Brief Communications

Changes in Salivary Prostaglandin Levels During MenstrualMigraine With Associated Dysmenorrheahead_1657 844..851

Paul L. Durham, PhD; Carrie V. Vause, MS; Frederick Derosier, DO; Susan McDonald, MA;Roger Cady, MD; Vincent Martin, MD

Objective.To measure prostaglandin levels in the saliva of individuals during menstrual migraine associated withdysmenorrhea (MMaD) and in response to treatment with a single tablet combination of sumatriptan succinate and naproxensodium.

Background.Prostaglandins are thought to play a role in MMaD as elevated serum prostaglandin levels have beenreported during attacks of menstrual migraine and are increased in the menstrual fluid of women with dysmenorrhea. Whiletriptans are the primary line of migraine treatment, nonsteriodal anti-inflammatory drugs are the most commonly prescribedtherapy for dysmenorrhea symptoms. Data from recent clinical studies have provided evidence that treatment with a singletablet combination of sumatriptan and naproxen sodium is an effective abortive therapy for attacks of MMaD.

Methods.Women diagnosed with MMaD were treated with a sumatriptan succinate and naproxen sodium single tabletcombination or placebo at time of migraine attack. Saliva samples were collected at time of attack as well as 2 and 4 hours aftertreatment. PGD2, PGE2, PGF2, PGI2, and TXA2 levels were determined by enzyme-linked immunosorbent assay.

Results.Elevated levels of PGD2, PGF2, and TXA2 at 2 and 4 hours and PGE2 at 4 hours were found in saliva obtainedfrom placebo subjects when compared with onset of attack levels. However, in subjects treated with a single tablet combinationof sumatriptan and naproxen sodium, the levels of PGD2, PGF2, and PGE2 were not elevated at either time point while TXA2levels were still elevated at 4 hours.

Conclusions.Data from this pilot study provide evidence that saliva levels of several prostaglandins increase duringattacks of MMaD and that treatment with a single tablet combination of sumatriptan and naproxen sodium prevents elevationof prostaglandin levels.

Key words: dysmenorrhea, migraine, prostaglandins, saliva

Abbreviations: MMaD menstrual migraine associated with dysmenorrhea, NSAIDs nonsteroidal anti-inflammatory drugs,PGs prostaglandins, SumaRT/Nap sumatriptan succinate and naproxen sodium single tablet combination

(Headache 2010;50:844-851)

From the Center for Biomedical & Life Sciences, Missouri State University, Springfield, MO, USA (P.L. Durham and C.V. Vause);GlaxoSmithKline, RTP, NC, USA (F. Derosier and S. McDonald); Banyan Group, Inc., Springfield, MO, USA (R. Cady); Depart-ment of Internal Medicine, University of Cincinnati, Cincinnati, OH, USA (V. Martin).

NIH Registration Numbers: NCT00329459 and NCT00329355, ClinicalTrials.gov

Financial support: Supported by GlaxoSmithKline, Inc. (Substudies TRX105850/TRX105852) and sponsored by POZEN, Inc.

Address all correspondence to P.L. Durham, Center for Biomedical & Life Sciences, Missouri State University, Springfield, MO65806, USA.

Accepted for publication January 24, 2010.

Conflict of Interest: Frederick Derosier, DO, and Susan McDonald, MA, are employed by GlaxoSmithKline. Paul Durham, PhD,Roger Cady, MD, and Vince Martin, MD, receive support from GlaxoSmithKline. Carrie Vause, MS, has no conflict of interest.

ISSN 0017-8748doi: 10.1111/j.1526-4610.2010.01657.xPublished by Wiley Periodicals, Inc.

Headache 2010 the AuthorsJournal compilation 2010 American Headache Society

844

Menstrual migraine has been defined by theInternational Headache Society (IHS) as a predict-able migraine attack during perimenstrual timeperiods, confirmed in 2 out of 3 menstrual cycles.1

Menstrual migraine has a prevalence of 3% in thegeneral female population but can occur in 42-61% offemale migraineurs.2,3 Dysmenorrhea may be encoun-tered by 40-50% of all women4 and is characterizedby intense pain and uterine cramps during menstrua-tion.5 Furthermore, it has recently been suggestedthat menstrual migraine and dysmenorrhea mayshare a common pathogenesis.6

Prostaglandins (PGs) may provide the link toexplain the possible shared pathogenesis betweenthese 2 menstrually related disorders. Elevated serumPG levels are found during attacks of menstrualmigraine,7 and PG inhibitors abort and prevent men-strual migraine.8 Serum and salivary PG levels arealso increased during headache-free time periodswithin migraineurs.9 Likewise, levels of PGs areincreased in the menstrual fluid of those with dysmen-orrhea and PG inhibitors are effective abortive thera-pies for dysmenorrhea.10 Therefore, PGs may play animportant role in the pathogeneses of menstrualmigraine and dysmenorrhea.

While triptans are the primary line of migrainetreatment for most physicians, nonsteroidal anti-inflammatory drugs (NSAIDs) are the most com-monly prescribed therapy for dysmenorrheasymptoms. It was hypothesized that attacks of men-strual migraine associated with dysmenorrhea(MMaD) might represent a subgroup of attacks thatwere prostaglandin-mediated and thus moreresponsive to a triptan/NSAID combination therapy.Toward this end, 2 replicate studies were conductedto determine if a sumatriptan succinate and naproxensodium single tablet combination (SumaRT/Nap) wasan effective abortive therapy for attacks ofMMaD.11,12 These studies demonstrated significantlyhigher 2-hour pain-free rates in patients receiving theSumaRT/Nap therapy than in those receivingplacebo. There was also a reduction in dysmenorrheasymptoms in the SumaRT/Nap group when data werepooled from both studies.11

Although previous studies have shown thatPGE2, PGI2, and TXA2 were present in human saliva

during migraine,13,14 the levels of these and other PGsduring the attack phase and in response to treatmentof MMaD have not been investigated. Previousstudies involving analysis of human saliva proteinsand other molecules has provided evidence thatlevels of certain salivary molecules correlated withpathological condition and therapeutic response.15-17

The goal of this substudy was to evaluate PG levels insaliva obtained from individuals during MMaD aswell as determine the effects of SumaRT/Nap treat-ment on salivary PG levels.

METHODSPatients and Treatment.The subjects for this

substudy were selected from 2 large clinical trials(NCT00329459 and NCT00329355) involving ran-domized, multicenter, double-blind, parallel group,replicate studies to investigate the efficacy ofSumaRT/Nap for the treatment of menstrualmigraine. Subjects were female migraneurs meetingIHS criteria for migraine 1.1 or 1.2 who typicallyexperience symptoms of primary dysmenorrhea atthe onset of MMaD. Subjects could not be pregnantand had to agree to practice approved methods ofcontraception during the study or otherwise be physi-ologically incapable of becoming pregnant. Exclusioncriteria included a history of inflammatory boweldisease or other conditions that required dailyNSAID medications. Protocols for this study wereapproved and conducted in accordance with ethicalstandards of Human Subjects Institutional ReviewBoards at all study sites.

Women with MMaD were randomly assigned 1:1to receive a single tablet of sumatriptan succinate,85 mg, and naproxen sodium, 500 mg (SumaRT/Nap)or placebo. Subjects were instructed to treat withinthe first hour of MMaD attack onset while pain wasmild. Rescue medication was allowed 2 hours afterthe initial dose of SumaRT/Nap or placebo as pre-scribed by a physician. A second dose of SumaRT/Nap was allowed as a rescue medication.

Saliva Collection.Subjects collected salivasamples according to a previously published proto-col.16 Briefly, all subjects prepared the mouth byrinsing with tap water and expectorating. The mouthwas stimulated to release saliva by an application of

Headache 845

sterile 2% citric acid (Sigma, St. Louis, MO, USA) tothe tip and sides of the tongue with a cotton-tippedapplicator.The first 2 minutes of saliva was discardedto prevent mixing of stimulated and unstimulatedsaliva. Citric acid was continually applied every 30seconds for an additional 5 minutes or until 5 mL ofsaliva was collected. Subjects collected saliva byexpectorating into a cold 50 mL polypropyleneconical tube (MidSci, St. Louis, MO, USA).After col-lection, all samples were kept at -20C until prepara-tion for analysis. Samples were taken at 3 time points:MMaD baseline (pretreatment during an attack), 2hours after treatment, and 4 hours after treatment.

PG Analysis.Saliva samples were thawed slowlyat room temperature and placed on ice. The volumeof each sample was recorded. Samples were centri-fuged at 3000 rpm at 4C and the resultant superna-tant utilized for PG and total protein analysis. PGD2,PGE2, 8-isoprostane PGF2a (PGF2), 6-keto PGF1a, astable metabolite of the prostacyclin PGI2, andthromboxane B2, the stable metabolite of thrombox-aneA2, were analyzed by ELISA (Cayman Chemical,Ann Arbor, MI, USA) according to manufacturersinstructions.The amount of PG was normalized to thevolume of saliva collected and amount of totalprotein, which was determined using a standardprotein assay (Bio-Rad Laboratories, Hercules, CA,USA).

Statistical Analysis.Intra-subject statisticalanalysis was performed with the nonparametric Wil-coxon matched pairs signed-rank test, while between-group statistical analysis utilized the nonparametricMannWhitney U-test. All statistical tests were per-formed using Minitab15 statistical software. Resultsof statistical analyses were considered significant atP < .05. All values are reported relative to baselinelevels (set to 1) + SEM. Subjects who rescued with anadditional dose of SumaRT/Nap or another rescuemedication were not included in any statisticalanalysis.

RESULTSThe 15 subjects selected for this study included

white women age ranged from 23 years to 48 years.Subjects reported an average of 3.46 0.4 migrainesper month with a total average of 6.13 0.8 headache

days per month. All subjects reported having addi-tional migraines other than MMaD.

PGD2 levels were increased over attack onsetlevels at 2 and 4 hours after treatment with placebo(P < .05) but were not significantly increased in sub-jects treated with SumaRT/Nap (Fig. 1A). MeanPGD2 levels at attack were 355.37 88.6 nmol/mgtotal protein for placebo-treated subjects and214.18 36.3 nmol/mg total protein for SumaRT/

Fig 1.Effect of SumaRT/Nap treatment on salivary levels ofPGD2, PGF2, and TXA2 during a migraine attack. Placebo-treated subjects are represented by black boxes (n = 6) whileSumaRT/Nap-treated subjects are represented by light greyboxes (n = 9). Mean SEM are shown. Values for the placeboand SumaRT/Nap, whose means were made equal to 1, werenormalized to their respective levels during at attack. *P < .05relative subjects at attack, #P < .05 relative to placebo-treatedsubjects 2 or 4 hours post treatment. PG = prostaglandins;SumaRT/Nap = sumatriptan succinate and naproxen sodiumsingle tablet combination.

846 May 2010

Nap-treated subjects. Similarly, PGF2 levels, whichwere increased over attack onset levels at 2 and 4hours (P < .05) in the placebo group, were notincreased following treatment with SumaRT/Nap at 2and 4 hours (Fig. 1B). The mean attack onset levelsfor PGF2 were 4.51 0.74 nmol/mg total protein forplacebo and 2.14 0.45 nmol/mg total protein forSumaRT/Nap treatment. TXA2 levels, measured asthe metabolite TXB2, were similar to PGD2 and PGF2trends and were increased at 2 and 4 hours (P < .05)in placebo-treated subjects (Fig. 1C). However, whileTXA2 levels were not increased in SumaRT/Nap-treated patients at 2 hours, TXA2 levels were not sig-nificantly different from attack levels at the 4-hourtime point in response to SumaRT/Nap.Attack onsetlevels were 53.46 14.0 nmol/mg total protein forplacebo and 39.06 7.12 nmol/mg total protein forSumaRT/Nap treatment for TXA2. In summary, therewere significant differences between placebo andSumaRT/Nap treatment at 2 and 4 hours for PGD2and PGF2, as well as a significant difference in TXA2levels at 2 hours.

PGE2 and PGI2 levels did not follow the sametrend as seen with the other PGs. For example, levelsof PGE2 were only increased at the 4-hour time pointin placebo-treated subjects (P < .05) (Fig. 2A). Inaddition, while PGE2 levels were significantly belowplacebo-treated and attack levels 2 hours after treat-ment with SumaRT/Nap (P < .05), PGE2 levels weresimilar to attack levels 4 hours post treatment(Fig. 2A). The PGE2 levels during an attack were4.87 1.64 nmol/mg total protein for placebo sub-jects and 3.88 2.27 nmol/mg total protein for theSumaRT/Nap-treated subjects. PGI2 levels, measuredas the metabolite 6-keto PG F1a, were not significantlydifferent from attack onset in either group or timepoint (Fig. 2B).The PGI2 levels during an attack were17.5 5.34 nmol/mg total protein for placebo sub-jects and 14.94 5.91 nmol/mg total protein for theSumaRT/Nap-treated subjects.

DISCUSSIONBased on pharmacological studies, it is known

that triptans and NSAIDs target different mecha-nisms of migraine pathology,18,19 and both have beensuggested as acute and short-term prophylactic treat-

ments for menstrual migraine.20,21 Results fromrecently published studies provide evidence that asingle tablet of SumaRT/Nap may be more effectivein treating migraine headache than either compoundalone.11,12,22 It is thought that the addition of naproxensodium, which is known to inhibit PG synthesis, maycontribute to improved drug efficacy as PGs areimplicated in the pathology of MMaD.7,10 A novelfinding of our study was that while salivary levels ofPGD2 and PGF2 increased significantly after migraineonset in placebo-treatedMMaD subjects, the levels ofthese PGs were not elevated in subjects treated withSumaRT/Nap. Although TXA2 showed a similartrend with elevated levels 2 and 4 hours after onset ofattack in placebo-treated subjects, the levels of TXA2

Fig 2.Effect of SumaRT/Nap treatment on salivary levels ofPGE2 and PGI2 during a migraine attack. Placebo-treated sub-jects are represented by black boxes (n = 6) while SumaRT/Nap-treated subjects are represented by light grey boxes(n = 9). Mean SEM are shown. Values for the placebo andSumaRT/Nap, whose means were made equal to 1, were nor-malized to their respective levels during at attack. *P < .05relative subjects at attack, #P < .05 relative to placebo-treatedsubjects 2 or 4 hours post treatment. PG = prostaglandins;SumaRT/Nap = sumatriptan succinate and naproxen sodiumsingle tablet combination.

Headache 847

were elevated after SumaRT/Nap at the 4-hour timepoint. Interestingly, PGE2 levels were significantlybelow attack levels in response to SumaRT/Nap at 2hours but not significantly different from placebo-treated controls at 4 hours.We found that PGI2 levelsdid not increase in saliva in placebo or SumaRT/Nap-treated subjects at either time point. Data from ourstudy provide the first evidence, to our knowledge,that levels of several PG decrease with SumaRT/Naptreatment, a finding that correlates with pain-freeoutcome.

Our results with respect to levels of PGE2, TXA2during MMaD, PGI2 are similar to previouslyreported findings for these PGs during a migraineattack when compared with interictal time periods.We found that while salivary levels of PGD2 andPGF2 reached peak amounts by 2 hours after theonset of the attack, TXA2 and PGE2 levels reachedpeak amounts by 4 hours. In agreement, other inves-tigators have reported elevated levels of PGE2 duringmigraine attacks. For example, Nattero et al7 demon-strated that plasma concentrations of PGE2 wereincreased during attacks of pure menstrual migrainewhen compared with other days of the menstrualcycle. Similarly, in a study by Sarchielli et al,23 theplasma levels of PGE2 were reported to increaseduring a migraine attack and reached peak levels by2-4 hours after onset. In addition, Vardi et al24 dem-onstrated that salivary PGE2 was increased during amigraine attack when compared with an interictaltime period. Furthermore, Tuca et al25 showed thatsalivary levels of PGE2 and thromboxane B2 (eg,metabolite of TXA2) were increased during7 amigraine attack compared with an interictal timeperiod.We also found elevated levels of TXB2 duringan attack, a finding in contrast to the study by Natteroet al7 in which TXB2 levels were not increased duringattack. Our finding that 6-keto-PGF1a (eg, metaboliteof PGI2) levels were not elevated during an attack arein agreement with results from a previous study.7

Although data from our study have shown thatthe levels of several PGs are elevated during attack,we can only speculate on the role of the PGs inMMaD pathology at this time. PGs are foundthroughout the nervous system and are known to playan important role in homeostatic as well as pathologic

states.26 It is known that PGs are generated duringperimenstrual time periods as a result of hormonewithdrawal. Declining serum levels of progesteroneup-regulate cyclooxygenase-2 enzymes within epithe-lial cells of the uterus.27 The subsequent synthesis ofPGs by the uterus and their release into the circula-tion could theoretically trigger and/or modulateattacks of MMaD as well as symptoms of dysmenor-rhea. It is also possible that declining progesteronelevels might up-regulate PG synthesis within thetrigeminovascular system and cause sensitization ofsensory trigeminal afferents and, thus, lower theirthreshold of activation. In support of this notion,increased levels of PGs have been found to modulateneurotransmission within sensory neurons of thetrigeminovascular system.Direct application of PGD2and PGE2 to adult rat trigeminal neurons wasreported to cause increased release of CGRP.28 Inaddition, PGD2, which can be released from mastcells, is known to modulate nociception,29,30 and is alsoinvolved in cellcell communication and neuronalsensitization through neurohormonal function-ing.31,32 Similarly, PGE2 has been shown to enhancethe excitability state of small diameter trigeminalganglion cells.33 Importantly, PGF2 is thought to playan important role in dysmenorrhea, but is alsoinvolved in hyperalgesia and allodynia.34 Takentogether, it is unlikely that the actions of PGs on thetrigeminovascular system are mediated through theeffects of 1 PG subtype alone.Thus, the net effect ofPGs in MMaD is likely mediated by the cumulativeaction of several PGs. Furthermore, PGs, which aresecreted by endometrial tissue and immune cells, mayact to sensitize trigeminal nerves, and thus facilitate amigraine attack by lowering their threshold of activa-tion to other inflammatory molecules.

Multiple studies have detailed the advantages ofusing saliva as a readily available, noninvasive meansof detecting biomarkers for orofacial and other sys-temic disorders.16,35-37 For example, results from recentstudies have demonstrated the diagnostic potential ofusing salivary biomarkers for studying the progres-sion and treatment of a variety of diseases, includingtype II diabetes, breast cancer, and oral cancer as wellas determining human immunodeficiency virus (HIV)load and the onset of preterm labor.36,38 In addition,

848 May 2010

we have recently found that salivary levels of CGRPcorrelate with therapeutic response to triptan treat-ment.16,17 Importantly, in the study by Cady et al,17 itwas shown that CGRP levels were elevated overinterictal levels during the premonitory period as wellas the mild and moderate headache phases ofmigraine in subjects responsive to rizatriptan. In con-trast, saliva CGRP levels were not elevated overbaseline at any phase of a migraine attack in subjectswho did not respond to rizatriptan.17 Thus, successfulresponse to rizatriptan was found to correlate withreturn of CGRP levels to interictal levels. It will be ofclinical importance to determine whether changes inthe level of PGs shown to be elevated in this pilotstudy will be reliable and specific indicators ofMMaD pathology and predictive of responsivenessto SumaRT/Nap as well as other new migrainetherapies.

In conclusion, data from this substudy provideevidence to support that changes in salivary levels ofthe PGs, PGD2 and PGF2, and possibly PGE2 andTXA2, correlate with headache pain associated withMMaD attacks and may be predictive of the effec-tiveness of SumaRT/Nap to treat this type ofmigraine. Additional larger studies are required tofurther clarify the role of PGs in MMaD, as well asidentify and validate the use of saliva PG levels asdiagnostic and therapeutic indexes for MMaD andother types of headaches.

Unlabeled Use of Products Disclosure: Treximethas been approved for use in adults for the acutetreatment of migraine with or without aura.

Acknowledgments: The authors would like to

acknowledge Dr. Lisa Mannix and Dr. Merle Diamond

for their contributions to study design and data collection

for the sub-study.

STATEMENT OF AUTHORSHIP

Category 1(a) Conception and Design

Paul L. Durham; Carrie V. Vause; Roger Cady;Vincent Martin

(b) Acquisition of DataPaul L. Durham; Carrie V. Vause; Roger Cady;Vincent Martin

(c) Analysis and Interpretation of DataPaul L. Durham; Carrie V. Vause; FrederickDerosier; Susan McDonald; Roger Cady;VincentMartin

Category 2(a) Drafting the Article

Paul L. Durham; Carrie V. Vause; Roger Cady;Vincent Martin

(b) Revising It for Intellectual ContentPaul L. Durham; Carrie V. Vause; FrederickDerosier; Susan McDonald; Roger Cady;VincentMartin

Category 3(a) Final Approval of the Completed Article

Paul L. Durham; Carrie V. Vause; FrederickDerosier; Susan McDonald; Roger Cady;VincentMartin

REFERENCES

1. International Headache Society. The internationalclassification of headache disorders: 2nd edition.Cephalalgia. 2004;24(Suppl. 1):1-160.

2. Martin VT, Behbehani M. Ovarian hormones andmigraine headache: Understanding mechanismsand pathogenesisPart 2. Headache. 2006;46:365-386.

3. Martin VT, Wernke S, Mandell K, et al. Symptomsof premenstrual syndrome and their associationwith migraine headache. Headache. 2006;46:125-137.

4. Ylikorkala O, Dawood MY. New concepts in dys-menorrhea.Am J Obstet Gynecol. 1978;130:833-847.

5. Baker FC,Driver HS,Rogers GG,Paiker J,MitchellD.High nocturnal body temperatures and disturbedsleep in women with primary dysmenorrhea. Am JPhysiol. 1999;277:E1013-E1021.

6. Mannix LK. Menstrual-related pain conditions:Dysmenorrhea and migraine. J Womens Health.2008;17:879-891.

7. Nattero G,Allais G,De Lorenzo C, et al. Relevanceof prostaglandins in true menstrual migraine.Head-ache. 1989;29:233-238.

8. Al-Waili NS. Treatment of menstrual migraine withprostaglandin synthesis inhibitor mefenamic acid:Double-blind study with placebo. Eur J Med Res.2000;5:176-182.

Headache 849

9. Mohammadian P,Hummel T,Arora C,Carpenter T.Peripheral levels of inflammatry mediators inmigraineurs druing headache-free periods. Head-ache. 2001;41:867-872.

10. Milsom I, Minic M, Dawood MY, et al. Comparisonof the efficacy and safety of nonprescription doses ofnaproxen and naproxen sodium with ibuprofen,acetaminophen, and placebo in the treatment ofprimary dysmenorrhea: a pooled analysis of fivestudies. Clin Ther. 2002;24:1384-1400.

11. Martin V, Mannix L, Ballard J, Derosier F, Lener S,McDonald S. Relief of menstrual symptoms andmigraine after early treatment with a fixedsingle-tablet formulation of sumatriptan with RTtechnology and naproxen sodium (SumaRT/Nap).Neurology. 2008;70:A355.

12. Mannix LK, Martin VT, Cady RK, et al. Combina-tion treatment for menstrual migraine and dys-menorrhea using sumatriptan-naproxen: Two ran-domized controlled trials. Obstet Gynecol.2009;114:106-113.

13. Mohammadian P,Hummel T,Arora C,Carpenter T.Peripheral levels of inflammatory mediators inmigraineurs during headache-free periods. Head-ache. 2001;41:867-872.

14. Puig-Parellada P, Planas JM, Gimenez J, et al.Plasma and saliva levels of PGI2 and TXA2 in theheadache-free period of classical migraine patients.The effects of nicardipine. Headache. 1991;31:156-158.

15. Nicolodi M, Del Bianco E. Sensory neuropeptides(substance P, calcitonin gene-related peptide) andvasoactive intestinal polypeptide in human saliva:Their pattern in migraine and cluster headache.Cephalalgia. 1990;10:39-50.

16. Bellamy J, Cady R, Durham P. Salivary levels ofCGRP and VIP in rhinosinusitis and migrainepatients.Headache. 2006;46:24-33.

17. Cady R, Vause C, Ho T, Bigal M, Durham P.Elevated saliva calcitonin gene-related peptidelevels during acute migraine predict therapeuticresponse to rizatriptan. Headache. 2009;49:1258-1266.

18. Brandes JL,Smith T,DiamondM,AmesMH.Open-label, long-term tolerability of naratriptan for short-term prevention of menstrually related migraine.Headache. 2007;47:886-894.

19. Durham P, Russo A. Stimulation of the calcitoningene-related peptide enhancer by mitogen-activatedprotein kinases and repression by an antimigraine

drug in trigeminal ganglia neurons. J Neurosci.2003;23:807-815.

20. Allais G, Bussone G,De Lorenzo C, et al. Naproxensodium in short-term prophylaxis of pure menstrualmigraine: Pathophysiological and clinical consider-ations. Neurol Sci. 2007;28(Suppl. 2):S225-S228.

21. Evers S, Afra J, Frese A, et al. EFNS guideline onthe drug treatment of migraineRevised report ofan EFNS task force. Eur J Neurol. 2009;16:968-981.

22. Brandes JL,KudrowD,Stark SR,et al. Sumatriptan-naproxen for acute treatment of migraine: A ran-domized trial. JAMA. 2007;297:1443-1454.

23. Sarchielli P, Alberti A, Codini M, Floridi A, GallaiV. Nitric oxide metabolites, prostaglandins andtrigeminal vasoactive peptides in internal jugularvein blood during spontaneous migraine attacks.Cephalalgia. 2000;20:907-918.

24. Vardi J, Flechter S, Alguati A, Regev I, Ayalon D.ProstaglandinE2 levels in the saliva of commonmigrainous women.Headache. 1983;23:59-61.

25. Tuca JO, Planas JM, Parellada PP. Increase in PGE2and TXA2 in the saliva of common migrainepatients. Action of calcium channel blockers.Head-ache. 1989;29:498-501.

26. Simmons DL, Botting RM, Hla T. Cyclooxygenaseisozymes: The biology of prostaglandin synthesisand inhibition. Pharmacol Rev. 2004;56:387-437.

27. Critchley HO, Jones RL, Lea RG, et al. Role ofinflammatory mediators in human endometriumduring progesterone withdrawal and early preg-nancy. J Clin Endocrinol Metab. 1999;84:240-248.

28. Jenkins DW, Feniuk W, Humphrey PP. Character-ization of the prostanoid receptor types involved inmediating calcitonin gene-related peptide releasefrom cultured rat trigeminal neurones.Br J Pharma-col. 2001;134:1296-1302.

29. Abdel-Halim MS,Hamberg M, Sjoquist B, AnggardE. Identification of prostaglandin D2 as a majorprostaglandin in homogenates of rat brain. Prostag-landins. 1977;14:633-643.

30. Liang X, Wu L, Hand T, Andreasson K. Prostaglan-din D2 mediates neuronal protection via the DP1receptor. J Neurochem. 2005;92:477-486.

31. Urade Y, Hayaishi O. Prostaglandin D synthase:Structure and function.VitamHorm. 2000;58:89-120.

32. Vesin MF, Barakat-Walter I, Droz B. Preferentialsynthesis of prostaglandin D2 by neurons and pros-taglandin E2 by fibroblasts and nonneuronal cells inchick dorsal root ganglia. J Neurochem. 1991;57:167-174.

850 May 2010

33. Kadoi J, Takeda M,Matsumoto S. Prostaglandin E2potentiates the excitability of small diametertrigeminal root ganglion neurons projecting ontothe superficial layer of the cervical dorsal horn inrats. Exp Brain Res. 2007;176:227-236.

34. Jabbour HN, Sales KJ. Prostaglandin receptor sig-nalling and function in human endometrial pathol-ogy. Trends Endocrinol Metab. 2004;15:398-404.

35. Kaufman E,Lamster IB.The diagnostic applicationsof salivaA review. Crit Rev Oral Biol Med. 2002;13:197-212.

36. Lawrence HP. Salivary markers of systemic disease:Noninvasive diagnosis of disease and monitoring ofgeneral health. J Can Dent Assoc. 2002;68:170-174.

37. Li Y,Denny P,Ho CM,et al. The Oral FluidMEMS/NEMS Chip (OFMNC): Diagnostic and transla-tional applications. Adv Dent Res. 2005;18:3-5.

38. Segal A, Wong DT. Salivary diagnostics: Enhancingdisease detection and making medicine better.Eur JDent Educ. 2008;12(Suppl. 1):22-29.

SUPPORTING INFORMATIONAdditional Supporting Information may be found in the

online version of this article:

Appendix S1.Headache Basic Science ArticleChecklist.

Please note: Wiley-Blackwell is not responsible for the

content or functionality of any supporting materials sup-

plied by the authors. Any queries (other than missing

material) should be directed to the corresponding

author for the article.

Headache 851

Copyright of Headache: The Journal of Head & Face Pain is the property of Wiley-Blackwell and its contentmay not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's expresswritten permission. However, users may print, download, or email articles for individual use.