Stroke Rounds: PFO Closure Works Long-Term Only for Some

You can find out how good your doctor is by how long or whether s/he ever tells you about this study.  Good luck.
http://www.medpagetoday.com/Cardiology/Strokes/54184?
Long-term follow-up of the RESPECT trial continued to show no overall outcome advantage of patent foramen ovale (PFO) closure, although the procedure did pan out for reducing recurrent cryptogenic strokes, particularly for younger adults.
With an addition of about 2.5 years of follow-up on average, risk of all-cause strokes was no different for patients who had PFO closure with the investigative Amplatzer device than for those on medical management alone (P=0.16), John D. Carroll, MD, of the University of Colorado Denver, and colleagues found.
Because other components of the primary composite endpoint -- recurrent nonfatal ischemic stroke, fatal ischemic stroke, or early death after randomization -- were likewise no different between treatment arms, the conclusion was the same as in the main intent-to-treat result at an average of 3 years of follow-up.
"This is a study that started in 2003 and there have been a lot of lessons learned," Carroll said, adding, "Things we learned over the 10 years were not all recurrent events are going to be due to paradoxical embolism, and when one out of five patients are now over the age of 60, we cannot look at PFO closure as curative."
But he pointed to side analyses that suggested the device did what it should have been expected to do.
For example, there was a 54% relative reduction in recurrent cryptogenic stroke in the PFO closure group in the intent-to-treat analysis compared with medical management group in long-term follow-up (P=0.042).
Looking just at patients with a device in place versus those who never got a device in the trial, the relative reduction was 70% (P=0.004), Carroll noted. Eleven percent of the control group sought off-study PFO closure at some point.
In sensitivity analysis, people under age 60 had a 52% relative risk reduction of any recurrent stroke in the intent-to-treat analysis, which was significant (P=0.035).
One-third of the recurrent strokes had a known mechanism that PFO closure would not have been able to prevent, from things like small vessel disease, cardioembolism from endocarditis, and atrial fibrillation.
During extended follow-up, one in five patients who reached the over-60 age group were excluded from the trial in the initial enrollment criteria, because of the higher likelihood of other confounding etiologies for stroke with older age.
"In those patients who have recurrent strokes that were not PFO mediated and had a separate mechanism, it really stresses another important take-home message...that there has to be attention to modification of other stroke risk factors," Carroll told reporters at a press conference.
Carroll reported the findings at a late-breaking clinical trial presentation at the Transcatheter Cardiovascular Therapeutics (TCT) meeting in San Francisco.
And for the 620 patients with substantial shunts or atrial-septal abnormalities, the benefit of PFO closure in preventing recurrent cryptogenic stroke was also significant in the intent-to-treat analysis (hazard ratio 0.245, P=0.007).
However, extended follow-up out to an average of 5.5 years in the PFO occluder arm and 4.9 years in the medical management arm included a declining proportion of patients as follow-up went on, noted press conference discussion panelist Roxana Mehran, MD, of Mount Sinai School of Medicine in New York City. "By the second or third year...your number at risk is half of the original population. And then when you get out to 10 years, you're down to 15 patients or so."
And considering that the initial intent-to-treat results were negative, the subgroup results have to be taken "with a grain of salt," she argued. "There's no question the curves are separating, and it's good to see that. And I do believe that those patients in the original subgroup with a significant shunt and the atrial-septal aneurysm are probably the best patients to think about for sure for recurrent cryptogenic stroke."
Carroll responded that there were still almost 600 patients being followed at 4 or 5 years and that this is probably the best evidence that will become available. "We're never going to have a trial like this again in terms of such extensive follow-up of so many patients."
There is another trial coming -- the REDUCE trial with a different PFO closure device -- though with a somewhat more modest 664 patients and 24-month primary endpoint.
While all the panel said, via a hand-raising vote, that they were convinced by the data for specific subgroups and would want the device themselves in such a circumstance, FDA approval will be a challenge with such data, predicted press conference moderator and conference co-director Ajay J. Kirtane, MD.
With such a low event rate in RESPECT, 600 out of almost 900 "doesn't really pass muster for most clinical trials in terms of follow-up," said Kirtane, of New York-Presbyterian Hospital/Columbia University Medical Center in New York City.
On the other hand, "this device looks incredibly safe," said panelist Howard C. Herrmann, MD, of the University of Pennsylvania in Philadelphia.
In adjudicated safety findings, there were no intra-procedure strokes and no cases of device embolism, thrombosis, or erosion. The rates of major vascular complications and device explant were "very low" at 0.9% and 0.4%, respectively.
The deep vein thrombosis or pulmonary embolism rate was higher with the PFO occluder than with medical management (0.61 versus 0.12 per 100 patient-years).
The difference was of "unclear significance" because of no association with procedure or access site, lack of thrombophilia evaluation in the trial, and allowance of warfarin use in the medical management group. Also, Carroll noted, the difference could have been due to differential follow-up as more participants dropped out in the medical management group.