Modelling the cascade of biomarker changes in GRN-related frontotemporal dementia

Abstract

Objective: Progranulin-related frontotemporal dementia (FTD-GRN) is a fast progressive disease. Modelling the cascade of multimodal biomarker changes aids in understanding the aetiology of this disease and enables monitoring of individual mutation carriers. In this cross-sectional study, we estimated the temporal cascade of biomarker changes for FTD-GRN, in a data-driven way.

Methods: We included 56 presymptomatic and 35 symptomatic GRN mutation carriers, and 35 healthy non-carriers. Selected biomarkers were neurofilament light chain (NfL), grey matter volume, white matter microstructure and cognitive domains. We used discriminative event-based modelling to infer the cascade of biomarker changes in FTD-GRN and estimated individual disease severity through cross-validation. We derived the biomarker cascades in non-fluent variant primary progressive aphasia (nfvPPA) and behavioural variant FTD (bvFTD) to understand the differences between these phenotypes.

Results: Language functioning and NfL were the earliest abnormal biomarkers in FTD-GRN. White matter tracts were affected before grey matter volume, and the left hemisphere degenerated before the right. Based on individual disease severities, presymptomatic carriers could be delineated from symptomatic carriers with a sensitivity of 100% and specificity of 96.1%. The estimated disease severity strongly correlated with functional severity in nfvPPA, but not in bvFTD. In addition, the biomarker cascade in bvFTD showed more uncertainty than nfvPPA.

Conclusion: Degeneration of axons and language deficits are indicated to be the earliest biomarkers in FTD-GRN, with bvFTD being more heterogeneous in disease progression than nfvPPA. Our data-driven model could help identify presymptomatic GRN mutation carriers at risk of conversion to the clinical stage.

Figure 1

Cascade of biomarker changes in FTD-GRN along with the uncertainty associated with it. (A) Non-imaging biomarkers. (B) Multimodal biomarkers with Siamese GMM. (C) Multimodal biomarkers without Siamese GMM. The biomarkers are ordered based on the position in the estimated cascade. The colour map is based on the number of times a biomarker is at a position in 100 repetitions of bootstrapping. FTD-GRN, progranulin-related frontotemporal dementia; GMM, Gaussian mixture modelling.

Figure 2

Gaussian mixture modelling (GMM) distributions. The histogram bins are divided in three colours, where the green part shows the proportion of non-carriers, the yellow part shows the proportion of presymptomatic carriers and the red part shows the proportion of symptomatic carriers. The Gaussians shown here are the ones that were estimated using GMM, where the green Gaussian is the normal one estimated using non-carriers and the red Gaussian is the abnormal one estimated using the carriers. The amplitudes of these Gaussians are based on the estimated mixing parameter. The grey curve shows the total estimated distribution, which is the summation of green and red Gaussians.

Figure 3

Frequency of occurrence of subjects with different disease severities, estimated using cross-validation. (A) Results using non-imaging biomarkers in discriminative event-based modelling (DEBM). (B) Results using multimodal biomarkers in DEBM.

Figure 4

Correlation of disease severity (as estimated by multimodal DEBM using cross-validation) with years since onset and FTD-CDR-SOB. The 2D scatter plots in subfigures A and C show the correlations of disease severity with years since onset, for symptomatic nfvPPA and bvFTD subjects, respectively. The 2D scatter plot in subfigures B and D show the correlations of disease severity with FTD-CDR-SOB. The plot on top of each subfigure shows the probability density function of the disease stages. The plots on the right of subfigures A and C show the probability density functions of years since symptom onset. The plots on the right of subfigures B and D show the probability density function of FTD-CDR-SOB. 2D, two-dimensional; bvFTD, behavioural variant frontotemporal dementia; DEBM, discriminative event-based modelling; FTD-CDR-SOB, Frontotemporal Lobar Degeneration Clinical Dementia Rating Scale Sum of Boxes; nfvPPA, non-fluent variant primary progressive aphasia.

Figure 5

Cascade of multimodal biomarker changes in nfvPPA (A) and bvFTD (B) subjects along with the uncertainty associated with it. The biomarkers are ordered based on the position in the estimated cascade. The colour map is based on the number of times a biomarker is at a position in 100 repetitions of bootstrapping. bvFTD, behavioural variant frontotemporal dementia; nfvPPA, non-fluent variant primary progressive aphasia.