Huntington’s disease is an inherited, autosomal dominant neurodegenerative disorder caused by an expanded stretch of CAG trinucleotide repeats in the HTT gene. The toxic repeat expansion gradually disables and then kills vulnerable neurons, particularly in the striatum and basal ganglia, before spreading to involve the cortex. Clinically, patients experience a triad of motor symptoms — chorea and dystonia early on, bradykinesia and rigidity later — alongside cognitive decline and psychiatric or behavioral disturbances that often precede the first visible movement problems. Although CAG repeat length strongly predicts age at onset, it explains only up to about 70 percent of the variability in when symptoms begin and how quickly they worsen, leaving considerable room for genetic modifiers and other factors to shape each patient’s trajectory.

That heterogeneity is precisely why biomarkers matter. Clinical assessment is especially difficult during the premanifest phase, when subtle symptoms may evolve over decades and the transition to manifest disease is hard to define. Reliable biofluid markers could support precision medicine by helping select patients for clinical trials, enriching study cohorts, and monitoring whether experimental disease-modifying therapies actually change the underlying biology. Until now, the leading candidate has been neurofilament light chain (NfL), a well-established but non-specific marker of neuroaxonal injury that correlates strongly with disease activity in Huntington’s. Proenkephalin has recently emerged as a more striatum-specific complement. Neither, however, directly captures the immune pathways increasingly implicated in the disease process.

To search for such pathways, the team behind the new study turned to a high-sensitivity multiplex proteomic technology called proximity extension assay, or PEA. Using the Olink Explore Neurology panels, they quantified 734 proteins simultaneously in small volumes of cerebrospinal fluid, a dramatic advance over older approaches that relied on one-at-a-time ELISA tests or laborious mass spectrometry. The platform reports results as normalized protein expression values, a log2-transformed relative measure, with built-in internal controls and stringent quality thresholds. Of the 734 assays run across 136 samples, 20 failed quality control and analysis focused on 442 assays with the strongest, most reliable signals.

The study drew on the Uppsala Huntington’s disease CSF cohort, an ongoing longitudinal effort at Uppsala University Hospital, supplemented by a validation cohort recruited from Karolinska Institute in Stockholm and Sahlgrenska University Hospital in Gothenburg. In total, the analysis included 61 HTT gene expansion carriers, 54 neurologically unaffected controls, and 21 longitudinal samples from gene carriers followed with repeat lumbar punctures one to nearly eight years apart. Disease stage was classified using the Huntington’s Disease Integrated Staging System, a biological framework spanning stages 0 through 3, with neurofilament light used as a proxy for the earliest stage where imaging data were unavailable. Genetic burden was quantified with the normalized CAG-Age-Product score, where a value of 100 corresponds to predicted motor onset.

The results were striking. Sixteen proteins were nominally dysregulated in gene carriers compared with controls, but after rigorous correction for multiple testing — a threshold set at a p value of 0.00024 based on 205 independent protein groups — only two survived: NfL, the expected veteran, and TNFRSF8, the newcomer. TNFRSF8 showed the larger effect, a roughly twofold decrease that remained highly significant after adjustment for age and sex (p = 7.6 × 10⁻⁸). Remarkably, TNFRSF8 displayed the strongest association with CAG repeat length of all 442 proteins measured, with a correlation of −0.59, yet showed no association with age — an unusual profile suggesting the immune protein is tied not just to diagnosis but to the size of the genetic expansion itself.

Longitudinal data added a crucial dimension. Among 21 gene carriers with repeated samples, a linear mixed-effects model showed that TNFRSF8 declined significantly over time, dropping about 0.11 normalized protein expression units per year (p = 0.017). Subgroup analysis revealed the decline was pronounced in manifest patients but absent in premanifest carriers, who were on average nearly 18 years from predicted onset. In contrast, NfL levels were essentially flat over the same intervals — consistent with its known pattern of early elevation followed by plateauing in later disease. This suggests TNFRSF8 may be a more sensitive tracker of ongoing progression in advanced stages, continuously declining where NfL has already exhausted its dynamic range.

The link to clinical status held up under scrutiny. In the discovery cohort, TNFRSF8 correlated with the composite Unified HD Rating Scale, an integrated measure of motor, cognitive, and functional impairment, and remained nominally significant even after adjustment for age, sex, and CAG repeat length. In the independent validation cohort, where full composite scores were not available, reduced TNFRSF8 still correlated strongly with Total Functional Capacity (Spearman rho = 0.65, p = 0.006), and the reduction versus controls persisted after adjusting for age, sex, and even differences between collection sites. Two recent studies had hinted at the connection — one small proteomic analysis flagged TNFRSF8 as a top candidate, and another combining MRI and proteomics proposed that striatal atrophy triggers secondary immune dysregulation indexed by falling sCD30 — but the new work is the first to confirm the biomarker with controls, comprehensive clinical ratings, multiplicity correction, longitudinal sampling, and predefined validation.

What does the falling protein mean biologically? TNFRSF8, or CD30, is a cell surface receptor expressed mainly on activated B and T lymphocytes, and its soluble form is shed from the membrane by proteolytic cleavage. Binding of CD30 by its ligand activates the NF-κB signaling pathway, which has been linked to neurodegeneration in Huntington’s disease. Elevated soluble CD30 is classically seen in lymphomas and systemic inflammation, and in multiple sclerosis it rises during relapses and remission — so a decline below normal levels is unusual and poorly understood. The authors speculate that reduced TNFRSF8 may reflect failing immune-regulatory signaling, a breakdown in the brain’s ability to keep neuroinflammation in check as mutant huntingtin drives progressive immune dysregulation through both direct and indirect mechanisms.

Other dysregulated proteins point to complementary pathways. MFGE8, a glycoprotein involved in efferocytosis and microglial phagocytosis, and GPR101, an orphan G protein-coupled receptor, were both reduced in gene carriers, while SFRP1, which modulates astrocyte-to-microglia crosstalk in neuroinflammation, and WASHC3, a component of the endosomal sorting machinery, rose with symptom severity — hinting at converging mechanisms of chronic inflammation, cellular stress, and endosomal-lysosomal dysfunction. The study has limitations: the PEA panel is restricted to preselected proteins, results are relative rather than absolute concentrations, MRI data were unavailable, and the sample size is moderate. Even so, the convergence of discovery and validation cohorts, longitudinal confirmation, and clinical correlation makes a compelling case. If larger studies and cell and animal models map the mechanism, the TNFRSF8 axis could become both a progression biomarker for trials and a genuine drug target for the neuroimmune dysfunction at the heart of Huntington’s disease.

Subject of Research: Cerebrospinal fluid proteomic biomarkers of Huntington's disease progression, focusing on the immune protein TNFRSF8 (sCD30)

Article Title: Decreased cerebrospinal fluid TNFRSF8 (sCD30) confirmed as a biomarker of Huntington’s disease progression

Article References: Grétarsdóttir, H. M., Cunningham, J. L., Rasmusson, A., Burman, J., Kultima, K., Paucar, M., Svenningsson, P., Constantinescu, R., & Niemelä, V. (2026). Decreased cerebrospinal fluid TNFRSF8 (sCD30) confirmed as a biomarker of Huntington’s disease progression. Journal of Neurology, 273(10), Article 593. https://doi.org/10.1007/s00415-026-14112-5

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14112-5

Keywords: Huntington's disease, TNFRSF8, sCD30, biomarkers, cerebrospinal fluid, proteomics, neuroinflammation, neurofilament light chain, HTT gene, proximity extension assay, disease progression, NF-kB signaling