Paediatric bone lesions: diagnostic accuracy of imaging correlation and CT-guided needle biopsy for differentiating...

To determine the sensitivity, specificity and accuracy of CT-guided needle biopsy (CT-NB) for distinguishing benign and malignant lesions in children with suspected primary bone tumours, and to assess the correlation between imaging diagnosis and final diagnosis.

Retrospective review of children who underwent CT-NB of a suspected primary bone tumour between October 2016 and October 2019. Data collected included anatomical location, imaging diagnosis, type of needle, type of biopsy sample, CT-NB diagnosis, final diagnosis and post-procedural complications. The final diagnosis was established based on surgical histology or clinical/imaging follow-up.

125 patients met the inclusion criteria (68M, 57F: mean age 11 years; range 10 months-18 years). Biopsy was performed using a 10 cm Jamshidi needle (10G n = 96; 13 G n = 8); 14G Tru-Cut needle (n = 18); 14G Temno needle (n = 3). The commonest anatomical locations were the femur (n = 40), tibia (n = 25) and humerus (n = 16), while the commonest diagnoses were osteosarcoma (n = 35), CRMO (n = 15) and LCH (n = 14). A benign tumour was correctly identified on imaging in 100% of cases, and a malignant tumour in 95.8%. Sensitivity, specificity and diagnostic accuracy of CT-NB for distinguishing malignant from benign lesions were 98%, 100 and 99%. Of 24 indeterminate biopsy results, all that had a non-aggressive radiological appearance were benign. No immediate complications were recorded.

CT-NB represents a safe and very effective tool for differentiating benign and malignant lesions in children presenting with a suspected primary bone tumour. Suspected radiological diagnosis plays a pivotal role in the management of indeterminate biopsy results.

Paediatric bone tumours pose a significant diagnostic and therapeutic challenge. The interpretation of the imaging findings is essential for the successful management of indeterminate histological results.