Abstract / Summary
Background: The optimal minimally invasive strategy for contiguous two-level central lumbar stenosis is uncertain. We evaluated two-incision unilateral biportal endoscopy (UBE) ipsilateral hemilaminectomy using finite-element (FE) and clinical analyses. Methods: A single-subject L2–S1 FE model simulated L3–4/L4–5 ipsilateral hemilaminectomy (M1), two-level laminotomy (M2), and an idealized posterior-element resection reference (M3) under normal-bone and osteoporotic conditions. The clinical component was a single-center retrospective cohort of 36 patients with contiguous two-level central stenosis involving L2–3/L3–4, L3–4/L4–5, or L4–5/L5–S1 (19 hemilaminectomy, 17 laminotomy), treated nonrandomly and followed for a mean 14.7 months. Perioperative, patient-reported, and radiographic outcomes were assessed; longitudinal Oswestry Disability Index (ODI) and visual analog scale (VAS) outcomes were analyzed with mixed-effects models. Results: M1 and M2 produced smaller immediate range of motion (ROM) and stress perturbations than M3. Baseline imbalance was substantial for BMI, spinal epidural lipomatosis severity, and corticosteroid exposure (absolute standardized differences 0.89, 1.33, and 0.70). Hemilaminectomy was associated with shorter incisions and operative time and lower estimated blood loss and drainage. Group-by-time interactions were absent for low-back pain and leg-pain VAS but present for ODI (P < 0.001); the largest between-group difference in ODI change was 3.54 percentage points (95% CI, 1.67–5.41). Composite MCID (>12-percentage-point ODI improvement plus >3-point improvements in both VAS domains) was achieved by 14/19 (73.7%) after hemilaminectomy and 13/17 (76.5%) after laminotomy. No clear short-term radiographic differences were detected. Conclusions: The FE findings indicate smaller immediate mechanical perturbations with the UBE constructs than with M3. In this selected cohort treated by an experienced UBE surgeon, two-incision hemilaminectomy showed short-term feasibility but did not establish comparative effectiveness, noninferiority, or long-term stability. Prospective validation is required.