Imaging biomarkers for malignant peripheral nerve sheath tumors in neurofibromatosis type 1

Objective To determine the utility of quantitative metrics obtained from fMRI using diffusion-weighted imaging (DWI)/apparent diffusion coefficient (ADC) mapping compared with metabolic (F-18-fluorodeoxyglucose [FDG]-PET/CT) imaging in patients with neurofibromatosis type 1 (NF1) for the characterization of peripheral nerve sheath tumors (PNSTs) as benign or malignant. Methods This Institutional Review Board-approved, Health Insurance Portability and Accountability Act-compliant study retrospectively reviewed imaging of 55 PNSTs in 21 patients with NF1. Imaging included anatomic (unenhanced T1, fluid-sensitive, contrast-enhanced T1-weighted), functional DWI (b = 50, 400, 800 s/mm(2)) and ADC mapping, magnetic resonance sequences, and FDG-PET/CT imaging. Anatomic (size), functional (minimum ADC values), and metabolic (maximum standardized uptake values [SUVmax]) imaging characteristics were recorded. ADC values were correlated with SUVmax. With histologic correlation for all malignant PNSTs (MPNSTs) or clinical or imaging stability (>12 months) for benign lesions used as reference standards, diagnostic accuracy was calculated. Results Of 55 PNSTs, there were 19 (35%) malignant and 36 (65%) benign PNSTs. Benign PNSTs were overall smaller than MPNSTs (largest diameter 4.3 +/- 1.3 vs 8.2 +/- 3.3 cm, respectively, p = 0.014). Benign PNSTs had higher ADCmin (x10(-3) mm(2)/s) than MPNSTs (1.6 +/- 0.4 vs 0.6 +/- 0.2, respectively, p < 0.0001) and lower SUVmax than MPNSTs (3.2 +/- 1.8 vs 8 +/- 3.9, p < 0.0001, respectively). ADCmin correlated inversely with SUVmax (correlation coefficient r = -0.0.58, p < 0.0001). Maintaining a sensitivity of 100% with threshold values of ADCmin <= 1 or SUVmax >3.2, DWI yielded a specificity of 94% while FDG-PET/CT offered a specificity of 83%. Conclusions Both quantitative metabolic imaging and functional imaging offer high sensitivity for the characterization of PNSTs in NF1; however, DWI/ADC mapping offers increased specificity and may be a more useful modality. Classification of evidence This study provides Class II evidence that for patients with NF1, MRI using DWI/ADC mapping accurately distinguishes malignant and benign PNSTs.