Brain Metabolic Changes in the Presymptomatic Stage of Frontotemporal Dementia Associated with GRN Mutations
Dario Saracino1, Leila Sellami1, Hugo Boniface2, Maxime Locatelli2, Mélanie Pelegrini-Issac3, Aurélie Funkiewiez4, Marion Houot1, Daisy Rinaldi1, Karim Dorgham5, Florence Pasquier6, Mathieu Chastan7, Anne Hitzel8, Jérémie Pariente9, Amandine Pallardy10, Eric Guedj11, Mira Didic11, Aurelie Kas12, Marie-Odile Habert13, Isabelle Le Ber1
1Paris Brain Institute / Sorbonne Université / Hôpital Pitié-Salpêtrière, 2CATI Neuroimaging / Hôpital Pitié-Salpêtrière, 3LIB / Sorbonne Université / Hôpital Pitié-Salpêtrière, 4Institute of Memory and Alzheimer Disease / Hôpital Pitié-Salpêtrière, 5CIMI-Paris / Hôpital Pitié-Salpêtrière, 6Hopital Roger Salengro / CHU Lille, 7CHU Charles Nicolle, 8CHU Toulouse, 9CHU Purpan, 10CHU Nantes, 11APHM Hôpital Timone, 12Département de Médecine Nucléaire / Hôpital Pitié-Salpêtrière, 13CATI Neuroimaging / LIB / Département de Médecine Nucléaire / Hôpital Pitié-Salpêtrière

To investigate the longitudinal changes of multimodal biomarkers during the presymptomatic phase of GRN-associated frontotemporal dementia (FTD), and in particular to assess the usefulness of FDG-PET to identify early regional metabolic alterations.

The presymptomatic phase of genetic FTD offers an interesting perspective to study pathophysiological changes and test disease-modifying therapies. Previous studies aimed at tracking preclinical changes in biochemical or imaging biomarkers. Glucose metabolism may be a sensitive marker of early brain changes, and its alterations may significantly precede structural changes. However, there is a lack of longitudinal investigations assessing brain metabolic changes in genetic FTD, and in particular in GRN-associated forms.

This prospective study analyzed 58 asymptomatic first-degree relatives of GRN patients recruited in the Predict-PGRN study cohort (NCT04014673). Twenty-seven were mutation carriers and 31 non-carriers. They were longitudinally evaluated over a 5-year period with cognitive/behavioral assessments, plasma samples, brain MRI and FDG-PET imaging. PET data were analyzed with three approaches: voxel-wise comparisons, metabolic percent annual changes maps (PET-PAC) and regions of interest (ROIs) method.


Demographic and cognitive characteristics were comparable between GRN carriers and non-carriers at baseline. The mean age at inclusion for carriers was 42 years, approximately 17 years before expected onset. Carriers displayed a cluster of significant hypometabolism in the left middle temporal gyrus compared to non-carriers, in absence of cortical atrophy. Some ROIs, namely right middle and superior temporal gyri, left inferior parietal, and bilateral precuneus, exhibited greater annualized metabolic decline in carriers (up to ~20%) compared to non-carriers during follow-up. These changes were paralleled by higher plasma neurofilaments increase in some of the carriers.


Brain metabolic changes are useful to monitor the earliest phases of GRN disease, long before the occurrence of structural damage. Longitudinal modifications in selected areas may become a valuable biomarker to assess the efficacy of therapeutic trials in presymptomatic carriers.