OBJECTIVES: To describe the radiological features of primitive neuroectodermal tumour (PNET) of the chest wall (Askin tumour) at diagnosis and to analyse the radiological changes occurring as a consequence of treatment and during follow-up. MATERIALS AND METHODS: Nine children with histologically proven PNET were studied. At diagnosis, all patients underwent chest X-ray (CXR), chest CT and bone scintigraphy; three patients also had MR and three had US. During treatment and follow-up, CT was performed in all patients. RESULTS: CT demonstrated a solid heterogeneous chest wall mass in all children at diagnosis and six had a rib lesion. Small nodular densities in the extra-pleural fat were identified in three patients at diagnosis. US, performed in three patients, excluded tumour infiltration of the lung or diaphragm, which had been suspected on CT. On MR, the lesions showed high signal intensity in T1-weighted/proton-density images and intermediate/high signal intensity in T2-weighted images compared with muscle. Minimal chest wall involvement was demonstrated in one case by MRI. Extensive necrosis of tumour mass with pseudo-cystic appearance was documented in the five patients who underwent chemotherapy. Macroscopically complete resection was performed in five patients but there was early local recurrence after surgery in two, identified by CT in one and by MR in the other. CONCLUSIONS: PNET of the chest wall should be considered in a child with a chest wall mass. CT is valuable for evaluating tumour extension at diagnosis, the effects of chemotherapy and assessing tumour recurrence after surgery. However, CT can overestimate pleural, lung or diaphragmatic infiltration, which are better evaluated by US. MR was superior to CT in the evaluation of tumour extension in one of three patients and may be considered complementary to CT, particularly in very large chest wall tumours.
OBJECTIVES: To describe the radiological features of primitive neuroectodermal tumour (PNET) of the chest wall (Askin tumour) at diagnosis and to analyse the radiological changes occurring as a consequence of treatment and during follow-up. MATERIALS AND METHODS: Nine children with histologically proven PNET were studied. At diagnosis, all patients underwent chest X-ray (CXR), chest CT and bone scintigraphy; three patients also had MR and three had US. During treatment and follow-up, CT was performed in all patients. RESULTS: CT demonstrated a solid heterogeneous chest wall mass in all children at diagnosis and six had a rib lesion. Small nodular densities in the extra-pleural fat were identified in three patients at diagnosis. US, performed in three patients, excluded tumour infiltration of the lung or diaphragm, which had been suspected on CT. On MR, the lesions showed high signal intensity in T1-weighted/proton-density images and intermediate/high signal intensity in T2-weighted images compared with muscle. Minimal chest wall involvement was demonstrated in one case by MRI. Extensive necrosis of tumour mass with pseudo-cystic appearance was documented in the five patients who underwent chemotherapy. Macroscopically complete resection was performed in five patients but there was early local recurrence after surgery in two, identified by CT in one and by MR in the other. CONCLUSIONS: PNET of the chest wall should be considered in a child with a chest wall mass. CT is valuable for evaluating tumour extension at diagnosis, the effects of chemotherapy and assessing tumour recurrence after surgery. However, CT can overestimate pleural, lung or diaphragmatic infiltration, which are better evaluated by US. MR was superior to CT in the evaluation of tumour extension in one of three patients and may be considered complementary to CT, particularly in very large chest wall tumours.