Neuroinflammation in Semantic Variant Primary Progressive Aphasia

Introduction: Semantic variant primary progressive aphasia (svPPA) has a characteristic clinical phenotype defined by cognitive and imaging findings. Anomia and single word comprehension impairment define the initial stages of the aphasic syndrome, which does not affect speech production or the ability to repeat. Despite epidemiological and genetic data linking svPPA to inflammation, the topography of neuroinflammation in svPPA is unclear. svPPA pathology starts at the tip of the left temporal lobe, where a strong signal appears with the tau positron emission tomography (PET) ligand [¹⁸F]flortaucipir. However, svPPA is not typically associated with aggregates of tau, but TDP-43 protein instead. To test the hypothesis that inflammation contributes to the [¹⁸F]flortaucipir signal in svPPA, we characterized the topography of inflammation in svPPA using the PET ligand for translocator protein (TSPO), [¹¹C]PBR28. Methods: Eight amyloid-PET-negative svPPA patients (age 66 ± 8, 5 women) confirmed with [¹¹C]PiB or [¹⁸F]florbetapir underwent a 90-min dynamic [¹¹C]PBR28 and a static [¹⁸F]flortaucipir PET scan. Healthy controls underwent [¹¹C]PBR28 PET (n=12, age 69 ± 8, 6 women) or [¹⁸F]flortaucipir PET (n=12, age 67 ± 7, 6 women). [¹¹C]PBR28 binding was calculated as V_T by using a metabolite-corrected arterial input function. [¹⁸F]flortaucipir uptake was calculated as SUVR referenced to cerebellum. In one patient, to confirm that increased uptake of [¹⁸F]flortaucipir was not binding to monoamine oxidase B (MAO-B), four oral doses of selegiline (total dose: 20 mg) were administered 40, 28, 18, and 6 hours before repeating [¹⁸F]flortaucipir scanning. [¹¹C]PBR28 and [¹⁸F]flortaucipir scans were analyzed based on Hammers N30R83 atlas and also SPM. The rs6971 genotype was used as a covariate to analyze [¹¹C]PBR28 scans. Postmortem samples from additional two svPPA patients who didn’t have the PET cans were studied by H&E staining, and immunohistochemistry for phospho-TDP43, phospho-tau, and the microglial marker Iba1.

Results: [¹¹C]PBR28 uptake was mostly cortical, but [¹⁸F]flortaucipir uptake was greatest in the white matter. The uptake of both ligands was significantly increased in the left temporal lobe and in the right temporal pole, as well as in regions adjoining left temporal pole. However, within the left temporal lobe, the distribution of these ligands was markedly different. Peak uptake of [¹⁸F]flortaucipir localized to the left temporal pole, the epicenter of pathology, whereas the peak of [¹¹C]PBR28 localized to a more posterior, mid-temporal region, and the left insula in the periphery of damage core (Fig.). The uptake of the two radioligands was not correlated. MAO-B blockade did not change the [¹⁸F]flortaucipir signal. Postmortem studies showed neuronal loss and gliosis in the left temporal pole. TDP-43-positive neurites were abundant in the cortex, but phospho-tau was negative, with the exception of a single microscopic focus in one patient. Neither tau nor TDP-43 pathology was present in the white matter.

Conclusions: Neuroinflammation is greatest in the areas of progression of the pathologic process in svPPA, i.e., mid-temporal region, and left insula but not at the epicenter. [¹⁸F]flortaucipir uptake in white matter is unlikely to be due to increased TDP-43. Neuroinflammation should be further studied as a possible target for immunotherapies to slow disease progression. PET imaging of neuroinflammation is expected to be a critical method to monitor the efficacy of therapies.

• Download figure
• Open in new tab
• Download powerpoint