Inflammation in Migraine…or Not…: A Critical Evaluation of the Evidence

Andrew Charles MD; Sinifunanya E. Nwaobi MD, PhD; Peter Goadsby MD, PhD


Headache. 2021;61(10):1575-1578. 

In This Article

Neuroimaging—What are we Looking at?

Neuroimaging provides evidence of inflammation including edema, vascular permeability, and tissue damage. With the exception of rare cases of hemiplegic migraine or migraine with prolonged aura, however, contrast enhancement and signal changes consistent with edema or increased vascular permeability are not typically seen in migraine. Magnetic resonance imaging (MRI) and positron emission tomography (PET) studies show no increased permeability of the blood–brain barrier with migraine.[23–25] An MRI approach to identify macrophage-mediated inflammation using ultrasmall superparamagnetic iron oxide–enhanced imaging found no increased signal on the pain side versus the nonpain side during migraine attacks,[26] indicating that migraine without aura is not associated with macrophage-mediated inflammation.

Other imaging studies have investigated cellular activation indicative of an inflammatory response using PET labeling with [(11)C]PBR28, which binds to the 18 kDa translocator protein (TSPO), whose expression is increased in the setting of tissue injury or neurodegeneration. Because changes in the morphology and number of microglia and astrocytes are associated with increased expression of TSPO, increased TSPO expression has been considered a surrogate marker for "glial cell activation," and by extension, "neuroinflammation." Two recent PET/MRI studies investigated [(11)C]PBR28 binding in individuals with migraine versus controls. One reported increased binding in multiple brain regions; some correlated with migraine attack frequency.[27] The second reported persistent, interictal extra-axial binding of the TSPO ligand in meninges and calvarial bone overlying the occipital lobe in individuals with migraine with visual aura.[28] There are a number of issues regarding TSPO imaging that raise questions about the interpretation of these findings. Among these are vascular artifacts and nondisplaceable PET ligand binding that may confound interpretation,[29,30] genetic interindividual variation of PET ligand binding caused by polymorphisms in the TSPO gene,[31] and lack of specificity of TSPO levels as a marker of pathological glial cell activation. Physiological or supraphysiological activation of neurons may also cause increased neuronal expression of TSPO.[32] Finally, although the term "glial cell activation" is now commonly used to denote an inflammatory process, it is important to recognize that glial cells are activated as part of normal function of the nervous system, playing roles in neurovascular coupling, modulation of synaptic activity, and remodeling of neuronal processes. Questions surrounding TSPO and its measurement with PET, therefore warrant significant caution in the interpretation of results with this approach.[33]