To review the evidence for using triptans for acute migraine in clinical practice.

All pharmaceutical companies that manufactured triptans were approached for IPD from published and unpublished trials of triptans. The authors also stated that they conducted a systematic review of English language literature for non-company-sponsored triptan trials; details of the search (databases, dates and keywords) were not reported. The principal investigators of any trials identified from the literature search were approached for IPD.

Study designs of evaluations included in the review

Double-blind randomised controlled trials were eligible for inclusion. For six of the seven triptans, the authors were able to obtain individual patient data (IPD). Summary data from publications were used for frovatriptan.

Specific interventions included in the review

Trials that compared an oral triptan, used at a recommended clinical dose, with a placebo or active drug were eligible for inclusion. The included trials investigated various doses of sumatriptan, zolmitriptan, naratriptan, rizatriptan, eletriptan, almotriptan and frovatriptan, compared with placebo or each other. Treatment had to be administered within 8 hours of the onset of a migraine attack.

Participants included in the review

Trials of patients aged 18 to 65 years with moderate or severe migraine, defined according to the International Headache Society criteria, were eligible for inclusion.

Outcomes assessed in the review

To be included in the review, the studies had to measure headache on a 4-point pain scale. The outcomes reported in the review were: headache response at 2 hours (the primary outcome); pain-free at 2 hours; recurrence rates; sustained pain-free at 24 hours; adverse events; and intra-patient consistency. If the results for efficacy at 2 hours were not provided, the trial was included for data on adverse events only.

How were decisions on the relevance of primary studies made?

The authors contacted all relevant drug companies and principal investigators of non-company-sponsored trials, but did not state how many reviewers were involved in deciding whether the trials met the inclusion criteria.

The authors cross-checked all data with published or presented (summary) data where possible. Companies that provided data were sent the results of the meta-analyses prior to publication and asked to check the data for accuracy. It was not stated whether the principal investigators of non-company-sponsored trials were also contacted. The authors did not state how many of them were involved in making decisions about the validity of the trials. The reasons for trials being excluded were reported.

Data extracted for each individual patient related only to their first migraine attack during the course of the study.

How were the studies combined?

Relative risks (RRs) and 95% confidence intervals (CIs) were calculated for the combined data using the random-effects model of DerSimonian and Laird. The authors stated that this method was likely to overestimate the background prevalence ratio (as migraine was not a rare disease) and that, as placebo response rates could be as high as 50% in migraine treatment trials, the maximum RR would be limited. Therefore, they also calculated the placebo-subtracted proportion or therapeutic gain (by subtracting for each study the placebo response from the response to the active drug). This equates to the absolute relative risk, or risk difference (RD). The number-needed-to-treat was derived from this. The authors also seemed to have calculated odds ratios for some outcomes and hazard ratios for time-to-event outcomes.

How were differences between studies investigated?

Statistical heterogeneity was assessed using chi-squared tests. The authors also commented on clinical heterogeneity between the study designs and eligibility criteria.

Fifty-three RCTs (n=31,797) were included in the review. IPD were not available for 5 trials (n=2,892) of frovatriptan versus placebo, so the data were taken from published conference abstracts.

The authors stated that the study designs and eligibility criteria were similar across the triptan trials.

Headache response at 2 hours.

Sumatriptan 100 mg was more effective than placebo (mean RD 29%, 95% CI: 26, 34). The authors reported that eletriptan 80mg (mean RD 42%, 95% CI: 36, 48) was significantly better than sumatriptan 100 mg, while frovatriptan (mean RD 17%, 95% CI: 13, 20) was significantly worse. These appeared to be indirect comparisons.

Pain-free at 2 hours.

Sumatriptan 100 mg was significantly more effective than placebo (mean RD 19%, 95% CI: 17, 22). The authors stated that the mean RDs were significantly higher for rizatriptan 10 mg and eletriptan 80 mg, but these also appeared to be indirect comparisons.

Recurrence.

The recurrence rates with sumatriptan 100 mg were significantly lower than with placebo (RR 30%, 95% CI: 27, 33). The recurrence rates were reported to be lower for eletriptan 40 mg and 80 mg, and higher for rizatriptan 5 mg and 10 mg; these seemed to be indirect comparisons.

Adverse effects.

Sumatriptan 100 mg had a mean RD of any adverse effect of 13% (95% CI: 8, 18). The authors stated that the rates for other triptans overlapped, except for the lower values of naratriptan 2.5 mg and almotriptan 12.5 mg, which did not differ from placebo.

Direct comparisons of triptans.

Compared with sumatriptan 100 mg, cafergot 2 mg showed lower efficacy but fewer central nervous system adverse effects; zolmitriptan 5 mg showed no differences; naratriptan showed lower efficacy at 4 hours and fewer adverse effects; rizatriptan 10 mg was superior for the pain-free outcome; eletriptan 40 mg and 80 mg were both superior, although the 80 mg dose showed more adverse effects; almotriptan 12.5 mg showed no difference on efficacy end points but caused fewer adverse effects.

Triptans at marketed doses were well tolerated and effective. Differences among them were relatively small. Rizatriptan 10 mg, eletriptan 80 mg and almotriptan 12.5 mg seemed to provide the highest likelihood of consistent success.

This was a large meta-analysis of IPD. The review question was clearly stated and the inclusion criteria were appropriate. The authors identified trials by approaching manufacturers of triptans for IPD, and obtained some unpublished data. They also stated that they searched English language literature to identify non-company-sponsored trials but, since details of this search were not provided, it was unclear whether the search was adequate or whether some studies might have been missed. The authors did not provide many details of the validity assessment, although they did state that the data were checked against published versions and the meta-analyses were returned to the companies for checking. Details of the process of the review (how many reviewers were involved in making decisions at each stage) were not provided, therefore it was unclear whether steps were taken to reduce error and bias.

The methods used to combine the data seemed appropriate. However, it was unclear whether the reviewers used odds ratios or RRs for the summary estimates, and some of the results were only presented in graphical form. The authors' conclusions seem reasonable given that they were based on such a large data set.

The authors did not state any implications for practice or further research.

Ferrari MD, Roon KI, Lipton RB, Goadsby PJ. Oral triptans (serotonin 5HT 1B/1D agonists) in acute migraine treatment: a meta-analysis of 53 trials. Lancet 2001;358:1668-75.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.