The analysis of segmental lordosis restored by oblique lumbar interbody fusion and related factors: building up preoperative predicting model

Abstract

Background: Oblique lumbar interbody fusion (OLIF) procedures have the potential to increase the segmental lordosis by inserting lordotic cages, however, the amount of segmental lordosis (SL) changes can vary and is likely influenced by several factors, such as patient characteristics, radiographic parameters, and surgical techniques. The objective of this study was to analyze the impact of related factors on the amount of SL changes in OLIF procedures and to build up predictive model for SL changes.

Methods: This is a retrospective study involving prospectively enrolled patients. A total of 119 patients with 174 segments undergoing OLIF procedure were included and analyzed. The lordotic cages used in all cases had 6-degree angle. Radiographic parameters including preoperative and postoperative segmental disc angle (SDA, preSDA and postSDA), SDA changes on flexion-extension views (ΔSDA-FE), CageLocation and CageInclination were measured by two observers. Interobserver reliability of measurements were ensured by analysis of interclass correlation coefficient (ICC > 0.75). Pearson correlation coefficient analysis and multivariate linear regression were employed to identify factors related to SDA changes and to build up predictive model for SDA changes.

Results: The average change of segmental disc angle (ΔSDA, postSDA-preSDA) was 3.9° ± 4.8° (95% confidence interval [CI]: 3.1°-4.6°) with preSDA 5.3° ± 5.0°. ΔSDA was 10.8° ± 3.2° with negative preSDA (kyphotic), 5.0° ± 3.7° with preSDA ranging from 0° to 6°, and 1.0° ± 4.1° with preSDA> 6°. Correlation analysis revealed a significant negative correlation between ΔSDA and preSDA (r = - 0.713, P < 0.001), CageLocation (r = - 0.183, P = 0.016) and ΔSDA-FE (r = - 0.153, P = 0.044). In the multivariate linear regression, preSDA and CageLocation were included in the predictive model, resulting in minimal adjusted R² change (0.017) by including CageLocation. Therefore, the recommended predictive model was ΔSDA = 7.9-0.8 × preSDA with acceptable fit. (adjusted R² = 0.508, n = 174, P < 0.001).

Conclusions: The restoration of segmental lordosis through OLIF largely depends on the preoperative segmental lordosis. The predictive model, which utilized preoperative segmental lordosis, facilitates preoperative planning for corrective surgery using the OLIF procedure.

Keywords: Deformity correction; Lumbar interbody fusion; OLIF; Predictive model; Segmental lordosis.