NEW YORK (Reuters Health) – While observational data indicate that closure of a patent foramen ovale (PFO) is better than medical therapy in preventing stroke recurrence, the results of a randomized trial show that there is no difference between the two strategies, according to a systematic review published online December 15 in Stroke.

“The case of PFO closure highlights the risks of over-reliance on observational evidence, particularly when RCTs are feasible and ethically justifiable,” comment the authors of the report.

Dr. David M. Kent, with the Institute for Clinical Research and Health Policy Studies at Tufts Medical Center in Boston, Massachusetts, and colleagues explain that mechanical closure of a PFO discovered in patients who have had a cryptogenic stroke has been widely adopted, based on observational reports of its safety and efficacy in preventing recurrence.  However, the recently completed CLOSURE I trial comparing PFO closure with medical therapy did not identify any significant advantage of closure.

To quantify the differences between the observational and randomized results, the team identified and synthesized data from 53 relevant single-arm studies and 7 comparative non-randomized studies, as well as the CLOSURE I data.

Based on the pooled non-randomized data, the incidence rate of recurrent cerebrovascular events following PFO closure was 0.36 events per 100 person-years compared to 2.53 events per 100 person-years with medical therapy, the report indicates.

On the other hand, the incident rate in the CLOSURE I randomized trial was 1.34 versus 1.41 events per 100 person-years for the closure and medical arms, respectively – a nonsignificant difference.

In other words, the authors point out, the incident rate of recurrent stroke in patients undergoing closure was almost 4-fold higher (1.34 vs 0.36) in the randomized trial than in the observational studies.

“The discrepancy between the findings of these 2 types of studies may be a reflection of inherent limitations of observational evidence,” they comment.  “It is well-appreciated that confounding by indication can create noncomparable populations in the medical and closure treatment arms, in terms of the underlying risk of the outcomes of interest for the patients.”

Summing up, Dr. Kent and colleagues conclude that the randomized data “challenge the credibility” of the observational evidence, although they acknowledge that further randomized trial data are needed.

Indeed, they point out that if the 7000-plus cryptogenic stroke patients undergoing PFO closure in the observational studies had instead been enrolled in a randomized trial, “then it is likely that the best way to prevent recurrent strokes in this population would now be known.

SOURCE:

Stroke 2012.