PURPOSE: To determine the nature of the imaging findings following reconstructive surgery using massive allografts in children with malignant bone tumours. MATERIALS AND METHODS: A retrospective review of the imaging studies and medical charts of 25 consecutive children who received an allograft as part of the management of a malignant bone tumour. RESULTS: Uncomplicated allografts were sclerotic relative to native bone on radiographs and showed a typical 'tramline' appearance on bone scintigraphy. On MR, the medullary canal of the allograft showed low signal, similar to or greater than skeletal muscle, but less than subcutaneous fat, on 91 % of T1-weighted images. On short-tau inversion recovery images, the medullary canal was inhomogeneous and hyperintense to subcutaneous fat in 70 % and hyperintense to muscle in the remainder. Complications occurred in 68 % of patients and included allograft fractures (36 %), recurrent tumour (20 %), infection (8 %), and non-union or delayed union (8 %). The radiographic findings alone permitted accurate diagnosis of most serious complications. Infection and rejection were difficult to distinguish with any technique. All complications were suspected on clinical and/or radiological grounds before being shown by MR or scintigraphy. CONCLUSIONS: Allografts, whether normal or complicated, have characteristic imaging findings, except that infection and bone resorption related to rejection and revascularisation are difficult to distinguish. Routine MR and bone scintigraphy appear to contribute little to the management of these patients.
PURPOSE: To determine the nature of the imaging findings following reconstructive surgery using massive allografts in children with malignant bone tumours. MATERIALS AND METHODS: A retrospective review of the imaging studies and medical charts of 25 consecutive children who received an allograft as part of the management of a malignant bone tumour. RESULTS: Uncomplicated allografts were sclerotic relative to native bone on radiographs and showed a typical 'tramline' appearance on bone scintigraphy. On MR, the medullary canal of the allograft showed low signal, similar to or greater than skeletal muscle, but less than subcutaneous fat, on 91 % of T1-weighted images. On short-tau inversion recovery images, the medullary canal was inhomogeneous and hyperintense to subcutaneous fat in 70 % and hyperintense to muscle in the remainder. Complications occurred in 68 % of patients and included allograft fractures (36 %), recurrent tumour (20 %), infection (8 %), and non-union or delayed union (8 %). The radiographic findings alone permitted accurate diagnosis of most serious complications. Infection and rejection were difficult to distinguish with any technique. All complications were suspected on clinical and/or radiological grounds before being shown by MR or scintigraphy. CONCLUSIONS: Allografts, whether normal or complicated, have characteristic imaging findings, except that infection and bone resorption related to rejection and revascularisation are difficult to distinguish. Routine MR and bone scintigraphy appear to contribute little to the management of these patients.