Magnetic resonance imaging appearances following hindquarter amputation for pelvic musculoskeletal malignancy

Abstract

To describe our experience of the magnetic resonance imaging (MRI) appearances after hindquarter amputation, focusing on the normal pattern of time-related changes in signal intensity within the amputation flap and appearances of local disease recurrence. This was a retrospective review of the records of patients identified on our orthopaedic oncology database who had undergone hindquarter amputation at our centre. Examinations were divided into one group consisting of 'normal' postoperative findings and another group demonstrating local recurrence. 'Normal' postoperative findings were assessed for signal intensity, morphology and presence of the 'muscle texture sign' within the amputation flap. Examinations were grouped according to time from surgery. Examinations demonstrating local recurrence were assessed for site, signal intensity, morphology and axial dimensions of recurrent tumour. Comparison was made to pre-operative imaging. Thirty-five patients with postoperative MR images were identified. Seventy-seven examinations were reviewed (range 1-18 per patient). The 'normal group' comprised 18 patients (seven female, 11 male, mean age 44 years, range 16-75 years), with 47 examinations reviewed. The 'muscle texture sign' was preserved in all examinations in this group. Up to 6 months after surgery, 100% (10) of examinations showed hyperintense T2- and short-tau inversion recovery (STIR-weighted) signals within the muscle flap, indicating oedema. Only after 4 years was the muscle flap signal isointense in all cases (13). Up to 6 months after surgery, 50% (5 of 10) of examinations demonstrated flap swelling, but after 1 year, 100% (28) showed atrophy. Twenty percent (2 of 10) of examinations 0-6 months from surgery showed hyperintense T1-weighted signal within the muscle flap, consistent with fatty infiltration, but after 2 years, 100% (20) showed a hyperintense T1-weighted signal. The 'local recurrence' group comprised 20 patients (six female, 14 male, mean age 45.7 years, range 14-69 years), with 30 examinations demonstrating recurrent disease or metastasis. There were 23 recurrent tumours, of which 43% (10) were in the muscle component of the amputation flap, 13% (3 of 23) were in subcutaneous tissues of the flap, 13% (3 of 23) were at the posterior margin of the bone resection and 13% (3 of 23) were paraspinal. All recurrent tumours demonstrated hyperintense T2 and STIR signals. The signal intensity pattern of recurrence closely followed that of the original tumour. The 'muscle texture sign' was absent in all examinations demonstrating recurrence in muscle flap. A pattern of time-related signal intensity and morphological changes is observed within normal hindquarter amputation flaps on MRI. Recurrent disease is most likely to occur within the muscle flap, typically of high T2-weighted signal intensity and characteristically follows the signal pattern of the original tumour.