Assessing utility of imaging based biomarker to differentiate pathologically classified frontotemporal lobar degeneration syndromes

Abstract

Background Frontotemporal lobar degeneration (FTLD) comprises distinct neuropathological entities that present as several overlapping clinical syndromes. Two main pathological inclusions are identified in FTLD – TDP-43, and tau itself further subdivided into 4 repeat (e.g., PSP-tau and CBD-tau) groups. Clinicopathological correlation is poor, at ∼60%. In vivo sensitive and specific biomarkers are needed to improve pathological diagnosis for clinical trials.

The midbrain-pons ratio (MP) is an imaging biomarker that has been shown to differentiate PSP-tau from Parkinson’s disease and multisystems atrophy, as well as clinically defined progressive supranulclear palsy from other FTLD syndromes.

We examined whether the MP at presentation could distinguish PSP-tau from controls and Alzheimer’s disease (AD), as well as from other FTLD pathologies.

Method Participants in a clinical cohort-brain bank program based at Brain and Mind Centre, Sydney, Australia, with FTLD pathologies were included, as were matched healthy Controls and clinically diagnosed AD patients. The midbrain and pons were manually traced from the sagittal MRI slice at the midline at the initial clinical assessment of all the participants for comparison.

Result In total, 65 participants were included in the study – 42 with FTLD (13 PSP-tau, 6 CBD-tau, 23 TDP-43), 11 AD and 12 control participants; no patients with Pick’s disease were available for inclusion. Only 6/13 PSP-tau patients, 1/6 CBD-tau patients and 18/23 individuals with TDP-43 were clinically diagnosed with PSP, CBS, and FTD respectively. The MP differed significantly in the PSP-tau group compared to the control, AD, and other FTLD groups, including the CBD-tau group. Using relevant cut off scores, positive predictive values of 90-100% were achieved in distinguishing PSP-tau from controls, AD, and other FTLD pathologies.

Conclusion The MP is a useful in vivo biomarker to identify individuals with PSP-tau. It can distinguish PSP-tau from controls, AD, and other FTLD pathologies with high PPVs, even CBD-tau, another important tauopathy often causing a similar clinical presentation. The PPVs achieved in using the MP were similar to those achieved using CSF and PET biomarkers of amyloid to predict AD pathology. The MP should be considered for routine use in clinical practice and selection into trials of anti-PSP-tau therapies.