INTRODUCTION

The diagnostic yield of and best approaches for imaging-guided percutaneous biopsy for vertebral osteomyelitis is controversial. Early studies suggest yields of up to 90%; however, recent evidence shows lower yields of 30-40%. We aim to determine yield and predictors of yield in percutaneous CT-guided biopsies in vertebral osteomyelitis.

METHODS

We conducted a retrospective observational single-centre study cohort study of all patients presenting for vertebral biopsy or aspiration between 2014 and 2018. Only patients undergoing biopsy for suspected infection were included. Patients with malignant indications were excluded. Comprehensive review of medical records was performed for clinical presentation, comorbidities, imaging, biomarkers, microbiology and treatment.

RESULTS

Overall, 40 out of 88 biopsies were performed for suspected infection, in 36 patients. Mean age was 59 ± 18 years; 29 (81%) were male. Of the 40 samples, an organism was identified in 14 samples (35%). Gram-positive organisms were most commonly identified; Staphylococcus aureus was cultured in 7 (50%) of samples. Mean admission CRP was significantly higher in patients with identified organisms compared to those without (137 ± 106 vs 54 ± 78, P = 0.008). Aspiration was a strong independent predictor of positive microbiological growth on multivariate analysis (OR 6.52 [1.25-34.02], P = 0.026). Biopsy or aspiration aided clinical decision-making in half of cases.

CONCLUSIONS

Percutaneous CT-guided biopsy has a modest yield for identifying the culprit organism in suspected cases of vertebral osteomyelitis. Elevated CRP and aspiration of fluid collections are associated with improved microbiological yield and should be considered in deciding when and where to biopsy.