br The progestogen OH P
The progestogen 21OH-P4 (also known as deoxycorticosterone), el-evated in patients with long survival, is converted into CORT  and ALDO in the adrenal gland, but the absence of the enzymes involved in these conversions (CYP11B1 and CYP11B2) indicates that this route in corticosteroid metabolism is absent in EC. Similarly, intracrine forma-tion of 11DOC and 9(S)-HODE from 17OH-P4 seems negligible in EC due to the absence of the responsible enzymes (CYP21A1 and CYP11B1). How-ever, HSD11B1 and HSD11B2, controlling the balance between active cortisol and inactive cortisone, are present in EC as shown earlier . The role of these steroids in EC is poorly studied in previous investigations.
In this study we investigated whether a difference in gene expres-sion could be observed between patients with long and short survival in a cohort with otherwise similar characteristics. Several genes were differentially expressed between the two groups, and for many, such differences were also validated in a larger cohort. Some of these genes have earlier been linked to cancer and should be further explored as po-tential biomarkers in endometrial cancer [39–41]. As E2 drives prolifer-ation in endometrial cancer we also explored whether high estradiol levels were linked to a specific gene expression profile in tumor. We ob-served an enrichment of gene sets associated with estrogen related sig-naling in tumors from patients with high plasma levels of E2 compared to patients with low levels, indicating that high E2 concentration in blood contribute to increased transcription of ER regulated genes in the tumor.
This study was designed to investigate whether differences in blood (circulating) steroids, tumor gene expression and local steroid metabo-lism could be observed between two similar patients groups with differ-ent survival times. We identified 19 patients in our large cohort that fulfilled the inclusion criteria for the short survival group, and these were matched with 19 patients with long survival. This highly selected, small population is not representative for the whole population of EC patients. Also, the patients in this study were matched for specific criteria (FIGO stage, histologic type and grade, age, BMI and parity) but not for depth of myometrial infiltration or ERα and AR status. Al-though these parameters did not differ significantly between groups,
Fig. 2. Systemic steroids and local metabolism in EC. A. The levels of all measured steroids in the blood are shown in the left panel (blood steroids). Pico molar rages are indicated in red, nano molar ranges in green and micro molar ranges are indicated in blue. Mean concentration values of all 38 subjects are shown. At the right, the potential steroid metabolism and conversion are indicated. The mean expression level of each gene is indicated by the color (based on the heat map and levels in Supplementary Fig. S2). The sulfatase pathway consisting of STS and SULTs is not shown. *Adiol = androstenediol and androstanediols were not measured in this study. B. Heat map of the levels of the various analyzed steroids in patients with long compared to short survival.
patients with short survival tended to have deeper myometrial infiltra-tion, and loss of ERα or AR. Both the small population, and the patient selection supports that our findings should be validated in a large pop-ulation based series to further investigate the role of blood steroids in endometrial cancer patients.
GSEA (curated gene sets) of genes upregulated in patients with high level of estradiol.
Ranka Gene set name (selected gene sets) Genes in P-value FDR
a Complete list is given in Supplementary Table 3.
Despite being a hormone dependent cancer, there is limited knowledge regarding the relation between level of steroids in blood and prognosis for endometrial cancer (EC) patients. In this study we have identified steroids that might be promising biomarkers in blood samples from patients with EC, and we also ob-serve an interesting association between E2 levels in blood and fat distribution. The sample size in this study is however limited and our findings should be validated in a larger population based patient cohort.
Supplementary data to this article can be found online at https://doi.
We thank Kadri Madisoo, Ellen Valen, Ann-Helen Pridesis and Karlijn Cornel for technical assistance.
The Genomics Core Facility (GCF) at the University of Bergen, which is a part of the NorSeq consortium, provided services on high through-put sequencing and bioinformatics analysis; GCF is supported in part by major grants from the Research Council of Norway (grant no. 245979/F50) and Bergen Research Foundation (BFS) (grant no. BFS2016-genom).