Ph.D Thesis

Ph.D StudentKisler Lee-Bareket
SubjectPain Neuromatrix: Pain Perception and Modulation in Migraine
and the Effect and Prediction of Migraine
Preventive Treatment
DepartmentDepartment of Medicine
Supervisors PROF. David Yarnitsky
DR. Irit Weisman-Foegel


Migraine is a chronic pain condition, characterized by severe unilateral throbbing pain, nausea, photophobia and phonophobia. The available prophylactic treatments are often ineffective, which emphasizes the need for a mechanistic based framework to individually tailor the proper treatment. 

            One of the main mechanisms that shape our pain experience is descending analgesic pathways that act to reduce pain through serotoninergic and noradrenergic mechanisms of endogenous-analgesia (EA). EA-efficiency measured psychophysically in migraine is controversial. It has not been tested concurrent with imaging methods of functional MRI (fMRI) which could provide additional important data on EA in migraine.  The serotonin norepinephrine reuptake inhibitor (SNRI) duloxetine was shown to alleviate pain, and improve EA in chronic non-migraine pain patients. Further, duloxetine effect on pain-modulatory brain areas and EA in migraine is not known. Therefore, this study used behavioral pain assessment (quantitative sensory testing (QST)), imaging-based brain activity and anatomical measures (gray matter volume (GMV)) in healthy-controls and migraineurs to determine the following:

1) brain-activity associated with EA by fMRI in attack-free migraineurs.

2) the clinical effects of the SNRI duloxetine in migraine prevention and the use of QST to predict duloxetine efficacy, in a randomized, double-blind, placebo-controlled study.

3) duloxetine’s effects on the EA-efficiency and related GMV changes in migraine prevention.

            Both healthy-controls and migraineurs went through QST assessment of response to painful stimuli, including EA-efficiency, assessed by the conditioned pain modulation (CPM) paradigm. In the following week, anatomical and fMRI (during CPM-assessment) were performed. Thereafter, the patients underwent a treatment stage of 8 weeks of either duloxetine or placebo. Towards the end of the treatment period, patients repeated the QST and the MRI sessions. Treatment effect was examined compared to migraine pre-treatment levels. Treatment prediction by QST measures and changes in EA-efficiency and related GMV were also examined.

1) Migraineurs showed similar EA-efficiency as healthy-controls, revealed in psychophysical CPM and related brain-activity. 2) Compared to placebo, duloxetine patients reported greater estimated migraine-improvement that was predicted by increased pre-treatment sensitivity to tonic heat-pain in duloxetine, yet not in placebo. 3) Compared to placebo, the duloxetine group showed higher CPM-efficiency, at treatment completion, that was associated to increased GMV in the anterior cingulate cortex and right cerebellum. 

            One of the suggested mechanisms associated with migraine-characteristics is a dysfunctional EA. However, as revealed psychophysically and supported by fMRI, as a group, interictal migraineurs show normal EA. Nonetheless, compared to placebo, the SNRI duloxetine, which act to improve EA-function and reduce pain-sensitivity, resulted in greater subjective treatment-related improvement, that was predicted by greater pre-treatment pro-nociceptivity. Further, as suggested by the drug mechanism of action, duloxetine resulted in greater EA-efficiency (i.e. CPM), related to increased GMV in pain modulatory brain areas. Therefore, though as a group interictal migraineurs do not show impaired EA-function, the treatment part of this study supports the involvement of EA-disfunction in migraine severity. Our results might also point on the existence of subgroups in the migraine population with different EA characteristics, that should be considered when contemplating preventive treatment for the individual patient.