Introduction

Non-contrast CT imaging may be helpful in distinguishing cystic from solid lesions with a proposed cutoff of <20 HU. Current guidelines suggest that renal lesions <20HU on pre-contrast study require no further evaluation as they are most commonly benign cysts. We evaluated the frequency of pRCC presenting with low pre-contrast attenuation that might otherwise be considered radiographically benign, as well as the relationship of this metric to histologic subtype of pathologically proven pRCC.

Methods

The prospectively maintained Fox Chase Cancer Center kidney cancer database was reviewed for pT1 or T2 pRCC between 2003-2015. Patients were categorized by papillary subtype. Preoperative non-contrast CT images were analyzed. Maximum tumor diameter was measured in centimeters. Attenuation was calculated as the average Hounsfield Units (HU) from 6 distinct axial regions. Low attenuation was defined as ? 20 HU and high attenuation as >20 HU. We assessed the relationship between pre-operative renal mass attenuation and pRCC subtype using logistic regression controlling for stage, age, gender and laterality.

Results

58 patients were identified with pT1 or pT2 pRCC in whom preoperative non-contrast CT images were evaluable. 24 (41%) had type 1 and 34 (59%) had type 2 pRCC. No significant differences were noted in age (median 66.75 vs. 63.41 years, p = 0.23) or tumor diameter (median 5.27 vs. 6.32 cm, p = 0.18). 27 patients (47%) exhibited an average preoperative non-contrast ROI ? 20 HU, of which 6 patients (22%) were type 1 pRCC and 21 patients (78%) were type 2 pRCC. Type 1 pRCC demonstrated a higher attenuation than type 2 tumors using both average and max attenuation (29.6 vs. 20.6 HU, p <0.01; and 37.3 vs 26.3 HU, p = <0.01, respectively). After adjustment, HU was an independent predictor of pRCC subtype. Relative to low-density tumors, high-density tumors were 80% less likely to be type 2 pRCC (OR = 0.20, 95% CI 0.06-0.70, p=0.01).

Conclusions

Nearly half of our evaluated pRCC presented with low attenuation lesions (<20HU) on unenhanced CT and would have been dismissed as benign lesions under current guidelines. Importantly, we demonstrate that not only were these low attenuation lesions not benign, but they were associated with the more aggressive type 2 subtype. These data contradict the opinion that low attenuation renal lesions require no further evaluation, and suggest that attenuation on non-contrast CT imaging is insufficient as a single parameter to rule out malignancy.

Funding

none