Fixing a Hole in the Heart: Percutaneous Closure of the PFO
The patent foramen ovale (PFO) is frequently associated with cryptogenic stroke. PFO is present in about 25 percent of the population, but is not harmful to the vast majority. Even as the evidence surrounding this congenital condition expands, practitioners find themselves clashing on if and how to treat this common heart defect.

Currently no devices are approved in the U.S. for PFO closure, but off-label devices approved for atrial septal defect (ASD) are being widely used. Percutaneous PFO closure has been generally limited to recurrent cryptogenic stroke patients who have failed conventional drug therapy; however, recent evidence for other indications has arisen, but has not yet been proven with randomized controlled trials.

This field is continuously surrounded by debatable questions such as: what is the most optimal treatment strategy for a PFO? What role does the PFO play in cryptogenic stroke? Should a PFO be closed if discovered incidentally during surgery? And, should PFO closure be used for migraine relief? These questions remain unanswered.

To close or not to close?

The prospective, randomized CLOSURE I trial in 2009 randomized 909 cryptogenic stroke or transient ischemic attack (TIA) patients with a PFO to receive closure with the STARFlex septal closure device plus medical therapy (447 patients) or medical therapy—24-month aspirin or warfarin or combination—alone (462 patients) to measure the prevention of recurrent stroke or TIA (Stroke 2010;41:2872-2883).

CLOSURE I researchers found no increased benefit of combining PFO closure and medical therapy and reported that two-year primary composite endpoints were similar for both groups, 5.9 percent in the STARFlex arm vs. 7.7 percent in the medical therapy arm.

The lackluster results of CLOSURE I have left some clinicians questioning whether percutaneous closure of the PFO is advisable in any situation due the high rates of vascular complications, atrial fibrillation (AF) and major bleeds found in the device arm. The rates of AF and bleeding for percutaneous closure and medical therapy compared with medical therapy alone were 5.7 vs. 0.7 percent and 2.6 vs. 1.1 percent, respectively. Major vascular complications occurred at a rate of 3.2 percent in the device arm vs. 0 percent in the medical therapy arm.

Despite the higher complication rates in the device group, Anthony J. Furlan, MD, principal investigator of CLOSURE I and chair of the department of neurology at the University Hospitals Case Medical Center in Cleveland, says, "It would be incorrect to say that CLOSURE I proved that there is no indication for closure, but the indications may be very limited."

Even while PFO closure may have limited indications and its benefits go unverified, a 2008 letter by Opotwsky et al in JAMA noted that PFO/ASD closure is common for these defects and has outpaced surgery as the method of choice (JAMA 2008;299(5):521-522).

In fact, the researchers reported that percutaneous PFO/ASD closure procedures increased 58-fold over the six-year study period, while the annual number of PFO/ASD closure procedures increased 50-fold. The percentage of closure procedures performed for these congenital heart defects increased from 18.5 percent to 92.5 percent.

The authors deemed the increase in PFO closure "concerning," due to the lack of data, "especially in regards to PFO closure for cryptogenic stroke and migraine headache."

But, the increased interest for PFO closure may spawn from the fact that more than 90 percent of patients go home within the same day of procedure, and there is no need for life-long anticoagulation therapy like warfarin, where frequent blood checks are necessary.

Richard Krasuski, MD, director of Adult Congenital Heart Disease Services at the Cleveland Clinic, evaluates four to five patients per week referred to him with PFO, and in 80 percent of cases, he recommends against closure.

In most cases, Krasuski prescribes PFO patients a baby aspirin to reduce clotting risk, particularly if patients have no other alarming symptoms. However, he says that there is currently no evidence to support or refute this strategy in patients without a prior event.

Only about 20 percent of the patients with PFOs treated at Cleveland Clinic will undergo percutaneous closure, due to the fact that in most cases patients fall outside guidelines and don't have symptoms suggesting a prior neurological event, thus making device closure an unnecessary method of treatment, says Krasuski.

Similarly, John D. Carroll, MD, an interventional cardiologist and medical director of the Cardiac and Vascular Center at University of Colorado Hospital in Aurora, Colo., sees 150 PFO patients per year and refers only half for percutaneous closure. This includes enrolling patients into clinical trials to better study PFO closure.

"This is an area where many uncertainties remain and the evidence base for a PFO closure is incomplete," says Carroll.

But, who are the patients who might benefit from a percutaneous PFO closure?

Krasuski offers that patients who are being considered for closure should be on appropriate medical therapy—either aspirin or blood thinners (warfarin)—and be experiencing recurring neurologic events, where other causes other than PFO have been excluded.

"As long as these patients do not have clotting disorders or other contraindications, they are probably eligible to undergo device closure," he says.

Krasuski says there are three options for patients presenting with a first neurologic event and a PFO:
  • Medical therapy;
  • Off-label device closure; and
  • Enrollment into randomized controlled clinical PFO trials.

However, Carroll says there are some instances when treating PFO patients with percutaneous closure may be necessary, including:
  • A pre-lung transplant patient with a significant PFO. Patients with moderate to large PFOs post-transplant often experience respiratory difficulties including rejection and are at risk of having shunting through their PFO causing blood oxygen levels to fall to dangerously low levels; thus, patients cannot undergo transplant unless the PFO is closed; or
  • A patient with cryptogenic stroke placed on warfarin by his or her primary care physician or neurologist who then has a brain MRI study indicating multiple infarcts despite the medical therapy. PFO closure could then be considered.

"I don't think there is any way that you can unequivocally say that PFO closure is superior to medical therapy, but there are patient modifiers that come into the decision-making process," Carroll notes. "These factors may point to the fact that the strategy of PFO closure as opposed to life-long anticoagulation is something that could be considered for this individual patient."

What to do with incidentally found PFOs?

In 2005, researchers at Columbia University in New York City, surveyed cardiothoracic surgeons on what they would do if they found an incidental PFO during surgery (J Cardiothor Vasc An 2005;19 [2]:150-154).

About 1 percent of respondents said that if a PFO was found during an off-pump procedure/bypass surgery, they would convert to an on-pump surgery to close the PFO. Additionally, 27.9 percent of the respondents said they always close an intraoperatively diagnosed PFO compared with 10.2 percent who said they would never close an intraoperatively diagnosed PFO.

These results showed inconsistencies in the way PFOs are managed during open heart surgery.  

Krasuski calls the findings "remarkable," because "when you are consenting a patient for a certain surgery and he or she is expecting that surgery, to alter the surgical plan based on something found incidentally is very controversial, particularly with the PFO, which may not have any additional consequences and is so incredibly common."

Krasuski says that some literature argues that closing an incidental PFO during surgery can help facilitate recovery even if not related to stroke, and can aid in a more rapid removal of a ventilator or shorten a hospital stay. Given these controversies, Krasuski and his Cleveland Clinic colleagues conducted a clinical trial that examined 13,092 patients undergoing open heart surgery with no prior PFO diagnosis (JAMA 2009;302(3):290-297). In the 17 percent of patients with incidentally discovered PFO, 28 percent underwent operative closure of the PFO.  

Patients with a PFO were no more likely to experience stroke or die after surgery and hospital stays were similar, according to the study. Closure of the PFO did not reduce hospital stay and did not improve long-term survival. In fact, patients undergoing closure were 2.5 times more likely to have a post-operative stroke than those who were left unrepaired.

"We shouldn't have a knee-jerk reaction to close an incidental PFO during surgery," Krasuski says. "It's better to take a step back and think and not just instinctively close the PFO because there could potentially be adverse outcomes associated with that strategy."

Similarly, Carroll says "PFOs are not abnormal structures and they are incidentally found frequently." He says that the "big challenge from an investigative and clinical viewpoint is to distinguish an incidental PFO from a PFO that has been potentially involved in a clinical event, like a stroke."

Carroll says that closing incidentally found PFOs may make sense, particularly in cases where a surgeon is performing a mitral valve operation where they enter through the atrial septum. "When they are done, it does make sense to sew up the PFO; however, this shouldn't be a clear-cut indication to place a device to close the PFO."  

While Carroll offers that PFO closure is "relatively straightforward," he says that the PFO anatomy varies and there are a number of procedural complications. While he estimates that these devices should have complications less than 1 percent of the time, device complications such as atrial arrhythmias, thrombi forming on the device, among others, are always a concern.

Can PFO closure relieve migraine stresses?

Recent data have surfaced revealing that there may be a distinct association between the PFO and migrain e headache. Also, questions remain as to whether or not treating this patient population is recommended or beneficial, leaving clinicians drawing a line in the sand.

In the 2008 MIST trial, Dowson et al assessed the effects of PFO closure for migraine in patients who failed two or more classes of prophylactic treatments and had moderate to large right-to-left shunts with the presence of a PFO [Circulation 2008; 117 [11]: 1397-1404].

A total of 147 patients were randomized to receive either transcatheter PFO closure with the STARFlex device (74 patients) or a sham procedure (73 patients).  The researchers found no significant effect for primary (cessation of migraine headache) or its secondary endpoints. Larger trials are underway for further investigation.

While studies have shown a link between the PFO and migraine, Carroll says, "The migraine issue has been further compounded by a lack of understanding of how the PFO plays into the pathophysiology of migraines. That is where the field has reverted to going back to the basics as to how the presence of PFO can cause or trigger a migraine."

He notes that further randomized and sham-controlled migraine trials will be necessary to understand the potential role of PFO closure in migraineurs.

Krasuski says that the MIST trial failed to show a resolution of migraines, but there were some signals that it reduced the severity of migraines in some patients. However, he urges that only patients with migraines enrolled in randomized controlled trials (RCTs) should undergo these closure procedures.

Carroll calls closure of PFO in migraine patients an "experimental act," and says that interventional cardiologists should "back off" and leave these types of closures to be managed by neurologists managing patients within RCTs. While PFO closure for migraine may be an "innovative and promising treatment," he says we should not apply a "gunslinger approach."

While the evidence base for PFO closure varies on whether the benefits outweigh the risks, the spike in the number of percutaneous PFO closures warrants further investigation, particularly because these procedures are performed with off-label devices. Prior to CLOSURE I, PFO closures have been reimbursed under the same ICD-9 code as ASD; however, due to the bleak results of CLOSURE I, Furlan says reimbursement criteria may change.

"If a physician is implanting devices in more than half of his or her PFO patients, he or she is probably not following the existing evidence," concludes Krasuski. "You can cause a world of harm if you don't understand the procedure or don't have enough expertise with percutaneous closure, even though it is not a particularly challenging procedure."

The indications for percutaneous PFO closure need to be further refined with larger clinical trials, particularly as new evidence, such as that to improve migraines, continues to arise. While some patients may benefit from PFO closure, until RCTs prove that closure benefits surpass the risk of complication, medical therapy will likely remain the first-line treatment strategy for most PFO patients.