Latozinemab, a novel progranulin-elevating therapy for frontotemporal dementia

Abstract

Background: Heterozygous loss-of-function mutations in the progranulin (PGRN) gene (GRN) cause a reduction in PGRN and lead to the development of frontotemporal dementia (FTD-GRN). PGRN is a secreted lysosomal chaperone, immune regulator, and neuronal survival factor that is shuttled to the lysosome through multiple receptors, including sortilin. Here, we report the characterization of latozinemab, a human monoclonal antibody that decreases the levels of sortilin, which is expressed on myeloid and neuronal cells and shuttles PGRN to the lysosome for degradation, and blocks its interaction with PGRN.

Methods: In vitro characterization studies were first performed to assess the mechanism of action of latozinemab. After the in vitro studies, a series of in vivo studies were performed to assess the efficacy of a mouse-cross reactive anti-sortilin antibody and the pharmacokinetics, pharmacodynamics, and safety of latozinemab in nonhuman primates and humans.

Results: In a mouse model of FTD-GRN, the rodent cross-reactive anti-sortilin antibody, S15JG, decreased total sortilin levels in white blood cell (WBC) lysates, restored PGRN to normal levels in plasma, and rescued a behavioral deficit. In cynomolgus monkeys, latozinemab decreased sortilin levels in WBCs and concomitantly increased plasma and cerebrospinal fluid (CSF) PGRN by 2- to threefold. Finally, in a first-in-human phase 1 clinical trial, a single infusion of latozinemab caused a reduction in WBC sortilin, tripled plasma PGRN and doubled CSF PGRN in healthy volunteers, and restored PGRN to physiological levels in asymptomatic GRN mutation carriers.

Conclusions: These findings support the development of latozinemab for the treatment of FTD-GRN and other neurodegenerative diseases where elevation of PGRN may be beneficial. Trial registration ClinicalTrials.gov, NCT03636204. Registered on 17 August 2018, https://clinicaltrials.gov/ct2/show/NCT03636204 .

Keywords: Frontotemporal dementia; Latozinemab; Neuroprotection; Pharmacodynamics; Pharmacokinetics; Phase 1 clinical trial; Progranulin; Safety; Sortilin.

Fig. 1

In vitro characterization of latozinemab. A Binding curve of sortilin-expressing HEK293T cells by latozinemab. n = 4 technical replicates for latozinemab, n = 5 technical replicates for isotype control. B Percentage of dose-dependent blocking of human PGRN binding to sortilin by latozinemab. n = 3 technical replicates for latozinemab, n = 2 technical replicates for isotype control. C Dose-dependent decrease in surface sortilin expression on U251 cells. n = 3 technical replicates for latozinemab, n = 2 technical replicates for isotype control. D Extracellular PGRN is increased relative to baseline (dashed line) in the presence of latozinemab. n = 3 technical replicates for latozinemab, n = 2 technical replicates for isotype control. All data represent the mean ± SD. 2–5 technical replicates shown for each experiment, each experiment was repeated N > 3, representative data are shown. PE, phycoerythrin; PGRN, progranulin

Fig. 2

In vivo assessment of an anti-sortilin antibody in a mouse model of GRN haploinsufficiency. A Cell lysate sortilin levels and B plasma PGRN levels from WT and Grn^+/− mice that received weekly i.p. injections of S15JG or isotype control for 4.5 weeks. C PGRN levels measured from ISF 33–48 h post injection demonstrated a significant main effect of antibody treatment (P = 0.0089) by Mann–Whitney U tests. (A-C) N = 4–6 per group. All data represent the mean ± SEM. P values are by planned post hoc Tukey’s tests and are adjusted for multiple comparisons. D Schematic of social dominance test. E–H Ratio of the number of matches won by each animal in a given matchup. Male and female pairs were tested at 19–21 months of age (n = 9–12 per group). P values are by Mann–Whitney U tests. All data represent the mean ± SEM. Anti-sort, anti-sortilin antibody S15JG; Ctl, control; ISF, interstitial fluid; PGRN, progranulin

Fig. 3

Latozinemab decreases sortilin levels in WBCs and increases PGRN levels in plasma and CSF of cynomolgus monkeys. A–B Plasma A and CSF B concentrations of latozinemab antibody as a function of time. C Sortilin concentrations measured as a function of time in WBCs. D–E Plasma D and CSF E PGRN levels measured as a function of time. Male monkeys, 24–50 months of age, were used (n = 3 per group). All data represent the mean ± SEM. Dashed lines in C-E indicate baseline. CSF, cerebrospinal fluid; PGRN, progranulin

Fig. 4

Pharmacokinetics of latozinemab in HVs and aFTD-GRN participants. A–B Mean ± SD serum concentrations of latozinemab plotted as a function of time for HVs in the SAD groups A and aFTD-GRN participants who received the 60-mg/kg dose B. For the HV SAD groups A, n = 4–13 participants/group/timepoint, and for the aFTD-GRN group B, n = 8 at all timepoints, except for days 15 (n = 7 at 4 h post dose), 85 (n = 4), and 141 (n = 3). C–D Mean ± SD CSF concentrations of latozinemab plotted as a function of time for HVs in the SAD groups C and aFTD-GRN participants D; note that CSF samples were not collected for HVs in the 2-mg/kg and 6-mg/kg dose groups. For the HV SAD groups C, n = 6 participants/group/timepoint, except for the 60-mg/kg dose group prior to dosing (n = 13) and on day 13 (n = 5). For the aFTD-GRN group D, n = 5 at all timepoints, except for day 25 (n = 1). Nominal time after dose is relative to the end of infusion. Placebo-treated subjects were excluded from the PK population. aFTD-GRN, asymptomatic carrier of GRN mutations causative of frontotemporal dementia; CSF, cerebrospinal fluid; HV, healthy volunteer

Fig. 5

Latozinemab decreases sortilin in WBCs and increases PGRN levels in the plasma and CSF of HVs and aFTD-GRN participants. A–B Median percentage change from baseline in sortilin in WBCs from HVs who were administered a SAD of latozinemab A and from aFTD-GRN participants who received a single injection of 60-mg/kg latozinemab B. For the HV SAD groups A, n = 4–13 participants/group/timepoint and for the aFTD-GRN group B, n = 3–5 participants/timepont. C–D Median percentage change from baseline in PGRN levels in plasma from HVs C and from the aFTD-GRN group D. For the HV SAD groups C, n = 4–13 participants/group/timepoint and for the aFTD-GRN group D, n = 5 at all timepoints, except on days 2 (30 h post dose), 57, 85, and 113 (n = 3) E–F Median percent change from baseline in PGRN levels in CSF from HVs E and from the aFTD-GRN group F; note that CSF samples were not collected for HVs in the 2-mg/kg and 6-mg/kg dose groups. For the HV SAD groups E, n = 2–13 participants/group/timepoint and for the aFTD-GRN group F, n = 5 at baseline and day 13 and n = 1 at day 25. Nominal timepoints are relative to the end of infusion. The percent change from baseline was calculated for each individual based on their own predose value. Dashed lines indicate baseline. aFTD-GRN, asymptomatic carriers of GRN mutations causative of frontotemporal dementia; HV, healthy volunteer; PGRN, progranulin