AF With Stroke Despite OAC (I)

Although the use of oral anticoagulant (OAC) can prevent the occurrence of ischemic stroke in patients with atrial fibrillation (AF), OAC itself is never a panacea. Some AF patients may still have a stroke despite taking OAC. It may possess a difficult clinical scenario for the in-charge physicians about what to do next. 

A recent article by Seiffge D, et al, addressed this important clinical issue using dataset that pooling multiple prospective registry in Europe, Japan, and the United States. Yesterday, I share this paper with my colleagues in Taiwan as an online lecture. After a nice discussion after the lecture, I try to make a summary in serial article in my bloggers. What you see now is the first one.

One of the co-authors, Professor Masatoshi Koga, is a friend of mine for many years. He now works in Japan's National Cerebrovascular Center, Suita, Osaka, Japan.

I summarized Methods and Results as the following:

In the Results, the authors stated that

Are those percentages correct? Typo?

In the primary analysis (OAC prior vs. naive), patients in the OAC prior group were older and have more comorbidities. However, the authors did not provide the percentage of AF known before / diagnosed after stroke in each group. That's a major limitation, though I guess the percentage of AFDAS should be higher in the OAC naive group.

Univariate analysis with K-M curve showed that OAC prior had higher risk of stroke recurrence compared with OAC naive.

In multivariate analysis, OAC prior group had 60% more risk of stroke recurrence, compared to OAC naive. Co-variate in the models included the timing of AF diagnosis. It's very strange that they didn't show that information in Table 1.

In the secondary analysis comparing risks of stroke recurrence in OAC changed vs. unchanged, patients in the OAC changed group had fewer ischemic stroke as the index event (74.6% vs. 88.7%, p = 0.002).

However, the authors stated in the manuscript that “Patients with OACchanged were older, more often female, had more often an ischemic stroke as index event,…” Is it a typo here?

K-M curve showed that there's no difference between OAC changed vs. unchanged.

There're also no differences of risks for primary and secondary outcome in multivariate analysis, too.

The authors concluded that "patients having a stroke despite being on therapy with an oral anticoagulant are at high risk of recurrent ischemic strokes". Besides, "changing the type of anticoagulant after the index event was not associated with a decreased risk of further ischemic strokes." Those patients may require "better prevention strategies, for example, adding antiplatelet therapy or left atrial appendage occlusion to OAC."

(To be continued)