Magnetic resonance imaging‑based preoperative planning of pedicle screw length in traumatic lumbar spine injuries: a prospective study in a resource‑limited setting

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Background: Accurate preoperative determination of pedicle screw length is essential for safe and effective spinal stabilisation in traumatic lumbar injuries. In resource‑limited settings where computed tomography (CT) may be unavailable, magnetic resonance imaging (MRI) offers a radiation‑free alternative. This prospective study evaluated the utility of MRI for predicting pedicle screw length, vertebral body purchase, and screw accuracy in traumatic lumbar spine injuries. Methods: A prospective study was conducted at National Orthopaedic Hospital Dala, Kano, from January 2025 to December 2025. Two hundred pedicle screws (6.5 mm diameter; lengths 40 mm, 45 mm, and 50 mm) were inserted in patients with traumatic lumbar fractures (L1-L5). Preoperative MRI measured the pedicle‑vertebral body length (PVBL). Screw length was selected to achieve ≥75% vertebral body purchase. Postoperative CT assessed screw accuracy using the Gertzbein‑Robbins classification, with inter‑observer agreement calculated. Results: Mean PVBL was 44.8±3.6 mm (range 38-54 mm), with significant variation across lumbar levels: L1 42.4±3.1 mm, L2 43.8±3.4 mm, L3 45.2±3.5 mm, L4 46.8±3.7 mm, and L5 45.6±3.9 mm. MRI‑predicted PVBL showed strong correlation with actual screw length (Pearson’s r=0.89, p<0.001). Vertebral body purchase ≥75% was achieved in 172 screws (86.0%). Screw accuracy (Gertzbein‑Robbins grades A/B) was 92.0% (184/200), with inter‑observer agreement of κ=0.87 (95% CI: 0.82-0.92). The 50 mm screws achieved the highest purchase (78.4±4.2%) but had a slightly higher breach rate (6.0%) compared with 45 mm screws (4.5%). Conclusions: MRI‑based preoperative planning effectively predicts pedicle screw length and achieves ≥75% vertebral body purchase in 86% of traumatic lumbar injuries. The strong correlation between MRI‑predicted PVBL and actual screw length supports MRI as a viable alternative to CT in resource‑limited settings, with high screw accuracy confirmed on postoperative CT.