Hassanain Toma, MD

Nakamura A, Ago T, Kamouchi M, Hata J, Matsuo R, Kuroda J, et al. Intensity of anticoagulation and clinical outcomes in acute cardioembolic stroke: the Fukuoka Stroke Registry. Stroke. 2013

Nakamura et al. sought to clarify whether INR levels in patients with cardioemoblic strokes are associated with the severity of neurological deficits (NIHSS) on admission, and functional outcomes (modified Rankin Scale) at discharge.  They performed a retrospective cohort study of 602 cardioembolic-stroke patient from their multicenter hospital based stroke registry in Japan. They stratified their cohort according to admission INR levels: <1.50, >1.50-1.99, and ≥2.00.

Their analysis revealed that the odds ratio of developing severe neurological deficits (NIHSS>10) when comparing the reference group (INR <1.5) with the INR of 1.50-1.99, and ≥2.00 groups, were 0.43 (95% CI, 0.43–1.00;), and 0.41 (95% CI, 0.20–0.83), respectively.  Furthermore, the odds ratio of developing poor functional outcome (mRS ≥4) was 0.2 (95% CI, 0.06–0.55) in patients with INR ≥2.00 compared to the INR <1.5 group.  They concluded that the severity of neurological deficits were inversely associated with INR on admission.

This is certainly good news, in terms of prognosis, for acute stroke patients that present to the hospital with an INR of ≥2.00, especially since tPA is contraindicated in this group. But how about the unfortunate patients that present with an INR of 1.50-1.99? According to this study, although their NIHSS at admission is expected to be relatively low, their functional outcome will not differ from those who present with an INR of <1.50.  This begs the question of whether one should treat this population with tPa, which highlights a significant shortcoming in this study; the authors should have compared a cutoff INR of 1.7. The 2013 guidelines for use of tPA precludes patients with an INR ≥ 1.7.  Assuming the authors administered tPA based on these guidelines, this may explain why functional outcome was not significant in patients that presented with INR of 1.50-1.99.  It would be interesting to compare functional outcome in patients that presented with an INR of ≥1.7, as this may lead to a change in future tPA-administration guidelines.

Furthermore, the data suggests a possible oversight of the authors in making additional valuable conclusions.  Their data revealed that a significant number of patients were bridged with heparin. They also revealed that gastrointestinal bleeding occurred in 11 patients with PT-INR <2.0, whereas no patient with PT-INR ≥2.0 developed it. Furthermore symptomatic hemorrhagic infarction occurred in 27 patients with PT-INR <2.0 and one patient with PT-INR ≥2.0. Assuming they applied the standard North American practice of bridging patients who’s INR <2.0, they could have provided an odds ratio of bleeding complication when patients are bridged with heparin. I guess for this we will have to defer to the results of the BRIDGE Study.