Objectives: To evaluate the usefulness of normalising intra-tumour tracer accumulation on (18)F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) to reference tissue uptake for characterisation of peripheral nerve sheath tumours (PNSTs) in neurofibromatosis type 1 (NF1) compared with the established maximum standardised uptake value (SUVmax) cut-off of >3.5.
Methods: Forty-nine patients underwent FDG PET/CT. Intra-tumour tracer uptake (SUVmax) was normalised to three different reference tissues (tumour-to-liver, tumour-to-muscle and tumour-to-fat ratios). Receiver operating characteristic (ROC) analyses were used out to assess the diagnostic performance. Histopathology and follow-up served as the reference standard.
Results: Intra-tumour tracer uptake correlated significantly with liver uptake (rs= 0.58, P = 0.016). On ROC analysis, the optimum threshold for tumour-to-liver ratio was >2.6 (AUC = 0.9735). Both the SUVmax cut-off value of >3.5 and a tumour-to-liver ratio >2.6 provided a sensitivity of 100 %, but specificity was significantly higher for the latter (90.3% vs 79.8%; P = 0.013).
Conclusions: In patients with NF1, quantitative (18)F-FDG PET imaging may identify malignant change in neurofibromas with high accuracy. Specificity could be significantly increased by using the tumour-to-liver ratio. The authors recommend further evaluation of a tumour-to-liver ratio cut-off value of >2.6 for diagnostic intervention planning.
Key points: • (18)F-FDG PET/CT is used for detecting malignancy in PNSTs in NF1 patients • An SUV max cut-off value may give false-positive results for benign plexiform neurofibromas • Specificity can be significantly increased using a tumour-to-liver ratio.