Cerebral oedema after reperfusion therapy in patients with ischaemic stroke : predictors, outcomes and treatment

Abstract

Introduction: Reperfusion therapy by intravenous thrombolysis (IVT) and/or endovascular thrombectomy (EVT) are established treatments in ischaemic stroke. Cerebral oedema (COED), caused by dysfunction the blood brain barrier (BBB), is common early after acute ischaemic stroke (AIS), can aggravate the symptoms and worsen the prognosis. Data on predictors and the effect of recanalization on early COED is limited. A large infarction with COED involving the middle cerebral artery (MCA) can be life-threatening. Decompressive hemicraniectomy (DHC) reduces mortality and may have a positive effect on functional outcome in younger patients. Animal data suggest that imatinib, a tyrosine kinase inhibitor, may restore BBB integrity, thereby reducing haemorrhagic transformation (HT) and COED. The aim of this doctoral thesis was to contribute to the understanding of clinical aspects of COED in patients with AIS of the anterior circulation.

Methods: Paper I, II and IV reported retrospective, observational studies using data from the Safe Implementations of Treatments in Stroke (SITS) International Stroke Registry, a prospective, multinational registry. These studies included patient data recorded using the SITS Registry data collection protocols for IVT and EVT, and to some extent general stroke, in time periods between 2002 and 2019. All patients had presumed ischaemic stroke. Paper III reported a phase 2, randomized, open-label, pilot study of imatinib in patients who received IVT after ischaemic stroke at 5 hospitals in Stockholm 2011-2014. All papers evaluated COED using the SITS COED scale (no, mild, moderate or severe COED). Outcomes at 3 months were functional outcome using the modified Rankin scale (mRS) score and death of any cause.

Results: In paper I, the most important predictors of COED after AIS were assessed. Among 42 187 patients (median age 70 years), 12.5% had mild COED on follow-up imaging (22-36 hours or any extra investigation) and 10.2% had moderate or severe COED. Baseline National Institutes of Health Stroke Scale (NIHSS) score, followed by hyperdense artery sign (HAS), were the strongest predictors for COED. Additionally, higher blood glucose, impaired level of consciousness and imaging signs of early infarction at baseline were predictors for COED. Increasing degree of COED at 22-36 hours was associated with increasing mortality and worse functional outcome at 3 months. In paper II, the effect of recanalization on COED was assessed. Reperfusion therapy was administered to the 22 184 patients (median age 71 years and NIHSS score 16): only IVT (82.6%), IVT and EVT (13.8%) or only EVT (3.6%). Overall, recanalization was associated with a 10.6% (p<0.001) absolute risk reduction of moderate to severe COED at 22-36 hours, relative risk (RR) 0.55 (95% CI 0.52-0.58). Two models with high predictive ability provided the following estimates: adjusted OR 0.52 (95% CI, 0.46-0.59) and, with additional adjustment for parenchymal haemorrhage (PH), OR 0.46 (95% CI, 0.41-0.52). Moreover, recanalization was associated with a 13.6% (p<0.001) absolute reduction of mortality at 3 months, RR 0.58 (95% CI 0.55-0.61), adjusted OR 0.48 (95% CI 0.45-0.53). In paper III, 60 patients were randomized (15 patients in low-dose, 14 patients in medium and high-dose and 17 patients in control). Four serious adverse events (2 in control and 2 in low-dose group) resulted in the death of 3 patients. Of the dead patients 2 were allocated to low-dose group but of these, 1 did not receive imatinib and 1 patient had received only 2 doses. In the per protocol analysis, there were 21 haemorrhagic infarctions (6 in control), 3 PH (1 in control) and 4 remote parenchymal haemorrhages (0 in control). There were 33 cases of COED with moderate to severe COED being less frequent with higher doses, and no cases of moderate to severe COED in the high-dose group. After adjustment for EVT, the mean improvement in the NIHSS score compared to controls was 2 points (p=0.259) for the low-dose group, 3 points (p=0.106) for the medium-dose groups and 5 points (p=0.012) for the high-dose group. Functional independence (mRS 0-2) at 3 months was observed in 61% of the control group and 72% of all imatinib-treated patients; OR, adjusted for EVT, was 2.33 (95% CI 0.48-11.44). Paper IV reported anterior circulation AIS patients that underwent DHC. In 684 patients from 35 countries median age was 56 years and NIHSS score at baseline 18 and 98.1% received reperfusion therapy. Moderate to severe COED was detected in in 76.0% and PH in 25.8% at 22-36 hours follow-up imaging scans. Surgery-related details, for example timing of DHC, were not registered. Mortality at 3 months was 32.7% (159/486). Among baseline variables, only increasing age was independently associated with death (OR 1.06, 95% CI 1.03-1.08). Good outcome (mRS 0-3) at 3 months was observed in 13.9% (66/475) and mRS 0-4 was observed in 39.4% (187/475). Outcomes differed between patients aged ≤60 years ≥61 years (25.2% versus 47.8% for mortality and 16.6% versus 8.4% for good outcome). Right-sided involvement of vascular territory was more common than left-sided.

Conclusions: The most important baseline predictors for early COED are NIHSS score, HAS, higher blood glucose, decreased level of consciousness, and signs of acute infarction at baseline. This finding can be used to improve selection and monitoring of patients for drug or surgical treatment. In patients with AIS, recanalization was associated with a lower risk for early COED even after adjustment for higher rate of PH in recanalized patients. Imatinib is safe and tolerable and may reduce neurological disability in patients treated with IVT after AIS. A confirmatory randomized trial is ongoing. DHC in routine clinical practice may have worse outcomes than randomized trials, although there are caveats due to short follow-up of the patients in this study. Right-sided arterial occlusions were more common than left-sided, which indicates a tendency to perform DHC in infarctions of the right hemisphere. In general, this doctoral thesis added new knowledge about several aspects of COED in AIS and a potential new pharmacological therapy for acute ischaemic stroke. Further research is required to confirm these results which are based on 3 retrospective observational studies and one phase 2 pilot study. In fact, an efficacy trial of imatinib is now ongoing.