Factors Impacting the Yield of Image-Guided Biopsy in Native Vertebral Osteomyelitis: A 10-Year Retrospective Study

Background Image-guided biopsies in patients with suspected native vertebral osteomyelitis (NVO) are recommended to establish the microbiological diagnosis and guide antibiotic therapy. Despite recent advances, the microbiological yield of this procedure remains between 48% and 52%. A better understanding of factors associated with this low yield may lead to improved microbiological diagnosis. Methods We retrospectively identified patients with suspected NVO undergoing image-guided biopsies from January 2011 to June 2021 at our institution. Two hundred nine patients undergoing 248 percutaneous biopsies were included. Demographic data, biopsy and microbiologic techniques, clinical characteristics, and antibiotic use were collected. Multivariable logistic regression analysis was conducted to determine factors associated with microbiological yield. Results A total of 110 of 209 (52.6%) initial image-guided biopsies revealed positive microbiological results. This number increased to 121 of 209 (57.9%) when repeat image-guided biopsies were included. In multivariable analysis, aspiration of fluid was associated with a 3-fold increased odds of yielding a positive result (odds ratio [OR], 3.13; 95% confidence interval [CI], 1.39–7.04; P = .006), whereas prior antibiotic use was associated with a 3-fold decreased yield (OR, 0.32; 95% CI, .16–.65; P = .002). A univariate subgroup analysis revealed a significant association between the length of the antibiotic-free period and microbiological yield, with the lowest rates of pathogen detection at 0–3 days and higher rates as duration increased (P = .017). Conclusions Prior antibiotic use in patients with suspected NVO was associated with a decrease in the microbiological yield of image-guided biopsies. An antibiotic-free period of at least 4 days is suggested to maximize yield. Successful fluid aspiration during the procedure also increases microbiological yield.