PFO Closure Reduces Strokes, and PPIs Increase Dementia Risk?

Hans-Christoph Diener, MD, PhD


October 17, 2017

Dear colleagues, I am Christoph Diener, senior professor of clinical neurosciences at the University Hospital in Essen, Germany. Today I have two very exciting new studies. One is about patent foramen ovale (PFO) and cryptogenic stroke, and the other one is about proton pump inhibitors and dementia.

PFO Closure and Stroke

Let me start with PFO. We have had a number of observational studies which indicated that in people who have a large PFO, are younger than 60 years, and had a cryptogenic stroke, perhaps PFO closure could prevent recurrent strokes, but the two studies that were done were unfortunately underpowered. They only showed a trend toward efficacy of PFO closure. Now, on September 14, the New England Journal of Medicine published online three studies which were finished at the same time. It is interesting that we have an identical result.

The first trial was coordinated out of France—the CLOSE study.[1] The CLOSE study recruited 663 patients who had a large PFO or an atrial septum aneurysm and a cryptogenic stroke. These patients were randomized to either PFO closure, antiplatelet therapy, or anticoagulation. The primary endpoint was stroke, and the follow-up was 5.3 years. In the closure group, there were no strokes among 238 patients; in the antiplatelet group, there were 14 strokes in 235 patients. The hazard ratio was 0.03. By the way, there was no difference between oral anticoagulation and antiplatelet therapy.

The REDUCE study,[2] which was coordinated out of the US, randomized patients with cryptogenic stroke and a medium or large PFO to PFO closure or antiplatelet therapy by a ratio of 2:1. They also performed brain imaging at baseline and after 2 years. The primary endpoint was freedom from ischemic stroke and freedom from recurrent strokes on imaging. Thus, the study recruited 664 patients, the follow-up was 3.2 years on average. The rate of ischemic stroke was 1.4% in the PFO closure group compared with 5.4% in the antiplatelet group. There was a hazard ratio of 0.23, and rates of new brain infarcts were 5.7%, and 11.3% in favor of PFO closure. This translates into a relative risk of 0.51.

The third study was RESPECT,[3] which compared PFO closure or antiplatelet therapy in patients with large PFOs who were between 18 and 60 years of age. The endpoint was a composite of fatal or nonfatal ischemic stroke in early tests. There were 900 patients in the study and the follow-up was almost 6 years. There were 18 events out of 499 patients in the PFO group and 28 out of 481 patients in the antiplatelet group. This translates into a hazard ratio of 0.55.

Ladies and gentlemen, these are very convincing data from three independently performed randomized trials. PFO closure in patients below the age of 60 is highly efficient and is clearly superior to antiplatelet therapy. The procedure has a few side effects and risks. For example, novel atrial fibrillation was observed in 1%-5% of patients, and in some patients there was a problem with the device. I think these results will definitely change our guidelines.

PPIs and Dementia

The other topic is the relationship between the intake of proton pump inhibitors and dementia. In 2016, JAMA Neurology[4] published an analysis from Germany. This was based on insurance data. The authors looked at 73,679 patients above the age of 60, and 2950 of those took proton pump inhibitors. Hence, they looked at the diagnosis of dementia in the database. They found a hazard ratio of 1.44, which would indicate that proton pump inhibitors increase the risk for dementia.

Many of us did not believe the data, for several reasons. The sample size was small and the author unfortunately did not correct for confounders. Now we have two new papers looking at this, one in Gastroenterology and the other one in the American Journal of Gastroenterology. The first study[5] was an analysis of the Nurses' Health Study. This is a prospective study; the analysis is based on almost 14,000 nurses who were prospectively followed and had a neuropsychological test battery. Over time, there was no association of the intake of proton pump inhibitors and cognitive function.

The other study[6] is even more convincing. The authors identified more than 70,000 patients with dementia and compared them with 282,000 controls. The odds ratio for dementia with the intake of proton pump inhibitors, with a time lag of 3 years, was 1.03 and clearly not significant. Although, if they looked at longer time intervals, they could not find a relationship between the intake of proton pump inhibitors and dementia.

Why is this important? The study published in JAMA Neurology caused a lot of problems in the public press. This came out basically as evidence that proton pump inhibitors lead to dementia, and, unfortunately, many patients who definitely need these drugs stopped taking them. I think this shows again that we should be extremely careful to draw conclusions about therapy of patients from underpowered observational studies.

Ladies and gentlemen, these are two exciting developments this month. One is that closure of PFO is highly effective in patients with cryptogenic stroke. The other is that proton pump inhibitors most likely do not lead to dementia.

Thank you very much. I am Christoph Diener, senior professor of clinical neurosciences from Essen, Germany.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.