World Stroke Congress
October 17-20, 2018

Ravinder-Jeet Singh, MBBS, DM

The session on advances in stroke imaging was very interesting and informative. Three speakers — Dr. Ken Butcher from Edmonton and Drs. Gregory Albers and Chitra Venkatasubramanian from Stanford — discussed some of the recent changes in the field of stroke imaging and how it is influencing the practice of stroke care.

Dr. Butcher spoke about multimodal CT based imaging paradigms in acute ischemic stroke, especially use of CTA and CTP in patient selection. It is of note that two of the “early window” EVT trials (SWIFT PRIME and EXTEND IA) and both the late window trials (DAWN and DEFUSE 3) used perfusion imaging for patient selection. He discussed various software platforms available for processing and interpreting CTP data. Options include inbuilt software provided by CT vendors and other commercially available software including RAPID (iSchemaView Inc.), MIstar (Apollo Medical Imaging Technology), and OleaSphere (Olea Medical). The key message was that clinicians should be careful in interpreting CTP images and also the outputs (core, penumbra volumes, and mismatch ratios) provided by these softwares. It is essential to look at arterial input function and venous output function curves to assess the quality of CTP data. These curves should display rapid upslopes and downslopes. Presence of any truncation of these curves should also be noted, which could lead to misestimation of core and penumbra size. In addition, the presence of motion artifacts can severely degrade image quality and, therefore, its interpretation. He also pointed out another important area in which CTP could be helpful — the stroke mimics — where a normal CTP would be reassuring in the appropriate clinical context.

Dr. Albers talked about imaging-based patient selection into the recent endovascular trials. He initially emphasized the variation in the rates of good functional outcomes among the landmark “early window” EVT trials and explained it based on the difference in the imaging paradigms used for patient selection in these trials. For example, MR CLEAN did not use any core size for patient exclusion and showed smaller of the effect size compared to the ESCAPE trial, which used ASPECTS for core estimation, vs. SWIFT PRIME and EXTEND IA, which encouraged/employed CTP for core size calculation and patient exclusion and, therefore, demonstrated the larger effect size. All the late window trials (DAWN and DEFUSE 3) employed CTP to determine penumbra using mismatch paradigms either clinical-core or perfusion-core mismatch and thus the patient selection. Core size, the definition of which varied in both trials, was used to exclude patients. Interestingly, treatment effect size was even larger in the late window trials than the early window trials, termed “late window paradox” (published by Dr. Albers in Stroke 2018;49:768-771). He explained this phenomenon by the behavior of control arm in the late window trials, which did not receive intravenous thrombolysis and thus performed poorly. This widened the difference (effect size) between the good outcome seen in the EVT vs. the control arm. The favorable outcomes (modified Rankin score 0-2) in the control arm was seen in 27% of patients in the early window EVT trials (meta-analysis, HERMES collaboration) compared to only 15% in the late window trials. He discussed AURORA (a meta-analysis of late presenters in ESCAPE and patients in the DAWN and DEFUSE 3 trials) and showed that late presenting patients did well irrespective of whether the selection was based on a clinical mismatch or a target mismatch. Interestingly, late presenting patients in the ESCAPE trial, which did not use perfusion imaging, also showed treatment benefit. He finally alluded to future opportunities of trials in 4.5-24 hours window TNK trial using DEFUSE 3 selection criteria and EVT for low NIHSS, distal occlusion, and large core.

Dr Venkatasubramanian discussed the utility of some of the newer techniques in determining the etiology of intracerebral hemorrhage (ICH) in patients having secondary ICH. She outlined clinical and radiological features which should raise suspicion for secondary ICH. Salient clinical features were: ICH at a young age, female gender, the absence of conventional risk factors (especially hypertension, coagulopathy, smoking), family history of ICH and clinical circumstances (fever, smoldering neurological symptoms, prior stroke/TIA, history of malignancy). Radiological clues included location (lobar, infratentorial, SAH), enlarged vessel or calcification at the margins of hemorrhage, unusual shape of hematoma, edema out of proportion to ICH, multiple ICH, other brain masses and dense venous sinuses. Among the special imaging, she discussed dual-energy CT and florbeta PET and also arterial spin labelling (ASL) perfusion. Dual-energy CT is especially helpful to distinguish post-EVT contrast staining of brain parenchymal/subarachnoid spaces vs. hemorrhage. By using iodine mask, the two can be very easily differentiated. This might be very helpful in a situation in which clinician needs to consider early antithrombotic, for example, carotid stenting during EVT. She also discussed the DASH study, which assessed the utility of MRI in investigating etiology of spontaneous ICH/IVH and demonstrated the additive clinical benefit of early routine MRI. Finally, the ASL technique could help in suggesting high flow lesion — for example, dural AV fistula — as an etiology of ICH even when conventional MR sequences are not of much help.

All in all, the session was very useful in understanding how the imaging is helping in the better selection of ischemic stroke patients to reperfusion therapies and also to determine the etiology of secondary ICH.