MIST Trial: PFO Closure With STARFlex vs Sham for Refractory Migraine With Aura (Dowson et al., 2008)
In patients with migraine with aura refractory to at least 2 classes of preventive medication and a moderate or large right-to-left shunt, does STARFlex PFO closure stop migraine compared with a sham procedure?
Dowson A et al. (Circulation 2008;117(11):1397-1404) · doi:10.1161/CIRCULATIONAHA.107.727271 · 147 patients
Population
Included
- Migraine with aura
- Failure of at least 2 classes of preventive medication
- Moderate or large right-to-left shunt
Verifier-confirmed inclusion
- Migraine with aura
- Failure of at least 2 classes of preventive medication
- Moderate or large right-to-left shunt
Primary Outcome
Cessation of migraine, days 91 to 180 after the procedure
PRIMARY ENDPOINT NOT MET. Migraine cessation occurred in 3 of 74 with closure and 3 of 73 with sham (P=0.51). The pre-specified secondary endpoints were also negative. The one positive figure attached to MIST, a reduction in total migraine days, came from an exploratory analysis that deleted 2 outlier patients; a formal Correction was later published, and two members of the trial steering committee refused authorship over that exclusion.
Negligible absolute difference
Migraine cessation, days 91 to 180
No confidence interval was recoverable for any MIST result, so no interval, no absolute risk difference and no NNT are shown. The two arms are drawn at the same height because the same number of events occurred in each arm, 3 and 3; the denominators differ slightly (74 and 73).
Study Arms
- Agent
- STARFlex septal repair implant
- Route
- Percutaneous transcatheter device implantation
- Frequency
- Single procedure
STARFlex was subsequently discontinued and the sponsor ceased operations in 2011, so this arm does not describe Amplatzer-era hardware.
- Route
- Sham
- Frequency
- Single procedure
MIST is one of only two sham-controlled trials in this literature. The other is PREMIUM.
Trial Design
Type
- Prospective, multicenter, double-blind, SHAM-CONTROLLED randomized trial
- Transcatheter PFO closure with the STARFlex septal repair implant vs a sham procedure
- Enrolled migraine with aura refractory to at least 2 classes of preventive medication, with a moderate or large right-to-left shunt
- Primary endpoint: cessation of migraine 91 to 180 days after the procedure
- Primary endpoint NOT met (3 of 74 vs 3 of 73, P=0.51); secondary endpoints also negative
- A formal Correction was published (Circulation 2009, doi:10.1161/CIRCULATIONAHA.109.192626)
Timeline
6-month follow-up; published Circulation 2008;117(11):1397-1404
N
147
Enrollment
147 patients randomized 74 closure : 73 sham. STARFlex septal repair implant, a device later discontinued; its sponsor ceased operations in 2011, so MIST does not speak to any device in current use. Published Circulation 2008; Correction published 2009.
Bedside Pearl
MIST (Circulation 2008) is the sham-controlled STARFlex trial and it MISSED its primary endpoint: migraine cessation days 91 to 180 was 3 of 74 with closure vs 3 of 73 with sham, P=0.51. Secondary endpoints were negative too. The reduction in migraine days at P=0.027 that gets quoted is an exploratory post-hoc analysis that deleted 2 outlier patients; a formal Correction was published and two members of the trial steering committee refused authorship over that deletion. STARFlex is discontinued. No NNT.