5 CHAPTER 5 106 not different from the spontaneous migraine attack study and thus could have picked up a similar rise in PGE2 levels. In addition, two studies found that PGE2 levels in plasma, 39 (18 cases and 12 controls) and saliva 40 (6 cases and 9 controls) in subjects with migraine were lower compared to controls outside attacks and increased during a spontaneous attack surpassing the levels found in controls. Although this was not our primary question, we tested this and did not find a difference in baseline PGE2 levels between cases and controls. Giving the small number of participants in previous studies, our larger study should have been able to reveal differences in PGE2 levels during GTN-induced migraine-like attacks. Another reason for the discrepancy with earlier studies might be in the measuring techniques used and/or the matching and correction of data. For our study we used a highly reliable, standardized technique for measuring PGE2 levels and additionally have minimized external effects on PGE2 levels, by careful matching and correcting for multiple factors to single out the effect of PGE2 on a migraine attack. Whereas such external effects do not seem to have affected our results, they might have played a role in earlier studies. Another possibility is that spontaneous attacks are not always the same as provoked attacks (e.g. GTN provocation in migraine patients with aura leads to a migraine-like attack, but not an aura). This may indicate that in spontaneous attacks different pathways may be initiated depending on headache (sub)type, none the less these pathways ultimately lead to the same migraine headache. We envisage several possible explanations why we found no evidence for a change in PGE2 levels over the course of a GTN-induced attack. PGE2 acts via four distinct G protein–coupled receptors EP1, EP2, EP3 and EP4. Ligand binding to the different EP receptors leads to the activation of distinct downstream signaling pathways, resulting in distinct biological outcomes,31, 41 one of these second messengers being cyclic adenosine monophosphate (cAMP).31 Via its receptors, PGE2 is known to play a role in nociceptive pain processing and inflammation,42, 43 exerting both damaging pro-inflammatory and protective anti-inflammatory effects in the brain.44-46 Thus, the PGE 2 response is dependent on the array of receptors cells express as well as on intracellular pathways to which they are coupled 46, 47. Hence, any involvement of PGE 2 in the pathogenesis of migraine may be very complex. As mentioned previously, the immediate headache is thought to be the result of vasodilation via the NO-cGMP pathway,7 independent of CGRP release 8, whereas the delayed migrainelike attack is thought to be the result of trigeminovascular activation mediated via CGRP.5, 7, 9 However, there likely is extensive cross talk between both pathways (for details see Figure 2). For instance, on a cellular level multiple components in the migraine pathway are known to be vasodilators, but can also lead to migraine attacks. As exemplified by ATP-sensitive potassium (KATP) channel openers (levcromakalim) and (big)-conductance calcium-activated K+ (BK Ca) channel opener (MaxiPost), both activated via the NO-cGMP pathway, which is known to play a role in the immediate headache, but activation of these channels can also induce migraine-like attacks.33, 34 However, the rather long delay of several hours between infusion of levcromakalim/ MaxiProst and the occurrence of a migraine-like attack (with a median time of 3 hours) indicates
RkJQdWJsaXNoZXIy MTk4NDMw