Distinguishing characteristics of carpal osteoid osteoma versus osteoblastoma: a comparative study

Osteoid osteoma and osteoblastoma can be difficult to differentiate without biopsy, and clinical management can differ. Therefore, we asked: (1) what are the differences in clinical symptoms and imaging characteristics between carpal osteoid osteoma and osteoblastoma; and (2) which radiological measurements can differentiate carpal osteoid osteoma from osteoblastoma, and what are their optimal cut-off values?

This retrospective single-centre study included 20 surgically treated patients with carpal osteoid osteoma (16 patients) or osteoblastoma (four patients) from January 2012 to December 2025. Demographic, clinical and imaging data were collected. Clinical assessment included night pain worsening score, rest pain, activity-related pain, pain after non-steroidal anti-inflammatory drug treatment, local tenderness and swelling. Preoperative computed tomography images were used for three-dimensional segmentation analysis to measure lesion volume, mineralized volume, lesion diameter, lesion-to-carpal volume ratio and mineralization-to-lesion ratio. Bone marrow and soft tissue oedema were evaluated on magnetic resonance imaging when available. Diagnostic performance and measurement reliability were also assessed.

Baseline demographic characteristics were similar between groups. Compared with osteoblastoma, osteoid osteoma showed greater night pain worsening scores (2.0 (IQR: 1.0 to 2.0) vs. 0.5 (IQR: 0 to 1.0)) and lower pain visual analogue scale scores after non-steroidal anti-inflammatory drug treatment (3.5 (IQR: 3.0 to 4.0) vs. 6.0 (IQR: 5.8 to 6.0)). Rest pain, activity-related pain, local tenderness and swelling were similar between groups. Osteoblastoma had a larger lesion diameter (18 mm (IQR: 17 to 20) vs. 5.8 mm (IQR: 5.5 to 6.7)), lesion volume (1300 mm³ (IQR: 920 to 1700) vs. 72 mm³ (IQR: 48 to 120)), and lesion-to-carpal volume ratio (40% (IQR: 36 to 48) vs. 2.7% (IQR: 2.5 to 2.9)), whereas osteoid osteoma had a higher mineralization-to-lesion ratio (46% (IQR: 40 to 56) vs. 3.9% (IQR: 2.8 to 4.3)). The cut-off values were 460 mm³ for lesion volume, 16% for lesion-to-carpal volume ratio, and 9.4% for mineralization-to-lesion ratio. Magnetic resonance imaging oedema scores were similar between the two groups.

Worsening nocturnal pain, relief following non-steroidal anti-inflammatory drug administration, a smaller lesion size and a higher mineralization-to-lesion ratio are suggestive of osteoid osteoma, while a larger lesion-to-carpal volume ratio and lower internal mineralization support a diagnosis of osteoblastoma.

IV.