Direct mechanical thrombectomy fails to show non-inferiority compared with IV t-PA plus thrombectomy for LVOs

swift-direct trial thrombectomy
Urs Fischer (L) and Jan Gralla (R)

The preliminary results of the SWIFT-DIRECT (Solitaire with the intention for thrombectomy plus IV t-PA versus direct Solitaire stent-retriever thrombectomy in acute anterior circulation stroke) trial have indicated that, in the treatment of large vessel occlusions (LVOs), direct mechanical thrombectomy failed to demonstrate non-inferiority to intravenous tissue-plasminogen activator (IV t-PA) plus thrombectomy—or “bridging thrombolysis”. These results were presented by Jan Gralla, chair and director of the Department of Diagnostic and Interventional Neuroradiology at the Inselspital, University Hospital of Bern (Bern, Switzerland), and Urs Fischer, chairman and head of the Department of Neurology at the University Hospital of Basel (Basel, Switzerland), at the 7th European Stroke Organisation Conference (ESOC 2021; 1–3 September, virtual).

“The quality of data in this trial is very good and, as an interventionist, it was clear to see that the preintervention reperfusion rate with intravenous thrombolysis was very low—this is basically what we expected based on our clinical experience,” Gralla said. “Nonetheless, the postintervention reperfusion rate was much higher in the IV t-PA arm, so, from my point of view—in the current setting—there is no reason to skip intravenous thrombolysis in LVOs.”

“In terms of what we have shown, we have not proven the superiority of bridging thrombolysis over direct mechanical thrombectomy, but this can be seen as a shift towards bridging thrombolysis,” Fischer added. He did, however, also note that there are probably some subgroups within the SWIFT-DIRECT patient population who will see a benefit from direct mechanical thrombectomy, in addition to those who benefit from receiving IV t-PA prior to the thrombectomy procedure. “Therefore, the aim in the future will be an individualised treatment approach in acute stroke therapy,” he stated.

In presenting these results, the researchers relayed that their primary aim going into the trial was to determine whether patients eligible for IV t-PA, who are experiencing an ischaemic stroke due to an internal carotid artery (ICA) or M1 occlusion, and have been referred to a stroke centre, will have a non-inferior functional outcome at 90 days when treated with direct mechanical thrombectomy, compared to those treated with combined IV t-PA plus thrombectomy. The non-inferiority margin in the trial was set at 12%, they noted.

The SWIFT-DIRECT trial was a prospective, multicentre, 1:1 randomised, open-label, blinded-endpoint study in which patients in the standard group received intravenous alteplase (0.9mg/kg) plus mechanical thrombectomy using Solitaire devices (Medtronic), and patients in the intervention group received thrombectomy with Solitaire devices alone. The trial’s primary efficacy endpoint was functional independence—measured via a modified Rankin Scale [mRS] score of 0–2—at 90 days, with secondary outcomes being mortality at 90 days, mRS shift, successful reperfusion, and both symptomatic and asymptomatic intracerebral haemorrhages.

A total of 423 patients were recruited at 48 sites across eight different countries within Europe and North America. Some 408 of these patients were included in the final analysis, with 201 being randomised to receive direct mechanical thrombectomy (52% female, median age=73 years), and 207 being randomised to receive IV t-PA plus thrombectomy (50% female, median age=72 years). In these groups, 198 and 204 patients received their allocated intervention, respectively.

“Overall, it is extremely important to note that outcomes were very good in both treatment arms, with good functional outcomes—indicated by an mRS score from 0–2—of 62%,” Fischer said. “In terms of the primary outcome, 57% of patients in the direct mechanical thrombectomy group had a good functional outcome, compared to 65% of patients in the bridging thrombolysis cohort.” He went on to report a risk difference of -7.3% between the two groups, and a -15.1% lower limit of one-sided 95% confidence interval—which fell outside of the 12% non-inferiority margin, and led the researchers to conclude that SWIFT-DIRECT did not show statistical non-inferiority of direct mechanical thrombectomy when compared to IV t-PA plus thrombectomy.

Regarding secondary outcomes, the researchers recounted that there were no differences in mRS shift, mortality at 90 days, or changes in median National Institutes of Health Stroke Scale (NIHSS) scores at 24 hours, in the two groups—and there were very low rates of successful preintervention reperfusion rates in both groups as well. However, while successful postintervention reperfusion rates were above 90% in both treatment arms, they were “significantly higher” in the IV t-PA plus thrombectomy cohort, at 97% compared to 91%. Symptomatic intracerebral haemorrhage rates were very low in both groups too—but, at 1.5% compared to 4.9%, were significantly lower in the direct mechanical thrombectomy group.

Gralla qualified the preliminary nature of these results by stating that the database for the trial is yet to be closed, meaning the numbers may change slightly in the future, but that all of the clinical outcomes are “100% clear”, and the patterns these data indicate “will not change now”. In a discussion following their joint presentation, Fischer noted that the SWIFT-DIRECT researchers are now hoping to pool their data with those from the Improving Reperfusion Strategies in Ischaemic Stroke (IRIS) group—which will begin collaborating in early 2022—in order to identify subgroups who may benefit from either one of these two treatment options for LVO stroke.


Please enter your comment!
Please enter your name here