PHILADELPHIA – according to an investigation presented at the annual meeting of the American Academy of Neurology. These biomarkers appear to reflect the earliest detectable changes in the natural history of Huntington’s disease, but the longitudinal prognostic value of changes in these biomarkers requires further investigation, the researchers said.
Huntington’s disease has a long prodromal phase and is associated with long survival. Investigators still need well-validated biomarkers of disease progression, prognosis, and pharmacodynamics to aid drug development, said Filipe B. Rodrigues, MD, clinical research fellow at University College London. After several years of study, Dr. Rodrigues and colleagues found mutant huntingtin and neurofilament light (NfL) to be the most promising potential biomarkers in Huntington’s disease. They sought to understand how these two biomarkers compare with each other, what their predictive ability is, and how they change longitudinally.
To this end, Dr. Rodrigues and colleagues designed the HD-CSF study, a prospective, observational, longitudinal cohort study with a 2-year follow-up. They recruited 20 healthy controls, 20 patients with premanifest Huntington’s disease, and 40 patients with manifest Huntington’s disease. All participants underwent regular clinical assessments and standardized collections of cerebrospinal fluid (CSF) and blood. They also had the option of undergoing brain MRI scans.
The investigators analyzed their data using multiple linear regression models, Pearson’s correlations, receiver operating characteristic curves, and sample size calculations. They used an event-based model to evaluate the temporal sequence of changes in Huntington’s disease-related biomarkers.
Dr. Rodrigues and colleagues first observed that all three biomarkers successfully distinguished between healthy controls, patients with premanifest Huntington’s disease, and patients with Huntington’s disease. Mutant huntingtin, the pathogenic agent in Huntington’s disease, discriminated perfectly between healthy controls and mutation carriers, as the researchers had expected. CSF and plasma levels of NfL also discriminated well between healthy controls and mutation carriers. These biomarkers had areas under the ROC curve greater than 0.9. NfL in plasma and CSF also distinguished well between patients with premanifest Huntington’s disease and those with manifest Huntington’s disease, with areas under the curve greater than 0.9. Their discriminative ability in this regard was significantly better than that of mutant huntingtin.
When the researchers examined the relationship between the three biomarkers, they found that CSF levels of NfL were strongly correlated in a linear fashion with plasma levels of NfL. CSF levels of mutant huntingtin were moderately associated with CSF levels of NfL.
Levels of all three biomarkers increased significantly as the disease progressed and were associated with all clinical scales and imaging measures. CSF and plasma levels of NfL had superior predictive ability for clinical and imaging measures, compared with mutant huntingtin. CSF and plasma NfL were associated with brain volume, but mutant huntingtin was not.
All three biomarkers were stable during a 6-week period. Dr. Rodrigues and colleagues calculated sample sizes for a two-arm interventional trial involving various hypothetical therapeutic effects. They found that the required sample sizes were small enough to be incorporated easily into ongoing and future clinical trials.
In silico modeling suggested among the markers measured in the HD-CSF study, the three biofluid biomarkers were the first factors to be altered in the course of Huntington’s disease. Alterations in the biomarkers were followed by changes in imaging markers, and then by changes in clinical markers (for example, motor and cognitive function).
Finally, Dr. Rodrigues and colleagues found preliminary evidence that levels of NfL in CSF and plasma increase over time at different rates in patients with Huntington’s disease, compared with healthy controls. NfL appears to be more useful than mutant huntingtin for evaluating the rate of disease progression than for gauging response to treatment, said Dr. Rodrigues. “If [we] can prove that we can assess response to treatment by measuring NfL, I think that would be great.”
The investigators are currently analyzing the longitudinal predictive value of changes in these biomarkers. They also have begun analyzing other markers such as tau and brain-derived neurotrophic factor.
This study was funded by the Medical Research Council UK, the CHDI Foundation, and F. Hoffmann-La Roche.
This article was updated 6/18/19.
SOURCE: Rodrigues FB et al. AAN 2019, Abstract S16.003.