Validity of the Bone-RADS to evaluate incidental bone lesions on computed tomography or magnetic resonance

Abstract

Reporting and Data Systems (RADS) have played a pivotal role in radiology, providing uniform terminology, report organization and a structured approach to interpreting imaging findings. Recognizing the absence of standardized procedures for managing incidental solitary bone lesions detected on computed tomography (CT) and magnetic resonance imaging (MRI), the Society of Skeletal Radiology introduced the Bone-Reporting and Data System (Bone-RADS) guidelines in 2022. These guidelines offer four algorithms that allow the user to arrive at four diagnostic management recommendations for incidentally encountered solitary bone lesions on CT and MRI. Evaluating the accuracy of Bone-RADS is the initial step towards its integration into clinical practice. OBJECTIVE: The main objective is to evaluate the diagnostic accuracy of the Bone-RADS guidelines, in separating between non-aggressive (benign) from aggressive (malignant) bone lesions incidentally encountered on either CT or MRI in adults compared with histopathological diagnoses or a one-year follow up considered as gold standard. STUDY DESIGN AND PARTICIPANTS: This study adopts an observational, descriptive, and prospective design to evaluate diagnostic accuracy. A total of 462 adult subjects with incidentally identified solitary bone lesions on CT or MRI will be included using consecutive sampling, and the recruitment period is set for 2 years. The sample will be derived from Hospital Universitari Doctor Josep Trueta and Hospital Santa Caterina from Girona, Spain. METHODS: Three experienced musculoskeletal radiologists, blinded and previously trained, from Institut de Diagnòstic per la lmatge (IDI) will independently use the Bone-RADS guidelines to categorize the incidental solitary bone lesions identified on CT or MRI into four management options (Bone-RADS1 – 4). These results will be compared to the histopathological diagnosis or one-year follow-up, established as the reference standard. Bone-RADS will be dichotomized as non-aggressive (Bone-RADS1) and potentially aggressive (Bone-RADS2-4) to calculate accuracy parameters (sensitivity, specificity, global accuracy, positive and negative predictive value, and likelihood ratio)