Performance of Computed Tomography-Guided Spine Biopsy for the Diagnosis of Malignancy and Infection

Computed tomography (CT)-guided percutaneous biopsies are used to guide treatment in vertebral osteomyelitis and spinal malignancy, but the efficacy of this study remains unclear.

To investigate the performance of CT-guided spinal biopsy, and factors that may influence its success.

Retrospective study of all consecutive patients who underwent a CT-guided spine biopsy at a UK teaching hospital between April 2012 and February 2019. Biopsies were performed by 3 consultant neuroradiologists for a lesion suggestive of either malignancy or infection. Data collection included patient factors, biopsy factors, further investigations required, and diagnosis. Data were analyzed using contingency tables, analysis of variance, unpaired t-test, chi-squared test, and Fisher's exact test.

A total of 124 percutaneous biopsies were performed on 109 patients with a mean follow-up of 34.5 mo (range 4-86 mo) and a mean age of 66 yr (range 27-93). Approximately 32.3% (n = 40) of the biopsies investigated possible infection, and 67.7% investigated malignancy. The sensitivity for infected cases was 37.0%, and for malignancy 72.7%. The diagnostic accuracy was 57.5% and 78.6%, respectively. Complication rate was 1.6%. In our study, neither needle gauge, anatomic level of the biopsy, or bone quality significantly affected the rate of positive biopsy.

Both in our study and in the wider literature, CT-guided biopsy has a vastly superior sensitivity for malignancy compared with suspected infection. These procedures may be painful, poorly tolerated, and are not entirely risk free. As such we advocate judicious use of this modality particularly in cases of suspected infection.