Recurrent Stroke Rates Low 15 Years After PFO Closure

Transcatheter patent foramen ovale (PFO) closure for cryptogenic embolism holds up well a decade-and-a-half after the procedure, with a low incidence of recurrent clinical stroke, observational data show.

Through a median follow-up of 15.0 years, stroke and TIA occurred at rates of 0.27 and 0.46 per 100 patient-years, respectively, researchers led by Siu-Fung Wong, MBChB (University Hospitals Sussex NHS Foundation Trust, Brighton, England, and United Christian Hospital, Hong Kong), report.

Moreover, outcomes were largely similar when comparing patients treated with antithrombotic therapy for only 6 months after closure and those treated with a longer-term regimen, albeit with a hint toward less bleeding with the more limited approach.

There is some controversy about whether long-term antithrombotic therapy is needed after PFO closure, Wong told TCTMD, noting that a 6-month regimen is standard at his center. For treatment beyond 6 months, it’s an “evidence-free zone,” he added.

In that context, the current study, published online recently in Catheterization and Cardiovascular Interventions, provide “reassuring results that, especially for these relatively young patients, a long-term commitment to an antithrombotic agent, usually an aspirin only, may not be necessary for them,” Wong said.

Cosmo Godino, MD (San Raffaele Hospital, Milan, Italy), who was not involved in the study, agreed in an email to TCTMD.

“The most important implication is that successful PFO closure should not automatically translate into years of antiplatelet therapy,” he said. “In appropriately selected patients—successful closure, no relevant residual shunt, no alternative indication for antithrombotic treatment, and particularly a high probability that the index event was PFO-related—discontinuation of antiplatelet therapy after the early postprocedural period appears increasingly reasonable.”

A Long-term Look

Although transcatheter PFO closure has been shown to prevent recurrent ischemic stroke when compared with antithrombotic therapy in patients with cryptogenic embolism, there have been only a few studies reporting outcomes beyond 10 years of follow-up.

Wong said that because most patients undergoing PFO closure are relatively young, with diagnosis typically occurring when they’re around 40 to 50 years old, “it is more clinically relevant to know how these patients are doing in the longer term, like over 10 or even 20 years, to see how this therapy holds up.”

Previously, the Brighton research group published results of PFO closure in a cohort of 301 patients (mean age 48.6 years; 54.4% men) treated at their tertiary care center between May 2004 and January 2013, with follow-up lasting up to 6 years (mean 40 months).

The current study more than doubles the length of follow-up in the same group of patients, with 98.3% having complete data for all-cause death and 96.7% complete data for other clinical endpoints at 15 years. Over that span, 3.4% of patients had a stroke, 6.2% had a TIA, 6.5% developed venous thromboembolism (VTE), and 9.6% developed incident atrial fibrillation (AF). One out of every 10 patients died, and 3.8% had a bleeding event.

Wong noted that the incidence of VTE was higher than has been seen in other observational studies (about 2% to 3%). One possible reason is that “we do not have very stringent thrombophilia screening for those individuals having a cryptogenic stroke,” he said. “Some of them may be underdiagnosed or not diagnosed to have thrombophilia and instead have PFO closure performed.”

The influence of the duration of antithrombotic therapy—which involved a single antiplatelet in more than 85% of the cohort—was assessed in 280 patients who did not require long-term antithrombotics for another reason: 149 stopped taking antithrombotics at 6 months and 131 extended treatment beyond 6 months.

When compared with short-term therapy, longer-term treatment was associated with significantly higher risks of any bleeding (6.1% vs 1.3%; P = 0.03) and major bleeding (3.8% vs 0; P = 0.02). After propensity-score matching, both endpoints remained numerically higher in the long-term group, although the differences were no longer significant: any bleeding (5.1% vs 1.0%; P = 0.21) and major bleeding (2.0% vs 0; P = 0.50). Rates of other outcomes, including clinical stroke, TIA, VTE, AF, and all-cause death, did not vary based on the length of antithrombotic therapy after PFO closure, either before or after propensity-score matching.

Reassuring Results

Godino said this study “gives important reassurance that the protection afforded by PFO closure appears durable over time,” with a recurrent stroke rate comparable to what has been seen in both randomized trials and observational studies.

As for the findings regarding antithrombotic therapy duration, he said, “I would interpret the study as providing reassuring, rather than definitive, evidence that prolonged antithrombotic therapy may offer little additional ischemic protection after successful PFO closure while exposing patients to cumulative bleeding risk.”

Overall, Godino said its results are “remarkably consistent” with those seen in a paper on the PROLONG registry, for which he was senior author.

“There are important methodological differences between the studies—Wong et al used a 6-month antithrombotic strategy and a single-center historical cohort, whereas PROLONG compared discontinuation within 12 months with continuation beyond 12 months in a larger multicenter population—but the direction of the findings is strikingly similar,” Godino said.

“For me,” he added, “the fact that two independent cohorts from different countries and clinical settings converge on essentially the same clinical message is particularly reassuring.”

When considering the duration of antithrombotic therapy, there is a need to look at two phases, he said. 

“During the first months after implantation, antiplatelet therapy is primarily intended to protect against device-related thrombosis while endothelialization is occurring. This process is generally considered to take approximately 3 to 6 months, and in PROLONG, device-related thrombosis was extremely uncommon and occurred exclusively during the first 6 months,” Godino said. “Beyond that period, the rationale for continuing antiplatelet therapy becomes much less related to the device itself and more related to the patient’s underlying vascular risk.”

He said he prescribes dual antiplatelet therapy during the initial postprocedural period and then switches patients to single antiplatelet therapy, after which he assesses the need for prolonged medication.

“In a patient with successful PFO closure, no significant residual shunt, no other indication for antithrombotic therapy, and particularly a high RoPE [Risk of Paradoxical Embolism] score, I would strongly consider discontinuing antiplatelet therapy within 6 to 12 months,” Godino said.

On the other hand, in the presence of a lower RoPE score or other vascular risk factors, he said, “I would be more cautious, because the original neurological event may have resulted from a mechanism unrelated to the PFO and antiplatelet therapy may still have a role as secondary prevention.”

In those patients, either long-term single or dual antiplatelet therapy may be warranted, he said.

Godino cautioned, however, that both the current study and the PROLONG registry are retrospective, observational studies, which are subject to possible residual confounding, with a relatively small number of events.

“PROLONG itself explicitly considered the findings exploratory and hypothesis-generating and called for randomized confirmation,” he said. “I would not say that these studies establish a new guideline today. But taken together, they provide a strong clinical and pathophysiological rationale for a randomized trial testing a tailored strategy—potentially using RoPE score and procedural success to identify patients in whom antiplatelet therapy can safely be discontinued early.”