Original title: Closure of Patent Foramen Ovale versus Medical Therapy after Cryptogenic Stroke. Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment (RESPECT). Reference: John D. Carroll et al. N Engl J Med 2013;368:1092-100.

It is still unclear if the closure of patent foramen ovale (PFO) is effective in preventing recurrent stroke (CVA) after cryptogenic ischemic stroke. 

Observational studies and meta-analyses suggest benefits, however the CLOSURE 1 randomized trial failed to show any superiority. This prospective, multicenter, controlled trial included patients with a history of cryptogenic stroke and the presence of PFO identified by TEE. Patients were excluded if they had an identified mechanism that explains stroke such as carotid disease, cardio embolism, lacunar stroke, (small vessel disease), or a hyper-coagulable state. The primary efficacy end point was a composite of non-fatal recurrent ischemic stroke, fatal ischemic stroke and early death.

Between 2003 and 2011, 980 randomized (1:1) patients were included either to percutaneous PFO closure, (Amplatzer PFO Occluder), or medical treatment, (aspirin, warfarin, clopidogrel or aspirin / dipyridamole). The device was successfully implanted in 96.1% of patients. The mean follow-up was 2.6 ± 2 years of the primary analysis with the intention to treat nine patients in the device closure group and sixteen patients of the medical treatment group who had a recurrent stroke, (OR 0.49, 95% IC 0.22 to 1.11, p = 0.08). When conducting the secondary analysis, (pre-specified by protocol), according to the treatment they actually received, five patients in the device closure group and sixteen patients in the medical group had recurrent stroke; this difference was significant, (OR 0.27, 95% IC , 0.1 to 0.75, p = 0.007).

Recurrent stroke size was different between groups with 69% of the patients in the medical group with moderate, large or massive infarcts versus 14% in the closure group, (p = 0.06). The rate of serious adverse events was not significantly different between the device closure and medical treatment groups, (23% versus 21.6% respectively, p = 0.65).

Conclusion: 

The closure of the foramen ovale with an Amplatzer PFO Occluder device in patients with cryptogenic stroke was superior to medical treatment in the secondary analysis, (pre-specified by protocol), but this was not so in the primary crude analysis by intention to treatment, where the difference of events between groups was not significant.

Editorial Comment: 

This study has some limitations that make interpretation difficult. For example, the loss of patients at follow-up was high and significantly higher in the medical treatment group, (17.2% versus 9.2%, P = 0.009), which reach the criterion for non-validity for the crude cohort by intention to treat. To this we must add the nine patients who had recurrent stroke in the group assigned to device closure, three patients actually had not received the device at the time of the event. Simultaneously with this work, another was published in NEJM, (it included half the patients and had different inclusion criteria), which also failed to demonstrate significant differences on the analysis by intention to treat. Beyond these, methodological discussions benefit seems marginal and there are probably groups who benefit most as the greatest degree of shunt and atrial septal aneurysm. 

SOLACI.ORG