Quantitative Relationship Between Tominersen Concentrations in Cerebrospinal Fluid and Biomarker Changes in Huntington's Disease Patients

Trial design and analysis

Pharmacometrics

CNS

Genetic

Br J Clin Pharmacol. 2025 Nov 25. Online ahead of print.

Yamamoto Y, Silber Baumann HE, Björnsson MA, Grimsey P, Hawellek DJ, Anderson K, Leavitt BR, Landwehrmeyer BG, Tabrizi S, McColgan P, Wild EJ, Sanwald Ducray P.

This publication detail the quantitative relationship between tominersen and cerebrospinal fluid biomarkers in Huntington's disease. As a progressive neurodegenerative disease with no current treatments, it presents a significant challenge to drug developers. Tominersen, an antisense oligonucleotide, aims to slow the progression of Huntington’s disease by lowering levels of huntingtin gene products, including mutant huntingtin protein.

The central role of this pharmacometric work was to provide a data-driven rationale for dose optimization following earlier safety signals observed with higher doses in clinical trials. To achieve this goal, a non-linear mixed effect PKPD modeling and Exposure-Response analysis on clinical data was used, yielding crucial insights for dose optimization:

The modeling confirmed the quantitative relationship between tominersen concentration in CSF and mutant huntingtin reduction.
Exposure-Response analysis provided a data-driven rationale for investigating lower exposures. This approach demonstrated that lower exposures were associated with a favorable biomarker profile, avoiding elevations in neuroinjury and inflammation markers (like NfL) seen at higher doses.

This work demonstrates that Model-Informed Drug Development is an essential tool for decision-making in complex neurological landscapes. Using quantitative frameworks to interpret complex signals, this collaboration successfully identified the optimal dose range for the ongoing GENERATION HD2 study.

We are honored to support Roche in this important research described in this publication co-authored by our colleague Marcus Björnsson, in collaboration with Yumi Yamamoto, Hanna E Silber Baumann, David J Hawellek, Patricia Sanwald Ducray, Paul Grimsey and Peter McColgan from Roche, Karen Anderson from Georgetown University School of Medicine, Blair R Leavitt from University of British Columbia, Bernhard G Landwehrmeyer from Ulm University, and Sarah Tabrizi and Edward J Wild from Queen Square Institute of Neurology, University College London.