Epidural Fibrosis Raises Thresholds but Preserves Recruitment in Conventional Spinal Cord Stimulation: Lateral...

Epidural fibrosis (scar) around chronically implanted spinal cord stimulation (SCS) leads contributes to the late loss of efficacy known as the tolerance phenomenon, yet how maturing scar reshapes the intraspinal electric field, and which programming paradigms resist it, remain poorly understood. We combined a three-dimensional volume-conductor model of the thoracic spinal cord with multicompartment myelinated axon models and performed population-level uncertainty quantification across a virtual cohort (N = 200) varying in anatomy and scar properties. Four configurations-monopolar, longitudinal guarded tripole, transverse lateral-balance, and a guard-dominant steered tripole-were evaluated for threshold shift, spatial fidelity of activation, and recoverability. For monopolar and guarded tripole, the most common clinical programs, maturing scar behaved as a near-pure scalar attenuator, preserving the spatial activation pattern (median correlation > 0.99) so that amplitude re-titration restored the original modelled recruitment. Transverse lateral steering was the sole exception: failure was governed primarily by lead mediolateral offset. Near-centred leads were fragile (correlation 0.57; threshold rise +74%), whereas large offsets approached a less scar-sensitive quasi-monopolar drive (correlation 0.95; +7%). A guard-dominant configuration with limited lateral bias retained robustness even under asymmetric scar (correlation > 0.995). For most SCS programs, therefore, scar detunes amplitude rather than spatial target; lateral steering is intrinsically fragile under fibrosis, its reliability dictated by lead centring; and guard-dominant programming outperforms pure lateral balance when steering is required.