A proposed "Radiological Evaluation Score for Bone Tumors" (REST): An objective system for assessment of a radiograph in patients with suspected bone tumor

Although radiographs have been widely used in the evaluation of patients with suspected bone tumors, the lack of an objective radiological assessment method leads to a challenge in reaching correct diagnosis. The study aimed to propose a Radiological Evaluation Score for Bone Tumors (REST) which includes eight radiological factors [characteristics, content, cortical breach, distinctiveness, distribution, periosteal reaction, fracture, and soft tissue swelling] to form a single score along with its validation by multidisciplinary observers. We reviewed the radiographs of 100 patients with a primary bone tumor which were selected at random from the database between January 2017 and January 2019 of a tertiary cancer center. Four reviewers (two orthopedic oncologists and two surgical oncologists) independently assessed the radiographs, based on the reporting system of REST. We constituted two groups according to the probable diagnosis of bone tumor (suspected benign tumor and suspected malignant tumor). The mean score in the suspected benign tumor group was 1.1 (range 0-3, 95% CI 0.8-1.3) and in malignant tumor group was 6.1 (range 2-8, 95% CI 5.8-6.4). A receiver operator characteristic (ROC) curve for REST was with a cutoff of 3.5, with the most diagnostic value area under curve (AUC) of 0.99. The sensitivity was 98% and specificity was 100% with a positive predictive value of 100% and a negative predictive value of 98%. The inter-observer correlation coefficient was 0.985 (p value