WSC 2025 | Primary findings from the TRIMIS study: endovascular thrombectomy for mild ischemic stroke

Thrombectomy in mild ischemic stroke with a visible intracranial occlusion, or TRIMIS, was an international collaboration, multi-center comparative cohort that studied the role of endovascular thrombectomy in treatment of patients with a target vessel occlusion who present with a mild neurological syndrome as captured by NIH stroke scale of equal or less than five at the time of presentation to the first hospital. Just a little bit of background, about somewhere between eight to a quarter of patients with mild stroke have a target vessel occlusion. And this wide variation is relevant to A, the definition of target occlusion, which is historically described as presence of a large vessel occlusion rather than any target occlusion being distal or medium vessel occlusion. And the other issue has been the fact that conventionally the publications that gave us this percentage came from studies of patients presenting early on under six hours with mild neurological symptoms, whereas majority of mild neurological patients with mild neurological symptoms actually present late to the hospital. But despite these variations in definition and exactly the prevalence of a target occlusion in this population, what has been consistently shown in the literature is that presence of a target occlusion, however distal, is a predictor of poor outcomes. So it’s conceivable that study that tried reperfusion therapies and specifically as it comes to TRIMIS endovascular treatment for this population. And unfortunately, patients with mild neurological symptoms were either underrepresented or completely excluded from the pivotal randomized trials of endovascular thrombectomy. And that includes the recently published DMVO trials, where mild stroke was actually underrepresented in those trials. And so our current international guidelines are reflective of the lack of sort of solid evidence in that area. And as such, multi-center collaborations like TRIMIS is necessary to fill this gap.

So a little bit about what TRIMIS included. So obviously our inclusion criteria was patients presenting with mild neurological symptoms and mild defined by NIH stroke scale of five equal or less than five, inclusive of NIH stroke scale of zero. As long as there was some neurological symptoms that would be relevant to the target occlusion. And we were very pragmatic in terms of definition of a target occlusion. Any visible occlusion, whether in the anterior or posterior circulation, could have been qualified for TRIMIS. And we were also very permissive in terms of time. So we allowed patients presenting within 24 hours from last seen normal or anyone if there was no last known normal. So in those without a last known normal, anyone that had that was, again, amenable to endovascular treatment could have been included in TRIMIS. An important few issues to mention regarding exclusion criteria. It was fairly standard for any endovascular studies or trials. We excluded in hospital cases or excluded pregnant patients. An important exclusion criteria for TRIMIS was that seeing that TRIMIS ran, the study ran in parallel with many of the DMVO trials, was that the patients could not have been enrolled in any ongoing mild clinical trials, such as Endo-Low or Extremis-Most, or any of the parallel DMVO trials, for example, Escape-MeVo, if the site was participating in any of these trials. So TRIMIS was registered in clinicaltrials.gov, and the study ran from January 2018 to September 2024. 43 sites from 18 countries across four continents provided data for TRIMIS.

So in total, we had 4,831 patients. So we excluded patients that did not have either a visible target occlusion on review of the data or their primary presenting NIH stroke scale at the time of presentation was not mild. And that is conceivable because sometimes people have mild neurological presentation and rapidly deteriorate, but at the time of first presentation to the treating hospital, they were already of an NIH stroke scale of over five. So we wanted to really capture those patients that at the time of presentation, A, had intracranial target occlusion, and B, had mild neurological symptoms, as mentioned, and the decision was then whether or not to treat with endovascular thrombectomy immediately or do maximal medical therapy. So then with this description, we went down to 4,440 patients for the intention to treat analysis of TRIMIS. Of those, 2,230 patients received immediate endovascular, I’m sorry, 2,230 patients received maximal medical therapy. That’s just over 50% of patients in TRIMIS. And then 2,210 patients, that’s 49.7% of TRIMIS received immediate endovascular therapy. So we were almost 50-50. Now, of those 2,230 patients that at the decision of the treating physician had received best medical therapy. Amongst those 2,000 patients, another 404 patients, so that’s 18% of those patients that initially were treated with best medical management subsequently received rescue endovascular treatment. So that’s important to keep in mind that the primary analysis of TRIMIS, which I’m about to tell you in the primary results, is based on the intention to treat analysis of immediate endovascular thrombectomy versus best medical management.

A little bit about the baseline characteristics of patients. Again, I want to highlight that the intention to treat analysis of TRIMIS was looking at those whose initial treatment was a decision to proceed with best medical management that is inclusive of receiving intravenous thrombolysis versus immediately going for endovascular therapy. The decision for immediate endovascular therapy could have been made after a postural challenge. So that wouldn’t be considered a best medical treatment. And so as expected, a little bit about the baseline characteristics of patients that received the best medical treatment versus immediate endovascular therapy, as expected for any non-randomized studies that were some imbalances in baseline characteristics. Importantly, to keep in mind some highlights here, is more female in patients who received immediate endovascular therapy. Most of those patients who received immediate endovascular therapy had a lower percentage of vascular risk factors. There wasn’t a difference between age, and that was an interesting finding. Again, unsurprisingly, some differences in time. So onset to presentation was slightly shorter for the best medical treatments earlier on, just over 3.3 hours, whereas immediate endovascular thrombectomy is 4.2 hours. More patients that received immediate endovascular thrombectomy had been transferred from a primary or a non-thrombectomy capable center to a comprehensive or thrombectomy capable center. So 46% transfer in the immediate EVT group versus 30% transfers. The rival NIH stroke scale, though capped at NIH of 5, as mentioned, obviously, for the study, was still significantly higher in the immediate endovascular thrombectomy, median NIH of 4 in the immediate EVT category versus median NIH of 3 in the best medical treatment group. Again, remember we said that patients who received best medical therapy had arrived significantly earlier to the hospital and the percentage of intravenous thrombolysis was higher in that category, 46% versus 35%. In terms of radiographic differences, ASPECT score was indifferent between the two. There was, again, significant differences in terms of the location of occlusion. 40% of patients receiving immediate endovascular thrombectomy had an anterior circulation large vessel occlusion, that is a distal internal carotid artery occlusion, or an M1 occlusion, as opposed to 25% of those allocated or received best medical therapy. In contrast, 68.3% of patients who received best medical management had a distal or medium vessel occlusion as compared to 50% in the immediate EVT group.

So with that, we’re on to the primary outcome of TRIMIS, which is modified Rankin scale of 0 to 1 at 90 days. Again, comparison between immediately endovascularly treated categories versus best medical treatment. And the numbers were very similar. So 54.4% of those who received best medical management achieved a modified Rankin scale of zero to one. So excellent functional outcome, 90 days, as compared to 54.2%, a difference that was not different in unadjusted or adjusted analyses, accounting for all the baseline variables. Also, we did not show, obviously, this is an audio only recording, so I can’t show you the slides, but we did not show a shift in improvement of or changes in mRS across the two categories.

Now, moving on to the primary safety outcome, that was symptomatic intracerebral hemorrhage defined by Heidelberg classification. A total of 186 patients suffered from symptomatic intracerebral hemorrhage, 6% in immediate EVT category versus 3% in best medical treated category. A difference that was statistically significant in favor of best medical treatment, both in the unadjusted as well as in the adjusted analysis. Now moving on to the 90-day mortality, we had 227 deaths within the first 90 days. In TRIMIS, again, 4% in best medical treatment arm versus 6% in those who underwent immediate endovascular thrombectomy. This difference did not reach a statistically significant difference, either in the unadjusted or the adjusted models.

So we had some subgroup analyses, pre-specified subgroup analyses for the primary outcome of excellent neurological outcome at 90 days. And I’ll just go over two of the subgroup analysis. One is looking at best medical management versus immediate endovascular therapy in subgroups that received IV thrombolysis versus did not. There was no heterogeneity and differences in outcomes. However, what we found, and that was interesting, was a clear benefit in favor of immediate endovascular thrombectomy in mild stroke patients by occlusion sites, a favor for patients to receive immediate endovascular thrombectomy to achieve that mRS of 0 to 1 at 90 days if they had indeed an anterior circulation, large vessel occlusion, again, a distal ICA or an M1 occlusion, there was a significant difference in favor of immediate EVT. Moving on to the subgroup analysis for primary safety outcome of symptomatic intracerebral hemorrhage. Again, there was no heterogeneity across patients who received IV thrombolysis versus not, whereas there was heterogeneity for occlusion site, clear sort of issue regarding harm with increase or worse outcomes in terms of receiving symptomatic intracerebral hemorrhage in patients with a distal and medium vessel occlusion. Again, the difference reached statistical significance.

So in conclusion, here are the sort of take-home messages from TRIMIS. Amongst patients with mild ischemic stroke, so that’s defined as a NIH stroke scale of equal or less than five, and a target vessel occlusion, number one, 50% of these patients received immediate endovascular thrombectomy in routine care. And one in five patients who initially had received best medical management for one reason or another ended up receiving rescue endovascular therapy. In the intention to treat analysis, comparing best medical management to immediate endovascular thrombectomy, we did not find that there was a benefit from immediate endovascular thrombectomy and achieving excellent functional outcomes at 90 days. And immediate EVT was associated with significantly higher rates of symptomatic intracerebral hemorrhage. However, mild stroke patients with an anterior circulation large vessel occlusion had higher odds of achieving excellent functional outcomes with immediate EVT. So definitely a benefit for occlusion site being large vessel occlusion in the anterior circulation. And immediate EVT in patients with a distal and medium vessel occlusion and mild neurological symptoms carried sort of a signal of harm in our subgroup analysis and was associated with significantly higher odds of symptomatic intracerebral hemorrhage.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.