br Estrone sulphate E S
Estrone-sulphate (E1S) was determined using 300 μl of plasma sam-ples that were pre-processed by adding 650 μl of acetonitrile and 20 ng of the internal standard (d4-E1S). E1S was measured by Thermo Scien-tific TSQ Vantage system (Thermo Scientific, Breda, The Netherlands) equipped with a HESI-2 ion source in the negative ion mode (ion spray voltage was 3500 V, vaporizer temperature 380 °C, capillary tem-perature 320 °C and S-lens RF 50) in a targeted SIM (selective ion mon-itoring) mode. E1S at m/z 349.1 and d4-E1S at m/z 353.1. The reversed phase column Acclaim 120 (C18, 3 μm, 2.1 × 150 mm; Thermo Scientific, Breda, The Netherlands) was used at a flow rate of 200 μl/min and a lin-ear gradient from 20% B to 80% B in 10 min. The mobile phase was water with 0.05% ammoniumhydroxide (solvent A) and acetonitrile with 0.05% ammoniumhydroxide (solvent B). At the retention time of six mi-nutes the analytes of interest eluted. 5 μl of standards or samples were injected. The isotope ratio was calculated from the area A349 (E1S) and A353 (d4-E1S). The correction factor f1 was determined for the pure E1S (A353/A349) and f2 for d4-E1S (A349/A353). The calculation of the mass ratios resulted in the following formula: (A349 − f1
2.3. Image analysis on CT scans
Diagnostic abdominal contrast-enhanced Computer Tomography (CT) scans (n = 20) were evaluated for assessment of abdominal fat volume. Using the software iNtuition (TeraRecon Inc.; San Mateo, CA, USA), cross-sectional images were analyzed consecutively from the upper right 28094-15-7 to L5/S1-level, using a semi-automated method for volumetric quantification of abdominal fat . This method is based on segmentation of pixels with values for Hounsfield units (HU) corresponding to adipose tissue (−195 to −45 HU) . The correct segmentation between visceral and subcutaneous fat compartments was adjusted by the operator if necessary. Both the visceral abdominal fat volume (VAV; cm3) and the subcutaneous abdominal fat volumes (SAV; cm3) were estimated, and the sum of these was the total abdom-inal fat volume (TAV; cm3). The percentage of visceral fat was calculated ([VAV / TAV] × 100; VAV%).
2.4. Gene expression analysis
For both microarray and RNA sequencing RNA was extracted from fresh frozen tissue from primary EC tumors using the RNeasy Mini Kit (Qiagen). Gene expression was analyzed using microarrays for 256 samples as previously described. The samples were hybridized to Agilent Whole Human Genome Microarrays 44 k (Cat. No G4112F),
scanned and normalized as previously described [24,25]. cDNA libraries were prepared and sequencing performed by Illumina HighSeq 4000 (paired end, 75 bp). Raw RNA-Seq reads were aligned to human ge-nome hg19 using hisat 2.0.5 with Gencode v26 transcriptome reference. Aligned files were processed using Samtools. Furthermore, reads aligned in the coding regions of the genome were counted using Feature Counts. Finally, read counts were normalized using DESeq2, then nor-malized expression values were subjected to differential analysis (mean based fold change) and statistical testing using the Students t-test in the R/Bioconductor programming environment. Gene set enrich-ment analysis (GSEA) (http://www.broadinstitute.org/gsea)  was performed using the Molecular Signatures Database (MSigDB, version 5.1) dataset C2 (Curated gene sets) (www.broadinstitute.org/gsea/ msigdb).
2.5. Statistical analysis
Statistical analysis were performed using the software package SPSS
24.0 (SPSS Inc., Chicago, IL), and R-studio. Probability of b0.05 was de-fined as statistical significant. For categorical variables, the Pearsonχ2 or Fisher's exact test was used to evaluate associations between groups, and for continuous variables the Mann-Whitney U test was used. Corre-lations were assessed by Spearman's rank correlation (ρ = rho). Univar-iate analyses were done using the Kaplan-Meier (product-limit) method. Entry date was the date of primary surgery, and time to death due to EC was the endpoint (disease specific survival). Survival between groups was compared using the log-rank test (mantel-cox). All analyses were done blinded for patient characteristics.
3.1. Plasma concentrations of progesterone (P4) and progesterone metabo-lites are increased in patients with long survival compared with patients with short survival
The plasma concentration of nineteen steroids including estrogens (E2, E1, E1S), progestogens (P4, 17OH-P4, 21OH-P4, P5, 17OH-P5), an-drogens (T, DHT, A4, AN), glucocorticoids (cortisol, cortisone, ALDO, 11-deoxycortisol, CORT) and adrenal precursors (DHEA, DHEAS) was analyzed in 38 postmenopausal patients with EC. Of these patients, 19 had short survival, and died due to the cancer within three years after they were diagnosed, while 19 were long time survivors. The two groups were matched for FIGO stage, histologic type and grade, age, BMI and parity (Table 1). Although patients were not matched for myometrial infiltration or hormone receptor status, there was no statis-tical difference in these parameters between the two groups, however a tendency towards more patients with deep myometrial infiltration (N50%) and loss of ER or AR was observed for the group with short sur-vival. Differences in circulating steroid levels between groups, analyzed by Mann-Whitney U test, revealed that five compounds had a signifi-cantly different concentration in the patient group with short survival compared with the patient group with long survival (Table 2). The con-centrations of P4, of the adrenal precursors DHEAS and DHEA as well as the corticosteroid precursor 21OH-P4 were significantly higher in the patient group with long survival compared with the group with short survival (Fig. 1). With regard to circulating estrogens, no significant dif-ference in the concentration of unconjugated compound was seen, but E1S was increased in the group with long survival compared with the group with short survival (Fig. 1C and D). When analyses were re-stricted to patients with endometrioid only or non-endometrioid only EC, high levels of circulating AN, DHEA, A4 and E1 were associated with long survival, however the number of subjects included in these sub-analyses were very small (Supplementary Table S1). Steroid levels in blood were associated with other patient characteristics; DHEA and DHEAS levels inversely correlated with age (Supplementary Fig. S1),