Abstract PR15: The prognostic power of stromal reactivity: An integrated approach to prostate cancer evolution

Abstract

Abstract Prostate cancer (PCa) is the fifth most common cancer worldwide. Gleason score is the standard metric for defining prostate cancer aggressiveness. Still, prostate cancer is difficult to prognose. Stroma is known to play a major role in tumor progression to malignancy and upon activation (reactive) produces growth factors (GFs) necessary for tissue repair. The prognostic significance of reactive stroma grading in prostate biopsies remains uncertain, except for recurrence in Gleason 7 patients. Moreover, how stromal reactivity and tumor aggressiveness drives progression in PCa remains unclear. To better understand PCa progression, we investigated the interplay between stroma, tumor and microenvironmental heterogeneity in driving PCa aggressiveness. We used an integrated approach combining 7200 simulated prostate cancers over 20 years, 870 prostate cancer patients with 20 years of follow-up (recurrence and death) as well as in-vivo experimentation. We designed a multiscale computational model that incorporates, chemical and physical properties of the microenvironment that modulate the dynamic behavior of hundreds of thousands of cells. This model was simulated on an anatomically reconstructed tissue domain that represents most of the largest segment of the prostate where cancer occurs most frequently (peripheral zone). Each simulated tumor was initiated by seeding a heterogeneous distribution of GF producing tumor cells along with 3 different degrees of reactive stroma. We corroborated our in silico results with whole mount slides from clinical radical prostatectomy specimens. Histopathological analysis included stage, primary and secondary Gleason grades with associated outcomes information. In addition, in vivo experiments were performed combining an initiated epithelial cell line with fibroblasts from patients with pathologically-defined degrees of stromagenesis. The recombinant tissues were grown, stained and the aggressiveness of the tumor was recorded. Our findings reveal that the degree of reactive stroma significantly correlates with the level of aggressiveness of the tumor. Crucially, the degree of stromal reactivity appears to be a significant selection pressure in evolving aggressive tumors. Counter intuitively, a low degree of reactive stroma correlates with a strong selection pressure for higher GF production in the tumor cells - due to the restricted levels of GF found in the microenvironment. This results in a slow progression towards increasingly more aggressive tumor phenotypes. In contrast, a high degree of reactive stroma correlates with a weaker selection pressure for GF producing tumor cells and fast progression - since the GFs are readily accessible to the tumor via the microenvironment. Based on these findings, we conclude that the degree of stromal reactivity is a significant prognostic indicator of recurrence and death for the risk level of low, intermediate and high Gleason grades. For example, Gleason 6 is considered to have a low risk prognosis; however, in cases where the stroma is found to be highly reactive the risk should be considered as high. Patients with Gleason scores 9 or 10 already have a high risk prognosis even when there is a low degree of reactive stroma, but their risk increases even more, when there is a high degree of reactive stroma. In brief, using an integrated computational, experimental and clinical data approach we conclude that the stroma surrounding prostate tumors plays a substantial role in driving PCa evolution toward aggressiveness and prognosis towards higher risk. This abstract is also presented as Poster A66. Citation Format: Ziv Frankenstein, David Basanta, Omar Franco, Douglas Strand, Simon Hayward, Gustavo Ayala, Alexander Anderson. The prognostic power of stromal reactivity: An integrated approach to prostate cancer evolution. [abstract]. In: Abstracts: AACR Special Conference on Cellular Heterogeneity in the Tumor Microenvironment; 2014 Feb 26-Mar 1; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(1 Suppl):Abstract nr PR15. doi:10.1158/1538-7445.CHTME14-PR15