abstract
- OBJECTIVES: Segmental enhancement inversion (SEI) is a controversial imaging finding reportedly specific for the diagnosis of renal oncocytoma. The purpose of this study was to re-evaluate SEI using biphasic CT and multiphase MRI. METHODS: With research ethics board approval, a retrospective analysis of patients with resection or biopsy of oncocytoma or chromophobe renal cell carcinoma (Ch-RCC) between 2008-2012 was performed. Twenty-four patients with oncocytoma and 13 patients with Ch-RCC underwent CT, while 13 patients with oncocytoma and 10 patients with Ch-RCC underwent MRI. Two blinded radiologists reviewed the CT and MRI studies independently in separate sessions to assess for SEI. A third radiologist established consensus. Interobserver variability was calculated and diagnostic accuracy was compared using ROC and the Fisher exact test. RESULTS: There was no difference in detection of SEI between oncocytoma and Ch-RCC at CT [both readers (p = 0.65, 0.5) and consensus review (p = 0.29)] or MRI [both readers (p = 0.64, 0.74) and consensus review (p = 0.53)]. The interobserver variability at CT (K = 0.28-0.33) and MRI (K = 0.25-0.44) was fair. The sensitivity and specificity for diagnosis of oncocytoma were 21 % and 92 % at CT and 15 % and 90 % at MRI. CONCLUSION: SEI is not useful for the diagnosis of renal oncocytoma with CT or MRI. KEY POINTS: • SEI was detected in a minority of renal oncocytomas and chromophobe RCC. • Interobserver agreement for segmental enhancement inversion was only fair. • SEI is not useful for diagnosing renal oncocytoma with CT or MRI.