Abstract 4013: A novel quantitative histomorphological tool to assess multiple biomarkers to predict prostate cancer aggression

Abstract

Abstract Introduction and Objective: This study was designed to develop an integrated, quantitative histomorphometric and molecular biomarker-based predictor for prostate cancer (PCa) progression based on a tissue microarray (TMA) useful for small volume active surveillance biopsies. Methods: Two TMAs with 80 PCa cases of different Gleason scores were used and each PCa case includes quadruplicates of cancer and cancer-adjacent benign areas. Also, 25% of the 80 cases had recurrence. Multiplex Tissue Immunoblotting (MTI) was used to detect and pre-qualify the expression of five biomarkers on a single 5 micron section: CACNA1D, Periostin, Her2/neu, EZH2 and (-7)ProPSA. Immunofluorescence was determined using excitation at 633 nm which produced the Cy5 fluorescence, while excitation at 488 nm produced FITC fluorescence for the biomarkers. The scanned images were analyzed using ImageQuant 5.2 software and the final data was normalized by total protein for each biomarker. The results were graphically and statistically analyzed to study the relationship with PCa grades and recurrence risk. The same approach was applied to active surveillance biopsies. RESULTS: The biomarkers showed that for CACNA1D, an epithelial expressed Ca++ transporter glycoprotein biomarker, demonstrated a decrease with increasing Gleason score in both cancer and benign areas. The ROC (Receiver Operating Characteristic) curve showed a continuous increase in the AUC (Area Under the Curve) and statistical analysis also showed clear difference by comparing group of Gleason score 3+3 with 8 and above. The expression of Periostin, a stromal glycoprotein biomarker indicated a decrease with increasing Gleason score from 3+3 to 3+4 and 4+3, but the expression rebounded in Gleason score 8 and above. A field effect was shown in benign cancer-adjacent area of prostate tissues for most biomarkers. (-7)ProPSA, an epithelial biomarker shown to be significantly elevated with increasing Gleason score in PCa, which was supported by the AUC-ROC in both cancer and benign area. Applying Logistic Regression analysis, CACNA1D and Her2/neu biomarkers predicted recurrence yielding a 0.87 AUC-ROC. Additionally, (-7)ProPSA, Her2/neu, CACNA1D and Periostin separated Gleason score up to 7.0 vs. above 7.0 with a AUC 0.97. CONCLUSIONS: Our results suggest that several molecular biomarkers may be useful to predict a PCa aggressive phenotype (i.e. CACNA1D, Periostin, Her2/neu, and (-7)ProPSA). Further, this MTI tool is currently being used to characterize aggressive PCa of active surveillance patients. Citation Format: Zhi Liu, Christhunesa Christudass, Hui Zhang, Joon-Yong Chung, Stephen Hewitt, Jonathan Epstein, Robert W. Veltri. A novel quantitative histomorphological tool to assess multiple biomarkers to predict prostate cancer aggression. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4013. doi:10.1158/1538-7445.AM2014-4013