SEMORINEMAB PHARMACOKINETICS AND THE EFFECT ON PLASMA TOTAL TAU PHARMACODYNAMICS IN CLINICAL STUDIES

 

V. Ramakrishnan¹, B. Bender¹, J. Langenhorst², M.O. Magnusson², M. Dolton³, J. Shim¹, R.N. Fuji¹, C. Monteiro¹, E. Teng¹, N. Kassir¹, J. Jin¹

 

1. Genentech, Inc., South San Francisco, California, USA; 2. Pharmetheus AB, Uppsala, Sweden; 3. Roche Products Australia Pty Ltd, Sydney, New South Wales, Australia

Corresponding Author: Vidya Ramakrishnan, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, Tel: +1 (650) 225-6154, ramakrishnan.vidya@gene.com

J Prev Alz Dis 2024;
Published online July 17, 2024, http://dx.doi.org/10.14283/jpad.2024.146

 

---

Abstract

BACKGROUND: Semorinemab is a monoclonal antibody that targets the N-terminal domain of the tau protein that is in clinical development for the treatment of Alzheimer’s disease.
OBJECTIVES: To perform model-based evaluations of the observed pharmacokinetics in serum and the total plasma tau target-engagement dynamics from clinical studies evaluating semorinemab.
DESIGN: The observed semorinemab pharmacokinetics and plasma tau target engagement from phase 1 and 2 clinical studies were modeled using a non-linear mixed effect target-mediated drug disposition model. The model was simulated to understand target engagement at clinical dose levels.
SETTINGS AND PARTICIPANTS: The clinical studies testing semorinemab were evaluated in healthy volunteers, subjects with prodromal-to-mild Alzheimer’s disease, and subjects with mild-to-moderate Alzheimer’s disease. The data included a total of 8430 semorinemab serum concentrations and 4772 total tau protein plasma concentrations from 463 subjects treated with a range of single and multiple doses of semorinemab.
MEASUREMENTS: Serum concentrations of semorinemab and the total plasma tau concentrations were measured after administration of a range of doses of semorinemab to subjects with Alzheimer’s disease. A sensitivity analysis was performed wherein key target-mediated drug disposition model parameters were estimated separately between healthy volunteers, subjects with prodromal-to-mild Alzheimer’s disease, and subjects with mild-to-moderate Alzheimer’s disease.
RESULTS: Serum concentrations of semorinemab were consistent across studies and showed a dose-proportional increase across the evaluated dose range. The pharmacokinetic profile was comparable between healthy volunteers and subjects with Alzheimer’s disease. Total plasma tau concentrations increased in a dose-dependent non-linear manner upon semorinemab administration. The target-mediated drug disposition model adequately described the serum pharmacokinetics and protein dynamics with an estimated antibody-ligand binding strength, Kss, of 42.7 nM. The estimated values of clearance and central volume of distribution were 0.109 L/day/70 kg and 2.95 L/70 kg, respectively, and were consistent with typical values for IgG mAbs. In the sensitivity analysis, Kss (32 nM) and baseline tau protein (0.30 µM) were estimated to be lower for healthy volunteers compared to subjects with Alzheimer’s disease but were comparable between subjects with Alzheimer’s disease of different severities (Kss: 52-57 nM, baseline tau: 0.44-0.47 µM). The models suggested that peripheral target engagement was over 90% at the clinical doses in each of the diagnostic subgroups.
CONCLUSION: Our target-mediated drug disposition model adequately described the serum pharmacokinetics and the peripheral non-linear increase with dose of the total tau. The model confirmed that these dose-response relationships were consistent across populations of healthy volunteers and subjects with different severities of Alzheimer’s disease.

Key words: Pharmacokinetics, Total tau protein, Target-mediated drug disposition model.

---

 

Background

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by progressive cognitive and functional decline and the presence of two key pathologies in the brain: the extracellular plaques comprised of β-amyloid (Aβ) peptide and intracellular neurofibrillary tangles comprised of hyperphosphorylated tau (1, 2). Tau protein is produced within the neurons and released in the extracellular space upon synaptic activity (3-5), with subsequent uptake by other postsynaptic neurons (6).
Semorinemab is a fully humanized monoclonal antibody with an immunoglobulin G4 (IgG4) isotype backbone. It preferentially binds to the N-terminal domain of tau (amino acid residues 6-23) and is known to bind all forms of full-length tau including phosphorylated and aggregated species of tau. Semorinemab reduces tau accumulation in a dose-dependent manner, both in in vitro cell culture and in vivo in a transgenic mouse model (7).
Semorinemab serum pharmacokinetics (PK) have been evaluated in three clinical studies so far. A phase 1 study (GN39058) enrolled healthy volunteers and subjects with mild-to-moderate AD and assessed the safety and serum PK of semorinemab administered at doses ranging from 225 mg to 16800 mg (7). Subsequently, the two phase 2 studies, GN39763 (Tauriel) and GN40040 (Lauriet) tested the efficacy and safety of semorinemab in subjects with prodromal-to-mild and mild-to-moderate AD, respectively (8, 9). Three clinical doses (1500, 4500, and 8100 mg) administered intravenously were tested in GN39763 while a single clinical dose of 4500 mg administered intravenously was tested in GN40040 (7-9). In GN39763, treatment with semorinemab failed to slow clinical disease progression compared to placebo at any of the three dose-levels. The results from GN40040 were mixed; subjects treated with semorinemab 4500 mg for 48 or 60 weeks had significantly slower cognitive decline relative to those treated with placebo as measured by the co-primary cognitive outcome measure, ADAS-Cog11; however, no treatment effects were observed on the co-primary functional outcome measure (ADCS-ADL), or on secondary cognitive (MMSE) and global (CDR-SB) outcome measures. All three studies evaluated the serum semorinemab concentrations (PK) and plasma total tau concentrations (pharmacodynamics – PD).
Monoclonal antibodies (mAbs) often exhibit target-mediated drug disposition (TMDD) kinetics upon binding to their target, thereby influencing the observed longitudinal PK and PD profiles. The observed data can be used to inform a mathematical model that estimates peripheral target engagement dynamics via parameters representing the binding affinity, target turnover, and mAb-target complex turnover. Such models allow for prediction/simulation of the longitudinal concentrations of free target species, which are difficult to directly quantify given the challenges associated with assay development. In addition, these models can enable simulations/predictions for novel dose-levels/regimens beyond those included in the clinical studies. This work reports the peripheral PK and PD observed in the three clinical studies of semorinemab. We developed a TMDD model that simultaneously describes serum semorinemab PK and plasma total tau (mid-domain) concentrations to quantify target-engagement in a clinical trial setting.

 

Methods

Study designs and subjects

Our analyses included semorinemab blood PK and PD data collected from healthy volunteers and subjects with AD enrolled in one phase 1 (GN39058) and two phase 2 [GN39763 (Tauriel) and GN40040 (Lauriet)] studies. The healthy volunteers were only included in the phase 1 (GN39058) study; subjects with AD were included in all three studies (GN39058, GN39763, and GN40040). The PK data included the serum total semorinemab concentrations and the PD data included plasma total tau concentrations. The details of the studies describing the methodology, randomization, and study conduct are available in the published literature (7-9). Supplementary Table 1 provides an overview of the semorinemab studies.
The phase 1 (GN39058) study included healthy volunteers and subjects with mild-to-moderate AD with a flat-dosing scheme in the single ascending dose (SAD) and multiple dosing (MD) cohorts. The SAD cohorts included semorinemab doses of 225, 675, 2100, 4200, 8400, 16800 mg or placebo administered intravenously. The cohorts included 8 healthy volunteers (6 semorinemab, 2 placebo), with the exception of the 225 mg cohort which included only 3 subjects (2 semorinemab and 1 placebo). The study also included a separate bioavailability assessment cohort that received 1200 mg of semorinemab administered subcutaneously. The MD cohort, including 10 healthy volunteers and 10 subjects with mild-to-moderate AD, received 8400 mg of intravenously administered semorinemab or placebo (8 semorinemab, 2 placebo) once weekly every 4 weeks for up to four doses. In the SAD cohorts, including the subcutaneous cohort, blood samples for measurement of semorinemab and tau concentrations were collected pre-dosing, immediately after-dosing, 4, 8, 12, 24, 48 hours after-dosing. Blood samples were also collected at the follow-up visits 1, 2, 4, 6, 8, 12, and 16 weeks after the single dose. In the MD cohort, blood samples were collected at pre-dosing, immediately after-dosing, 4, 8, 12, 24 (after first and fourth doses), and 48 hours after first dose. Blood samples were also collected at the follow-up visits 1, 2, 4, 6, 8, 12, and 16 weeks after the final dose.
GN39763 (Tauriel) was a phase 2 clinical trial evaluating the safety and efficacy of semorinemab in prodromal-to-mild AD subjects. It included administration of multiple intravenous doses of placebo or semorinemab at three dose levels, 1500 mg, 4500 mg, and 8100 mg. The first three doses were administered every two weeks to rapidly achieve steady-state concentrations, with subsequent doses administered every four weeks. The subjects were randomized in 2:3:2:3 ratio to the three dose-levels in ascending order and placebo respectively. Serum PK samples and blood samples for tau biomarker measurements were collected at weeks 1, 3, 5, 9, 13, 17, 33, 49, 65, and 73. Per protocol week 1 serum PK samples were taken 0‒4 hours before the start of infusion and 1 hour (+15 minutes), 2 hours (+20 minutes), and 4 hours (+30 minutes) after the end of infusion. All other serum PK samples were taken 0‒4 hours before the start of infusion and 0‒30 minutes after the end of infusion. Week 1 and 17 blood biomarker samples were taken 0‒4 hours before the start of infusion and 0‒30 minutes after the end of infusion.
GN40040 (Lauriet) was a phase 2 clinical trial evaluating the safety and efficacy of semorinemab in subjects with mild-to-moderate AD. Semorinemab was administered at a dose of 4500 mg or placebo (1:1 randomization). As with the Tauriel study, the first three doses were administered every two weeks to achieve steady-state concentrations, with subsequent doses administred every four weeks. Due to the COVID-19 pandemic, the double-blind treatment duration was extended to 60 weeks for subjects who missed at least one dose (cohort 2); all other subjects completed the double-blind treatment period in 48 weeks (cohort 1). Serum PK samples and blood samples for tau biomarker measurements were collected at weeks 1, 2, 5, 9, 13, 25, 37, 49 (for all subjects), and 61 (for cohort 2 only subjects). As per the protocol, serum PK samples were taken 0‒4 hours before the start of infusion and 0‒30 minutes after the end of infusion. The blood biomarker samples were taken 0‒4 hours before the start of infusion.

PK/PD assay

Serum PK

Total semorinemab concentrations (free and tau-bound species) were quantified using a validated ELISA-based ligand binding assay. The semorinemab quantification range was 50-2200 ng/mL with a lower limit of quantification (LLOQ) of 50 ng/mL. The serum PK assay was performed at PPD Laboratories (Richmond, VA).

Plasma tau (PD)

Total Plasma Tau concentrations (free and semorinemab-bound tau) were measured at Microcoat (Bernried am Starnberger See, Germany) using a Cobas e 411 instrument (Roche Diagnostics, Rotkreuz, Switzerland) with an Elecsys Robust Prototype Assay. The lower limit of detection of the assay was 1 pg/mL. The assay captures and detects the mid-domain of tau (amino acids 159 to 224), a region conserved between cynomolgus and human, present in all known tau isoforms, and detectable by the assay irrespective of phosphorylation state and presence of semorinemab.

PK/PD analysis

The semorinemab serum PK and plasma tau dynamics data from all subjects who received at least one dose of semorinemab and had evaluable measurements of PK and PD from the three studies were included in the analyses. The data was analyzed using nonlinear mixed effects modeling with NONMEM (version 7.5., ICON Development Solutions, Ellicott City, MD, USA). The complete dataset included a total of 8430 PK observations and 4772 PD observations from a total of 463 subjects, including 65 healthy volunteers and 398 subjects with AD. Only a negligible percentage of the PK or PD measurements were below the lower limit of quantification. The serum PK concentrations of semorinemab showed biphasic distribution kinetics and a two-compartment model was used to describe the distribution of semorinemab. A target-mediated drug disposition model with quasi-steady state approximation (10) was used to simultaneously describe the PK/PD relationship, the total semorinemab serum concentrations, and the total plasma tau concentrations (Fig. 1). A transit compartment was included to fit the absorption phase for the subcutaneous PK data. The model also included baseline subject body weight as a covariate on the pharmacokinetic parameters of clearance and volume of distribution. From the final model based on the full data, the consistency of key TMDD parameters was verified across populations as a sensitivity analysis: healthy volunteers (sub-population of GN39058) versus prodromal-to-mild AD (GN39763) versus mild-to-moderate AD (part of GN39058 and GN40040). To achieve this, TAU0 (baseline unbound tau concentration) and Kss (steady-state constant in the quasi-steady state approximation TMDD model) including uncertainty were estimated for each sub-population and compared. In addition, predicted peripheral target engagement was estimated for each dose-level and compared to evaluate the potential for clinical impact.

According to the quasi-steady approximation, the free drug, the target, and the complex can be assumed to be in a quasi-steady state, where the binding rate (kon) is balanced by the sum of the dissociation (koff) and internalization rates (KCOMP).

 

Results

Serum semorinemab PK concentrations

The serum PK profile of semorinemab was consistent with that of other therapeutic IgG molecules and showed a biphasic distribution after administration of the first dose (Fig. 2A). The serum concentrations of semorinemab were consistent across the phase 1 and phase 2 studies and showed a dose-proportional increase across the dose ranges (225–8400 mg) evaluated (Fig. 2B). The serum PK profile in healthy volunteers and subjects with AD was also comparable (Fig. 2A).

Semorinemab concentrations versus time since most recent dose stratified by dosing occasion (A) and versus time since first dose (B), based on the analysis data set. Each thin line represents the data for one subject and the thick lines are medians shown for each nominal time with at least 4 observations. Lines are colored by treatment group and line-type distinguishes subjects with Alzheimer’s disease and healthy volunteers.

 

Total plasma tau concentrations

Total plasma tau concentrations increased upon semorinemab administration, suggesting peripheral target engagement. The increase in peripheral total tau was delayed relative to peak serum PK concentrations and the maximum increase was observed around ~ 4 weeks after semorinemab administration. The increase in total tau was dose-dependent, but not dose-proportional. The additional increase was minimal above ~ 4200 mg dose (60 mg/kg, assuming a 70-kg subject) suggesting saturation of the target in the periphery. The increase in peripheral total tau concentrations was similar across the phase 1 and phase 2 studies (Fig. 3A).

Tau protein concentration (mAb-bound and free tau) profiles stratified by the treatment groups and disease status, based on the analysis data set (A) versus time since first dose, where each thin line represents the data for one subject and the thick lines are medians shown for each nominal time with at least 4 observations. (B) geometric mean of absolute concentrations and mean change from baseline (CFB), with the associated 90 % confidence intervals, versus nominal dose, stratified by sampling time in the analysis data set subset to nominal times 14 and 28 days; error bars present the 5 % and 95 % quantiles of the observed data.

 

In the phase 1 (GN39058) study it was observed that subjects with AD had a higher baseline of total tau concentrations in the plasma compared to the healthy volunteers consistent with observations of higher central nervous system (CNS) tau burden in subjects with AD compared to healthy volunteers. The data from phase 2 studies (GN39763 and GN40040) included subjects with AD and showed greater total tau concentrations in the plasma compared to the phase 1 study which mainly included healthy volunteers. The total tau concentrations in subjects with AD in phase 1 (GN39058) and phase 2 (GN39763 and GN40040) studies were consistent. The change from baseline tau concentrations across GN39058 and GN39763/GN40040 were comparable and were dose-dependent, similar to absolute concentrations (Fig. 3B). No substantial differences were observed between different AD subpopulations¬—prodromal-to-mild AD (GN39763) versus mild-to-moderate AD (part of GN39058 and GN40040).

 

PK-PD peripheral target engagement modeling

The serum PK concentrations and the plasma total tau PD concentrations were simultaneously modeled to describe the target engagement dynamics in the periphery. A target-mediated drug disposition (TMDD) model using the quasi-steady state approximation (Fig. 1) described the observed serum PK and plasma PD data from the phase 1 and phase 2 studies as seen in the prediction-corrected visual predictive checks for both PK and PD data (Fig. 4). The inclusion of a transit compartment to account for the subcutaneously dosed cohort in GN39058 improved model performance as observed by a reduction in the objective function upon model convergence (∆OFV=-321). The lower concentrations and variability of the baseline total tau protein (and less variable) in healthy volunteers compared to subjects with AD required the estimation of separate tau formation rate constant (Kin) values (∆OFV=-101) and associated inter-individual variability (∆OFV=-55). This model is described as the ‘final model’ in the rest of the text. Model diagnostic plots are available in Supplementary Figure S1.

Prediction corrected visual predictive check of semorinemab concentrations and Tau concentrations versus time since first dose for the final model, based on 200 simulations on a semi-logarithmic scale.

 

The estimated values of clearance (CL) and central volume of distribution (Vc) were 0.109 L/day/70 kg and 2.95 L/70 kg, respectively, and were consistent with typical values for IgG mAbs (11, 12). The production rate of tau in plasma in subjects with AD was estimated to be 1.04 nM/day compared to a value of 0.703 nM/day in healthy volunteers. All model parameters were estimated with good confidence and are listed in Supplementary Table 2.
In order to evaluate any potential differences related to disease status (i.e., healthy volunteers versus prodromal-to-mild AD versus mild-to-moderate AD) across the studies, a sensitivity analysis using a model estimating a separate Kss, the steady-state constant parameter that quantifies the antibody-ligand binding strength, and the derived parameter of baseline total tau concentration for each sub-population was performed. The results from this full-fixed effect model showed that the point estimate of the Kss value was lower in healthy volunteers, 32 nM, as compared to subjects with AD regardless of disease severity (prodromal-to-mild AD = 52 nM and mild-to-moderate AD = 57 nM) (Fig. 5A). The baseline total tau concentration point estimate was also lower in healthy volunteers compared to subjects with AD and was similar between the two different grades of the disease (Fig. 5A).

(A) Forest plots illustrating the parameter estimates of KSS and Tau0 for the final model and for the full-fixed effect model across different populations. Values are calculated based on 200 sampled parameter vectors from the distribution of parameter estimates of a 25-sample non-parametric bootstrap. The black solid symbols are the median point estimate, and the black solid line is the 90% confidence interval on the point estimate. (B) Forest plots illustrating the predicted peripheral target engagement for the final model and for the full-fixed effect model for different dosing regimens and populations. Values are calculated based on 200 sampled parameter vectors from the distribution of parameter estimates of a 25-sample non-parametric bootstrap. The black solid symbols are the median point estimate, and the black solid line is the 90% confidence interval on the point estimate.

 

Peripheral target engagement simulations at steady state with varying doses of semorinemab

Both versions of the TMDD model (final model and the full-fixed effect model) were used to simulate the mean target engagement in plasma at different semorinemab doses to illustrate the dose-response relationship with regards to peripheral target engagement. The results are shown in Fig. 5B. The simulations showed that the overall peripheral target engagement was high and that the separate Kss and baseline total tau values based on disease severity had a negligible influence on peripheral target engagement. In addition, the simulations also showed that peripheral target engagement was above 90% at the clinical doses of semorinemab in the clinical phase 2 studies (Tauriel and Lauriet) in subjects with AD.

 

Discussion

The work presented here included a pooled model-based analysis of a large dataset including serum PK and plasma PD data from three clinical studies of semorinemab, which included healthy volunteers and subjects with prodromal-to-mild and mild-to-moderate AD. The semorinemab PK and PK/PD as estimated through TMDD did not exhibit relevant differences across studies or subjects with differing disease severities.
The serum PK was shown to be dose-proportional across the doses evaluated (225–8400 mg once every 4 weeks (Q4W) intravenous (IV)) with the PK parameters (clearance, volume of distribution, and half-life) being similar to that of typical IgG-based antibodies. Baseline body weight was identified as a significant covariate affecting the PK parameters of clearance, central volume of distribution, and peripheral volume of distribution. The exponent estimate for the effect of body weight was such that an increase in body weight led to an increase in the PK parameters. Lower PK may introduce insufficient target attainment for low body weights, assuming the flat dosing. Supplementary Figure S2 illustrates that at the clinical dosing regimens the target attainment for both high and low body weights is expected to be similarly good. The model estimated that the bioavailability for subcutaneous administration was 67.9 %, similar to the reported bioavailability of 69 % from the non-compartmental analysis of observed data from the Phase 1 study.
The total tau (free tau + semorinemab bound tau) concentrations increased following administration of semorinemab suggesting peripheral target engagement by the antibody in the blood. Per the TMDD model, this result is a consequence of slower elimination rate of the semorinemab-tau complex as compared to the elimination rate of free tau in the periphery leading to a longer residence time of the tau species in the periphery upon binding with semorinemab. Alternatively, the increase in total tau in plasma could also be a result of efflux of the complex from the cerebrospinal fluid (CSF)/brain back into plasma, which would yield a different interpretation of the estimated degradation constant of the complex (Kcomp): Kcomp, estimated = Kcomp, true – Kefflux, csf-plasma. Of note, Kcomp, true could be, but is not necessarily, lower than kdeg. The hypothesis of complex efflux is supported by a previously identified correlation of increased plasma total tau with ISF tau upon administration of an anti-tau mAb in a mouse model (13). This increase in total tau was not dose-proportional and appeared to saturate above the ~ 4200 mg flat dose. One limitation in the current measurements of total tau species is that the immunoassay used to measure the total tau concentrations in these studies was specific to the measurement of the mid-domain region of the tau species which is distant from the N-terminal epitope to which semorinemab binds. Also, the tau species are known to undergo aggregation and/or fragmentation in biological fluids including the plasma and CSF, thus affecting the quantification of true target-engagement (14-16). However, it is believed that tau fragmentation in CSF may be more pronounced than in the plasma (15, 17). These limitations also likely contribute to the approximately 10-fold difference observed in the model-estimated Kss value versus the average dissociation equilibrium constant (Kd) value (3.8 nM) for the binding of semorinemab to human tau as measured by surface plasmon resonance technology as reported in Ayalon et al (7).
A TMDD model was developed to simultaneously describe the serum PK and total tau PD data. The biphasic distribution phase of the observed PK profile in the studies upon graphical exploration suggested that a two-compartment model was needed to describe the serum PK. Since fixed allometry cannot be assumed for a monoclonal antibody, weight exponents were estimated. The addition of a transit compartment to characterize the subcutaneous PK data from the phase 1 study improved the model performance. The observed baseline total tau concentrations in healthy volunteers were markedly lower with a higher spread compared to subjects with AD. This observation along with the biological plausibility of different treatment-naïve tau kinetics between these populations, prompted the need for separate estimation of the formation rate constant for the tau species in the model and its associated IIV. The model described the linear PK and the non-linear increase in the PD response (total tau) simultaneously with increasing doses of semorinemab. The model confirmed that the dose-response relationships in all three studies were consistent.
Our sensitivity analysis exploring whether the TMDD system differed across disease status showed that the estimates of Kss (Fig. 5A) and baseline Tau (as derived from Kin) were lower in healthy volunteers compared to subjects with AD but very similar between prodromal-to-mild and mild-to-moderate disease severities. The peripheral target engagement simulations performed using both the final model and aforementioned full-fixed effect model showed greater than 90% engagement at steady-state at the high doses of semorinemab used in the clinical phase 2 studies in subjects with AD. This confirmed that, a) based on plasma target engagement kinetics, semorinemab binds to the target as intended and b) separate Kss or Kin estimates have a negligible effect on quantified peripheral target engagement across the studies. Overall, the point estimate differences between healthy volunteers and subjects with AD were in the same fold-level and hence had negligible influence on the simulated peripheral target engagement at the different dose-levels.
The work presented in this paper is specific to peripheral target engagement kinetics and will require additional considerations/assumptions for understanding target engagement kinetics in CSF and/or brain. Although we have an understanding of semorinemab’s partitioning in the CSF, the tau burden, the tau turnover kinetics, and transport processes governing plasma-CSF-brain partitioning of the tau species remain unclear and are important for deciphering the kinetics of target engagement in the site of action, i.e., brain. The simultaneous dynamics of turnover rates of each of the species in the TMDD model and the transport between plasma-CSF-brain makes extrapolations of target engagement in brain/CSF challenging and will remain to be a limitation of this modeling exercise until further understanding emerges. As such, these data do not address the discrepancy between the absence of clinical efficacy in the Tauriel study and the mixed effects on clinical end-points observed in the Lauriet study.
Our TMDD model has several potential applications, including but not limited to: a) the simulation of peripheral target engagement with doses beyond those tested in these clinical trials and a variety of fixed and/or intermittent dosing schedules, b) the extension of the model to predict CSF/brain target engagement while accounting for partitioning of the mAb to the CSF/brain and the total tau dynamics/concentrations in the CSF/brain, and c) incorporating tau fragmentation as specific tau species and assessing binding differences in target engagement thereafter. The utility of these applications will be dependent on a number of factors, including the specificity of the tau assay to measure the binding epitope, the development of an assay to measure free tau in plasma/CSF to quantify the absolute extent of target engagement, and differences in binding kinetics in the different biological regions (e.g., blood versus CSF versus brain interstitial fluid), etc.

 

Conclusion

Semorinemab serum PK and the dynamics of target-engagement of the tau protein in the plasma upon semorinemab administration were characterized in this study with a model-based quantitative approach. Our model verified that the target-engagement was similar across the three clinical studies and confirmed similar kinetics of target-engagement in healthy volunteers and subjects with Alzheimer’s disease of varying severity. This model can be utilized to perform simulations of peripheral target dynamics upon administration of semorinemab at different doses and treatment regimens.

 

Ethics approval and consent to participate: All studies were conducted in accordance with the ethical principles of the Declaration of Helsinki and complied with Good Clinical Practice. A central investigational review board and individual site institutional review boards reviewed and provided approval for the protocols as well as informed consent forms. All subjects provided informed consent and consent for publication before start of the study.

Consent for publication: All authors of this manuscript provide consent for the publication of this manuscript and all data associated with it in The Journal of Prevention of Alzheimer’s Disease.

Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests: VR, BB, MD, JS, RF, CM, ET, NK, and JJ are employees of Genentech Inc. at the time of the study completion and own Roche stocks. JL and MM were paid consultants, employed by Pharmetheus, at the time of the study completion. MM own Pharmetheus stocks.

Funding: This study was funded by F. Hoffmann-La Roche Ltd. The study sponsor was responsible for the overall study management, drug supply, data management, statistical analysis, PK and PD analysis, and the drug safety process. The study sponsor was involved in the design of the study, data analysis and interpretation, and in writing the manuscript.

Authors’ contributions: VR, BB, MD, RF, CM, ET, JJ contributed to the conception and design of the study, and conducting the experiments. VR, BB, JL, MM, MD, NK, and JJ contributed to interpretation of data and data analysis. VR, BB, JL, MM, MD, RF, CM, ET, and JJ contributed to intellectual contribution to writing. VR, BB, JL, MM, MD, JS, RF, CM, ET, NK, and JJ contributed to final approval of the version to be submitted.

Acknowledgements: Support for third-party writing assistance was provided by Anshin Biosolutions Corp.

Trial registrations: GN39058 ClinicalTrials.gov, NCT02820896, July 1, 2016, GN39763 (Tauriel): ClinicalTrials.gov, NCT03289143, Registered September 20, 2017. GN40040 (Lauriet): ClinicalTrials.gov, NCT03828747, Registered February 4, 2019.

 

SUPPLEMENTAARY MATERIAL

 

References

1. Knopman DS, Amieva H, Petersen RC, Chetelat G, Holtzman DM, Hyman BT, et al. Alzheimer disease. Nat Rev Dis Primers. 2021;7(1):33 doi:10.1038/s41572-021-00269-y.
2. Long JM, Holtzman DM. Alzheimer Disease: An Update on Pathobiology and Treatment Strategies. Cell. 2019;179(2):312-39 doi:10.1016/j.cell.2019.09.001.
3. Gallardo G, Holtzman DM. Amyloid-beta and Tau at the Crossroads of Alzheimer’s Disease. Adv Exp Med Biol. 2019;1184:187-203 doi:10.1007/978-981-32-9358-8_16.
4. Wu JW, Hussaini SA, Bastille IM, Rodriguez GA, Mrejeru A, Rilett K, et al. Neuronal activity enhances tau propagation and tau pathology in vivo. Nat Neurosci. 2016;19(8):1085-92 doi:10.1038/nn.4328.
5. Yamada K, Holth JK, Liao F, Stewart FR, Mahan TE, Jiang H, et al. Neuronal activity regulates extracellular tau in vivo. J Exp Med. 2014;211(3):387-93 doi:10.1084/jem.20131685.
6. de Calignon A, Polydoro M, Suarez-Calvet M, William C, Adamowicz DH, Kopeikina KJ, et al. Propagation of tau pathology in a model of early Alzheimer’s disease. Neuron. 2012;73(4):685-97 doi:10.1016/j.neuron.2011.11.033.
7. Ayalon G, Lee SH, Adolfsson O, Foo-Atkins C, Atwal JK, Blendstrup M, et al. Antibody semorinemab reduces tau pathology in a transgenic mouse model and engages tau in patients with Alzheimer’s disease. Sci Transl Med. 2021;13(593) doi:10.1126/scitranslmed.abb2639.
8. Monteiro C, Toth B, Brunstein F, Bobbala A, Datta S, Ceniceros R, et al. Randomized Phase II Study of the Safety and Efficacy of Semorinemab in Participants With Mild-to-Moderate Alzheimer Disease: Lauriet. Neurology. 2023;101(14):e1391-e401 doi:10.1212/WNL.0000000000207663.
9. Teng E, Manser PT, Pickthorn K, Brunstein F, Blendstrup M, Sanabria Bohorquez S, et al. Safety and Efficacy of Semorinemab in Individuals With Prodromal to Mild Alzheimer Disease: A Randomized Clinical Trial. JAMA Neurol. 2022;79(8):758-67 doi:10.1001/jamaneurol.2022.1375.
10. Gibiansky L, Gibiansky E, Kakkar T, Ma P. Approximations of the target-mediated drug disposition model and identifiability of model parameters. J Pharmacokinet Pharmacodyn. 2008;35(5):573-91 doi:10.1007/s10928-008-9102-8.
11. Betts A, Keunecke A, van Steeg TJ, van der Graaf PH, Avery LB, Jones H, et al. Linear pharmacokinetic parameters for monoclonal antibodies are similar within a species and across different pharmacological targets: A comparison between human, cynomolgus monkey and hFcRn Tg32 transgenic mouse using a population-modeling approach. MAbs. 2018;10(5):751-64 doi:10.1080/19420862.2018.1462429.
12. Deng R, Iyer S, Theil FP, Mortensen DL, Fielder PJ, Prabhu S. Projecting human pharmacokinetics of therapeutic antibodies from nonclinical data: what have we learned? MAbs. 2011;3(1):61-6 doi:10.4161/mabs.3.1.13799.
13. Yanamandra K, Patel TK, Jiang H, Schindler S, Ulrich JD, Boxer AL, et al. Anti-tau antibody administration increases plasma tau in transgenic mice and patients with tauopathy. Sci Transl Med. 2017;9(386) doi:10.1126/scitranslmed.aal2029.
14. Cicognola C, Brinkmalm G, Wahlgren J, Portelius E, Gobom J, Cullen NC, et al. Novel tau fragments in cerebrospinal fluid: relation to tangle pathology and cognitive decline in Alzheimer’s disease. Acta Neuropathol. 2019;137(2):279-96 doi:10.1007/s00401-018-1948-2.
15. Inekci D, Jonesco DS, Kennard S, Karsdal MA, Henriksen K. The potential of pathological protein fragmentation in blood-based biomarker development for dementia – with emphasis on Alzheimer’s disease. Front Neurol. 2015;6:90 doi:10.3389/fneur.2015.00090.
16. Meredith JE, Jr., Sankaranarayanan S, Guss V, Lanzetti AJ, Berisha F, Neely RJ, et al. Characterization of novel CSF Tau and ptau biomarkers for Alzheimer’s disease. PLoS One. 2013;8(10):e76523 doi:10.1371/journal.pone.0076523.
17. Zetterberg H, Wilson D, Andreasson U, Minthon L, Blennow K, Randall J, et al. Plasma tau levels in Alzheimer’s disease. Alzheimers Res Ther. 2013;5(2):9 doi:10.1186/alzrt163.

© Serdi 2024