Endovascular Trials Analysis Confirms Benefit Across Groups

February 25, 2016

LOS ANGELES — The first patient-level meta-analysis of the new endovascular thrombectomy trials in acute stroke has given more precise estimates of the benefits of such treatment and extends evidence to many different subgroups.

Results show that for every 100 patients with a large-vessel anterior-circulation ischemic stroke treated with endovascular thrombectomy, 38 will have a less disabled outcome than with best medical management alone, and 20 more will achieve functional independence.

Coauthor Tudor G. Jovin, MD, University of Pittsburgh Medical Center Stroke Institute, Pennsylvania, commented to Medscape Medical News: "This is a highly efficacious treatment. Very few therapies in medicine approach this level of benefit."

Dr Tudor G. Jovin

The meta-analysis also shows that endovascular treatment is beneficial in certain subgroups of special interest, including patients older than age80 years, those treated more than 300 minutes after symptom onset, those who did not receive tissue plasminogen activator (tPA), and those with moderate infarct volumes.

Dr Michael Hill

It also suggests that the intervention is worthwhile in patients up to 8 hours after symptom onset.

Results from the meta-analysis were presented here at last week's International Stroke Conference (ISC) by Dr Jovin, representing the REVASCAT investigators, and Michael Hill, MD, University of Calgary, Alberta, Canada, representing ESCAPE trial investigators.

The findings were also published online simultaneously with the presentation in The Lancet on February 18.

HERMES Collaboration

The meta-analysis includes data from the first randomized trials reported in 2015 — MR CLEAN, ESCAPE, REVASCAT, SWIFT PRIME, and EXTEND IA — all of which showed benefits of endovascular thrombectomy in patients with acute ischemic stroke caused by a large anterior occlusion.

The result of a collaboration known as HERMES (Highly Eff ective Reperfusion evaluated in Multiple Endovascular Stroke Trials) between the investigators of the five trials, the new meta-analysis pools individual-patient data from the studies to address remaining questions about whether the therapy is efficacious across the diverse populations included and specific subgroups for which the individual trials alone could not provide definitive results.

Dr Jovin said, "This has come about because of an exemplary collaboration between the five sets of researchers from the individual trials. It is difficult to obtain all the individual-patient data and analyze it all together. This was a remarkable team effort."

The authors note that although some study-level meta-analyses have been reported, these are considered less informative than patient-level meta-analytical approaches because of their inability to adjust for confounding baseline variables, which leads to less precise estimates of treatment effect.

"The consistent results across different patient populations suggest that benefit from thrombectomy is generalisable to a broad range of patients with large-vessel ischemic stroke," the authors state.

"Our study provides clear evidence that in clinical practice, endovascular therapy for stroke should not be withheld on the basis of advanced age, moderately extensive early ischaemic changes on baseline CT [computed tomography], and moderate or severe clinical deficit," they add.

For the meta-analysis, the researchers analyzed individual data for 1287 patients (634 assigned to endovascular thrombectomy and 653 assigned to control).

The primary outcome — modified Rankin Scale (mRS) shift analysis at 90 days — showed that endovascular thrombectomy led to significantly reduced disability at 90 days (adjusted odds ratio [OR], 2.49; P < .0001), with a number needed to treat to reduce disability by at least one level on the mRS score for one patient of 2.6.

Table. Efficacy Outcomes From the Pooled Data

Outcome Endovascular Intervention (%) Control (%) Adjusted OR P Value
mRS score 0 - 1 at 90 d 26.9 12.9 2.72 <.0001
mRS score 0 - 2 at 9 d 46.0 26.5 2.71 <.0001
National Institutes of Health Stroke Scale score 0 - 2 at 24 h 21.0 8.3 3.77 <0001
Early neurologic recovery at 24 50.2 21.2 4.36 <.0001


Mortality at 90 days and risk for parenchymal hematoma and symptomatic intracranial hemorrhage did not differ between groups.

Subgroup analysis of the primary endpoint showed no heterogeneity of treatment effect across all prespecified variables, including age, sex, National Institutes of Health Stroke Scale (NIHSS) score, site of intracranial occlusion, intravenous alteplase received or ineligible for this therapy, Alberta Stroke Program Early CT score (ASPECTS), time from stroke onset, and presence of tandem cervical carotid occlusion.

The direction of effect favored endovascular treatment across all subgroups, although the adjusted odds ratios were not significant for patients younger than 50 years, those with a low ASPECTS or NIHSS score, and in those with an M2 segment thrombus.

Significant benefit was seen with endovascular thrombectomy in several groups of special interest, including patients aged 80 years or older (OR, 3.68), those randomly assigned more than 300 minutes after symptom onset (OR, 1.76), and those not eligible for intravenous tPA (OR, 2.43).

Dr Jovin explained that it has been widely thought previously that patients older than age 80 years do not benefit from such intervention.

"But our data clearly show that they do benefit, and the benefit may actually be greater in these older patients," he said. "This group has a particularly dismal outcome without treatment, so even though the good outcome rate is low with intervention, it was much lower in the control group."

Data from the individual trials did not give enough information to permit firm recommendations about thrombectomy for patients who had not received tPA. "But our meta-analysis include 188 such patients and gives level 1 randomized data in favor of intervention in this group," Dr Jovin said.

He noted that another unresolved issue has been "how large is too large" in terms of irreversible damage, as shown by the size of the baseline core infarct.

"We thought before that patients would need to have only a small core infarct to see benefit with thrombectomy, but our analysis has shown benefit also for those with moderate infarct sizes — ASPECTS scores down to 5." On patients with large infarcts (ASPECTS score below 5), Dr Jovin said there were too few patients to permit any conclusions.

Benefit Up to 8 Hours

Another important question addressed by the meta-analysis is how long after symptom onset can thrombectomy produce benefits.

Presenting data on this from the pooled analysis in a separate presentation here, Jeffrey Saver, MD, University of California, Los Angeles, reported that benefit is clear before 6 hours and there is a strong trend toward benefit for those receiving thrombectomy between 6 and 8 hours from symptom onset.

Dr Jeffrey Saver

"I would say there is likely benefit up to 8 hours. Beyond this we don't know from these trials as very few patients were included later than this."

Dr Saver added that further information from patients presenting later will come from other ongoing trials.

"Our data and that from further trials on the effect of time on benefit of thrombectomy is crucial to help us rationally design systems of care so the emergency medical personnel can make the correct decisions as to which patients should be sent to comprehensive stroke centers to receive this intervention," he added.

Author disclosures are available with the published report.

Lancet. Published online February 18, 2016. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.