T2 hyperintensity extent and approach-specific outcomes in cervical OPLL

To determine whether the longitudinal extent of intramedullary T2-weighted hyperintensity modifies the adjusted comparative outcomes of anterior versus posterior decompression in cervical ossification of the posterior longitudinal ligament (OPLL) with gait disturbance. In this single-center observational cohort, consecutive patients with cervical OPLL, gait abnormality, and MRI-confirmed intramedullary T2 hyperintensity (signal change ratio [SCR] > 1.2) underwent anterior or posterior decompression between January 2020 and January 2022 and were followed for 3 years. Hyperintensity extent was prespecified as focal (≤ 2 segments) or long segment (> 2 segments). Serial MRI (SCR and T2 hyperintensity length), plantar-pressure gait assessment, and neurological evaluations were obtained at predefined intervals. The primary functional endpoint was medial foot pressure (MFP, %). Longitudinal trajectories were analyzed using mixed-effects models adjusted for baseline clinical and radiographic covariates. Because surgical approach selection was not randomized, comparative estimates were interpreted as adjusted associations rather than causal effects. Among 312 eligible patients, 190 had focal and 122 had long-segment hyperintensity; 159 underwent anterior and 153 posterior surgery. In the focal cohort, anterior surgery was associated with greater 3-year improvement in MFP, MRI signal regression, and neurological recovery than posterior surgery, with an adjusted posterior–anterior MFP contrast of − 4.50% (95% CI − 5.66 to − 3.34). In the long-segment cohort, both approaches were associated with substantial improvement, with no significant between-approach difference at 3 years (adjusted posterior–anterior contrast, − 0.90%; 95% CI − 2.33 to 0.53). The prespecified time × approach × extent interaction for MFP was significant. Overall complication rates were comparable, although dysphagia/hoarseness was more frequent after anterior surgery and C5 palsy was more frequent after posterior surgery. Among cervical OPLL patients with gait disturbance and preoperative intramedullary T2 hyperintensity, focal signal extent was associated with a larger adjusted anterior–posterior difference in gait, neurological, and MRI recovery, whereas long-segment hyperintensity appeared to attenuate the measurable between-approach difference, consistent with more diffuse or chronic cord involvement. These findings are restricted to patients with SCR > 1.2 and should be interpreted as adjusted associations within an observational cohort. T2 signal extent should be regarded as a prognostic and effect-modifying imaging feature rather than a stand-alone determinant of surgical approach.