(18F)FDG-PET/CT metabolic characterisation of lung nodules in paediatric patients

Abstract

1180

Objectives: PET/CT is successfully used in metabolic characterisation of solitary lung nodule in adult patients. An SUV max of 2.5 is generally accepted to distinguish benign from malignant nodules. Very few studies have been performed in paediatric patients. Aim of this study was to identify the lowest dimension (visual analysis) and the lowest semi-quantitative value capable to metabolically characterise lung nodules in paediatric patients.

Methods: 63 patients were enrolled: 32 patients: 12 osteosarcomas, 13 Ewing’s sarcomas, 2 HL, 1 NHL, 1 EBV-related Lymphohistiocytosis, 1 pinealoblastoma, 1 ALL. 84 lung nodules have been studied. PET/CT results have been compared to histology and clinical-radiological follow-up.

Results: Visual analysis: sensitivity=90,1%; specificity=97,5%; PPV=97,5%; NPV=90,7%; accuracy=94%(p<0,05). Semi-quantitative analysis with an SUVmax cut-off >= 1 sensitivity 93,3%; specificity 87,1%; PPV 89,3%; NPV 91,8%; accuray 90,4%. Restricting the same analysis (SUVmax >= 1) to nodules with a diameter higher than 5 mm the sensitivity was >99%; specificity 69,2%; accuracy 92,7%; no statistically significant results obtained for nodules with diameter < 5mm. The same analysis performed with an SUVratio>=1 (SUVratio= (SUVmax lesion)/(SUVmax mediastinic) measured on aortic arch) obtained similar results.

Conclusions: PET/CT seems accurate and sensitive for the metabolic characterisation of lung nodules higher than 5 mm of diameter in paediatric patients. An SUVmax and an SUVratio >= 1 seem an useful semi-quantitative cut-off, capable to improve the metabolic characterisation of lung nodules.