An international consortium studying genetic frontotemporal dementia has identified the specific factors that drive wide variation in how long patients survive after symptoms begin, a step researchers say could sharpen how clinical trials are designed and how families are counselled after a genetic diagnosis. The peer reviewed cohort study, published in The Lancet Neurology and drawn from the 32 site Genetic Frontotemporal dementia Initiative (GENFI), followed 278 participants carrying a causal C9orf72 expansion or a mutation in the GRN or MAPT gene. Median survival from symptom onset for the whole group was 6.94 years (95% CI 6.59 to 7.80); by gene group, median survival was 6.63 years for GRN mutation carriers and 7.04 years for those with a C9orf72 expansion, according to the paper. Of the 278 participants, 162 (58%) had died and 116 (42%) remained alive as of the June 1, 2024 censoring date.
Using Cox proportional hazards models, the researchers found that survival was independently predicted by older age at symptom onset, shorter disease duration between onset and enrolment in the study, a clinical presentation of frontotemporal dementia with amyotrophic lateral sclerosis, a first symptom in the motor or language domain rather than behavioural or other domains, and residence in southern or central Europe versus other regions, the paper reports. The team combined these factors into a composite score, termed GENFI-SI, and found that patients scoring above a threshold of 0.22 had a hazard ratio for death of 5.08 (95% CI 3.60 to 7.17) compared with those below it, meaning risk of death at any point during follow-up was reduced by roughly 80% in the lower scoring group, the authors report. The difference between the two groups was also visible on Kaplan-Meier survival curves (log rank test, p less than 0.0001).
The researchers then tested the GENFI-SI score in a separate, independent cohort to see whether it held up outside the original dataset, a step intended to guard against a score that only fits the data it was built from. That validation cohort included 167 people, 95 of whom (57%) had died and 72 (43%) of whom were alive as of its own censoring date of February 1, 2025, and comprised 85 C9orf72 expansion carriers and 74 GRN mutation carriers, according to the paper; the authors describe this external validation as supporting the score's applicability beyond the original GENFI sample. The authors write that clarifying what drives survival heterogeneity matters for the design of therapeutic trials and for evaluating whether a treatment is working, since trials that do not account for these predictors risk comparing groups with very different expected survival by chance.
The study's authors caution that survival time is only one measure of disease burden and does not capture quality of life, caregiving needs, or functional decline before death. They also note that genetic frontotemporal dementia specifically was examined here; findings from this cohort of gene mutation carriers may not generalize to the more common sporadic forms of frontotemporal dementia, which were not part of this study.