Presentation Authors: Brandon Mahal, Boston, MA, Mohamed Alshalalfa*, San Francisco, CA, Shuang Zhao, Ann Arbor, MI, William Chen, Elai Davicioni, Felix Feng, San Francisco, CA, Paul Nguyen, Boston, MA
Introduction: Progression of prostate cancer is a complex multistep process that involves molecular alterations in cells of the tumor and microenvironment, however the significance of stromal infiltration in primary prostate cancer is not well defined. Therefore, we performed genomic expression analyses of stromal infiltration markers and sought to determine the prognostic significance thereof in primary prostate cancer.
Methods: Genome-wide expression profiles of formalin-fixed paraffin-embedded RP tumor samples were evaluated from a prospective registry cohort (n=5,239) and three retrospective institutional cohorts with long follow-up for metastasis outcome (n=1,135) for a total of 6,374 patients. Additionally we used the TCGA-prostate cohort for validation (N=498). We used the ESTIMATE algorithm of 141 stromal genes to infer stromal infiltration from gene expression data. The association between stromal infiltration expression levels and metastasis-free survival was assessed using Cox proportional hazards regression.
Results: There was strong correlation between ESTIMATE stromal expression scores and key stromal markers (CAV1, VIM, TAGLN), basal activity, and CD3 and CD4 immune biomarkers (r>0.5 for all). ERG-fusion frequency was consistent across the deciles of stromal scores in the prospective cohort (about 40% for all deciles). The top decile of stromal expression scores was associated with the highest rate of high Decipher genomic risk score (60%), and was enriched with 15% and 20% of high CAPRA-S and Gleason 9-10 disease, respectively. The top decile of stromal expression scores was independently associated with lower metastasis-free survival in three independent cohorts, including TCGA (p < 0.05 for all). Furthermore, based on a 24-gene radiation sensitivity signature, tumors with high sensitivity had higher stromal infiltration(p < 0.001).
Conclusions: High expression of stromal infiltration markers was associated with high-risk Decipher genomic-risk, high CAPRA-S score, and Gleason 9-10 disease. Furthermore, high expression of stromal infiltration markers was associated with a higher risk of metastases compared to lower expression of stromal infiltration markers. Expression of stromal microenvironment markers appears to be prognostic for prostate cancer outcomes. Future studies will be needed to determine whether genomics of stromal microenvironment should be incorporated into clinical decision-making and management recommendations.
Source of Funding: This study was supported by the PCF-ASTRO Career Development Award to End Prostate Cancer