Characterization of adrenal lesions using MDCT wash-out parameters: diagnostic accuracy of several combinations of intermediate and delayed phases

Purpose: To evaluate the diagnostic accuracy of wash-out parameters calculated using multiple intermediate and delayed phases. Materials and methods: This prospective study had institutional review board approval and informed consent was obtained from all patients. Between January 2012 and October 2016, 108 consecutive oncologic patients (59 males, 49 females, mean age 52.6 years; 129 diagnosed lesions) underwent multiphasic CT protocol including unenhanced (UE), arterial (AE), portal (PE), 5-min (DE-5) and the 15-min (DE-15) delayed phases of adrenal glands. All images were randomly reviewed in consensus by two radiologists experienced in abdominal CT, unaware of clinical or pathologic data. Location, size and density were recorded. Absolute wash-out, percentage wash-out (PWO) and percentage enhancement wash-out ratio were calculated. The thresholds yielding the best accuracy in differentiating adenomas from nonadenomas were retrospectively determined on the basis of ROC curves. The corresponding diagnostic accuracy values were calculated. Paired sample t test was used to assess differences among imaging parameters within subgroups. Student t test was applied to compare lesions between independent subgroups. p values ≤ 0.05 were considered significant. Results: The final diagnosis included 82 adenomas (62 lipid-rich and 20 lipid-poor) and 47 nonadenomas (42 metastases, 3 pheochromocytomas, 2 carcinomas). All the 62 lipid-rich adenomas were correctly diagnosed as benign lesions on the basis of their UE attenuation < 10 HU. The PEAK attenuation was achieved during AE phase for 51/129 lesions (39.5%) and at the time of PE phase in 78/129 lesions (60.5%). The best overall accuracy in diagnosing adenomas (97.6%; 126/129 lesions correctly diagnosed) was obtained using 40% threshold for calculating PWO from PEAK to DE-15 scan. Conclusions: If only an intermediate phase is available, the 15-min delayed scan should be acquired to avoid any drop in diagnostic accuracy. The availability of two intermediate phase may be used to easy CT schedule by obviating the need to acquire a longer delayed phase.