Efficacy of CT-guided biopsies of the spine in patients with spondylitis--an analysis of 164 procedures.

OBJECTIVE

To evaluate efficacy of CT-guided spinal biopsy (CTSB) in patients with spondylitis considering patient characteristics, technical issues, antibiotic therapy, histopathological, and microbiological findings.

MATERIALS AND

METHODS

All CTSB procedures performed between 1995 and 2009 in patients with proven spondylitis were re-evaluated. Patient sex and age, antibiotic treatment, biopsy approach, number of specimens, length of needle path, laboratory results (CRP, WBC), and histopathological/microbiological findings were documented and compared to the final diagnosis of spondylitis. Statistical analysis was performed using Chi-square test and Student's t-test. The p-value was set to 5%.

RESULTS

164 CTSB procedures were performed in 159 patients (mean age 65 years, 60% men) in which spondylitis was histopathologically verified in 95%. Neither patient sex nor age, positioning, localization of the spinal lesion, bioptic approach, number of specimens, or depth of the needle showed significant impact on the rate of positive histopathological findings. A causative germ was identified in 40/127 biopsies (32%) with Staphylococcus aureus being identified in 50%. Tuberculous spondylitis was diagnosed in ten cases (6%). CRP significantly correlated with bacterial growth (13.3±12.2 mg/dl versus 8.8±7.6 mg/dl; p=.015) whereas administration of antibiotics did not show any significant impact on bacterial growth (29% versus 36% in patients without antibiotics; p=0.428). Patients with histopathological signs of active spondylitis showed a significantly higher CRP (16.5±15.8 mg/dl versus 8.9±8.0 mg/dl, p<.001). Complication rate was 0.6% (one focal bleeding).

CONCLUSION

CTSB of the spine in suspected spondylitis is an effective and safe procedure for establishing final histopathological diagnosis. However, microbiological yield is low regardless of technical issues and antibiotic therapy. Other than CRP values, laboratory investigations added little useful information to diagnose infection in our study group. CRP values significantly correlated with bacterial growth and with histopathological signs of active spondylitis.