Abstract
Dear Sir In an editorial (1) Burstein and Jakubowski argue that the conclusion of our study (2), that headache intensity is a better predictor than allodynia of treatment outcome in migraine attacks, is not justified because of several methodological shortcomings. In doing so, they give the false impression that we were not aware of the limitations of our study and that we did not address them in the discussion section of our article. In fact, most of the methodological criticisms they make are discussed in our paper, and it is clearly stated that the conclusions must be taken with some reservations. We will focus our rebuttal on some of the criticisms with which we do not agree.
Admittedly, we included in the allodynic group attacks associated with an abnormal sensation on brushing the skin without clear unpleasantness or pain, which is beyond the strict definition of allodynia. However, many patients during the migraine attack report, on touching or combing their hair, an abnormal sensation that is not necessarily painful, often difficult to describe, which might correspond to the lowest allodynia ratings on visual analogue scales graded from 0 to 100 (3). The incidence of allodynia we found is very close to that reported with other methods, which suggests that we captured comparable sensory changes. With a stricter definition, allodynia in our study would have occurred in 18% of attacks, which would reduce it to an epiphenomenon with little relevance for migraine management. Compared to other studies using in-hospital sensory testing or assessments by questionnaires, auto-evaluation by patients in their usual surroundings minimizes hetero-suggestion and recall bias, but it explores, for obvious reasons, only one sensory modality and depends on the patients’ collaboration. Because of these limitations, it is clearly stated that the conclusions and correlations from our study cannot be generalized.
Comparing allodynic and non-allodynic attacks in the same patient, as suggested by Burstein and Jakubowski, might be a methodological advantage in a large study with a different protocol, but such a study still needs to be done. In our study design, this would have been methodologically incorrect, as drug treatment differed between the two attacks and, even if differently treated attacks would have been pooled, the number of analyzable patients would have been too small. Pooling attacks according to the presence or absence of allodynia and separating them by treatment is therefore methodologically correct and acceptable, the more so because the effect of allodynia on treatment outcome is supposed to be attack- rather than patient-related.
Burstein and Jakubowski state that we did not test the efficacy of treatment at a time when allodynia was fully established, that is, more than four hours after onset of headache. In fact, in one of their studies (4) mechanical allodynia reached its maximum two hours after the beginning of attacks with aura. The difference with the median time to treatment of 50 minutes observed in our study is thus small, although we agree that the peak incidence of allodynia may not have been reached in several patients. This shorter time to treatment, however, applies to most patients in the real world, who will not delay treatment up to four hours, and therefore provides useful indications about the relevance of allodynia for practical migraine management.
Contrary to the statement by Burstein and Jakubowski, we do not question in the discussion their finding that IV ketoralac aborted migraine attacks are associated with allodynia. This is our phrasing: “That the effect of allodynia was significant for 2-h pain-free rates when placebo, and not aceclofenac, was combined with the triptan may suggest that the NSAID is able to reduce the negative impact of allodynia. This suggestion, however, which was drawn from the open use of an intravenous NSAID (16), is not supported in our study for the other outcome measures and needs thus to be confirmed by adequate larger trials.”
To conclude, we maintain that our study, originally designed to prove the opposite, provides scientific evidence that headache intensity, and not cutaneous mechanical allodynia, determines the therapeutic response to almotriptan, combined or not with aceclofenac, in migraine attacks treated on average within 50 minutes after onset. From a neurobiological point of view, this is not surprising, as central sensitization, which underpins allodynia, is driven by the afferent nociceptive input, and thus pain. In fact, since our article was accepted for publication, several other early treatment trials of acute migraine have shown that the efficacy of triptans is not reduced by the presence of allodynia (5–7). Whether allodynia, and not headache intensity, is the predominant factor that reduces acute treatment efficacy in later stages of the migraine attack remains to proven in adequate controlled trials.