[18F]AV-1451 binding in vivo mirrors the expected distribution of TDP-43 pathology in the semantic variant of primary progressive aphasia

Abstract

Introduction: Semantic dementia, including the semantic variant of primary progressive aphasia (svPPA), is strongly associated with TAR-DNA binding protein 43 (TDP-43) type C pathology. It provides a useful model in which to test the specificity of in vivo binding of the putative tau ligand [¹⁸F]AV-1451, which is elevated in frontotemporal lobar degeneration tauopathies.

Methods and results: Seven patients (five with svPPA and two with 'right' semantic dementia) and 12 healthy controls underwent positron emission tomography brain imaging with [¹⁸F]AV-1451. Two independent preprocessing methods were used. For both methods, all patients had clearly elevated binding potential (BP_ND (non-displaceable binding potential)) in temporal lobes, lateralising according to their clinical syndrome and evident in raw images. Region of interest analyses confirmed that BP_ND was significantly increased in temporal regions, insula and fusiform gyrus, consistent with those areas known to be most affected in semantic dementia. Hierarchical cluster analysis, based on the distribution of [¹⁸F]AV-1451 binding potential, separated semantic dementia from controls with 86% sensitivity and 100% specificity.

Conclusions: [¹⁸F]AV-1451 binds in vivo regions that are likely to contain TDP-43 and not significant tau pathology. While this suggests a non-tau target for [¹⁸F]AV-1451, the pathological regions in semantic dementia do not normally contain significant levels of recently proposed 'off target' binding sites for [¹⁸F]AV-1451, such as neuronal monoamine oxidase or neuromelanin. Postmortem and longitudinal data will be useful to assess the utility of [¹⁸F]AV-1451 to differentiate and track different types of frontotemporal lobar degeneration.

Figure 1

Using data preprocessed by method 1. (Upper panel) Column 1: axial and sagittal views of T1-weighted images for each semantic dementia participant, column 2: raw axial and sagittal BP_ND maps for each patient, column 3: unthresholded surface-rendered regional T-maps for each subject against all controls and column 4: equivalent T-maps thresholded at q<0.05, corrected for false discovery rate. (Lower panel) T1-weighted images and BP_ND maps for a representative control and the group comparisons of all 7 patients versus 12 controls both uncorrected (column 3) and corrected for false discovery rate q<0.05 (column 4). The numbering of individual patients is consistent with figure 2 and table 1. BP_ND, non-displaceable binding potential. FDR, false discovery rate.

Figure 2

Two-dimensional scaling of non-parametric whole brain regional correlations using data preprocessed by method 1. Red dots represent patients, and green dots controls. Patient labels correspond to individuals in table 1.