The Huntington’s Disease Young Adult Study (HD-YAS) is tracking HD gene-expanded (HDGE) young adults nearly two decades prior to their predicted clinical motor diagnosis. After approximately 4.5 years of follow-up at two study timepoints, this unique cohort offers the clearest picture yet of when neurodegeneration begins in people with HD. The study shows that cerebrospinal fluid (CSF) biomarkers can robustly detect changes long before symptoms or signs appear in the clinic, which has important implications for the design of future prevention trials.
A Window into the Earliest Changes
The HD-YAS enrolled 154 participants from across the United Kingdom (UK) at a single study site, including 73 HDGE and 81 controls. Of these, 103 completed both baseline and follow-up visits over approximately 4.5 years. At baseline, 81% of the HDGE group were in HD Integrated Staging System (HD-ISS) stage 0, meaning their caudate and putamen volumes were within normal limits. Over the study period, approximately 23% moved from stage 0 to 1, currently defined by measurable striatal volume loss.
MRI findings showed faster atrophy of the caudate and putamen, along with global grey and white matter loss and subtle white matter microstructural changes. Despite these detectable brain changes in the HDGE group, cognitive and neuropsychiatric performance remained comparable to controls at a group level, which, importantly, indicates that function remains preserved at this stage of the disease course.
Spinal Fluid Biomarkers of Silent Pathology
CSF samples collected at both HD-YAS timepoints revealed three biomarkers that differed significantly between the HDGE and control groups: neurofilament light (NfL) chain , proenkephalin (PENK), and chitinase-3-like protein 1 (known as YKL-40).
Building on our earlier finding of elevated CSF NfL at baseline, we now show that CSF NfL levels in the HDGE group rose significantly faster over time than in controls, pointing to accelerating damage to brain cells at the very earliest stages of disease. Those who progressed from HD-ISS stage 0 to 1 had higher CSF NfL levels than those who did not progress. This contributes to evidence that injury to neurons occurs before the loss in striatal volume that currently defines stage 1 HD and underscores the need for sensitive markers of early disease.
Further, CSF levels of PENK, a biomarker that reflects the health of striatal medium spiny neurons, significantly declined in the HDGE group, compared to controls, and this decline was associated with a loss of caudate and putamen volume. This observation strengthens the potential to use CSF PENK as a surrogate marker of striatal health and integrity in people who are far from HD onset.
YKL-40, an astrocytic marker, also rose significantly over time in the HDGE group, compared with controls. In contrast, other inflammatory markers, including interleukin-6 (IL-6), IL-8 and the astrocytic protein glial fibrillary acidic protein (GFAP), showed no significant group differences. This pattern points to early, selective astrocytic activation rather than a general inflammatory response in the earliest stages of the disease.
Of note, mutant huntingtin protein (mHTT) was near or below the detection limit in most CSF samples, with only 38% exceeding the assay threshold. This finding reinforces the need for more sensitive CSF mHTT assays, especially if HTT-lowering therapies are to be tested in individuals at HD-ISS stage 0 or 1.
Genetic Drivers and Mechanistic Insights
Our research also showed that, over time, the CAG repeat in HTT undergoes somatic expansion during HD-ISS stages 0 and 1. We were able to measure this in blood, and to show that this lengthening of CAGs is linked to caudate and putamen atrophy, even after adjusting for age and CAG repeat length.
At baseline, somatic expansion was also strongly associated with CSF NfL and PENK concentrations, although these cross-sectional links weakened after controlling for age and CAG repeat length.
Together, these findings provide in vivo evidence that somatic expansion is already driving pathology in the earliest stages of HD and highlights somatic expansion as a compelling target for disease-modifying therapies.
Implications for Prevention Trials
Neuronal injury and striatal dysfunction can be measured decades before clinical motor diagnosis. For this reason, CSF NfL, PENK and volumetric MRI can provide practical surrogate endpoints when traditional clinical measures and function are still normal.
Our power calculations suggest that in a three-year trial with a 50% treatment effect, 104 participants would be needed if CSF NfL were an outcome, 126 for caudate atrophy and 146 for putamen atrophy — making prevention trials feasible and realistic.
The findings from HD-YAS support using CSF biomarkers alongside neuroimaging as surrogate outcome measures in prevention trial design. By capturing the biology of progression while function is preserved, these measures lay the groundwork for a practical framework to design trials aimed at delaying or preventing disease onset.
Looking Ahead
The HD-YAS cohort is now being invited back for a third study visit, approximately two years after their second, to build on and refine these findings. With well-characterized HD-ISS stage 0 and 1 cohorts and a growing understanding of robust biomarkers, the field is closer than ever to being able to design prevention trials.
Our detailed profiling of HD-ISS stages 0 and 1 could help design and enrich recruitment for prevention trials, and the ability to define fluid biomarker cut-offs and thresholds for stratifying stage 0 cohorts holds considerable potential for future work.
None of this would be possible without the commitment and generosity of our HD-YAS participants. Continued contributions from other longitudinal studies, including HDClarity, will be important to strengthen the evidence base for CSF biomarkers and move the field forward toward effective early interventions and prevention trials.