Validation of a Cellular Imaging‐Based Method as a Potential Biomarker for SPG4 Hereditary Spastic Paraplegia

BACKGROUND: Hereditary Spastic Paraplegia (HSP) comprises a group of rare genetic diseases characterized by length-dependent axonal degeneration of the corticospinal tracts and dorsal columns, whose main clinical feature is spastic gait. Pathogenic variants in the SPG4 gene cause Spastic Paraplegia Type 4 (SPG4-HSP), the most common form of HSP. SPG4/SPAST encodes spastin, a protein involved in microtubule regulation and lipid droplet behavior. SPG4-HSP shows extreme heterogeneity in both clinical manifestations and genetics. Although no cure is currently available, several strategies aimed at restoring spastin levels are emerging; however, SPG4-HSP still lacks accessible cellular readouts for clinical studies. This study evaluates a cell-imaging approach that quantifies the distance between nucleus and cell centroid (dcnc) in peripheral blood mononuclear cells (PBMCs) in a genetically and clinically heterogeneous SPG4-HSP cohort. METHODS: PBMCs from 48 molecularly confirmed SPG4-HSP patients and 21 healthy controls (HC) were analyzed by automated cell imaging. Dcnc and additional cytoskeletal and lipid droplet-related parameters were measured. Patient-level discrimination was assessed via cross-validated classification; correlations with molecular and clinical features were explored. RESULTS: At the patient level, dcnc distinguishes SPG4-HSP from HC, independently of mutation type and disease severity, supporting robust cross-validated classification and inverse correlation with spastin protein levels. CONCLUSION: Dcnc is a robust, non-invasive, and mutation-independent cellular candidate biomarker for SPG4-HSP. It also offers preliminary evidence consistent with spastin target engagement supporting evaluation in future clinical trials.