Study design and patient selection
In this retrospective study, we evaluated the clinical and radiographic outcomes of minimally invasive surgical debridement in patients with thoracolumbar pyogenic spondylodiscitis. The study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (B-2203-742-104) and conducted in accordance with the Declaration of Helsinki. The ethics committees of Seoul National University Bundang Hospital have waived the requirement to obtain informed consent due to the retrospective study design.
Between March 2017 and December 2021, we identified 128 patients who underwent surgery for thoracolumbar spinal infections. The inclusion criteria for this study were as follows: patients diagnosed with thoracolumbar spondylodiscitis, patients who underwent microscopic debridement and disc drainage procedures, patients aged 20 years or older, and patients who underwent follow-up for at least 24 months postoperatively. The exclusion criteria were as follows: (1) surgical site infections, (2) tuberculosis spondylitis, (3) intradural abscess, (4) infections that directly spread from internal organs, (5) undergoing initial stabilization surgery using a pedicle screw because of severe instability and bone destruction with significant neurological deficit, and (6) not undergoing discectomy for an epidural abscess.
We reviewed the medical records of 128 patients, of whom 88 were excluded because of surgical site infections (n = 48), tuberculosis spondylitis (n = 11), intradural abscess (n = 1), aortic graft-related infection (n = 1), initial fusion surgery (n = 7), and epidural abscess (n = 20). Therefore, only patients with spondylodiscitis who underwent microscopic debridement and disc drainage were included, resulting in a final cohort of 40 patients.
Data collection and outcome measures
Patient characteristics included age, sex, body mass index, Charlson Comorbidity Index, smoking status, alcohol intake, American Society of Anesthesiologists score, infected spinal segments, extent of infection, and cause of infection. The initial laboratory findings showed the white blood cell count (103/μL) and high-sensitivity C-reactive protein levels in mg/dL.
Surgical outcomes included operative time (min), intraoperative blood loss (mL), transfusion requirements (packs), postoperative drainage volume (mL), and length of hospital stay (days). Radiographic outcomes were evaluated using plain radiography, computed tomography (CT), or magnetic resonance imaging. All preoperative and final follow-up radiographs of all patients were analyzed to confirm the spread of infection, vertebral body destruction, instability, disc space height changes, and radiographic union at the final follow-up. Disc space height was measured at the midpoint of the disc space. Radiographic bone union was assessed using CT. Union was defined as the presence of a bone bridge on CT at the final follow-up15.
Clinical outcomes were assessed using patient-reported outcome measures at several time points: pre-operative (baseline) and at 3, 6, 12, and 24 months after surgery. Clinical assessments included the visual analog scale (VAS) for low back pain (VAS-LBP) and radiating pain in the lower extremities, the Oswestry Disability Index (ODI)16 for disabilities, the European Quality of Life-5 Dimensions (EQ-5D) value for quality of life (QOL)17, and painDETECT for neuropathic pain18. The VAS ranges from 0 (no pain) to 100 (severe pain). The ODI score indicates the level of disability in the daily activities of patients with low back pain. The EQ-5D values ranged from -0.066 to 1.000, with 1 indicating the highest QOL. The painDETECT questionnaire assesses neuropathic pain in the lower extremities, with scores ranging from -1 to 38; scores below 12 suggest that neuropathic pain is unlikely, while scores above 19 indicate a high likelihood of neuropathic pain.
Surgical technique and treatment protocol
The procedure was performed under general anesthesia, with the patient in a prone position on a radiolucent operating table. This technique is similar to classic unilateral partial laminectomy followed by discectomy. Initially, unilateral partial laminectomy was performed. In cases with multiple segments of epidural abscess, laminectomies were performed at the most distal and proximal segments to facilitate drainage of the epidural abscess. Following laminectomy, total discectomy was performed on the segments affected by discitis. During discectomy, care was taken to avoid injury to the facet joints. The infected disc material was meticulously removed using a pituitary rongeur and curette. To achieve thorough removal of disc material on both the ipsilateral and contralateral sides, we used angled curettes and an angled pituitary rongeur. After removing the infected tissues, the disc space was thoroughly irrigated with at least 5000 cc of saline. After irrigation, Hemovac drain lines (Zimmer Biomet, Warsaw, Indiana, USA) were inserted into all infected segments to facilitate postoperative abscess drainage. Abscesses in the paraspinal muscles were removed during the procedure. The surgical procedure was completed after confirming adequate drainage and clearance of the infected material (Fig. 1).
Initially, a unilateral partial laminectomy was conducted. Following the laminectomy, a total discectomy was performed at the segments affected by discitis. The infected disc material was meticulously removed using a pituitary rongeur and curettes. After the infected tissues were removed, the disc space was thoroughly irrigated with saline. After irrigation, hemovac drain lines were inserted at infected segments.
Postoperatively, antibiotic therapy was continued for at least 6 weeks. Antibiotics were administered until the CRP levels were normalized. If CRP levels did not normalize within 12 weeks, the therapy was switched to oral antibiotics. The initial treatment involved empirical antibiotics, which were later changed based on the culture results. The drainage line was removed when the drainage volume was less than 10 mL/day (Fig. 2).
A 78-year-old woman with severe spondylodiscitis at L3-4 level. (A) Initial plain radiographs showed L3 inferior and L4 superior endplate destruction with spondylolisthesis and instability. (B) Initial MRI showed L3-4 intradiscal, perineural, psoas muscle abscess and severely compressed neural tissue, which induced cauda equina syndrome. (C) After L3-4 microscopic debridement surgery and continuous intravenous antibiotics, she completely cured of spondylodiscitis. During follow-up, the disc space was collapsed, and completely fused at the postoperative 12-month after surgery.
Statistical analysis
All variables were summarized using descriptive statistics. Continuous variables are expressed as means and standard deviations, while categorical variables are summarized as frequencies and percentages. To analyze the results according to bone destruction, we divided patients into group 1 (without bone destruction) and group 2 (with bone destruction) according to the presence of bone destruction. To compare continuous variables between groups, the Shapiro–Wilk test was used to assess normality. An independent t-test was used to analyze normally distributed data. If the data were not normally distributed, the Mann–Whitney U test was used. For categorical variables, the chi-squared test or Fisher’s exact test was used. This determined the significance of the differences in proportions between the two groups.
We performed statistical analysis using generalized estimating equations (GEE) to analyze changes in clinical outcomes and laboratory results during the follow-up period. The GEE model was used to analyze repeated outcomes during follow-up between the two groups. The GEE model was specified with an exchangeable correlation structure and Gaussian family. The interaction term between bone destruction and time was included to evaluate the differential effects of bone destruction on clinical outcomes over time. Stata/MP 17.1 (StataCorp LLC, College Station, Texas, USA) was used for all analyses. All statistical tests were two-tailed, statistical significance was set at p < 0.05.
