Abstract
Dear Sir It is unusual, even in the contentious area of migraine pathophysiology, to see entirely contradictory results from experiments using similar methods. In the May issue of Cephalalgia, we Aurora et al. (1) report the results of a study in which the threshold for the induction of phosphenes by single-pulse magnetic stimulation over the occipital region was compared across migraine patients with and without aura, and controls. We found that phosphene threshold was significantly lower in both migraine groups compared with controls. In direct contrast, Bohotin et al. (2), in the same issue, report that phosphene threshold is significantly higher in migraine patients with and without aura, compared with controls.
A number of methodological differences may have given rise to this contradiction, and it is important to highlight where the studies differ in terms of patient selection, technical detail, procedure and some aspects of the results. We further suggest that investigators in this area might wish to discuss the possibility of adopting a common protocol for future studies.
Bohotin et al. (2) comment that we may have introduced a confounding variable in our previous study (3) by selecting patients who reported visual triggers. In our more recent study (1), only the migraine with aura group were selected on the basis of their visual sensitivity. The migraine without aura group was recruited prospectively without reference to triggers or any other interictal visual sensitivity. The phosphene threshold of the migraine without aura group did differ from controls when those with a personal or family history of migraine were excluded.
Commercially available stimulators differ in a number of respects: amongst the most important of these are the maximum magnetic field strength, the available coil designs, and the way in which current flows through the stimulating coil (monophasic or biphasic). Our study (1) and that of Bohotin et al. (2) differ in terms of the maximum magnetic field strength and the coil design; biphasic currents were common to the two studies. The figure-of-eight coil used by Bohtonin et al. may in fact be physiologically different. There may be the potential of producing an artificial hypoexcitability in baseline unstable cells if there is a potential of producing spreading cortical depression (SCD). Theoretically then, it may take 15–20 min for the cells to repolarize, making the neurones hypoexcitable.
Bohotin et al. (2) claim that the time of the next attack after the experimental procedure is ‘a crucial variable’ (p. 265) to control for in studies of phosphene threshold. They furthermore claim that menstrual phase should be controlled for, as it may influence cortical excitability. In fact, there is very little evidence for these assertions. The timing of headache, however, has not proved a factor in phosphene threshold when subjects have been studied in a longitudinal fashion (4). In addition, Young et al. (4) show there was no variability of the phosphene threshold with the menstrual period. This finding is also reported by Mulleners et al. (5). We think it unlikely that these issues of study timing relative to migraine attacks could have given rise to the entirely opposite pattern of data reported by ourselves and Bohotin et al.
There were also differences between the two studies with regard to the blinding of the investigator. In our study, the investigator was fully blinded to diagnostic status. Bohotin et al. (2) report (p. 266) that ‘the investigator…was blinded to the diagnostic subtype of migraine’. Presumably therefore the investigator knew which participants were controls and which were not. This is not an ideal situation, for obvious reasons.
Transcranial magnetic stimulation has been embraced with enthusiasm by researchers interested in migraine pathophysiology, and for good reason. It offers a relatively cheap, non-invasive method of stimulating the human cortex in vivo. It is fair to say that the majority of published studies examining phosphene threshold concur that this threshold is lowered in migraine, and interpret this as evidence in favour of cortical hyperexcitability. One group (that of Bohotin and colleagues) consistently finds in favour of a cortical hypoexcitability. It remains possible that differences of method are at the root of this disagreement. To re-emphasize, it may be of use to the headache research community to agree on a systematic protocol for future studies of the phosphene threshold.