Recurrent inverted papilloma: Diagnosis with pharmacokinetic dynamic gadolinium-enhanced MR imaging

BACKGROUND AND PURPOSE: Dynamic gadolinium-enhanced MR imaging has been use d successfully to identify post-treatment recurrence or postoperative chang es in rectal and cervical carcinoma. Our purpose was to evaluate the useful ness of dynamic gadolinium-enhanced MR imaging for distinguishing recurrent inverted papilloma (IP) from postoperative changes. METHODS: Fifteen patients with 20 pathologically proved lesions (recurrent IF, 12; fibrosis or granulation tissue, eight) were enrolled in the study. Three observers, blinded to pathologic results, independently evaluated con ventional MR images, including T1-weighted (unenhanced and postcontrast), p roton-density-weighted, and T2-weighted spin-echo images, Results then were determined by consensus, Dynamic images were obtained using fast spin-echo sequences at 5, 30, 60, 90, 120, 150, 180, and 300 seconds after the injec tion of gadolinium-diethylenetriamine penta-acetic acid. Time-signal intens ity curves of suspected lesions were analyzed by a pharmacokinetic model, T he calculated amplitude and tissue distribution time were used to character ize tissue, and their values were displayed as a color-coded overlay, RESULTS: T2-weighted images yielded a sensitivity of 67%, a specificity of 75%, and an accuracy of 70% in the diagnosis of recurrent IF. Contrast-enha nced T1-weighted images yielded a sensitivity of 75%, a specificity of 50%, and an accuracy of 65%, Pharmacokinetic analysis showed that recurrent IP had faster (distribution time, 41 versus 88 seconds) and higher (amplitude, 2.4 versus 1.2 arbitrary units) enhancement than did fibrosis or granulati on tissue, A cut-off of 65 seconds for distribution time and 1.6 units for amplitude yielded a sensitivity of 100% and a specificity of 100% for diagn osing recurrent IF. CONCLUSION: Dynamic MR imaging can differentiate accurately recurrent IP fr om postoperative changes and seems to be a valuable diagnostic tool.