MR-IMAGING EVALUATION OF RENAL-CELL CARCINOMA

Background: This study examines the minimally required imaging protoco l needed for detection and staging of renal cell carcinoma (RCC). Meth ods: In 81 patients (21 women, 60 men; mean age = 62 years) with 85 RC Cs, T1-weighted (T1WI), contrast-enhanced T1-weighted (Gd-T1WI), T2-we ighted (T2WI), and gradient recalled echo-fast low flip angle shot (GR E/FLASH) images were evaluated alone and in combination. Surgical-path ological findings were available in all patients and were considered t he standard of reference. Results: Tumor detection for lesions smaller than 3 cm was better on Gd-T1WI than on any other sequence, but only the comparison with noncontrast T1WI and GRE/ FLASH was statistically significant (detection: T1WI = 33%, Gd-T1WI = 80%, T2WI = 60%, GRE = 4 7%). The respective accuracies of T1WI, Gd-T1WI, T2WI, and GRE/FLASH i mages were 81%, 78%, 71%, and 62% for evaluating local tumor extension ; 90%, 88%, 89%, and 85% for lymphadenopathy; and 89%, 81%, 91%, and 9 5% for renal vein thrombus. The combination of T1WI and GRE sequences rendered the highest overall staging accuracy. Conclusion: For tumor d etection, contrast-enhanced T1WI is necessary for lesions smaller than 3 cm. For tumor staging, although the addition of GRE results in sign ificant improvement in the evaluation of venous thrombus, any combinat ion of two sequences will result in similar accuracy, and the use of m ultiple sequences is not necessary.