'No Reason to Drop tPA' in Stroke Thrombectomy: SWIFT DIRECT

September 02, 2021

A new study has failed to show noninferiority of direct thrombectomy compared to bridging therapy with thrombolysis before thrombectomy for patients with large-vessel occlusion ischemic stroke. However, outcomes were very good for both groups.

The trial, SWIFT DIRECT, was presented on September 1 at the virtual European Stroke Organisation Conference (ESOC) 2021.

This is the fifth major trial to compare these two strategies. Results have conflicted as to whether direct thrombectomy is noninferior.

The SWIFT DIRECT investigators also presented a meta-analysis of all five trials. Results suggested that the direct approach was noninferior if less stringent noninferiority margins were used, but not with more stringent margins.

Asked how these results guide clinical practice, SWIFT DIRECT investigator Jan Gralla, MD, University Hospital Bern, Bern, Switzerland, replied: "The outcomes were very good in both arms, but the postintervention recanalization rate was much higher in the thrombolysis arm, so from my point of view, the results from this trial suggest that there is no reason to skip thrombolysis in patients with large-vessel occlusions."

Co-investigator Urs Fischer, MD, University Hospital Bern, agreed that even with the results of the meta-analysis, "there is no reason to drop tPA [tissue plasminogen activator] at present."

But he added: "There will probably be subgroups who will have the benefit of the direct thrombectomy approach, whereas others may benefit from the bridging strategy, so the aim in the future is an individualized approach."

The researchers are planning an individual patient meta-analysis of the data from all the studies to investigate this further.

Commenting on the results for Medscape Medical News, Michael Hill, MD, University of Calgary Medical Center, Calgary, Canada, said there seems to be only a small difference between the two approaches for patients who can be brought to the cath lab quickly.

"Overall, it will probably be acceptable to go straight to the angio lab for thrombectomy and not give thrombolysis, but only if there is no delay in treatment," Hill suggested.

The SWIFT DIRECT trial addressed the issue of whether intravenous thrombolysis is needed before thrombectomy for patients with large-vessel occlusion ischemic stroke.

Introducing the trial, Gralla explained that potential advantages of performing thrombolysis before thrombectomy include the improvement of preinterventional and postinterventional reperfusion, which could lead to better outcomes. Potential disadvantages include the possibility of increased clot fragmentation and distal emboli, delay to thrombectomy, and increased bleeding and intracerebral hemorrhage (ICH) rates.

The study aimed to show noninferiority of direct thrombectomy, using Solitaire devices, to combined thrombolysis and thrombectomy (bridging) with regard to functional outcome at 90 days with a noninferiority margin of 12%.

For the trial, 423 patients were randomly assigned to either one or the other of the two strategies. Of these, 70% had M1 occlusion, and 29% had internal carotid artery occlusion (the same for both groups). The time from emergency department to thrombectomy was 75 minutes in the direct group and 80 minutes in the thrombolysis group.

Presenting the results, Fischer said the outcomes were very good in both treatment arms, but the point estimate was in favor of the bridging group. A good functional outcome of Modified Rankin Score (mRS) of 0–2 at 90 days was achieved in 57% of the direct thrombectomy group and in 65% of patients receiving bridging thrombolysis.

This gives a risk difference of -7.3% for direct thrombectomy. The lower 95% CI was -15.1%, which fell outside the 12% noninferiority margin aimed for in the trial.

Preinterventional reperfusion was very low in both arms (1% for direct and 2.4% for bridging). Postinterventional reperfusion was very high in both groups but was significantly higher in the bridging group (91% for direct and 97% for bridging).

Rates of symptomatic ICH were very low in both groups but were significantly lower in the direct thrombectomy group.


Fischer also presented a study-level meta-analysis of the five studies so far conducted in which direct thrombectomy was compared to bridging thrombolysis and thrombectomy. The analysis included results from the SWIFT DIRECT trial.

He summarized the five trials and their main results as follows:

  • DIRECT MT: noninferiority (NI) of direct thrombectomy shown (NI margin odds ratio [OR], 0.8)

  • SKIP: NI not shown but outcomes nominally similar (NI margin OR, 0.74)

  • DEVT: stopped early owing NI being met (NI margin for mRS of 0–2, 10%)

  • MR CLEAN NO IV: neither superiority nor NI shown, but outcomes were nominally similar (NI margin OR, 0.8)

  • SWIFT DIRECT: NI not shown (NI margin for mRS of 0–2, 12%)

Results of all the trials together show a risk difference of -1.3% for the endpoint of good functional outcome (mRS of 0–2) in favor of bridging with thrombolysis, with a lower noninferiority margin of -5.6.

There was no significant difference in the mRS shift analysis, which also trended toward a favorable effect with bridging (OR, 0.94; 95% CI, 0.81 – 1.10). A similar result was seen for mortality, with a risk difference of 1.3% favoring the bridging group (95% CI, -1.7 to 4.3).

Successful reperfusion was significantly more frequent in the bridging group (risk difference, -4.5%), as was symptomatic ICH (risk difference, -1.7%) and any ICH (risk difference, -8.3%).

Go Direct if No Delay?

Commenting for Medscape Medical News, Hill pointed out that there appears to be only a small difference in treatment outcomes between the bridging and direct thrombectomy approaches for the specific subset of patients who arrive at a major tertiary hospital and who receive thrombectomy immediately.

"It appears that SWIFT-DIRECT showed the largest benefit in favor of IV thrombolysis plus thrombectomy, compared to the other trials. However, the pooled data show the difference between treatments remains small," he said.

"This makes intuitive sense, given that the proportion of patients who achieve reperfusion between IV thrombolysis start and thrombectomy start is small ― just 1% to 2 %. Pharmacologically, the short time frame (~30–45 min) is simply not enough time for thrombolysis to work."

He believes that it will be acceptable to go straight to the cath lab without giving thrombolysis if there are no delays.

"Any potential delay, such as the neuro-interventional team has to come in from home or a delay in anesthesia or a procedural delay at the hospital level or the need to transfer from a primary hospital to a comprehensive stroke center, implies a need to treat concurrently with thrombolysis."

He noted that it is equally acceptable to continue to treat all patients with thrombolysis. "You might not be able to foresee a delay, or the thrombectomy procedure may not be successful and your best opportunity to treat with alteplase is early."

Hill also points out that the total number of patients in these studies is not large and that "more data will be needed to explicitly define what patients and situations warrant a direct to angio approach eschewing IV thrombolysis."

The SWIFT DIRECT trial was funded by Medtronic.

European Stroke Organisation Conference (ESOC) 2021: Presented September 1, 2021.

For more Medscape Neurology news, join us on Facebook and Twitter.


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.