Background: The course of frontotemporal dementia (FTD) is heterogeneous and no predictors of survival are currently available. Cerebrospinal fluid (CSF) tau dosage has been demonstrated to be useful in predicting outcome over time in a number of neurological disorders.
Objective: To assess CSF tau levels in FTD and to evaluate their prognostic value.
Methods: Seventy-seven FTD patients with no mutations in known causative genes were consecutively enrolled, and CSF tau and phospho-tau levels analysed. Each patient was reassessed over time, and survival (i.e. death/bedridden and otherwise) was evaluated. The survival analysis was carried out by Cox proportional hazards regression models.
Results: Patients with high CSF tau levels (≥400 pg/ml) had shorter survival than those with low CSF tau levels [hazard ratio (HR) = 3.406; 95% CI: 1.151-10.077; Wald χ(2) = 4.902; d.f. = 1; p = 0.027]. The association between tau levels and survival probability was confirmed after adjusting for age, gender, clinical phenotype and FTD clinical dementia rating at enrolment (HR = 3.769; 95% CI: 1.143-12.433; Wald χ(2) = 4.748; d.f. = 1; p = 0.029). Neither demographic or clinical characteristics nor CSF phospho-tau levels or apolipoprotein E genotype were significantly associated with prognosis.
Conclusions: This study argues that CSF tau levels may be considered in FTD to predict patients' outcome. Establishing in vivo prognostic biomarkers is mandatory to define homogeneous groups for inclusion in future clinical trials and to monitor the effectiveness of future therapeutic approaches.
Copyright © 2013 S. Karger AG, Basel.