By Alan Z. Segal, MD
Synopsis: Two recent trials from China studied the effect of post-thrombectomy, intra-arterial thrombolysis on neurological outcomes. One study used tenecteplase and the other used urokinase. Both studies showed a trend toward improvement that was not statistically significant, and there was an increase in brain hemorrhage. The benefit of adjunctive thrombolysis with mechanical thrombectomy is uncertain and requires more study.
Sources: Huang J, Yang J, Liu C, et al. Intra-arterial tenecteplase following endovascular reperfusion for large vessel occlusion acute ischemic stroke: The POST-TNK randomized clinical trial. JAMA. 2025;333(7):579-588.
Liu C, Guo C, Li F, et al. Intra-arterial urokinase after endovascular reperfusion for acute ischemic stroke: The POST-UK randomized clinical trial. JAMA. 2025;333(7):589-598.
Dippel DWJ, Khan CF, Schoon BA. Intra-arterial thrombolytics during thrombectomy for ischemic stroke — end of the story or a new beginning? JAMA. 2025;333(7):571-573.
In 2015, multiple landmark trials presented at the American Heart Association International Stroke Conference demonstrated the benefit of intra-arterial (IA) thrombectomy for acute ischemic stroke caused by large vessel occlusion. Since that time, investigators have explored multiple avenues to expand the indications for this therapy. In 2018, the pool of thrombectomy-eligible patients increased dramatically, with the DEFUSE-3 and DAWN trials demonstrating efficacy with treatment up to 24 hours in patients with a significant ischemic penumbra (defined by perfusion studies showing salvageable at-risk tissue out of proportion to the permanently damaged core).
Other successful attempts to identify candidate patients for thrombectomy have included treatment of patients with milder stroke severity (National Institutes of Health Stroke Scale [NIHSS] score < 6), patients with larger initial core infarcts (Alberta Stroke Program Early Computed Tomography Score [ASPECTS] 0-5), and patients with lesions in the posterior circulation (basilar artery). Most recently, two 2025 trials studying medium vessel occlusions (i.e., beyond the M1 branch of the middle cerebral artery [MCA]), ESCAPE-MeVO and DISTAL, were disappointingly negative, as published in the February 2025 issue of the New England Journal of Medicine.
It is sobering that no matter how many patients are treated with thrombectomy and how many can be revascularized successfully using the ever-expanding arsenal of catheter-based devices, more than 50% of thrombectomy patients do poorly, failing to reach functional independence. Attempts to augment the effect of thrombectomy have included adjunctive medications, such as drugs given during the thrombectomy procedure itself (such as IA antiplatelet or antithrombotic agents), but outcomes for these patients have been harmful or neutral at best. Similarly, agents aiming to promote neuroprotection have not found any footing despite favorable angiographic outcomes. Previously, these agents were thought to be fruitless in the setting of persistent vascular occlusion, but the thrombectomy era has not significantly changed this calculus.
Poor outcomes associated with thrombectomy may be driven by multiple factors, including the persistence of thrombi downstream from the target lesion. Such clots may be visualized angiographically in distal arteries or occur below the resolution of imaging at the level of the microcirculation. It has long been thought that when there is failure of tissue reperfusion despite recanalization — the so-called “no-reflow” phenomenon — microcirculatory failure is the underlying pathophysiology.
Most recently, studies have aimed to address distal/microcirculatory thrombi by adding an IA thrombolytic to the mechanical clot extraction process. Interestingly, prior to devices such as stent-retrievers, IA thrombolysis alone (using drugs such as tissue plasminogen activator [tPA] or urokinase) showed promise as the sole treatment for a large vessel occlusion (LVO). These drugs would not only open the culprit lesion but also could work in a more widespread fashion to lyse residual clots. In 1999, the PROACT trial showed benefit of IA pro-urokinase over medical therapy, but this was never further pursued or approved by the U.S. Food and Drug Administration. Randomized trials of IA tPA failed to show benefit.
In 2021, the CHOICE trial investigated the effect of IA tPA after mechanical thrombectomy and showed a strong trend toward benefit, but ultimately was terminated prematurely in part because of slow enrollment during the COVID pandemic. CHOICE also was limited by the exclusion of patients (60% of those initially eligible) who received an intravenous (IV) thrombolytic prior to thrombectomy. It was thought that any additional lytic given via an IA route would result in a prohibitively high hemorrhage risk. Also, CHOICE enrolled patients with suboptimal revascularization — thrombolysis in cerebral infarction (TICI) scores of 2b (50% to 89% revascularization), as well as those with more favorable TICI 2c (90% to 99% revascularization) and TICI 3 (complete revascularization). Restricting CHOICE to only its TICI 2c and 3 patients, the study did show benefit for the addition of IA lytic. An impressive 59% of those patients had a favorable outcome. CHOICE was convincingly positive also when outcomes were studied using “shift analysis” across modified Rankin Scale (mRS) categories rather than by strict dichotomization.
Two new investigations reviewed here — the POST-TNK and POST-UK studies — examined the role of either tenecteplase (TNK) or urokinase (UK) as adjunctive IA therapy following thrombectomy for LVO. In the POST-TNK trial, there were 540 patients enrolled, with a favorable outcome (mRS score 0-1) in 49% of treated patients compared to 44% of controls, a non-statistically significant benefit. The observed incidence of any radiographic intracranial hemorrhage was significantly higher in the IA TNK group than in the control group, but this did not affect mortality, which was 15% in treated patients compared to 19% of controls. The dose of TNK in this study was 0.0625 mg/kg, much lower than the IV dose of 0.25 mg/kg (a four-fold difference).
In POST-UK, there were 535 patients enrolled, with a favorable mRS score in 45% of treated patients compared to 40% of controls, again a non-statistically significant benefit. There were no differences in mortality or in rates of symptomatic hemorrhage. The dose of UK was 100,000 IU, while prior investigations of UK as an IA lytic without thrombectomy the dose was in the 500,000 to 1 million IU range. Both studies include patients with TICI scores of 2c or 3 and enrolled patients in the six- to 24-hour time window who had small core infarcts (ASPECTS score ≥ 7) according to the criteria defined by the DEFUSE-3 and DAWN trials.
Commentary
Despite extensive studies, including the CHOICE trial and the two investigations reviewed here, it has not yet been proven that adding an IA lytic to the treatment of patients receiving endovascular therapy (EVT) is of benefit. However, as the authors observed, treatment effects (while favorable) were likely less robust than expected and, therefore, the sample sizes lacked the power to reach statistical significance. Odds ratios (ORs) favoring treatment were of marginal magnitude in both POST-TNK and POST-UK — in the 1.2 range. Although this was non-significant, 95% confidence intervals (CIs) were highly skewed toward positivity (0.87 to 1.7). More profoundly, as was observed in the attached editorial, pooled analysis of CHOICE, TNK, and UK showed an OR of 1.46, more strongly favoring a treatment effect and barely extending over the lower limit of statistical significance (CI, 0.96 to 2.23).
A major flaw in these studies is that they cannot account for the two largest drivers of poor outcomes in EVT. The most crucial of these is re-occlusion of a previously opened vessel, which may occur only shortly after the patient is taken off the angiography table and may be more likely among the Asian patients studied here, with significant underlying intracranial atherosclerotic disease (ICAD). The other concern is that tissue ischemia remains time-sensitive. Even in situations where perfusion data suggest a large penumbra, the “tissue clock” cannot erase the ticking of time. At least part of the issue is that the available techniques, primarily computed tomography (CT) perfusion interpreted with widely used analytic software (such as RAPID or VIZ-AI), are flawed. This author has seen many cases where the CT perfusion study suggests a negligible core and large penumbra, but where the simple non-contrast head CT already shows a significant early hypodensity.
As with many trials, patient selection remains paramount, and the optimal cohort is as yet undetermined. If the result of thrombectomy is suboptimal (TICI 2b with 50% to 99% recanalization), there would be a meaningful treatment target for additional IA lytic, but the overall prognosis would be worse. Conversely, if the result of thrombectomy is excellent (TICI 2c or 3), the favorable outcome would be driven primarily by the primary therapy itself without adding a lytic to “mop up” any minimal residual clot.
One option might be to add a lytic only to patients who can be proven to have perfusion failure. Flat panel detector CT scans capable of doing this are being incorporated into an increasing number of angiography suites. Otherwise, transporting a patient with a sheath in place to a CT scanner located elsewhere is logistically challenging and time consuming. An upcoming trial (Tenecteplase for Non-Complete Reperfusion of Intracranial Occlusions [TECNO]) will explore perfusion outcomes with or without adjunctive IA TNK, with endpoints including early reperfusion (on angiography immediately after lytic administration) followed by 24-hour CT or magnetic resonance perfusion studies.
Using an imaging outcome in TECNO, while not capturing the highly valued standard clinical outcome in severe stroke, the mRS score may help in the detection of subtle improvements not recognized by coarse disability scales. The mRS has a strong emphasis on gait, a function as much dependent on subcortical structures as it is on the motor cortex itself. However, favorable outcomes in stroke, especially when the goal is treatment of distal emboli and the microcirculation affecting the cortex, may be assessed more importantly using other functions such as aphasia or visual-spatial neglect. These abilities may be more important to patients’ quality of life than merely whether they can or cannot walk.
Alan Z. Segal, MD, is Associate Professor of Neurology, Weill Cornell Medicine.
Two recent trials from China studied the effect of post-thrombectomy, intra-arterial thrombolysis on neurological outcomes. One study used tenecteplase and the other used urokinase. Both studies showed a trend toward improvement that was not statistically significant, and there was an increase in brain hemorrhage. The benefit of adjunctive thrombolysis with mechanical thrombectomy is uncertain and requires more study.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content