Preoperative imaging in renal masses: does size on computed tomography correlate with actual tumor size?

To evaluate the discrepancy between tumor sizes determined from preoperative computed tomography (CT) and surgical specimens and its clinical implications.

The charts of 86 patients who underwent surgical resection of a renal mass between 1995 and 2007 were reviewed retrospectively. Tumor size was determined both from preoperative CT and pathologic specimen. Histopathologic evaluation was done. Wilcoxon test was used to compare the mean radiographic tumor size on CT with the mean pathologic size. P < 0.05 was considered as the threshold for statistical significance.

The median age was 59 (21-84). Clinical stage was T1a in 13, T1b in 47, and ≥ T2 in 26; pathologic stage was T1a in 12, T1b in 45, and ≥ T2 in 29 patients. Histological subtype was clear cell, papillary, chromophobe, sarcomatoid, and oncocytic in 72, 7, 5, 1, and 1 patients, respectively. Mean radiographic and pathologic size was 6.33 and 6.43 cm, respectively (p = 0.342). On the average, radiographic measurement underestimated pathologic size by 1 mm. When subgroups of patients according to tumor size were formed as < 4, 4-7, and > 7 cm, mean radiographic size was 2.79, 5.44, and 9.57 cm, mean pathologic size was 3.47, 5.62, and 9.26 cm, respectively. In subgroups of < 4, 4-7, and > 7 cm; radiographic measurement underestimated pathologic size by 0.68 (p = 0.018) and 0.18 cm (p = 0.470) and overestimated by 0.31 cm (p = 0.454), respectively.

Overall discrepancy between radiographic and pathologic tumor sizes was 1 mm. No significant stage shift due to measurement error was detected. Our findings suggest that CT is an accurate method with which to estimate renal tumor size preoperatively.