Letter by Meinel et al Regarding Article, “Endovascular Treatment for Acute Ischemic Stroke in Patients on Oral Anticoagulants: Results From the MR CLEAN Registry”

We read with great interest the article by Goldhoorn et al¹ exploring the safety of endovascular stroke therapy in patients on oral anticoagulants (OAC). We recently performed a multicenter observational cohort study and meta-analysis including 7462 patients regarding this topic.² In our meta-analysis on vitamin K antagonists ([VKAs] common odds ratio, 1.62 [95% CI, 1.22–2.17]), but not on direct oral anticoagulants ([DOACs] common odds ratio, 1.1; 95% CI, 0.4–2.8), had an increased risk of symptomatic intracranial hemorrhage (sICH) after endovascular stroke therapy as compared to patients not on OACs. Consistently, after adjustment in the observational study including 1942 patients, the association of VKA with sICH (adjusted odds ratio, 2.6 [95% CI, 1.3–4.8]) remained tangible without safety signals in patients on DOACs (adjusted odds ratio, 0.9 [95% CI, 0.3–3.3]).²

In the article by Goldhoorn et al,¹ we found a discrepancy between the estimates for sICH in the Abstract (adjusted odds ratio, 0.63) and the Results section/Table 2 (0.79).

Furthermore, we think that the statement in the discussion that the association with lower risk of sICH for patients on OAC persisted after adjustment for intravenous thrombolysis (IVT) is misleading. Primarily because the CI includes 1 and, therefore, the association is not significant. For patients with DOAC, there might even be a plausible explanation for reduced sICH risk.³ For patients with VKA, however, this seems unlikely from a pathophysiological perspective. The authors discuss themselves that IVT was less frequently administered in patients on prior OAC. However, simple adjustments for IVT, cannot fully address this since there might be an interaction between IVT and OAC regarding sICH occurrence. Also, because there is high collinearity between IVT and OAC treatment thus masking a potential increase of sICH after IVT as protective effect of OAC treatment. Restricting the model to patients not receiving IVT might have been informative. The missing dose-dependency of INR in regard to sICH in patients with VKA, might similarly be explained by different rates of IVT and less severe stroke in patients on therapeutic VKAs.

Clinicians and researchers should reliably assess OAC activity by determining compliance and laboratory diagnosis of OAC activity, especially in patients on DOACs. Grouping both populations and associated uncertainty regarding therapeutic activity, potentially hampers conclusions regarding safety and efficacy. In our study, the therapy was proven to be in the therapeutic range in only 50% of patients on VKAs and 72% of those on DOACs.²

We agree with the authors that prior OAC use should not be a sole reason to refrain from endovascular treatment. However, the reported increased risk for sICH in patients with VKA has several implications for clinical practice and research.² There are several scenarios where the risk/benefit ratio is uncertain (eg, distal occlusion, low NIHSS, small tissue at risk). In those patients, VKA pretreatment could tip the scales towards withholding endovascular stroke therapy, whereas DOAC pretreatment should not influence treatment decisions.

The authors report that successful recanalization was not significantly different between patients on VKAs and DOACs, respectively. L’Allinec et al⁴ reported an increased rate of successful reperfusion in patients on DOACs as compared to patients on VKAs (92.0% versus 74.3%, P=0.001). This was confirmed by our study showing that endovascular stroke therapy in patients on DOACs yielded excellent reperfusion rates as compared to VKA and controls (93.2% versus 81.5% versus 84.2%).²