PERCUTANEOUS RENAL BIOPSY UNDERESTIMATES NUCLEAR GRADE

218

THE JOURNAL OF UROLOGY®

Inclusion criteria were an enhancing solid renal mass, less than 7 cm, and 1mm slice thickness CT scans available for review (n = 300). We used Oncocare software (Siemens Medical Solutions, Malvern, PA) for making tumor volumetric measurements, and assessing the various tumor characteristics on non-contrast (NC), vascular phase (VP), and parenchymal phase (PP) of the CT scans. Tumor volume, largest transverse diameter (RECIST diameter), largest orthogonal diameter, and a maximum ratio of intraparenchymal extension to total diameter were calculated on PP. Means and standard deviations (SDs) of attenuation in the entire tumor, abdominal aorta and normal renal parenchyma were calculated on all phases. Final pathology of the 300 renal tumors (mean size, 3.0 cm; range, 0.9-6.7 cm) included, 157 (52%) clear cell renal cell carcinoma (RCC), 56 (19%) papillary RCC, 27 (9%) chromophobe RCC, 28 (9%) angiomyolipoma, and 32 (11%) oncocytoma. Using randomly selected records of 200 patients (training set) we constructed 3 different models, based on discrimination analysis, artificial neural network, and classification-regression tree (CART) analysis, to predict 5 histological subtypes of renal tumor and applied each model to the remaining100 patients (validation set). RESULTS: CART analysis appeared to be performing best in the validation set to predict 5 subtypes of renal tumor on final pathology. Hierarchical divisions by gender, BMI, tumor volume, RECIST/orthogonal diameter ratio, attenuation of tumor, aorta and normal parenchyma correctly identified renal malignancy with 90% sensitivity and 73% specificity in overall patients. Pathological subtype analysis showed that hierarchy of divisions by these same parameters correctly identified 72% of 5 subtypes of renal tumor with moderate agreement (kappa = 0.57). CONCLUSIONS: Computer-aided models based on certain demographics and advanced CT image analyses may allow preoperative prediction of renal tumor histology. Source of Funding: None

610 EFFECT OF COLLECTING DUCT HISTOLOGY ON OUTCOMES IN RENAL CELL CARCINOMA

Vol. 181, No. 4, Supplement, Sunday, April 26, 2009

611 PERCUTANEOUS RENAL BIOPSY UNDERESTIMATES NUCLEAR GRADE Aaron J Blumenfeld*, Khurshid A Guru, Hyung L Kim, Buffalo, NY INTRODUCTION AND OBJECTIVE: Small renal masses can be safely observed in select patients who are poor surgical candidates. Renal biopsy may help identify patients who are candidates for observation. The accuracy of renal biopsy for predicting the final nuclear grade and histologic subtype was assessed. METHODS: 67 patients (29 female, 38 male) underwent biopsy of their renal mass with ultrasound or CT guidance. Indications for biopsy included prior history of nonrenal malignancy, tumor in a solitary kidney, and a central tumor suspicious for urothelial carcinoma. Percutaneous 18-gauge biopsy cores were obtained, and all patients subsequently underwent radical nephrectomy, partial nephrectomy or nephroureterectomy. Pre-operative biopsy results were compared to postoperative specimens. RESULTS: The mean tumor size was 5.9cm (range 1.5 -17). Overall, biopsy correctly identified 57/67 (85%) histologic subtypes. The preoperative biopsy correctly identified 47/48 (98%) clear cell renal carcinomas, 1/7 (14%) papillary carcinomas, 3/3 (100%) chromophobe carcinomas, and 1/5 (20%) urothelial carcinomas. One preoperative biopsy was nondiagnostic and final pathology was a chromophobe carcinoma. The pathologists were able to determine nuclear grade for 46 biopsies and nuclear grade was assigned to 56 of the final specimens. The biopsy correctly identified 16/56 (29%) final nuclear grades. The biopsy underestimated the nuclear grade in 28/46 (60%) cases. In 6/46 (13%) of cases, the biopsy nuclear grade increased by 2 when compared to the final grade. The biopsy rarely overestimated the nuclear grade; 2 cases (4%) that were assigned a grade 2 on biopsy were assigned a grade 1 after nephrectomy. CONCLUSIONS: Core biopsy for renal masses underestimates nuclear grade in the majority of cases; however histologic subtype is more reliably assessed, particularly for clear cell renal tumors. Source of Funding: None

Jonathan L Wright*, James Hotaling, Michael C Risk, Daniel W Lin, Seattle, WA INTRODUCTION AND OBJECTIVE: Collecting duct renal cell carcinoma (CDRCC) is a rare entity. Recently, two small, multi-institutional surgical case series from Japan and Europe reported on collecting duct carcinomas with conflicting results. In this study, we use a United States population-based dataset to describe the survival experience of patients with collecting duct tumors compared to those with clear cell renal cell carcinomas (CCRCC). METHODS: Cases of CDRCC were identified from the Surveillance, Epidemiology and End Results (SEER) Program. Cases of CCRCC over the same time interval were identified for a comparison cohort. Demographic and pathologic characteristics at the time of diagnosis were compared. Differences in disease specific survival were compared with univariate and multivariate Cox regression analysis adjusting for age, gender, race, TNM stage, grade, primary treatment, tumor size, tumor registry and year of diagnosis. RESULTS: A total of 160 cases of CDRCC. Over that time period, 33,252 CCRCC cases were diagnosed. CDRCC was more common in African-Americans (23% vs. 9%, p < 0.001). CDRCC were more commonly T3+ (33% vs. 18%, p < 0.001) and metastatic (28% vs. 17%, p = 0.001). Nephrectomy rates were similar (84% and 78%, p = 0.006). The 3-year disease-specific survival (DSS) rates were 58% for CDRCC and 79% for CCRCC. In multivariate analysis, there was an increased risk of mortality for patients with CDRCC compared to CCRCC (HR 2.42, 95% CI 1.72 - 3.39, p = 0.001). CONCLUSIONS: Compared to CCRCC, patients with CDRCC more often present at higher stage and more often occurs in AfricanAmericans. Even after adjusting for demographic, surgical and pathologic factors, the DSS is significantly worse for those with CDRCC compared to CCRCC. Further research into the biology of this rare tumor is required to explain these results. Source of Funding: None

612 SENTINEL NODE DETECTION IN RENAL CELL CARCINOMA (RCC) WITH PRE- AND INTRAOPERATIVE IMAGING. A FEASIBILITY STUDY. Axel Bex*, Geraldine de Windt, Warner Prevoo, Henk van der Poel, Simon Horenblas, Renato Valdes-Olmos, Amsterdam, Netherlands INTRODUCTION AND OBJECTIVE: Lymphatic drainage from RCC is unpredictable and the therapeutic benefit and extent of lymph node dissection (LND) are controversial. Objective was to evaluate the feasibility of intratumoral injection of radiolabeled tracer to image and sample draining lymph nodes in clinically non-metastatic RCC. METHODS: Ten patients with cT1-2 cN0 cM0 (<10 cm) RCC prospectively receive percutaneous intratumoral injections of 99mTcnanocolloid the day before surgery under ultrasound guidance (0.4 ml, 240 MBq at 1-4 intratumoral locations depending on tumor size). Lymphoscintigraphy is performed 20 minutes, 2 and 4 hours after injection. After 4-hour-scintigraphy, SPECT-CT is performed using a hybrid camera. After correction for tissue attenuation and scatter SPECT is fused to CT to determine the anatomical localization of the sentinel node. Surgery with sampling is performed the following day using a gamma probe and a portable mini gamma-camera. RESULTS: Currently, 8 patients are included (6 female, 2 male, mean age 55 years (range 45-77 years), 7 tumors right side, 1 left). Mean tumor size was 4 cm (range 3.5-6.5 cm). Six patients had sentinel nodes on scintigraphy (2 retrocaval, 4 interaortocaval) with one extraretroperitoneal location along the internal mammary chain. All of these nodes could be mapped and sampled. In 2 no draining nodes were visualized. RCC were of clear cell subtype with no lymph node metastases. CONCLUSIONS: In RCC, sentinel node identification using