Introduction  It is unclear what factors are associated with failed indirect decompression with OLIF surgery. We aim to investigate the clinical and radiological parameters associated with revision posterior decompression surgery.

Methods  From February 2015 to December 2019, 103 consecutive patients (206 levels) who underwent OLIF with or without posterior instrumentation were studied. Of these, 11 patients (18 levels) were excluded due to missing postoperative MRI scans. 92 patients (aged 69.4±8.5 years; 188 segments) were included for analysis. Radiographic variables measured pre- and postoperatively included disc height, segmental lordosis, foraminal height and area, diameter of the bulging disc, spinal canal diameter and area, dural sac diameter and area, left and right axial subarticular diameter, left and right ligamentum flavum (LF) thickness, and LF area. Cages with different heights (10mm, 12mm, or 14 mm) and different positions (anterior or posterior) were compared for radiographic results. Visual Analogue Scale (VAS), revision surgery and persistent neurological symptoms were recorded. Indirect decompression failure (IDF) was defined as revision surgeries within 6 months and persistent compressive symptoms 6 months after the surgery.

Results  Taller cages were associated with more shrinkage of the bulging disc (10-mm cages: -0.1±1.1 mm; 12-mm cages: -0.2±1.1 mm; 14-mm cages: -0.9±1.3 mm; p=0.020) and better dural sac diameter increase (10-mm cages: 1.4±1.5 mm; 12-mm cages: 1.6±1.7 mm; 14-mm cages: 2.7±2.1 mm; p=0.011). Anteriorly placed fusion cages are associated with better segmental lordosis correction (anterior cage: 1.2±4.3 degrees; posterior cage: -0.1±3.9 degrees; p=0.045), while surgical levels treated with cages placed in posterior positions showed larger postoperative left (anterior cage: 2.3±1.3 mm; posterior cage: 2.7±1.3 mm; p=0.048) and right subarticular diameter (anterior cage: 2.3±1.4 mm; posterior cage: 2.8±1.4 mm; p=0.013). 12 patients (16 levels) had IDF, among which 15 levels underwent reoperation for posterior decompression. Multivariate logistic regression showed that after adjusting for age, sex, and BMI, small preoperative dural sac cross-sectional area (CSA) (OR 0.967 [95% CI 0.933 to 0.995]; p = 0.037) and anteriorly positioned cage (OR 0.293 [95% CI 0.088 to 0.886]; p = 0.034) were independent risk factors for IDF. For preoperative dural sac CSA, ROC curve showed that the most appropriate threshold for preoperative dural sac CSA was 44.1 mm2 (sensitivity 91.3%, specificity 81.3%).After adjusting anteroposterior diameter of the lower endplate of each surgical level into a 0-10 scale, ROC curve showed that the most appropriate threshold for cage center position was 4.913 (sensitivity 62.5%, specificity 80.8%).

Discussion  In OLIF surgery, taller cages provide more stretch to the disc space and ligamentum flavum while posteriorly placed cages provide more symptom relief in patients with subarticular stenosis. To maximize symptom relief and to avoid reoperation, surgeons should aim to place the center of the cage at the posterior half of the disc space when performing OLIF. Surgical levels with a preoperative dural sac CSA < 44 mm² may not be suitable for indirect decompression.