MIST: Migraine Intervention with STARflex Technology Trial

Mary Thompson


July 20, 2006

Editorial Collaboration

Medscape &

Presenters: Peter Wilmshurst, MD (Royal Shrewsbury Hospital, Shrewsburg, UK); Andrew Dowson, MD (King's College Hospital, London), on Behalf of the MIST Investigators

Catheter-based closure of patent foramen ovale (PFO) has been proposed as a means of preventing severe migraine headaches associated with aura. Its use is based on the assumption that migraines may be caused when unfiltered venous blood is shunted through the PFO from the right to left side of the heart, allowing migraine-triggering substances, such as microscopic emboli, to reach the brain. Although this concept is somewhat controversial, several randomized trials are now under way to test the effect of PFO closure on migraine incidence and severity.

Previous observational studies have noted a relatively high rate (about 80%) of improvement or cure in migraine patients treated with PFO closure. However, these studies were conducted on patients who underwent PFO closure because they were deep-sea divers (divers are prone to migraines because decreased ventilation and breath-holding lead to elevated levels of carbon dioxide) or had experienced a transient ischemic attack (TIA) or stroke; their reports of migraine relief were unexpected and incidental and their baseline migraine burden was not measured.

The Migraine Intervention with STARflex Technology (MIST) trial is the first prospective, randomized, double-blind, placebo-controlled study of PFO closure for migraines to reach enrollment completion and 6-month follow-up. Although the study demonstrated a statistically significant benefit from PFO closure in terms of reduced migraine burden (the number of headaches multiplied by their length in hours) and reduction in headache days, it failed to reach its primary endpoint of complete elimination of migraine in 40% of closure patients. However, diagnostic data from the trial did appear to confirm the link between migraine and large PFO shunts. The trial investigators believe longer follow-up in the larger MIST II study, now under way, may help clarify these results.

Study Design and Methods

The MIST trial was designed to investigate the effect of catheter-based PFO closure on the incidence and severity of migraine headache. The trial was undertaken to investigate whether these results would carry over to a population of patients with a high migraine burden and no other indications for PFO closure.

MIST enrolled 147 migraine patients, aged 18-60 years (mean age, 44 years), at 13 centers in the United Kingdom between January 2005 and July 2005. To qualify for enrollment, patients were required to have experienced migraine with aura and suffered from frequent migraines (5+ days/month) for at least 1 year . They also were required to have a moderate to large PFO and to be refractory to 2 types of migraine medication. Patients with prior stroke/TIA and other cardiac anomalies, as well as patients who had overused migraine medication, were excluded from the trial. All patients underwent contrast transthoracic echocardiography (CTTE) prior to randomization to assess PFO shunt size.

To control for a placebo effect, patients were randomized via telephone in a 1:1 pattern to receive either PFO closure with the STARFlex transcatheter closure device (NMT Medical, Inc., Boston, Massachusetts) (n = 74) or a surgical sham procedure (n = 73) consisting only of an incision in the femoral artery. The majority of patients were female (84% in the treatment group and 85% in the sham group). The treatment group had a slightly higher baseline migraine burden.

Following their procedure, patients were given aspirin and clopidogrel for the first 3 months (the healing phase of the trial), and outcomes data were collected during Months 4 through 6 (the analysis phase of the trial) by a headache specialist who was blinded to the treatment.

The primary endpoint was complete cessation of migraine in 40% of PFO closure patients at 6 months, by intention-to-treat analysis. Secondary endpoints included reduction in headache burden (frequency of headache multiplied by duration) and 50% reduction in the number of headache days.


The results presented at the ACC meeting represent a preliminary analysis of the data and do not include an analysis of any residual shunt in the closure patients. However, it was clear from the available data that the study had failed to reach its primary endpoint, with 3 patients in both the treatment and sham groups achieving a 40% cessation of migraine.

Analysis of the secondary endpoints revealed more positive data, however. There was a statistically significant difference between the groups in terms of both headache burden and 50% reduction in headache days. Patients who received the PFO closure implant demonstrated a 37% reduction in headache burden from baseline, while those who received the sham procedure showed a 17% reduction in headache burden, a difference that was statistically significant (P = .033) (Figure 1). In addition, 42% of those in the implant group, compared with 23% in the sham group, demonstrated a 50% reduction in headache days (P = .038) (Figure 2).

Figure 1. MIST: 50% reduction in headache days.
Figure 2. MIST: reduction in headache burden (frequency x duration).

The researchers also noted that the treatment effect seen in headache frequency and duration appeared to improve over time, noting that better results were observed in Months 5 and 6 of the analysis period than in Month 4 (although this conceivably could be due to lingering, confounding effects of the aspirin and clopidogrel treatment given during Months 1-3).

In addition to these data, results of the CTTE exams revealed that the rate of large shunts in this patient population with a high migraine burden was 6 times that of the general population. According to the investigators, this confirms the previous observed shunt rate in migraine patients. Nearly 40% of the patients screened for the study had a large PFO, and over 43% were found to have large shunts of all types (Table).

Table. MIST: CTTE Results
Result Patients (N) %
Total studied (n) 432 100
Small shunts
(atrial and pulmonary)
72 16.7
Large pulmonary shunt 22 5.1
ASD 3 0.7
Large PFO 163 37.7
Large shunts (all types) 188 43.5
Total shunts 260 60.2
ASD = atrial septal defect; CTTE = contrast transthoracic echocardiography; PFO = patent foramen ovale
  1. Primary endpoint for the complete elimination of headache was not achieved.

  2. However, using more conventional migraine trial endpoints, such as reduction in headache burden and 50% reduction in headache days, significant differences were observed.

  3. Further analysis of MIST data may indicate which patients are most likely to show a significant treatment response.

  4. Treatment effect with regard to frequency and duration appears to improve over time.

  5. MIST II, with a longer follow-up period and larger patient population, may help validate the observed treatment trends seen in MIST.

  6. Significant prospective data from MIST may help strengthen MIST II study design.


Although the study failed to reach its primary endpoint, the investigators nevertheless called it a "success" because the results confirm a connection between migraine headaches and the presence of a PFO and demonstrate a link between PFO closure and migraine relief. This is the first time these observations have come to light in a randomized controlled trial. The study is likely to influence the design of current and future trials of PFO closure and migraine, leading to longer follow-up periods and more realistic endpoints, which, it appears, will be needed to unequivocally demonstrate efficacy of this treatment.

The MIST study was sponsored by NMT Medical, Inc. (Boston, Massachusetts).


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.