RESPECT "device-in-place" analysis: Benefits of PFO closure?

June 26, 2013

London, UK - A post hoc "device-in-place" analysis of a trial of patent foramen ovale (PFO) closure in the secondary prevention of stroke has shown a magnification of benefits in comparison with the intention-to-treat or per-protocol analyses presented previously [1].

The latest analysis from the Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment (RESPECT), which considers only patients who actually had the PFO-closure device fitted as being in the active-treatment group, was presented last month at the 2013 European Stroke Conference .

"The magnification of benefit seen in this analysis is what would be expected if the device was exerting a biological effect," said Dr Jeffrey Saver (University of California, Los Angeles), reporting these latest data.

Also adding to the PFO-closure controversy, a new meta-analysis published earlier this month in Heart[2] found that percutaneous closure of PFOs is not clearly superior to medical therapy (see sidebar).

The RESPECT trial included 980 patients who previously had a cryptogenic stroke and had a PFO. They were randomly assigned to PFO closure with the Amplatzer (St Jude Medical) device or to medical treatment. The main results, published in the New England Journal of Medicine earlier this year and reported by heartwire at that time, showed a nonsignificant difference in the primary end point (death or recurrent stroke) in the intention-to-treat analysis but a significant reduction in the per-protocol analysis with the closure device.

An editorial accompanying the published study suggested that the data were not strong enough to advocate routine PFO closure in patients who had had a cryptogenic stroke.

Saver explained that for the current "device-in-place" analysis, 24 patients who were randomly assigned to the device group but who did not actually receive the device are now considered as being in the medical-treatment group.

Three of these patients had a stroke. "So instead of there being 16 events in the medical arm vs nine in the device arm as in the intention-to-treat analysis, the 'device-in-place' analysis has 19 events in the medical arm vs six in the device group."

This gave a 70% risk reduction (p=0.006). "This makes a big difference to the results, substantially magnifying the benefit of the device," Saver commented.

"When you do this type of analysis you are removing some of the noise that happens by chance, and if the treatment is working, the effect should be magnified. This is exactly what we see," he added. "These results provide further evidence that there is a genuine biologic effect. The device gives a modest but real benefit over medical therapy. The event rates in this trial were low, so it is challenging for the device to demonstrate superiority. It has to be extremely effective and safe to do so."

Saver cautioned, however, that this "device-in-place" analysis can only be considered exploratory because this type of analysis can sometimes overestimate the treatment effect.

A second smaller trial of the Amplatzer device in this indication, the Clinical Trial Comparing Percutaneous Closure of the Patent Foramen Ovale Using the Amplatze r PFO Occluder with Medical Treatment in Patients with Cryptogenic Embolism (the PC trial), published in the same issue of the New England Journal of Medicine as RESPECT, has shown similar results.

Saver said that an individual-patient meta-analysis of the two trials is now planned.

New meta-analysis shows no benefit, but certain "signals"


A new meta-analysis looking only at controlled trials of PFO closure (both randomized and with historical controls) will add more uncertainty to the often-polarized discussion over PFO closure. In a study published online June 21, 2013 in Heart, Dr Mathias Wolfrum (Spital Zollikerberg, Zollikerberg, Switzerland) and colleagues combined 14 studies, including the three randomized trials—RESPECT, PC Trial, and CLOSURE I —looking at the risk of recurrent stroke.

They found no significant difference with regard to recurrent stroke in patients who underwent PFO closure vs patients treated medically (RR 0.66, 95% CI 0.37-1.19). When the randomized trials were excluded from the meta-analysis, stroke was significantly reduced following PFO closure, a finding also seen in observational studies. And finally, a "time-to-event" analysis that include the three randomized controlled trials, plus two nonrandomized studies that controlled for "strict multivariate adjustments," showed a "borderline significant" reduction in stroke (RR 0.58, 95% CI 0.33-0.99, p=0.047). Of note, Wolfrum et al also saw a significantly increased risk of atrial fibrillation in PFO-closure-treated patients.

"Unfortunately," the authors conclude, "this meta-analysis is not able to contribute to a better understanding of this issue."

-Shelley Wood

The PC and RESPECT studies were sponsored by St Jude Medical. Saver discloses employment with the University of California, which received research funds and payments for consulting services from St Jude Medical.




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.