A Regression Meta-Analysis for Myelopathy in Re-Irradiation of Non-Spinal Cord Tumours

Radiation-induced myelopathy is a severe complication following spinal reirradiation. This dose-response meta-analysis quantified the relationship between cumulative biologically effective dose (BED) and myelopathy incidence following photon- and proton-based reirradiation for non-spinal cord tumours.

Following PRISMA guidelines, we screened 11,220 studies through May 2024. Included studies reported myelopathy outcomes following photon or proton reirradiation where the spinal cord was within the radiation field. Studies on primary spinal cord tumours and brachytherapy were excluded. Dose-response meta-analysis evaluated the relationship between cumulative BED (α/β = 2) and myelopathy incidence using generalized linear mixed models (GLMM) RESULTS: Forty-three studies encompassing 2527 patients were included. Among photon reirradiation cases (40 studies, 2444 patients), 30 myelopathy events occurred, yielding a baseline incidence of 0.009% (95% CI: 0.001-8.58%). Patients developing myelopathy received significantly higher mean cumulative BED (140 Gy₂ versus 100 Gy₂, p < 0.001) and experienced substantially shorter median inter-treatment intervals (8.5 versus 23.5 months, p < 0.001). Meta-regression demonstrated a significant dose-response relationship (p = 0.039), with each 10 Gy₂ increase in cumulative BED associated with 9.3% absolute increase in myelopathy risk (95% CI: 0.5-18.1%). Each additional month between reirradiation courses reduced myelopathy odds by 16% (OR=0.84, 95% CI: 0.77-0.91, p < 0.0001; 27 studies, n = 1850). Three proton reirradiation studies (83 patients, 3.3%) with mean cumulative BED of 170.2 Gy₂ reported zero myelopathy events.

Radiation-induced myelopathy following spinal reirradiation demonstrates significant dose-dependent risk, increasing by 9.3% per 10 Gy₂, with each additional month of inter-treatment interval reducing odds by 16%. These findings support stringent dose constraints and adequate treatment spacing in spinal reirradiation planning. Proton reirradiation showed no myelopathy events despite higher cumulative doses, warranting further prospective evaluation. PROSPERO (registration number: CRD42024600408).