Background  Retroperitoneal extra-adrenal paraganglioma is relatively uncommon, which needs to be differentiated from retroperitoneal schwannoma, and its MRI features are relatively limited.

  Objective  To analyze the clinical characteristics and MR imaging features of retroperitoneal extra-adrenal paraganglioma and schwannoma, and explore the diagnostic value of MRI.

  Methods  From January 2016 to December 2020, clinical data and MRI features of 37 cases with retroperitoneal extra-adrenal paragangliomas and 17 cases with schwannomas confirmed by pathology admitted to the First Medical Center of Chinese PLA General Hospital were retrospectively reviewed.

  Results  Of the 37 patients with paraganglioma, 21 cases had hypertension, 5 cases had abdominal pain and abdominal discomfort, and 11 cases were discovered incidentally. Of the 17 patients with schwannoma, 2 cases had hypertension and 15 cases were discovered incidentally. There was no statistically significant difference between retroperitoneal extra-adrenal paraganglioma and schwannoma in lesion location, size, diameter ratio, boundary, shape, texture and average ADC value or signal intensity on Pre-LAVA T1WI (all P > 0.05), while they had significant differences in enhancement degree in arterial phase and percentage arterial enhancement (PAE) (P＜0.05, respectively). Paraganglioma showed two main contrast enhancement patterns, including mild or moderate to obvious enhancement in arterial phase, and persistent or “wash-out” enhancement in venous and delayed phase, while schwannomas presented mild to moderate enhancement in the arterial phase and showed persistent enhancement in the venous and delayed phases. Fluid-liquid level was found in 3 cases, and flow void or tortuous vascular shadows were found in or around the tumor in 4 cases.

  Conclusion  Retroperitoneal extra-adrenal paraganglioma and schwannoma show different MR imaging characteristics. Fluid-fluid level, flow void or tortuous vascular shadows, and arterial phase enhancement pattern are helpful in distinguishing the two tumors, and the diagnostic accuracy will be improved when combining with clinical features.