The loss-of-function mechanism in progranulin (PGRN) mutation carriers makes PGRN an interesting target for upregulation as a therapeutic approach in neurodegenerative diseases like frontotemporal lobar degeneration. This gives rise to several questions: (1) how stable are PGRN levels in blood and cerebrospinal fluid (CSF) in follow-up? (2) Is it necessary to measure PGRN levels in CSF to monitor a therapeutic effect? Therefore, concentrations of PGRN were measured in paired CSF and serum samples of 22 patients with behavioural variant frontotemporal dementia, including one GRN mutation carrier (c.349+1G>C), 16 patients with amyotrophic lateral sclerosis and 17 non-neurodegenerative patients, which included 22 follow-up levels. PGRN levels of 14 patients with isolated dysfunction of the blood-CSF barrier were measured and PGRN was correlated with albumin quotients as a marker for blood-CSF barrier function. The intrathecal fraction of PGRN was calculated on the basis of CSF-to-serum ratios and hydrodynamic properties. Follow-up measurements of CSF and serum PGRN levels did not show any significant change in diagnostic groups. Mean PGRN levels are 35 times higher in blood than in CSF. However, the CSF-to-serum PGRN ratio does not correlate with the albumin quotient even in patients with severe impairment of the blood-CSF barrier. The calculated intrathecal fraction of CSF PGRN levels ranged between 80 and 90 %. Assuming that CSF PGRN is either brain-derived or transported from the vascular compartment via receptor mediated mechanisms, we propose that monitoring CNS specific effects of PGRN modulating drugs should be done in CSF.
