Ammad Mahmood, MBChB
@AMahmoodNeuro

European Stroke Organisation-World Stroke Organization 2020 Virtual Conference
November 7-9, 2020

This session covered a number of novel techniques used in stroke imaging with a number of distinguished speakers. Cutting-edge advances in pre-hospital imaging in the ambulance, imaging used in making decisions about reperfusion therapies, and diagnostic imaging used in aetiological classification were all discussed.

Pre-hospital imaging – Jeffrey Saver
Advantages of triage in the field of patients into ischemic (LVO and non-LVO) and hemorrhagic stroke enables routing to the most appropriate facility, minimizing delays to treatment access and the need for inter-hospital transfers. Quasi-imaging approaches involving innovative approaches such as radio frequency, infrared, or EEG ‘stroke helmets’ may help identify patients with LVO in the future. Ultrasound assessment of vessel patency in the field yielded 91% accuracy in identifying LVO in one study. Mobile stroke units with CT capability are already in use and allow CT, CTA, and possibly CTP in the field. Automated image processing software can aid in interpretation of CT results quickly. Future technologies may include mobile stroke units in helicopters or small aircraft, mobile neuro-intervention suites, and mobile MRI imaging. Prof. Saver presented some cases from the Los Angeles mobile stroke unit experience highlighting the optimization of patient management achieved through early triage. Lastly, he highlighted the ongoing BEST-MSU trial, which will evaluate the benefit of MSUs in stroke management in several cities in the United States.

Automated image processing software – Bruce Campbell
The advantages of using processing software include speed, improved sensitivity, and standardization of analysis, particularly in settings where local expertise in image interpretation may not be available. Based on clinical trials which have utilized CT perfusion analysed volumes as inclusion criteria, software can help identify patients eligible for thrombolysis and thrombectomy, particularly late window patients. Treatment decisions in groups such as minor stroke or patients with co-morbidities who may not fit usual trial criteria can be made easier by looking for a perfusion mismatch. Prof. Campbell highlighted the capabilities of several packages, including automated ASPECTS assessment, hypodensity volume estimation, hyperdense artery identification, hemorrhage identification, LVO detection, collateral assessment, and ischemic core and penumbra prediction.

Carotid plaque imaging – Marie-Luise Mono
Asymptomatic moderate to severe carotid stenosis or symptomatic mild carotid stenosis are both challenging scenarios when determining stroke aetiology. Detailed imaging of plaques can reveal features of vulnerable plaques, such as intra-plaque hemorrhage, which contribute to the enlargement of the lipid-rich necrotic core and rapid plaque progression. Modalities used can include CT, MRI, and ultrasound; MRI and ultrasound provide the best imaging of plaque characteristics. Whilst most of these techniques remain research-based, T1 MRI imaging is widely available and could be used to identify high-risk plaque.

Comprehensive cardiac assessment by CT – Keith Muir
Secondary prevention beyond antiplatelets and statin depends upon the aetiological mechanism identified by investigations after stroke. The cause is cryptogenic in a significant proportion of cases. The standard of investigation which needs to be completed before declaring a case ‘cryptogenic’ is not universal, and access to modalities of investigation is variable worldwide. Delayed aetiological diagnosis leads to delayed management during the period of greatest risk for recurrent stroke. One option for rapid, universally available, combined assessment of cardioembolic and large vessel disease is contrast CT extended beyond the aortic arch to include cardiac imaging. Prof. Muir presented results of the D-CCIST pilot study from our center in Glasgow, which compared comprehensive cardiovascular imaging (CCI) which comprised of cardiac CT, CTA arch to vertex, and 3T MRI brain at a single visit against routine imaging for patients presenting within 7 days of TIA or acute stroke. A significantly higher proportion of patients with CCI had a definite aetiology identified by day 7 though not at day 30. Time to access investigations was not different between groups, but the time to identification of the aetiological cause was significantly shorter in the CCI group (51 hours) vs routine (82 hours). A larger trial of this approach to determine whether it can reduce recurrent clinical events is warranted.

When is arterial wall imaging useful? – Catherine Oppenheim
While imaging of the vessels with angiography is widely used and interpreted in acute stroke, vessel wall imaging remains a specialized assessment. Prof. Oppenheim highlighted 4 cases where MRI assessment of the vessel wall was used to make the diagnosis. The T1 MRI sequences used are possible on most scanner types. The cases included progressive carotid stenosis with intraplaque hemorrhage; transient perivascular inflammation of the carotid artery (TIPIC syndrome); dissection of the basilar artery; and VZV related cerebral vasculitis in a patient with HIV.

Imaging in wake-up stroke – Jochen Fiebach
Lastly, one of the authors of the WAKE-UP trial described the concept of DWI-FLAIR mismatch, which was developed from serial MRI imaging of patients with known times of symptom onset. In this cohort, it became clear that the majority of patients with DWI lesions did not develop FLAIR lesions until around 4.5 hours. This led to the WAKE-UP trial, which demonstrated a clear benefit in functional outcome for patients with DWI-FLAIR mismatch receiving thrombolysis versus placebo. A helpful online training tool from the WAKE-UP cohort has also been developed, allowing trainees to analyze a set of scans from the trial for DWI-FLAIR mismatch.

Ammad Mahmood, MBChB
@AMahmoodNeuro

European Stroke Organisation-World Stroke Organization 2020 Virtual Conference
November 7-9, 2020

Welcome
Profs. Jesse Dawson and Michael Brainin welcomed all the delegates to the virtual conference and a chance to bring stroke care back into focus after a difficult year focusing on the COVID-19 pandemic. They spoke about the hard work that has gone into the rearrangement of this year’s conference into a virtual experience, and we thank them and their teams for their efforts in bringing the conference to us in these difficult times.

Fluoxetine after stroke
The first two presentations focused on the use of fluoxetine after stroke to assess any benefit in stroke rehabilitation and functional outcome. These trials came from an international collaboration of 3 trials focusing on this topic. First, Erik Lundstrom from Sweden presented the results of the EFFECTS trial. 1500 patients, at 35 Swedish centers, with ischemic or hemorrhagic stroke in the last 2-15 days but no history of depression were randomized 1:1 to receive either fluoxetine or placebo. The primary outcome of modified Rankin scale at 6 months demonstrated a neutral result with odds ratio 0.94 (0.78-1.13) at 6 months. Secondary outcomes demonstrated a small decrease in depression after stroke but a small increase in rates of fractures. Next, Graeme Hankey from Australia presented the 12-month results from the AFFINITY trial, which had also shown neutral results when the main trial results for 6 months outcome were previously published. The 12-month outcomes also showed a neutral effect of fluoxetine in functional outcome showing there is no delayed benefit. A small reduction in the rate of recurrent ischemic stroke in the fluoxetine group was seen in the AFFINITY trial, but not in the two other larger trials; therefore, this was felt to be a random effect. Overall, the results of the EFFECTS, AFFINITY, and FOCUS trials demonstrated that there is no benefit of fluoxetine after stroke in improving rehabilitation and functional outcome.

Ticagrelor in minor stroke and TIA
Next, Clay Johnston from Texas presented the results of the THALES trial. Dual antiplatelet therapy with aspirin and clopidogrel has become common practice after the POINT and CHANCE trials. The SOCRATES trial did not show any benefit on ticagrelor monotherapy versus aspirin monotherapy, and the THALES trial now examined dual antiplatelet therapy with ticagrelor and aspirin versus aspirin alone in minor stroke (NIHSS ≤5) and high-risk TIA (ABCD2≥6 or >50% symptomatic arterial stenosis) within 24 hours. 11016 patients were randomized 1:1 at 414 sites in 28 countries. Results showed a benefit of ticagrelor and aspirin over aspirin alone in the primary outcome of reduction in recurrent stroke or death at 30 days with a hazard ratio of 0.83 (0.71-0.96); most of the benefit was attributable to reduction in recurrent stroke in the first 7 days. However, there was a significant additional risk of hemorrhage in the ticagrelor group with a hazard ratio of 3.99 (1.74-9.14), though a small absolute difference. Overall, the authors helpfully summarized the effect as “for every 1000 patients treated, 11 strokes or deaths were prevented but 4 severe hemorrhages were produced.”

EVT in basilar artery occlusion
Wouter Schonewille from the Netherlands presented the results of the BASICS trial tackling one of the most difficult clinical problems in acute stroke, basilar artery occlusion. The trial randomized patients to either endovascular treatment plus best medical management (EVT + BMM) or best medical management (BMM) alone within 6 hours of the estimated time of basilar artery occlusion. They allowed inclusion of cases with stuttering onset as the time of basilar artery occlusion was taken as the point of acute severe deterioration. 300 patients were randomized in Europe and Brazil. The primary outcome of favorable functional outcome was neutral with 44.2% of EVT + BMM and 37.7% of BMM achieving mRS 0-3 with a risk ratio of 1.18 (0.92-1.50). There was no significant difference in rate of symptomatic ICH or mortality. The absolute risk reduction of 6.5% in the EVT group suggested a trend towards benefit of EVT. In subgroup analysis, patients presenting with NIHSS >10 showed a significant benefit of EVT. The authors concluded that EVT was safe within 6 hours and may be a better option for patients with NIHSS>10. The role of EVT beyond 6 hours and with minor deficit remains unclear.

LVO – where to?
Planning of stroke services worldwide involves deciding whether patients should be taken by the ambulance initially to the nearest stroke center or direct to a comprehensive stroke center where EVT is available. Next, Natalia Perez de la Ossa and Marc Ribo from Catalonia presented the results of the RACECAT trial. 1401 patients who resided in areas served by non-EVT stroke centers with suspected LVO according to the RACE score were randomized while in the ambulance to either proceed to the local stroke center (drip and ship) or direct to the EVT-capable center (mothership). The primary outcome of functional outcome measured by mRS in patients with ischemic stroke only was neutral. Mean time from symptom onset to arrival at the first center was 142 minutes in the drip-and-ship model and 216 minutes in the mothership model. Two-thirds of patients identified by the RACE score had ischemic stroke and two-thirds of these had LVO. Symptom onset to IV thrombolysis and EVT was 120 minutes and 270 minutes, respectively, in the drip and ship patients, and 155 minutes and 214 minutes, respectively, in the mothership patients. Overall, this well-designed trial answers an important question, and although planning of stroke services will always depend upon local needs, the validation of both drip-and-ship and mothership models is valuable.

Cryptogenic stroke – PFO closure vs antiplatelets
Next, Scott Kasner presented long-term outcomes from the REDUCE study of PFO closure vs antiplatelet therapy. Patients aged 18-59 with cryptogenic stroke and a PFO confirmed by echocardiography were randomized to either PFO closure plus antiplatelet therapy or antiplatelet therapy alone. In keeping with the previously published results in earlier follow up, the PFO closure group had a lower rate of recurrent ischemic stroke with a hazard ratio of 0.31 (0.13-0.76) though the absolute numbers of recurrent events are small especially in the latter years of the 5-year follow-up period. Significantly, more patients in the PFO group developed atrial fibrillation in the follow-up period. Most AF was peri-procedure though 2.7% of the PFO group (vs 0.4% in antiplatelet only) developed persistent or permanent AF and required anticoagulation. 

Closing
A Q&A panel was held to close out the session with speakers exploring some of the points of interest. Overall, despite the neutral results of many of the trials presented, important take-home messages about issues that the stroke community grapples with on a day-to-day basis — such as ‘mothership vs drip-and-ship’ care pathways, optimal management of basilar artery occlusion, and cryptogenic stroke — all contributed to an informative opening session at the conference. Of note, whilst this year’s conference aimed for greater gender balance and wider geographic representation, the large clinical trials sessions were noticeably lacking in this regard and demonstrate areas of improvement for the stroke community. More female representation and experiences of stroke care in the developing world shared on the largest stage at global conferences should continue to be in our collective aims for future conferences.

Tolga D. Dittrich, MD

European Stroke Organisation-World Stroke Organization 2020 Virtual Conference
November 7-9, 2020

Scientific Session: “Artificial Intelligence in Stroke Imaging,” Sunday, November 8, 2020
Speakers: Susanne Wegener, Roland Wiest, Paul Bentley, Kim Mouridsen, Sook-Lei Liew
Chairs: Kim Mouridsen, Susanne Wegener

Machine learning (ML) methods as a component of artificial intelligence are a growing field in stroke imaging research. We are already familiar with such automated evaluation systems, such as ASPECT scoring or mismatch volume calculation. Nevertheless, clinicians are often confronted with a complex mixture of different clinical, laboratory, and radiological parameters that must be weighed against each other to make an individual therapeutic decision.

“Machine learning is a precise mathematical way in which we can do this in a reliable, objective manner,” said Paul Bentley, of Imperial College London. Unlike conventional image interpretation, an algorithm can evaluate radiological source data to derive applicable rules. ML approaches are particularly promising for objectifying imaging results and detecting subtle changes in the context of intricate radiological findings in acute ischemic stroke. However, to provide additional information, ML methods need a relatively large set of initial data. This limitation especially becomes relevant in the context of imaging-based prediction of stroke recovery and rehabilitation response, where imaging does not constitute a common clinical component, as Sook-Lei Liew from the University of Southern California emphasized in her lecture.

The potential of ML in both acute stroke and stroke rehabilitation imaging is broad. In the future, ML-based techniques, for example, in ischemic core imaging in the extended time window, could help us to identify better patients who could benefit from endovascular treatment.

Aurora Semerano, MD
@semerano_aurora

European Stroke Organisation-World Stroke Organization 2020 Virtual Conference
November 7-9, 2020

The Controversy sessions in the ESO-WSO 2020 Conference are intriguing live Q&A sessions focused on grey zones in real-world stroke care, with stroke experts defending their points of view and facing each other in interesting rounds of discussion. On the first day of this ESO-WSO 2020 Conference, the session addressed the following hot topics about endovascular thrombectomy.

Do We Need Perfusion Imaging to Guide MT In Extended Time Window?

YES: Götz Thomalla (Germany) presented the pros of using advanced perfusion imaging for patient eligibility to mechanical thrombectomy (MT) in the extended time window (>6h from onset). He invited us to remember that, first of all, we should rely on current evidence. According to the evidence, whereas perfusion techniques are not required in the early time window, the DAWN and DEFUSE III clinical trials firmly base on advanced imaging for patient selection in the late time window. As a consequence, also AHA/ASA guidelines, ESO consensus, and ESMINT guidelines recommend advanced imaging for patient eligibility. He also pointed out that we cannot reliably trust the ASPECTS score for estimating viable tissue. In addition, against the opinion that perfusion techniques are time-consuming and difficult to interpret, Prof. Thomalla reassured that just a few minutes are needed to perform reliable perfusion imaging and that simple parameters are required for map interpretation: Tmax> 6 seconds identifies penumbra and CBF<30% works for infarct core. This little extra time is well invested to effectively guide reperfusion treatments.