The effectiveness of computed tomography-guided biopsy for the diagnosis of spondylodiscitis: an analysis of variables affecting the outcome.

OBJECTIVE To analyze: (i) the effectiveness of CT-guided biopsy for the diagnosis of suspected spinal infections (spondylodiscitis); (ii) identify common causative microorganisms and assess factors that could affect the diagnostic yield. PATIENTS AND METHODS Forty-five patients undergoing CT-guided biopsy for suspected spinal infection between November 2012 and October 2014 were analyzed. The time from presentation to diagnosis, administration of antibiotics before biopsy, blood culture results, admission C-reactive protein (CRP)/white cell count, presence of fever or neurological deficits, and soft tissue collections on MRI were analyzed. Multivariable logistic regression was performed to determine variables independently associated with a positive biopsy. RESULTS Eleven (24.4%) patients had positive blood cultures. The first biopsy was positive in 19 (42.2%) patients. Thirty-eight (84.4%) patients had a single biopsy, while seven (15.5%) patients underwent repeat biopsy with a positive yield in one (14.2%) patient. Overall, causative microorganisms were identified in 26 (57.8%) cases. Admission CRP was significantly associated with isolating the causative pathogen from CT-guided biopsy (p<0.001). A soft tissue collection on MRI was associated with identification of a microorganism in blood cultures (p=0.001). CRP was the only independent variable associated with a positive yield on CT-guided biopsy (p=0.007, OR 1.042) and was more likely in patients with CRP>50 (p<0.001). Administration of empirical antibiotics before biopsy did not affect the yield (p=0.572). CONCLUSIONS A high CRP was a strong predictor of isolation of the causative organism. Repeat CT-guided biopsy was found to have limited value with a low positive yield (14.2%) in our study.