Study design
A retrospective study was conducted on 196 patients undergoing radical prostatectomy at the Durham Veterans Affairs Medical Center (DVAMC) between 1993 and 2004. None of the patients had hormonal ablation or radiation therapy prior to surgery. Four core needle biopsies of cancer of the formalin fixed and paraffin embedded radical prostatectomy specimens were used to create tissue microarrays (TMAs). Slices of the TMA were stained for expression of UGT2B15, UGT2B17, and UGT2B28. We used two indicators for the staining power of the biomarkers, the average percentage of positive staining cells, and H-score. H-score was calculated for each of the three biomarkers using the formula: 3*percent staining strong +2*percent staining medium +1*percent staining weak staining.
Patients were followed for a median of 118 months. Biochemical recurrence (BCR) was defined as a single PSA value of >0.2 ng/ml, two values of 0.2 ng/ml, or need for secondary treatment for elevated PSA in the post-operative period. We excluded 1 patient missing extra-capsular extension, 2 missing seminal vesicle, and 3 missing follow-up, leaving a total of 190 patients in our study cohort. Institutional Review Board approval was obtained at Duke University, North Carolina Central University and the DVAMC and all patients signed an informed consent at the DVAMC prior to enrollment.
Immunohistochemistry of tissue microarrays
For all antibodies, TMA slides from DVAMC were baked overnight in a 60 °C oven before IHC staining. The staining protocols are as follows: UGT2B15: Sections were deparaffinized with xylene and rehydrated through graded ethanol. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 10 min. Heat-induced epitope retrieval (HIER) was carried out in EDTA, pH 8.0 (Invitrogen, cat# 005501) using a vegetable steamer at 95 °C for 25 min. Mouse polyclonal anti-human UGT2B15 primary antibody (Abcam, ab89274) was diluted with BSA to a concentration of 1:50 and applied to the sections overnight at 4 °C. Slides were rinsed with PBS then Dakocytomation Envision System Labelled Polymer HRP anti-mouse (DakoCytomation, cat# K4001) secondary antibody was applied for 30 min at room temperature. The antibody was then visualized using the Betazoid DAB Chromogen Kit (BioCare Medical, cat# BDB2004L). Finally, sections were rinsed with water, counterstained with hematoxylin, dehydrated through graded ethanol, cleared with xylene, and coverslipped.
UGT2B17: Sections were deparaffinized with xylene and rehydrated through graded ethanol. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 10 min. Heat-induced epitope retrieval (HIER) was carried out in EDTA, pH 8.0 (Invitrogen, cat# 005501) using a vegetable steamer at 95 °C for 25 min. Rabbit polyclonal anti-human UGT2B17 primary antibody (Abcam, ab92610) was diluted with BSA to a concentration of 1:50 and applied to the sections overnight at 4 °C. Slides were rinsed with PBS then Dakocytomation Envision System Labelled Polymer HRP anti-rabbit (DakoCytomation, cat# 4003) secondary antibody was applied for 30 min at room temperature. The antibody was then visualized using the Betazoid DAB Chromogen Kit (BioCare Medical, cat# BDB2004L). DAB was rinsed away with distilled water, cupric sulfate was applied for 5 min, and the tissue was rinsed again with distilled water. Finally, sections were rinsed with water, counterstained with hematoxylin, dehydrated through graded ethanol, cleared with xylene, and coverslipped.
UGT2B28: Sections were deparaffinized with xylene and rehydrated through graded ethanol. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 10 min. Heat-induced epitope retrieval (HIER) was carried out in 0.01 M citrate buffer, pH 6.0 using a vegetable steamer at 95 °C for 25 min. Rabbit polycloncal anti-human UGT2B28 primary antibody (Abcam, ab156131) was diluted with BSA to a concentration of 1:50 and applied to the sections for 45 min at room temperature. Slides were rinsed with PBS then Dakocytomation Envision System Labelled Polymer HRP anti-rabbit (DakoCytomation, cat# 4003) secondary antibody was applied for 30 min at room temperature. The antibody was then visualized using the Betazoid DAB Chromogen Kit (BioCare Medical, cat# BDB2004L). Finally, sections were rinsed with water, counterstained with hematoxylin, dehydrated through graded ethanol, cleared with xylene, and coverslipped.
Image analysis
Slides were digitized on a ScanScope AT (Leica Biosystems, Inc., Vista, CA) and morphimetric analysis performed with Definiens’ Tissue Studio (Definiens Inc., Parsippany, NJ) to determine the percentage of UGT2B15, UGTB17, UGTB28 and EGFR positive cells in a non-biased method. Briefly, stain specific algorithims were created using the pre-defined cytoplasmic detection module and classification tool, positive and negative stained cells within each tissue region were identified. Thresholds were set to classify hematoxylin stain for nuclei and DAB stain for positive cytoplasmic staining. The data were exported to Excel for further statistical analysis. Scanning and analyses were performed through the Translational Pathology Core Laboratory, Department of Pathology and Laboratory Medicine, David Geffen School of Medicine at UCLA.
Statistical analysis
Patient characteristics were summarized using median, 25th percentile, and 75th percentile for continuous variables and count and percentage for categorical variables. We tested the association between each biomarker’s (UGT2B15, UGT2B17, UGT2B28 – all continuous) percent positive and H-score and the risk of BCR using separate Cox proportional hazards models. Both crude models and models adjusted for PSA (continuous, log-transformed), age (continuous), pathological Gleason score (2-6 vs. 3 + 4 vs. ≥4 + 3), race (black vs. non-black), positive surgical margins (yes vs. no), extracapsular extension (yes vs. no), and seminal vesicle invasion (yes vs. no) were fit. In these models, UGT2B15, UGT2B17, and UGT2B28 percent positive and H-score were modeled per 10-unit increase. In an exploratory analysis, we split UGT2B17 total positive and H-score at the median and tested the association between UGT2B17 group and risk of BCR.
We also tested whether UGT2B15, UGT2B17, or UGT2B28 H-scores were correlated with pathological Gleason score using the Kruskal-Wallis test or chi-square test. For each level of Gleason score (2-6 vs. 3 + 4 vs. ≥4 + 3), we reported median, 25th percentile, 75th percentile of the biomarker H-score, <median (low expression), and ≥median scores (high expression).