Endpoint
PFO closure
Medical therapy
HR (95%CI)
p-value
Death (all cause)
1.0
0
5.20 (0.25–107)
0.95
Cardiovascular death
0
0
Stroke
0.5
2.4
0.20 (0.02–1.72)
0.14
TIA
2.5
3.3
0.71 (0.23–2.24)
0.56
Peripheral embolism
0
0
Myocardial infarction
1.0
0.5
2.04 (0.19–22.5)
0.56
PFO-related hospitalization
6.4
6.2
1.02 (0.48–2.21)
0.95
Bleeding (all)
3.4
5.7
0.58 (0.23–1.47)
0.25
Atrial fibrillation
2.5
1.0
2.60 (0.50–13.4)
0.25
Stroke, TIA, or peripheral embolism
2.5
5.2
0.45 (0.16–1.29)
0.14
Stroke, TIA, serious bleeding or peripheral embolism
2.9
5.7
0.49 (0.19–1.32)
0.16
Procedural success
96.9
N/A
Effective closure
95.9
N/A
At the event rates observed in our trial, this relative risk reduction would result in a number-needed-to-treat of approximately 40 patients to prevent one stroke over 5 years. With an annual control group event rate observed in CLOSURE I [16], our observed 80 % relative risk reduction would result in a number-needed-to-treat of 17 to prevent one stroke over 5 years. Both NNTs would be clinically relevant in our view, especially since our observed relative risk reduction of stroke was confirmed in the larger RESPECT trial, which used stroke as the pre-specified primary endpoint. Our trial, CLOSURE I and RESPECT taken together suggest that PFO closure is more effective as a lifelong medical therapy with antiplatelets or oral anticoagulation, as ongoing meta-analyses will confirm.
A PFO is more than a cosmetic blemish. In the realm of major pulmonary embolism, the presence of a PFO increased mortality from 14 to 33 % and the risk of stroke from 2 to 13 % [17]. In keeping with that, silent brain infarctions were found in patients with pulmonary embolism in 33 % with a PFO and 2 % without whereas peripheral embolism was found in 20 and 0 %, respectively [18]. A large population study in Denmark found a 2.5 % incidence of stroke and MI within the first weeks after a thromboembolic clinical event [19]. Even though this was not examined, that might well be charged to the presence of a PFO in the usual 25 % of these patients, Finally, the significantly lower prevalence of a PFO in old compared to young people ascribed to ongoing spontaneous closure of the PFO [20] may equally well reflect increased selective mortality with a PFO.
It is highly likely that extended follow-up of the PC trial (and the RESPECT trial for that matter) will reveal a significant reduction of stroke with PFO closure. In a 10-year follow-up analysis of patients randomly assigned to PFO closure or medical treatment and cared for by neurologists thereafter, a significant mortality benefit driven by stroke reduction was demonstrated when comparing the time after to that before or without PFO closure [8].
