Abstract Background: Despite novel, targeted therapies, metastatic breast cancer patients have an extremely unfavourable prognosis. Prognostic and predictive factors for patients with advanced breast cancer are not well understood. Molecular assessment of the patient and the tumor in the metastatic situation is not routinely performed despite advances in molecular precision medicine indicating great benefit to this patient group. Here we present early findings from the first 142 patients of a prospective molecular breast cancer registry with completed transcriptomic profiling. Methods: The PRAEGNANT study proctocol (NCT02338767) is a molecular registry designed to provide an infrastructure for the real-time comprehensive analysis of tumor and patient molecular characteristics under study conditions. Formalin fixed paraffin embedded tumors have been used from this registry to identify molecular, transcriptomic predictors for overall survival (OS). Known clinical correlates for OS (e.g. hormone-receptor status, age at diagnosis, and BMI) were analyzed by Cox proportional hazard ratios, and compared to transcriptomic markers of outcomes. Transcriptomes for all patient tumors were sequenced on the Illumina sequencing platform, and analyzed by RSEM to estimate transcripts per million (TPM) values for each gene isoform. Log-TPM values were used in established (PAM50) and novel (hierarchical clustering) expression-based subtyping of tumor samples. Expression-based subtypes were demonstrated to be strong prognostic indicators by Cox analysis. A Lasso regression machine learning algorithm was used to develop an expression-based predictive model of OS. Results: Hormone receptor positivity (HR=0.7, p<0.006) and TNBC status (HR=1.4, p<0.01) were significantly associated with outcomes. PAM50 subtypes were also strong indicators of outcomes (e.g. Basal disease compared to Luminal-A subtype has HR=1.4, p<0.017). A novel expression-based high-risk cluster in this cohort was more indicative of poor prognosis than clinical variates or Basal-type, with a HR=2.7 (p<0.009) when compared to Luminal-A subtype. An expression-based survival prediction model achieved a concordance-index of 0.65 in an unseen validation cohort. Patients predicted as having the shortest survival times were in the high-risk cluster. Conclusions: Here we demonstrate using molecular profiling to develop prognostic signatures that out-perform standard clinical correlates of poor outcomes, even in a small subset of the total cohort. As the PRAEGNANT cohort expands these prognostic tools will continue to improve and supplement physician knowledge to improve patient outcomes. Citation Format: Szeto C, Benz S, Nguyen A, Rübner M, Wallwiener D, Tesch H, Hadji P, Fehm TN, Janni W, Overkamp F, Lueftner D, Lux MP, Wallwiener M, Beckmann MW, Huebner H, Ettl J, Hartkopf AD, Mueller V, Taran FA, Belleville E, Schneeweiss A, Soon-Shiong P, Rabizadeh S, Fasching PA. Developing prognostic indicators of poor outcomes in PRAEGNANT metastatic breast cancer cohort [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-20.