Magnetic Resonance Imaging/Positron Emission Tomography Provides a Roadmap for Surgical Planning and Serves as a Predictive Biomarker in Patients With Recurrent Gynecological Cancers Undergoing Pelvic Exenteration

Abstract

Magnetic resonance imaging (MRI) is the modality of choice for staging gynecological cancers owing to its superb soft tissue resolution, whereas F-fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT) allows the assessment of glycolytic activity within the tumor microenvironment. In this study, we evaluated the incremental value of fused MRI/PET over MRI or fluorodeoxyglucose PET/CT alone for assessing local disease extent in patients with recurrent gynecological cancers undergoing pelvic exenteration and determined the associations between imaging findings and clinical outcomes in this patient population. The institutional review board approved this retrospective, Health Insurance Portability and Accountability Act (HIPAA)-compliant study of 31 patients who underwent pelvic MRI and PET/CT 3 months or less before pelvic exenteration for recurrent cancers of the uterine cervix, corpus, or vulva/vagina. Using a 1 to 5 scale (1, definitely not present; 5, definitely present), 2 readers independently evaluated MRI, PET/CT, and fused MRI/PET images for the presence of bladder, rectum, and pelvic sidewall invasion. Surgical pathology constituted the reference standard. Measurements of diagnostic accuracy, interreader agreement, and associations between imaging findings and progression-free survival and overall survival were calculated. Compared with MRI or PET/CT, fused MRI/PET correctly improved readers' diagnostic confidence in detecting bladder, rectum, or pelvic sidewall invasion in up to 52% of patients. Interreader agreement was consistently in the highest ("almost perfect") range only for MRI/PET (κ = 0.84-1.0). The highest sensitivities (0.82-1.0), specificities (0.91-1.0), and predictive values (0.80-1.0) were consistently achieved with fused MRI/PET (although the differences were not statistically significant [P > 0.05]). Pelvic sidewall invasion on MRI/PET was the only finding significantly associated with both progression-free and overall survival for both readers (P = 0.0067-0.0440). In patients with recurrent gynecological cancers undergoing pelvic exenteration, fused MRI/PET served as a predictive biomarker and yielded greater diagnostic confidence and interreader agreement than either MRI or PET/CT.