Intravenous thrombolysis in patients taking direct oral anticoagulants (ESO IVT guidelines comment)

Dear Editors

With great interest, we read the publication of Berge, Whiteley et al “European Stroke Organisation guidelines on intravenous thrombolysis (IVT) for acute ischaemic stroke”. ¹ We congratulate the authors for their tremendous work in accomplishing this very important guideline.

We were particularly interested in PICO 12.3 on IVT in patients on prior non-Vitamin K antagonist oral anticoagulant (NOAC) therapy.

We would like to highlight that the authors focus on safety (i.e., the risk of symptomatic intracerebral haemorrhage) and not functional outcome (as outlined in the PICO). We fully agree with the authors that in the context of co-medication which potentially increases the risk of bleeding, safety should be the first priority. We would like to suggest reconsidering the PICO question accordingly in future updates.

As for many specific IVT-in-stroke settings, data to help guiding treatment decisions is scarce. Nevertheless, there is a significant amount of data available which has recently been summarized with a special interest in selection criteria. ²

In their expert consensus statement, the authors provide the following recommendation:

“For patients with acute ischaemic stroke of <4.5 h duration, who used a NOAC during the last 48 h before stroke onset, and who have an anti-Xa activity <0.5 U/ml (for factor Xa inhibitors) or thrombin time <60 s (for direct thrombin inhibitors) 7 of 9 group members suggest IVT with alteplase”.

We respectfully do not agree with this recommendation. Why did the experts decide to use uncalibrated anti-Xa activity with a cut-off of <0.5 U/ml? We could not find any publication that actually reported data supporting this approach. In contrast, studies ² that used calibrated anti-Xa activity instead are available. Therefore, we wonder why the authors recommend an unsupported approach but failed to include an already widely used one. We also find it difficult to understand why an anti-Xa activity cut-off of 0.5 U/ml was chosen, given that, based on extrapolation of data from surgeriy, a cut-off of 30 ng/ml was suggested for calibrated NOAC tests by others. Furthermore, recent data showed that a cut-off of 0.5 U/ml of uncalibrated anti-Xa activity is not sufficient to rule out NOAC plasma concentrations above either of the two thresholds (30 or 50 ng/ml) that are currently recommended by national and international guidlines. ³

If their recommendation is based on concerns about the availability of calibrated anti-Xa activity assays, and feasibility was their priority, we would like to highlight the available literature on point-of-care tests (including CoaguChek INR, Hemochron Signature Elite PT/INR and active clotting time, ACT+) to determine anticoagulant activity below recommended treatment thresholds in patients taking NOAC. ⁴

Regarding dabigatran, the authors use thrombin time in their recommendations. Thrombin time, however, is not specific for dabigatran. By using the highly sensitive thrombin time, many patients with dabigatran concentrations allowing IVT would be considered ineligible. Therefore, using dabigatran concentrations determined by appropriate assays should be recommended instead. ⁵ We disagree with the general recommendation that patients otherwise eligible for IVT should not be treated if last NOAC intake was within 48 hours and no specific coagulation test is available. As the authors outline, for dabigatran the fast acting reversal agent idarucizumab is broadly available, rendering any coagulation testing optional.

Additionally, we are surprised that 8 of 9 group members suggest the combination of idarucizumab and IVT in dabigatran patients as opposed to 9 of 9 group members suggesting against the combination of andexanet and IVT in patients with preceding Xa-inhibitor intake. The significant lack of efficacy and safety data on both approaches should be taken into account. The obvious vote in favour of one approach (idarucizumab) and against the other (andexanet) gives the perception of this recommendation being evidence-based which is not the case.

Looking back at several “historical” exclusion criteria for thrombolysis such as dual antiplatelet therapy, there is a good chance, that IVT might one day be considered acceptable even in patients with therapeutic plasma levels of NOAC.

In order to gather real-world data on the efficacy and safety of IVT in NOAC patients, we need an adequately sized international registry of NOAC patients who received IVT erroneously despite (retrospectively) confirmed high plasma levels, who received IVT after reversal treatment, and who received IVT after POCT-based selection by plasma levels. Such evaluation gains additional importance when considering the high costs associated with both reversal agents and increasing numbers of NOAC patients.

Again, we congratulate the authors for their extensive guideline on thrombolysis in acute ischemic stroke, but we hope that our letter will spark an academic debate on the important and challenging topic of IVT in patients receiving NOAC in daily clinical practice.