Migraine and olcegepant – Authors' reply

Abstract

Tony Ho and colleagues (Dec 20, p 2115) conclude that “telcagepant 300 mg is eff ective as an acute treatment for migraine with effi cacy comparable to that of zolmitriptan 5 mg, but with fewer associ ated adverse events”. But are all the eff ects of telcagepant 300 mg in this randomised trial clinically relevant? For pain freedom at 2 h—the primary effi cacy measure recommended by the International Headache Society—the results are somewhat disap pointing despite their highly signifi cant p val ues: 26·9% for tel cage pant 300 mg and 9·6% for placebo. Furthermore, all active treatments eliminated pain more eff ectively than placebo, starting from about 1 h after the dose (see original fi gure 2). It would have been nice to have numbers, but none are given. Inspection of the curves shows, however, that these early painfree responses after 1 h all occur in less than 10% of patients and thus, for most patients, they are not clinically relevant. Rescue medication 2 h after intake of the test drug—a clinically relevant effi cacy measure—is not reported. Instead, space is used for the peculiar parameter “sustained pain relief (mild or no headache) for 24 h and 48 h”, which is not recommended by the International Headache Society. Finally, comparison of various outcomes on the migraine-specifi c quality-of-life questionnaire bet ween active treatment and placebo showed highly signifi cant p values. However, the actual fi gures are similar: the mean score for work functioning in the preceding 24 h was 12·4 for telcagepant 300 mg, 12·6 for zolmitriptan 5 mg, and 11·4 for placebo. For migraine symptoms, the mean scores were 13·2, 13·2, and 11·7, respectively. Parts of these results might even have been caused by the use of rescue medication. In conclusion, minor, clinically irrelevant diff erences can result in small p values in very large random ised trials in migraine. Authors’ reply Although we respect Peer TfeltHansen’s view, we would like to stress that the effi cacy of telcagepant was better than that of placebo and comparable to that of a well established triptan, zolmitriptan 5 mg. Both the effi cacy and tolerability results for zolmitriptan in our trial were similar to those obtained in a large meta-analysis of 53 triptan trials, supporting the external validity of our trial. Telcagepant proved well tolerated, with an adverse event rate similar to that of placebo and lower than that for zolmitriptan. Thus, for migraine patients who are not satisfi ed with other acute treatments owing to limited effi cacy, poor tolerability, cardiovascular risk factors, or contraindications, telcagepant, if approved, might off er an additional treatment option. Tfelt-Hansen also has specifi c comments relating to sustained effi cacy measures and use of rescue medication. The defi nition of sustained pain freedom from 2 to 24 h includes “absence of use of rescue medication”. Thus, reporting the latter as a separate endpoint would seem redundant. We chose this outcome measure as the key secondary endpoint, in addition to I declare that I have no confl ict of interest.