Effect of baseline Alberta Stroke Program Early CT Score on safety and efficacy of intra-arterial treatment: a subgroup analysis of a randomised phase 3 trial (MR CLEAN)

Summary

Background

Whether infarct size modifies intra-arterial treatment effect is not certain, particularly in patients with large infarcts. We examined the effect of the baseline Alberta Stroke Program Early CT Score (ASPECTS) on the safety and efficacy of intra-arterial treatment in a subgroup analysis of the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN).

Methods

MR CLEAN was a randomised, controlled, open-label, phase 3 trial of intra-arterial treatment in patients (aged ≥18 years from the Netherlands) with proximal arterial occlusion of the anterior circulation, given intra-arterial treatment within 6 h of stroke onset. The primary outcome was 90 day modified Rankin Scale (mRS) score. We estimated the intra-arterial treatment effect for all patients in MR CLEAN who had ASPECTS graded by using multivariable ordinal logistic regression analysis (a proportional odds model) to calculate the adjusted common odds ratio for a shift towards a better functional outcome according to the mRS for intra-arterial treatment and usual care than for usual care alone. We entered an interaction term into the model to test for interaction with prespecified ASPECTS subgroups: 0–4 (large infarct) versus 5–7 (moderate infarct) versus 8–10 (small infarct). MR CLEAN is registered with the Netherlands Trial Registry, number NTR1804, and the ISRCTN Registry, number ISRCTN10888758.

Findings

496 patients—232 (47%) in the intra-arterial treatment and usual care group and 264 (53%) in the usual care alone group—were included in the analysis. We noted no significant difference in intra-arterial treatment effect between the ASPECTS subgroups according to 90 day ordinal mRS (adjusted common odds ratio interaction term relative to ASPECTS 8–10: ASPECTS 0–4: 0·79 [95% CI 0·20–3·19], p=0·740; and ASPECTS 5–7: 1·02 [0·44–2·35], p=0·966). Intra-arterial treatment did not cause a significant increase in the proportion of patients with at least one serious adverse event in any of the ASPECTS subgroups (ASPECTS 0–4: eight [73%] of 11 patients in treatment and usual care group vs 11 [58%] of 19 in usual care alone group, p=0·42; ASPECTS 5–7: 32 [59%] of 54 vs 19 [49%] of 39, p=0·31; ASPECTS 8–10: 70 [42%] of 167 vs 82 [40%] of 206; p=0·68). For death within 7 days or within 30 days and hemicraniectomy, the differences between the intra-arterial treatment and usual care versus usual care alone groups were not significant by ASPECTS subgroups. A significantly higher proportion of patients had recurrent ischaemic stroke in the intra-arterial treatment plus usual care group than in the usual care alone group in the ASPECTS 8–10 subgroup (eight [5%] vs one [<1%]; p=0·007).

Interpretation

Contrary to findings from previous studies suggesting that only patients with non-contrast CT ASPECTS of more than 7 benefit from intra-arterial treatment, data from this study suggest that patients with ASPECTS 5–7 should be treated. Further evidence is needed for patients with ASPECTS 0–4, for whom treatment might yield only marginal absolute benefit.

Funding

Dutch Heart Foundation, AngioCare, Medtronic/Covidien/EV3, Medac/Lamepro, Penumbra, Stryker, and Top Medical.

Introduction

Evidence suggests that infarct size at baseline can be used to predict functional outcome after intra-arterial treatment.1, 2 Patients with smaller infarcts at baseline have better long-term functional independence, lower incidence of reperfusion haemorrhage, and lower mortality than do those with larger infarcts.1, 2 Whether baseline infarct size modifies intra-arterial treatment effect, particularly whether patients with large infarcts benefit from catheter-based reperfusion therapy, is not clear.3, 4

Despite the availability of advanced imaging systems, non-contrast CT is widely used for patient assessment before intra-arterial treatment.5, 6 The Alberta Stroke Program Early CT Score (ASPECTS) is a semi-quantitative method of estimation of infarct size with non-contrast CT during the acute phase. Findings from a post-hoc analysis of the PROACT II trial ³ showed an interaction of baseline ASPECTS (>7 vs ≤7) with intra-arterial treatment effect. However, in a similar analysis of the IMS III trial,⁴ no interaction was noted between ASPECTS and intra-arterial treatment effect, which might have been confounded by the overall neutral trial results. Additionally, data have suggested that a larger infarct threshold (ASPECTS 0–4) might better define patients likely to have a poor response to intra-arterial treatment than the previous ASPECTS threshold of 7.⁷

Research in context

Evidence before this study

We searched the Cochrane Stroke Trials Registry and MEDLINE from Jan 1, 1990, to June 10, 2015, to identify all randomised controlled trials of intra-arterial treatment versus control in acute ischaemic stroke. We used the search terms “acute ischaemic stroke”, “intra-arterial”, “mechanical thrombectomy”, “randomised controlled trial”, “prospective”, or related keywords, with no language restrictions. We found 13 randomised controlled trials that included patients with anterior circulation strokes. Overall, 2826 patients were included. Alberta Stroke Program Early CT Score (ASPECTS) on baseline non-contrast CT was used as an exclusion criterion in three studies: REVASCAT excluded patients with ASPECTS of lower than 7 and ESCAPE and SWIFT-PRIME excluded those with ASPECTS of lower than 6 (with SWIFT-PRIME using this criterion after the first 71 patients were enrolled using perfusion imaging criteria). The other studies that used non-contrast CT for infarct assessment used less precise criteria for infarct extent (eg, less than one-third of the middle cerebral artery territory rule [IMS III and EXTEND-IA]) or more descriptive criteria (SYNTHESIS Expansion) for imaging-based exclusion than did the previous three studies. Notably, investigators of IMS III suggested use of ASPECTS less than or equal to 4 as a guide for identification of large infarcts, but this suggestion was not mandated. Also, IMS III quantified only clear hypodensity versus early ischaemic changes. Finally, MR RESCUE, SWIFT PRIME (for the first 71 patients), and EXTEND-IA used advanced imaging—either MRI or CT perfusion—to exclude large infarcts and verify the presence of an imaging mismatch (ie, penumbral marker). Retrospective analyses of PROACT II and IMS III assessed for an ASPECTS-by-treatment interaction, with conflicting results.

Added value of this study

MR CLEAN was the only randomised trial of mechanical thrombectomy that both showed a benefit of intra-arterial treatment and did not exclude patients with large infarcts. As such, it provides the best data to examine intra-arterial treatment effect across the entire range of infarct sizes. In this prespecified subgroup analysis, ASPECTS was graded on the qualifying baseline non-contrast CT by masked readers in a core imaging laboratory and trichotomised as small (ASPECTS 8–10), moderate (ASPECTS 5–7), or large (ASPECTS 0–4) infarction. Non-contrast CT is a standard imaging examination that is done by virtually all centres that provide intra-arterial treatment, and ASPECTS is a validated and reliable method of quantification of infarct burden. Therefore, these results are generalisable to clinical practice. Contrary to findings from previous studies suggesting that only patients with non-contrast CT ASPECTS of more than 7 benefit from intra-arterial treatment, data from this study suggest that patients with ASPECTS 5–7 should be treated.

Implications of all the available evidence

Further research is needed to examine whether treatment of large infarcts is cost-effective. Findings from this study showed no safety concerns with treatment of large infarcts. Further data are needed to better understand intra-arterial treatment effect in patients with ASPECTS 0–4 than at present. Based on the poor outcomes of patients in this subgroup, the absolute intra-arterial treatment benefit might be marginal.

The results from the Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands (MR CLEAN)8, 9 provide a unique opportunity to further examine use of non-contrast CT for intra-arterial treatment selection because of its randomised controlled design, the positive overall effect of intra-arterial treatment in the trial, and the absence of explicit imaging exclusion criteria for infarct size. We sought to examine the effect of baseline ASPECTS (0–4 vs 5–7 vs 8–10) on the safety and efficacy of intra-arterial treatment effect in this prespecified subgroup analysis of MR CLEAN.