Doubt Cast on Major Role for PFO Closure in Cryptogenic Stroke

View on the News

Multiple Flaws Make CLOSURE I Results Misleading

I find 11 reasons why the CLOSURE I results are misleading and should not deter physicians from performing a closure procedure in patients who have had a cryptogenic stroke and have a patent foramen ovale.

1. The study tested whether PFO closure is superior to medical therapy. This was an inappropriate design because the efficacy of medical therapy had never previously been studied in a trial.

2. Patients in the PFO closure arm also received aspirin therapy for 24 months. This was a design mistake because if medical therapy is effective then PFO closure would not be expected to add any additional benefit.

3. The trial should not have excluded patients with deep vein thrombosis or hypercoagulopathy because these patients should benefit most from PFO closure.

4. The trial enrolled patients very slowly, at a rate of about two patients per year at each of the participating centers, a pace that must have introduced a selection bias.

5. The total number of about 900 enrolled patients and the study design’s presumption that the event rates would be 6% in the medical arm and 2% in the closure arm were too optimistic to produce a statistically significant difference between the two treatments tested. The study’s original design called for 1,600 patients.

6. The 2-year follow-up was too short. The average age of the patients in the study was 46. Patients like this choose to have PFO closure to avoid 30 years of anticoagulation therapy. The advantages of PFO closure would have become clearer during longer follow-up, as patients on warfarin therapy typically abandon it after several years.

7. The control, medical-therapy arm showed some strange outcomes. The event rates were higher in patients with smaller PFOs, and in those without a septal aneurysm.

8. Some of the operators performed PFO closures while still relatively inexperienced, early in their learning curve.

9. The STARFlex closure device used in the study is now outdated. It results in higher rates of atrial fibrillation and clot formation than do other closure devices now available.

10. Long-term anticoagulation is not a viable option for most patients. About 70% of patients on a warfarin regimen stop taking the drug after 5 years.

11. The 17% complication rate in patients undergoing PFO closure was unusually high. In the reported, worldwide experience with PFO closure the serious complication rate is about 3%.

Even if patients and physicians believe the CLOSURE I results despite all these problems, the question remains of which treatment option seems best for patients who have had a cryptogenic stroke and have a PFO. Doing nothing is not a good option, and closing the PFO by open surgery is not attractive because the periprocedural mortality rate is 0.5%-1.0%. The CLOSURE I results showed that medical therapy is no better and no safer than is endovascular PFO closure, and the annual risk for a major bleed on warfarin is 0.5%. For the long term, endovascular closure of a PFO is the most reliable way to prevent recurrent paradoxical embolism, and it has no downside; it was as safe as medical therapy. Patients with a PFO will usually prefer endovascular closure because they then won’t have to face the prospect of having to take anticoagulation therapy for the rest of their lives.

Dr. Horst Sievert is director of the CardioVascular Center in Frankfurt, Germany. He said that he has received consulting fees, travel expenses, or honoraria from Access Closure, AGA, Ardian, Arstasis, Atritech, Atrium, Avinger, Bard, Boston Scientific, Bridgepoint, CardioKinetix, CardioMEMS, Coherex, Contego, CSI, EndoCross, Epitek, Evalve, ev3, FlowCardia, Gore, Guidant, Lumen Biomedical, HLT, Kyoto Medical, Lifetech, Lutonix, Medinol, Medtronic, NDC, NMT, Occlutech, Osprey, Ovalis, pfm, Medical Mepro GmbH, ReCor, Rox Medical, Sorin, Spectranetics, Square One, TriReme Medical, Trivascular, Veryan Medical, and Viacor. He said that he owns stock or has stock options in CardioKinetix, Access Closure, CoAptus, Lumen Biomedical, and Coherex. He made these comments during a talk in January at the annual International Symposium on Endovascular Therapy in Miami Beach.



CHICAGO – Closure-device problems did not explain why endovascular closure of patent foramen ovale in patients with a history of cryptogenic stroke failed to produce better outcomes than those of medically-treated patients. The device performed well, Dr. Anthony J. Furlan said at the annual scientific sessions of the American Heart Association.

The problem was patient selection.

"For the vast majority of patients who have had a cryptogenic stroke and a patent foramen ovale [PFO] medical therapy is a good first-choice," said Dr. Furlan, professor and chairman of neurology at Case Western Reserve University in Cleveland. "The challenge now to the endovascular community is to refine the patient selection criteria and not close these holes so liberally."

Based on the results from A Prospective, Multicenter, Randomized Controlled Trial to Evaluate the Safety and Efficacy of the STARFlex Septal Closure System Versus Best Medical Therapy in Patients With a Stroke and/or Transient Ischemic Attack Due to Presumed Paradoxical Embolism Through a Patent Foramen Ovale (CLOSURE I), the first randomized study to compare anticoagulant therapy and endovascular PFO closure, "we should see a significant drop in the number of PFOs that get closed," Dr. Furlan said.

The CLOSURE I investigators selected patients "who are being closed worldwide," he said. "Many patients who have a cryptogenic stroke also have a hole and nothing else, but people don’t look that hard. Assessments tend to stop when they find a PFO, he said.

The CLOSURE I results "clearly showed that in patients with stroke of unknown cause and PFO there is no need for systematic PFO closure because it is not effective and may be associated with major vascular complications, atrial fibrillation, or bleeding," said Dr. Pierre Amarenco, professor and chairman of the department of neurology and stroke center at Bichat Hospital in Paris. "Cryptogenic stroke likely includes a broad range of underlying etiologies that may have diluted the true, causal effect of PFO. These results should now be translated into practice for the million patients with stroke of unknown cause and PFO. On a case-by-case basis, after careful evaluation by a vascular neurologist, there may be some patients for whom closing the PFO can be deemed useful, but these cases are likely very rare."

CLOSURE I enrolled 18- to 60-year-olds who had a cryptogenic stroke or transient ischemic attack (TIA) within the past 6 months, and a PFO documented by transesophageal echocardiography. Investigators at 87 centers in the United States and Canada randomized 909 patients during June 2003–October 2008 to endovascular closure with the STARFlex device or to medical therapy. Patients in the closure group also received aspirin and clopidogrel for 6 months following their procedure, followed by aspirin alone for an additional 18 months. Patients in the medical-treatment arm received warfarin alone, dosed to a target international normalized ratio of 2.0-3.0 (139 patients), warfarin plus aspirin (80 patients), or aspirin alone (243 patients) throughout their follow-up. The average age of the patients was 46; 52% were men. Nearly three-quarters had an index cryptogenic stroke.

"We believe the study population was representative of patients aged 60 or younger with cryptogenic stroke or TIA and a PFO," Dr. Furlan said.

The study’s primary end points were the individual rates of stroke and TIA and the composite rate after 2 years of follow-up on an intent-to-treat basis. The composite rates were 5.9% among the 447 patients in the closure group and 7.7% among 462 patients in the medical-therapy group, a difference that was not statistically significant. The numeric difference was driven primarily by a difference in the TIA rate, 3.3% in the closure group and 4.6% in the medical therapy group, not significantly different. The stroke rates were nearly identical in the two groups, 3.1% and 3.4%. The composite end point showed no significant difference between the two treatment arms regardless of the shunt severity or whether or not the patients had an atrial septal aneurysm.

Roughly 80% of the strokes that occurred during follow-up in both arms of the study had no relationship to paradoxical embolism, suggesting that in many of these patients the PFO may be coincidental, Dr. Furlan said.

In the control arm, the outcomes did not significantly differ regardless of whether patients received warfarin plus aspirin, warfarin alone, or aspirin alone. Dr. Furlan noted that the study was not powered to find significant differences among these three subgroups.

The overall rate of major adverse events was similar, 17%, in both treatment arms (see table). Patients who underwent PFO closure had significantly more major vascular complications and atrial fibrillations. Vascular complications occurred completely secondary to the closures, with none in the medically-managed patients, Dr. Furlan said. The closure procedure achieved technical success in 90% of patients in that arm of the study, with 86% of the PFOs closed at 6 months after treatment based on follow-up echocardiography examinations.


Recommended Reading

Stroke History Did Not Alter Dabigatran's Safety, Efficacy
MDedge Cardiology
Speed of Stroke Treatment Lags for Most in U.S.
MDedge Cardiology
MRI Can Guide Treatment in Unclear-Onset Stroke
MDedge Cardiology
Rare Stroke Primarily Affects Women, the Young : Diagnosis and management of cerebral venous thromboembolism require 'a high level of suspicion.'
MDedge Cardiology
Treating Mild Stroke Could Reduce Disability and Costs
MDedge Cardiology