PT - JOURNAL ARTICLE AU - Julia S. Gerke AU - Martin F. Orth AU - Yuri Tolkach AU - Laura Romero-Pérez AU - Fabienne Wehweck AU - Stefanie Stein AU - Julian Musa AU - Maximilian M. L. Knott AU - Tilman L. B. Hölting AU - Jing Li AU - Giuseppina Sannino AU - Aruna Marchetto AU - Shunya Ohmura AU - Florencia Cidre-Aranaz AU - Martina Müller-Nurasyid AU - Konstantin Strauch AU - Christian Stief AU - Glen Kristiansen AU - Thomas Kirchner AU - Alexander Buchner AU - Thomas G. P. Grünewald TI - Biomarker-based outcome prediction in prostate adenocarcinoma depends on the <em>TMPRSS2-ERG</em> status AID - 10.1101/546200 DP - 2019 Jan 01 TA - bioRxiv PG - 546200 4099 - http://biorxiv.org/content/early/2019/02/10/546200.short 4100 - http://biorxiv.org/content/early/2019/02/10/546200.full AB - Background Prostate adenocarcinoma (PCa) with/without the TMPRSS2-ERG (T2E)-fusion represent distinct molecular subtypes.Objective To investigate gene-signatures associated with metastasis in T2E-positive and -negative PCa, and to identify and validate subtype-specific prognostic biomarkers.Design, setting and participants Gene expression and clinicopathological data of two discovery PCa cohorts (total n=783) were separately analyzed regarding the T2E-status. Selected subtype-specific biomarkers were validated in two additional cohorts (total n=405).Outcome measurements and statistical analysis From both discovery cohorts, we generated two gene lists ranked by their differential intratumoral expression in patients with/without metastases stratified by T2E-status, which were subjected to gene set enrichment and leading-edge analyses. The resulting top 20 gene-signatures of both gene lists associated with metastasis were analyzed for overlaps between T2E-positive and -negative cases. Genes shared by several functional gene-signatures were tested for their association with event-free survival using the Kaplan-Meier method in a validation cohort. Immunohistochemistry was performed in another validation cohort.Results and limitations Metastatic T2E-positive and -negative PCa are characterized by different gene-signatures. Five genes (ASPN, BGN, COL1A1, RRM2 and TYMS) were identified whose high expression was significantly associated with worse outcome exclusively in T2E-negative PCa. This was validated in an independent cohort for all genes and additionally for RRM2 by immunohistochemistry in a separate validation cohort. No prognostic biomarkers were identified exclusively for T2E-positive tumors.Conclusions Our study demonstrates that the prognostic value of biomarkers critically depends on the molecular subtype, i.e. the T2E-status, which should be considered when screening for and applying novel prognostic biomarkers for outcome prediction in PCa.Patient summary Outcome prediction for PCa is complex. The results of this study highlight that the validity of prognostic biomarkers depends on the molecular subtype, specifically the presence/absence of T2E. The reported new subtype-specific biomarkers exemplify that biomarker-based outcome prediction in PCa should consider the T2E-status.