Longitudinal whole-spine CT mapping of ossification of spinal ligaments: Distribution, progression, and blood biomarkers.

Abstract

Diffuse idiopathic skeletal hyperostosis (DISH) and ossification of the posterior longitudinal ligament (OPLL) are prototypical ectopic spinal ossification disorders, but objective methods for early activity detection and prognostication remain limited. We evaluated ossification trajectories, blood biomarkers, circulating microRNAs, and genetic susceptibility for progression stratification. We analyzed 694 health-screening participants who underwent CT at baseline and follow-up (≥5 years); DISH was scored level-by-level and OPLL was assessed across all levels, using age- and sex-matched controls. Baseline plasma small-RNA sequencing in 40 DISH participants with extreme progression phenotypes (top versus bottom 20 by score change) identified candidate miRNAs via nested cross-validation, repeating feature selection within each fold; genetic plausibility was assessed by two-sample Mendelian randomization (MR) using GTEx miRNA eQTLs and UK Biobank DISH GWAS. DISH showed thoracic-predominant progression; OPLL clustered in the cervical spine, predominantly at C4-C6. OPLL had substantially higher BMI (SMD 0.85); its only FDR-surviving biomarker difference, lower HDL, lost significance after BMI adjustment (adiposity-confounded). CRP/hs-CRP (DISH) and HbA1c (OPLL) were nominally higher but did not survive FDR adjustment. A four-miRNA plasma panel discriminated DISH progression status (nested AUC 0.953), unimproved by adding age, sex, and BMI (AUC 0.897). MR gave nominal genetic evidence linking the miR-15a locus (a single shared -3p/-5p signal) to the DISH flow score, unlike the progression-panel miRNAs. These findings distinguish DISH and OPLL ossification trajectories and biomarker profiles, yield a discovery-stage miRNA panel, and provide genetic evidence for miR-15a in susceptibility; independent cohort validation and orthogonal assay confirmation are required for clinical translation.