Research Presentation Session

Neurovascular diseases

Lectures

1
RPS 1111 - Vascular hyperintensities on a post-contrast 3D fast-spin-echo T1-weighted sequence: a sign of poor collateral pathways in sickle-cell disease cerebral vasculopathy?

RPS 1111 - Vascular hyperintensities on a post-contrast 3D fast-spin-echo T1-weighted sequence: a sign of poor collateral pathways in sickle-cell disease cerebral vasculopathy?

05:16C. Provost, Paris / FR

Purpose:

Sickle cell disease (SCD) cerebral vasculopathy is characterised by progressive occlusion of cerebral arteries with the development of extensive collateral pathways to compensate. We aimed to evaluate the prevalence and meaning of post-contrast vascular hyperintensities (PCVH) on a 3D fast-spin-echo T1-weighted sequence in SCD.

Methods and materials:

We included 19 SCD patients with cerebral artery occlusion and 50 SCD patients without occlusion, as detected on MR angiography at 3-Tesla. Two neuroradiologists blinded to clinical data and the other MRI sequences looked for PCVH on a post-contrast 3D fast-spin-echo T1-weighted sequence. We evaluated the concordance between the arterial distribution of occlusions and PCVH. Dynamic susceptibility contrast perfusion imaging was performed to measure Tmax values in the occluded arterial territory.

Results:

Inter and intraobserver agreements for the detection of PCVH were excellent. PCVH was found in 14 of 19 patients (74%), with cerebral artery occlusion and only 1 of 50 patients (2%) without occlusion (p<0.001). This latter patient had bilateral severe stenosis of the cervical internal carotid arteries. The concordance between the arterial distribution of occlusions and that of PCVH was good to excellent (κ=0.73, 0.76, and 0.80 for the anterior, middle, and posterior cerebral arteries, respectively). Tmax values in the occluded arterial territory were higher in PCVH-positive patients compared to PCVH-negative patients (3 seconds versus 2 seconds, p<0.001).

Conclusion:

PCVH in sickle cell disease is a sign of severe cerebral vasculopathy and is associated with marked hypoperfusion. It may represent poor collateral pathways.

Limitations:

The 3D fast-spin-echo T1-weighted sequence was only performed at 3-Tesla, thus limiting the generalisability of the findings.

Ethics committee approval

This study received ethics committee (Comité de Protection des Personnes Île-de-France Saint-Louis) approval.

Funding:

No funding was received for this work.

2
RPS 1111 - White matter lesion volume in subjects with prediabetes, subjects with diabetes, and normoglycemic control subjects

RPS 1111 - White matter lesion volume in subjects with prediabetes, subjects with diabetes, and normoglycemic control subjects

05:55S. Grosu, Munich / DE

Purpose:

As white matter lesions (WML) of the brain are associated with an increased risk of stroke, cognitive decline, and depression, elucidating the associated risk factors is important. In addition to age and hypertension, prediabetes and diabetes may play important roles in the development of WML. Previous studies have, however, shown conflicting results. We aimed to evaluate the association between WML volume and prediabetes/diabetes.

Methods and materials:

400 subjects of the epidemiological KORA study cohort underwent 3T-MRI. WML were manually segmented on 3D-FLAIR images. An oral glucose tolerance test (OGTT) was administered to all participants without previously diagnosed type 2 diabetes. Linear and logistic regression analyses of WML volume and measures of diabetes and diabetes status were conducted while controlling for cardiovascular risk factors.

Results:

The final study population consisted of 388 participants (57% male; age: 56.3±9.2 years). WML were found in 249 participants with a mean volume of 1755±5920 mm3. 98 subjects had prediabetes and 51 established diabetes. Serum glucose concentration determined by OGTT, but not fasting glucose or HbA1c levels, showed a significant association with WML volume after adjustment (p=0.004). WML volume was significantly higher in subjects with prediabetes (p=0.023) and diabetes (p=0.003) compared to normoglycemic control subjects. However, these associations became insignificant after adjustment.

Conclusion:

In this population without prior cardiovascular disease, we identified serum glucose concentration assessed by OGTT as a quantitative determinant of WML volume. Our data suggests that the effect of diabetes, and particularly prediabetes, on WML volume might be driven by impaired glucose tolerance.

Limitations:

The limited sample size. Unmeasured confounding variables cannot be fully ruled out.

Ethics committee approval

Approved by the ethics committee of Ludwig-Maximilian-University Hospital, Munich.

Funding:

German Research Center for Environmental Health and the German Research Foundation (DFG).

3
RPS 1111 - Dynamic computed tomography angiography (dCTA) for determining infarct size and collaterals in order to predict the clinical outcome after recanalisation of acute ischaemic stroke

RPS 1111 - Dynamic computed tomography angiography (dCTA) for determining infarct size and collaterals in order to predict the clinical outcome after recanalisation of acute ischaemic stroke

05:55E. Puglielli, Teramo / IT

Purpose:

To investigate the impact of dCTA weighting on the evaluation of collateral circulation using classic methods and their ability to predict infarct size and clinical outcome. The visualisation of pial collaterals may vary depending on whether the CTA is arterial (A), arterio-venous (AV), or venous (V) weighted.

Methods and materials:

190 consecutive patients (mean age=47 years, April 2009-September 2019) with classical onset underwent endovascular therapy in a single-centre using predominantly stent retrievers, neurotrombectomy devices, or thromboaspiration. Basal CT and dCTA were used for imaging. Two readers studied infarct volume on imaging of the collateral scores stratified by dCTA weighing with age, gender, occlusion aetiology, symptoms, national institutes of health stroke scale (NIHSS) median score at presentation, clot burden score (CBS), and modified Rankin scale (mRS) correlation.

Results:

The median NIHSS score at presentation was 18 (range 3-36) and the onset to treatment time was 123 (76-187) minutes. dCTA scans were A-weighted in 135/190 (71.05%), AV in 25/190 (13.15%), and v-weighted in 30/190 (15.78%). Poor collateralisation (OR 12.50; 95% CI (3.9, 43); p<0.0001) and longer time to peak of maximum arterial enhancement (OR 3.1; 95% CI (1.93, 6.1); p <0.0001) were positively associated with a late venous phase cortical vein filling. All collateral scores were related to infarct volume irrespective of dCTA weighting (p<0.005). No association was shown between dCTA weighting, collateral grade, and clinical outcome.

Conclusion:

dCTA weighting did not significantly impact collateral grade analysis with the three common collateral scores and there is a lack of evidence of their ability to predict final infarct size after recanalisation

Limitations:
The possible error due to a small series.
Ethics committee approval

n/a

Funding:

No funding was received for this work.

4
RPS 1111 - The arterial remodelling ratio and normalised wall index of the middle cerebral artery are associated with leptomeningeal collateral status: a study of high-resolution vessel wall imaging

RPS 1111 - The arterial remodelling ratio and normalised wall index of the middle cerebral artery are associated with leptomeningeal collateral status: a study of high-resolution vessel wall imaging

05:16R. Tang, Beijing / CN

Purpose:

Arterial stenosis is not sufficient for stratifying stroke risk. In this study, we evaluated atherosclerotic plaque features of the middle cerebral artery (MCA) using high-resolution vessel wall imaging (HR-VWI) and explored the correlation between plaque features and leptomeningeal collateral (LMC) status.

Methods and materials:

61 patients with intracranial atherosclerotic disease who underwent HR-VWI scans (September 2016-December 2018) were retrospectively included. All patients presented with symptoms of ischaemic stroke or transient ischaemic attack with confirmed stenosis >30% in M1-2 segments. Plaque features were measured at the cross-section with most lumen narrowing (MLN) on HR-VWI. LMC was graded per patient on strategically acquired gradient echo (STAGE) derived magnetic resonance angiography (MRA) images using an established five-point collateral score (CS) system. Patients were divided into a poor (CS=0-2) or good (CS=3-4) LMC group. Independent two-sample t-tests, Mann-Whitney U-tests, or one-way analysis of variance was used as appropriate. Univariate and multivariate logistic regressions were conducted to identify the plaque features correlated with LMC status.

Results:

Larger outer wall area (OWA), lumen area (LAMLN), wall area, less stenosis percent (S%), M1 segment, and inferior wall distribution were found in the good LMC group compared with the poor LMC group (all p<0.05). Significant associations between the arterial remodelling ratio (ARR, odds ratio [OR] 3488.108, 95% confidence interval [CI] 1.198-2.637, p=0.004) and normalised wall index (NWI, OR 132.686, 95% CI 0.693-0.978, p=0.027) with LMC status were found.

Conclusion:

Plaques in the poor or good LMC group exhibit different morphology and distribution features, among which ARR and NWI are associated with LMC status, which may help determine therapeutic decision-making in the clinical setting.

Limitations:

n/a

Ethics committee approval

n/a

Funding:

No funding was received for this work.

5
RPS 1111 - The role of diffusion-weighted brain MR imaging in predicting outcomes after comatose cardiac arrest

RPS 1111 - The role of diffusion-weighted brain MR imaging in predicting outcomes after comatose cardiac arrest

05:34K. Coursier, Genk / BE

Purpose:

To assess a qualitative and quantitative approach to diffusion-weighted brain MR imaging (DWI) in order to investigate its role in predicting clinical outcomes after cardiac arrest.

Methods and materials:

75 patients underwent brain MR imaging within 5±2 days after a cardiac arrest. Two observers independently used a four-point Likert scale scoring template per assessed brain area on high b-value diffusion-weighted images (DWI). Intra and interobserver variability were tested using the intraclass correlation coefficient. ROC/AUC analysis was used for the outcome correlation with the clinical outcome defined by the cerebral performance category status at 180 days. Quantitative parameters on DWI were the percentage of voxels with apparent diffusion coefficient (ADC) <0.650 x 10-3 mm2/s and the mean whole-brain ADC value.

Results:

Intraobserver variability was excellent for the total cerebral cortex (TCC) score and for the total grey nuclei (TGN) score. Interobserver variability was excellent for the TCC and good to excellent for the TGN score. ROC analysis of the qualitative DWI showed a good correlation with the outcome for the TCC and for the TGN (AUC=0.83) score, whereas the quantitative DWI parameters correlated very poorly with the outcome (AUC 0.57-0.60).

Conclusion:

Qualitative assessment of DWI acquired early after cardiac arrest provides valuable information to predict the outcome after cardiac arrest, while the correlation of the quantitative ADC assessment with the outcome was poor.

Limitations:

n/a

Ethics committee approval

Study is based on a clinical trial that was approved by the ethics committee of UZ Leuven K.U.L., s58017.

Funding:

Study is based on a clinical trial that was funded by a non-commercial TBM grant from the Flemish Government (IWT Flanders, Belgium).

6
RPS 1111 - Imaging of endovascularly treated intracranial aneurysms: comparing the novel technique of ASL-based silent-MR angiogram with 3D TOF-MR angiogram and digital subtraction angiogram

RPS 1111 - Imaging of endovascularly treated intracranial aneurysms: comparing the novel technique of ASL-based silent-MR angiogram with 3D TOF-MR angiogram and digital subtraction angiogram

05:49A. Rajendran, Chennai / IN

Purpose:

To compare silent-MRA and 3D TOF-MRA in the evaluation of follow-up imaging of endovascularly treated intracranial aneurysms in comparison with DSA as the criterion standard for the assessment of the residual aneurysms and stent patency.

Methods and materials:

A prospective analysis of consecutive patients with endovascularly treated aneurysms either by coiling or flow diverter who had follow-up imaging with TOF, silent-MRA, and DSA. The results were analysed by two neuro-radiologists independently. 3D TOF-MRA, silent-MRA, and DSA were evaluated for the size of the residual aneurysms and graded by the Raymond–Roy scale. The patency of the stent was also analysed. Images were analysed for the quality and graded on a 5-point scale. Kappa parameters were used to look for the agreement between the two methods. Scores were averaged and compared between techniques by using a Wilcoxon signed-rank test.

Results:

19 patients (4=stent-assisted coiling, 6=flow-diverters, and 9=simple or balloon-assisted coiling) were included.

The mean image quality score for residual aneurysm evaluation was 3.89+/-0.6 for TOF-MRA and 4.68+/-0.46 for silent-MRA. Scores of silent-MRA and 3D TOF-MRA differed significantly (P<0.05) according to the Wilcoxon signed-rank test.

The mean image quality score for the assessment of the stent or flow-diverter patency was 2.8+/-1.07 for TOF-MRA and 4.2+/-0.6 for silent-MRA. Scores of silent-MRA and 3D TOF-MRA differed significantly (P<0.05) for evaluating the stent or flow diverter patency.

Intermodality agreement values of 0.90 and 0.40 were obtained for DSA/silent-MRA and DSA/3D TOF-MRA, respectively.

Conclusion:

Silent-MRA provides more accurate imaging information and better image quality than conventional TOF-MRA in the evaluation of residual aneurysms and stent patency by reducing the magnetic susceptibility artefacts.

Limitations:

The limited number of cases.

Ethics committee approval

Ethics committee approval obtained.

Funding:

No funding was received for this work.

7
RPS 1111 - Follow-up of intracranial aneurysms treated with endovascular techniques: a comparison of the different methods with a review of the literature

RPS 1111 - Follow-up of intracranial aneurysms treated with endovascular techniques: a comparison of the different methods with a review of the literature

05:25E. Puglielli, Teramo / IT

Purpose:

To compare time-of-flight 3D-TOF magnetic resonance angiography (MRA), contrast-enhanced CE-MRA at 1.5-Tesla, computed tomography angiography (CTA), and digital subtraction angiography (DSA) for evaluating aneurysm occlusion and parent artery patency after coiling or stent-assisted coiling with a review of the recent literature.

Methods and materials:

in this retrospective single-centre study, 427 patients were included if they had an intracranial aneurysm treated by coiling/stent-assisted coiling (April 2009-June 2019) followed by MRA (3D-TOF-MRA and CE-MRA), CTA or DSA, performed in an interval of 6, 12, or 24 months. Pooled sensitivity and specificity were calculated using the aneurysm occlusion status as defined by the Raymond-Roy occlusion grading scale.

Results:

Vessels visualisation in the vicinity of the treated aneurysm was better in CTA, while regarding aneurysm occlusion, evaluation of the agreement with DSA was better for CE-MRA (K=0.51) than 3D-TOF-MRA (K=0.25). Diagnostic accuracies for the aneurysm remnant depiction were similar for 3D-TOF-MRA and CE-MRA (P=1). For TOF-MRA, the sensitivity and specificity of all aneurysms undergoing endovascular therapy were 87% and 93%, respectively, in accordance to the recent literature.

Conclusion:

After coiling or stent-assisted coiling treatment, 3D-TOF-MRA and CE-MRA demonstrated good accuracy in detecting the aneurysm remnant, although it tends to overestimate. Vessels visualisation in the vicinity of treated aneurysm was better in CTA. Despite CE-MRA, agreement with DSA was better; there was no statistical difference between 3D-TOF-MRA and CE-MRA accuracies. Both MRAs were unable to provide a precise evaluation of the in-stent status but could detect the parent vessel occlusion. While digital subtraction angiography (DSA) remains the gold standard, magnetic resonance angiography (MRA) is attractive as a non-invasive follow-up technique.

Limitations:

The small series.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

8
RPS 1111 - Clinical radiomics nomogram for the risk estimation of haematoma expansion after intracerebral haemorrhage

RPS 1111 - Clinical radiomics nomogram for the risk estimation of haematoma expansion after intracerebral haemorrhage

05:15Y. Yang, Wenzhou / CN

Purpose:

Early haematoma expansion (HE) following intracerebral haemorrhage (ICH) is strongly associated with a poor outcome. The purpose of this study was to develop and validate a clinical radiomics nomogram for predicting HE in patients with acute spontaneous ICH.

Methods and materials:

In total, 1,153 eligible patients were enrolled, of whom 864 (75%) were assigned to the derivation cohort and 289 (25%) to the validation cohort. Based on the LASSO algorithm or multivariate analysis, three models (the clinical model, the radiomics model, and the hybrid model incorporating both clinical and radiomics predictors) were constructed to predict HE. The Akaike information criterion (AIC) and likelihood ratio test (LRT) were used for comparing the goodness of fit of the three models, and the AUC was used to evaluate their discrimination ability for HE.

Results:

The hybrid model (AIC=681.426; χ2=128.779) was determined as the optimal model with the lowest AIC and highest Chi-square values compared with the radiomics model (AIC=767.979; χ2=110.234) or clinical model (AIC=753.757; χ2=56.448). The radiomics model had better discrimination ability for HE than the clinical model in both derivation (p=0.009) and validation (p=0.022) cohorts. In both datasets, the clinical radiomics nomogram showed satisfactory discrimination and calibration in predicting HE (AUC=0.771, sensitivity=87.0%; AUC=0.820, sensitivity=88.1%; respectively).

Conclusion:

Integrating CT-based radiomic signatures with clinical risk factors could optimise the predictive performance for HE. In the absence of CT angiography, the clinical-radiomics nomogram can offer an individualised tool for the risk stratification of HE with good sensitivity and may help select more ICH patients for anti-expansion clinical trials.

Limitations:

The nomogram is derived from a single-centre analysis, thus lacking external validation.

Ethics committee approval

n/a

Funding:

No funding was received for this work.

9
RPS 1111 - Accelerated time-of-flight magnetic resonance angiography using spiral imaging improves the conspicuity of intracranial arterial branches whilst reducing scan time

RPS 1111 - Accelerated time-of-flight magnetic resonance angiography using spiral imaging improves the conspicuity of intracranial arterial branches whilst reducing scan time

05:50N. Sollmann, Munich / DE

Purpose:

To compare time-of-flight magnetic resonance angiography (TOF-MRA) acquired with compressed SENSE (TOF-CS) to spiral imaging (TOF-Spiral) for the imaging of brain-feeding arteries.

Methods and materials:

71 patients (60.2±19.5 years, 28.2% with pathology) who underwent TOF-MRA after the implementation of new scanner software enabling spiral imaging were analysed retrospectively. TOF-CS (standard sequence; duration: ~4 min) and the new TOF-Spiral (duration: ~3 min) were acquired. Image evaluation (vessel image quality and detectability, diagnostic confidence [1: diagnosis very uncertain-5: diagnosis very certain], quantitative measurement of aneurysm diameter, or degree of stenosis according to NASCET criteria) was performed by two readers. Quantitative assessments of pathology were compared to computed tomography angiography (CTA) or digital subtraction angiography (DSA).

Results:

TOF-CS showed higher image quality for intraosseous and intradural segments of the internal carotid artery, while TOF-Spiral better depicted small intracranial vessels like the anterior choroidal artery. All vessel pathologies were correctly identified by both readers for TOF-CS and TOF-Spiral with high confidence (TOF-CS: 4.4±0.6 and 4.3±0.7; TOF-Spiral: 4.3±0.8 and 4.3±0.8) and good inter-reader agreement (Cohen’s kappa: >0.8). Quantitative assessments of the aneurysm size or stenosis did not significantly differ between TOF-CS or TOF-Spiral and CTA or DSA (p>0.05), with excellent correlations between the measurements of both readers (ICC>0.95).

Conclusion:

TOF-Spiral for the imaging of brain-feeding arteries enables reductions in scan time without drawbacks in diagnostic confidence. A combination of spiral imaging and CS may help to overcome the shortcomings of both sequences alone and could further reduce acquisition times.

Limitations:

The retrospective design and comparatively low number of vessel pathologies observed among the enrolled patients.

Ethics committee approval

This retrospective study was approved by the IRB.

Funding:

Philips Healthcare.

10
RPS 1111 - Acute symptomatic lacunar ischaemic stroke as the first presentation of small vessel disease: how common is it?

RPS 1111 - Acute symptomatic lacunar ischaemic stroke as the first presentation of small vessel disease: how common is it?

05:39A. Guarnera, Roma / IT

Purpose:

To verify the specificity of non-EPI DWI-MRI in patients operated on for middle ear cholesteatoma who showed positivity at imaging performed 6 to 9 months after surgery and underwent second-look surgery.

Methods and materials:

In a single-centre, a consecutive cohort of patients treated for cholesteatoma and undergoing 1.5T non-EPI DWI-MRI 6 to 9 months after surgery was evaluated. Patients showing a hyperintense signal in the middle ear underwent revision surgery and were included in the study, whilst the others were still under radiological follow-up and were excluded. Two radiologists independently placed an ROI inside the brightest part of the observed signal alteration on coronal HASTE-DWI images. The mean signal intensity (SI) and maximum (SImax) SI values were recorded for each patient. Signal intensity ratios (SIR) were calculated using the inferior temporal cortex (SIRT) and the background noise (SIRN) as references.

Results:

143 subjects (210 ears) were evaluated. 27 subjects (40 ears) showed a high signal lesion inside the middle ear and underwent revision surgery. 36 ears were confirmed to be affected by residual/recurrent cholesteatoma. In 4 ears, inflammatory tissue was found. According to the ROC analysis, SI, SIRT, and SIRTmax showed the best statistical values (AUC=1).

Conclusion:

Residual/recurrent cholesteatoma can be accurately detected using quantitative evaluation of non-EPI DWI MRI, which may avoid a revision surgery if negative.

Limitations:

The analysed cohort was relatively small. The design of this study excluded patients with normal DWI since false-negative DWI-MRI in post-surgical population is only 3% of cases.

Ethics committee approval

The study was approved by the Institutional Ethics Board and in line with the Declaration of Helsinki. Informed consent was obtained for MR examination.

Funding:

No funding was received for this work.

PEP Subscription Required

This course is only accessible for ESR Premium Education Package subscribers.