A right-handed 66-year-old Japanese man experienced right elbow pain and was unable to extend his right thumb and fingers for 1 month. He did not have associated history of trauma to his elbow or any remarkable medical history. A physical examination showed swelling of his right elbow and a palpable mass on the anterior aspect of his right elbow. Grip strengths of his right and left hands, as measured with a Jamar digital dynamometer (Takei Scientific Instruments Co., Ltd., Niigata, Japan), were 30.4 and 35.0 kg, respectively. The respective ranges of motion for his right and left extremities, as measured with a standard goniometer, were as follows: elbow flexion, 115° and 145°; elbow extension, − 15° and 0°; forearm pronation, 30° and 70°; forearm supination, 80° and 90°. The muscle strengths of his right upper extremity, as evaluated using the British Medical Research Council scale, were the following: triceps, 5; wrist extensor, 5; extensor pollicis longus, 3; extensor digitorum communis and extensor indicis proprius (EIP), 2. There was no sensory loss.
Plain radiographs of his right elbow showed osteoarthritic change with calcifications and ossicles anteriorly, posteriorly, and laterally (Fig. 1a, b). Plain computed tomography (CT) and three-dimensional CT scans (Activion 16; Toshiba Medical Systems Corp., Tokyo, Japan) showed clustered calcifications around the radial neck, coronoid fossa, radial fossa, and olecranon fossa. Plain magnetic resonance imaging (MRI) scans (EXCELART Vantage 1.5 Tesla, version 9.51; Toshiba Medical Systems Corp.) showed mass lesions around the radial neck, medial epicondyle, olecranon fossa, and coronoid fossa, with heterogeneous intensity on T1-weighted and T2-weighted images (Fig. 1c, d). A nerve conduction study (NCS) for his radial nerve was performed. For the motor NCS, which was recorded at the EIP, surface electrodes for stimulation were set at 8 cm proximal to the EIP, 5 cm proximal to the elbow crease, and posterior to the insertion of the deltoid. For the antidromic sensory NCS, it was recorded at the middle between the first and second metacarpal bones; surface electrode for stimulation was set at 14 cm proximal to the recording position. The respective right and left motor nerve conduction velocities were 70.2 and 57.7 m/s from the upper arm to the elbow and 29.9 and 66.1 m/s from the elbow to the forearm. The results of sensory nerve conduction velocities were 62.5 and 57.9 m/s, respectively.
The quality of the three-dimensional CT images was poor for detecting the tumor so we created three-dimensional reconstructed images of the tumor based on MRI to visualize the shape and location of the tumor and its relation to the radial nerve. Digital imaging and communications in medicine data obtained from plain magnetic resonance (MR) images were transferred to Mimics computer-aided design (CAD) software (Materialise Japan, Yokohama, Japan). The bone region was segmented semi-automatically with an intensity threshold segmentation technique, and the tumor and nerve (radial nerve, PIN, and superficial branch of radial nerve) intensity was contoured manually using the CAD software; then, three-dimensional reconstruction was performed (Fig. 2a–c). The three-dimensional images clearly showed the tumor location and morphology; we found a giant tumor around the radial head and neck and a large mass in the radial fossa, olecranon fossa, and medial to the coronoid fossa. Moreover, the running course of the PIN was extremely changed at the distal corner of the tumor in contrast to the superficial branch of the radial nerve.
We diagnosed our patient as having incomplete PIN palsy caused by synovial osteochondromatosis. We decided to resect the tumor from the anterolateral part and radial fossa and perform neurolysis of the PIN because our patient had no impairment in daily life, for example, limitations in range of motion were compensated by shoulder joint motion.
Surgical treatment was performed via the anterolateral (Henry) approach using an air tourniquet, with our patient under general anesthesia. The three-dimensional reconstruction images were used as a basis for the surgical exposure, and the PIN was compressed between the arcade of Frohse and the tumor, which was under the PIN (Fig. 3a). The PIN was kinked, especially at the corner of the tumor, as shown by the three-dimensional reconstruction image. After the incision of the arcade of Frohse was made, neurolysis of the PIN was performed. The tumor was covered by the joint capsule (Fig. 3b). When the joint capsule was incised, a white, cartilaginous tumor was found. The tumor, including the synovium, was resected (Fig. 3c). Results of a histological analysis showed synovial osteochondromatosis without malignancy. No major postoperative complications occurred. Our patient fully recovered from the PIN palsy 9 months postoperatively. One year postoperatively, his grip strengths were 36.7 and 33.7 kg for his right and left hands, respectively. The ranges of motion of his right extremity were as follows: elbow flexion, 125°; elbow extension, − 15°; forearm pronation, 65°; forearm supination, 90°. No recurrence of the lesion occurred 1 year postoperatively.