i17-01842 waived consent for this retrospective study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @eytanraz
Contributors ER and SD designed the study. SD and JG created and tuned the imaging protocol. AG-Y, AnD, AhD contributed to data collection. ER and AG-Y drafted the manuscript. All authors contributed to editing the manuscript and approved the final manuscript. SD is the guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests ER: stock holder for Siemens.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.