PROSTAGLANDIN-E2 LEVELS OVER THE COURSE OF GLYCERYL TRINITRATE PROVOKED MIGRAINE ATTACKS 105 5 Discussion We performed a GTN provocation study in subjects with migraine and healthy controls and found that 82% of migraine participants developed a delayed onset migraine-like attack. We prospectively assessed PGE2 levels at three time points selected over the course of provoked migraine-like attacks and compared these to those without provoked attacks and controls. We found no evidence that GTN-induced migraine-like headaches are characterized by changes in plasma PGE2 levels towards the (pre)ictal state.This suggests that a rise in PGE2 is not an essential step in the initiation of GTN-induced migraine-like attacks. PGE2 is able to induce rapid-onset migraine-like attacks in subjects with migraine within 90 minutes,14 in contrast to provocation with substances such as PACAP, CGRP and GTN that result in a delayed (after a few hours) onset of a migraine-like attack.14, 16 Thus, we hypothesized that PGE2 could be one of the molecules involved in a(n experimentally induced) migraine attack. Given that administration of PGE2 can cause a rapid-onset migraine-like attack, in contrast to the other provocative substances, PGE2 may perhaps serve as a marker for upcoming migraine attacks, albeit that the timing of blood sampling is important. In our study, we used the GTN provocation model to assess the role of PGE2. It has been hypothesized that the time it takes to develop delayed migraine-like attack is due to various processes that include the regulation of gene expression and proteins ultimately resulting in migraine-like attacks in subjects with migraine with a median attack onset of 3 to 6 hours, after infusion of the provocation substance. Afterall, in animal models of migraine, GTN activates the COX-2-PGE2 pathway in the brainstem not before 4 hours after GTN administration.37 However, based on our proposed mechanism and the PGE2 human provocation studies with rapid onset of provoked migraine-like headaches, we expected a rise in PGE2 to be close to the start of a migraine attack as an early marker of migraine, which would fit our time points of blood withdrawal. The alternative explanation that we did not find a rise in PGE2 levels might indicate that the pathway activated by GTN towards a migrainelike attack does not primarily act via PGE2. One can envisage that pathways, independent of PGE2 via for instance cGMP or cAMP, are more strongly activated than the PGE2-pathway when GTN is administered. Another explanation might be that a rise in PGE2 is very locally and hence not measurable in blood. To our knowledge no other study measured PGE2 levels over the course of GTN-induced migraine-like attack in subjects with migraine. Still, few studies reporting measurements of PGE2 levels during spontaneous migraine attacks suggested those to be elevated in blood,38, 39 and saliva.40 More specifically, in contrast to our study, a much smaller study of only five subjects with migraine reported an increase in PGE2 levels in jugular venous blood peaking between 2 and 6 hours after the start of a spontaneous migraine attack and normalizing towards the end of the attack.38 In our study the mean attack onset was ~192 minutes, hence many cases were over 2 hours into their delayed migraine-like attack at the ~320-minute time point, which suggests that our timing was
RkJQdWJsaXNoZXIy MTk4NDMw