594

Biomarkers that aid in detecting early stage ovarian cancer might improve overall survival in this disease. Much attention has been given to serum biomarkers, but relatively little to biomarkers in urine. We have used protein expression profiling to analyze urine samples from 228 women with epithelial ovarian cancer including 52 with early stage disease, 176 with late stage disease, and 74 with benign ovarian disease, in addition to 98 normal healthy controls. 15 μL of urine was added to 23 μL of denaturing buffer (9M urea/2% CHAPS) and incubated for 30 min at 4oC. Denatured samples were further diluted with 263ul of binding buffer (100mM Sodium Acetate pH 4 or 100mM Sodium Phosphate in 0.5M NaCl pH 7) prior to incubation in duplicate on the surface of a cationic (CM10) or metal affinity (IMAC-Cu++) ProteinChip array. One half of the diluted sample was incubated on the surface for 30 min at RT, removed and replaced with the remaining diluted sample for another 30 min incubation. Following standard buffer washes and addition of sinapinic acid, arrays were read in a PBSIIc TOF-MS or PCS4000 ProteinChip reader. Data were analyzed using the Mann-Whitney test. Approximately 250 peaks were detected across the two chip combinations. 32 peaks had a p value <.01 comparing benign ovarian disease and all stages of invasive epithelial ovarian cancer and 8 peaks had a p value <.01 comparing benign disease from early stage ovarian cancer. AUC values for these peaks ranged from 0.65 to 0.75. Likewise, a comparison of normal urine to early stage ovarian cancer revealed 59 peaks with a p value <.01 and 44 peaks with p values <.01. AUC values for these peaks ranged from 0.67 to 0.85. These peaks, in combination with serum markers such as CA125 may aid in detecting invasive epithelial ovarian cancer.

98th AACR Annual Meeting-- Apr 14-18, 2007; Los Angeles, CA