Radiographic Predictors of Lumbar Disc Re-Herniation Requiring Repeat Discectomy or Fusion - A Matched Retrospective Cohort Analysis.

Abstract

Lumbar discectomy is the standard of care for symptomatic lumbar disc herniation (LDH). However, recurrent LDH (rLDH) necessitates reoperation in 3-18% of patients, posing a substantial challenge for both patient counseling and healthcare economics. While studies have focused on demographic and clinical risk factors, the contribution of preoperative radiographic parameters to recurrence risk remains poorly defined. This study evaluated preoperative MRIs for patients undergoing one-level lumbar discectomy to determine combined clinical and radiographic factors associated with operative rLDH following a single-level discectomy. This retrospective cohort study included adult patients who underwent a primary, single-level lumbar microdiscectomy at a tertiary academic center. Adult patients undergoing index one-level lumbar discectomy were included. Patients were excluded for revision surgery, fusion, non-discal pathology, or unavailable preoperative MRI. Patients were considered to have operative rLDH if they underwent repeat discectomy or discectomy and fusion at the index level within 3 years of their primary surgery. Potential variables included demographics, comorbidities, and radiographic parameters from preoperative MRI, including herniation morphology, disc degeneration (modified Pfirrmann grade), endplate changes (Modic), and facet arthropathy. A multivariable Cox proportional hazards model was used on a matched cohort to identify independent variables associated with time to reoperation. Receiver operating characteristics curves and were developed based on mixed clinical and radiographic models. 2608 patients underwent single level lumbar discectomy between 2013 - 2024, of whom 129 (4.9%) required reoperation for rLDH. The matched cohort included 250 patients, of whom 129 (51.6%) required reoperation for rLDH within 3 years. After multivariable adjustment, several factors were independently associated with recurrence risk. Significant radiographic factors included the presence of Modic changes (HR, 1.86; 95% CI, 1.23-2.82; P = .003) and a higher facet degeneration index (HR, 1.43 per grade; 95% CI, 1.30-1.57; P < .001). Significant clinical variables included a higher Charlson Comorbidity Index (HR, 1.21 per point; 95% CI, 1.08-1.36; P = .002), younger age (HR, 0.98 per year; 95% CI, 0.96-1.00, P = .04), and male sex (HR, 1.58; 95% CI, 1.09-2.30; P = .02). Radiographic factors increased the AUC compared to clinical factors alone, particularly beyond 1 year. Higher systemic comorbidity burden, degenerative endplate changes, and worsened facet arthropathy are significantly associated with risk of operative rLDH and should be considered for patient counseling and surgical planning.