Plasma Neurofilament Light Chain Is Elevated in Adaptor Protein Complex 4-Related Hereditary Spastic Paraplegia

Abstract

Background: Adaptor protein complex 4-associated hereditary spastic paraplegia (AP-4-HSP) is caused by pathogenic biallelic variants in AP4B1, AP4M1, AP4E1, and AP4S1.

Objective: The aim was to explore blood markers of neuroaxonal damage in AP-4-HSP.

Methods: Plasma neurofilament light chain (pNfL) and glial fibrillary acidic protein (GFAP) levels were measured in samples from patients and age- and sex-matched controls (NfL: n = 46 vs. n = 46; GFAP: n = 14 vs. n = 21) using single-molecule array assays. Patients' phenotypes were systematically assessed using the AP-4-HSP natural history study questionnaires, the Spastic Paraplegia Rating Scale, and the SPATAX disability score.

Results: pNfL levels increased in AP-4-HSP patients, allowing differentiation from controls (Mann-Whitney U test: P = 3.0e-10; area under the curve = 0.87 with a 95% confidence interval of 0.80-0.94). Phenotypic cluster analyses revealed a subgroup of individuals with severe generalized-onset seizures and developmental stagnation, who showed differentially higher pNfL levels (Mann-Whitney U test between two identified clusters: P = 2.5e-6). Plasma GFAP levels were unchanged in patients with AP-4-HSP.

Conclusions: pNfL is a potential disease marker in AP-4-HSP and can help differentiate between phenotypic subgroups. © 2023 International Parkinson and Movement Disorder Society.

Keywords: SPG47; SPG50; SPG51; SPG52; adaptor protein complex-4; biomarker; hereditary spastic paraplegia; neurofilament light; phenotypic clustering; plasma.

FIG. 1.

Evaluation of plasma NfL (neurofilament light chain) and GFAP (glial fibrillary acidic protein) levels as disease markers in AP-4-HSP (adaptor protein complex 4-associated hereditary spastic paraplegia). (A) Comparison of pNfL levels in AP-4-HSP patients and age- and sex-matched controls. (B) Receiver operating characteristic (ROC) analysis for pNfL. ROC curve for the pooled pNfL levels of AP-4-HSP patients versus matched controls. (C) Comparison of pGFAP (plasma glial fibrillary acidic protein) levels in AP-4-HSP patients and age- and sex-matched controls. (D) Associations with dichotomous clinical findings for pNfL levels in patients using multivariate regression models adjusted for age and sex. Forest plot with respective regression coefficients ($\beta$), $P$-values, and percentage increase in pNfL levels in the case of symptom presence (using log₁₀-backtransformed regression coefficients). (E) Upset plot annotated with respective pNfL levels showing the permutation of all possible inclusive intersections for patients presenting with a history of generalized-onset seizures, status epilepticus, and never having achieved unsupported walking. (F) Volcano plot showing the odds ratios and respective $P$-values for clinical findings when comparing patient clusters 1 and 2. $P$-values were adjusted for multiple hypothesis testing using the Benjamini–Hochberg procedure. Dot sizes represent relative frequencies of the respective clinical findings within the patient cluster. (G) Comparison of pNfL levels in patients assigned to clusters 1 and 2.