STAR and AKR1B10 are down-regulated in high-grade endometrial cancer.

ABSTRACT

Endometrial cancer is the most frequent gynecological malignancy in the developed world. The majority of cases are estrogen dependent, and are associated with diminished protective effects of progesterone. Endometrial cancer is also related to enhanced inflammation and decreased differentiation. In our previous studies, we examined the expression of genes involved in estrogen and progesterone actions in inflammation and tumor differentiation, in tissue samples from endometrial cancer and adjacent control endometrium. The aims of the current study were to examine correlations between gene expression and several demographic characteristics, and to evaluate changes in gene expression with regard to histopathological and clinical characteristics of 51 patients. We studied correlations and differences in expression of 38 genes involved in five pathophysiological processes (i) estrogen-stimulated proliferation; (ii) estrogen-dependent carcinogenesis; (iii) diminished biosynthesis of progesterone (iv) enhanced formation of progesterone metabolites; and (v) increased inflammation and decreased differentiation. Spearman correlation coefficient analysis shows that expression of PAQR7 correlates with age, expression of SRD5A1, AKR1B1 and AKR1B10 correlate with body mass, while expression of SRD5A1 and AKR1B10 correlate with body mass index. When patients with endometrial cancer were stratified based on menopausal status, histological grade, myometrial invasion, lymphovascular invasion, and FIGO stage, Mann-Whitney U tests revealed significantly decreased expression of STAR (4.4-fold; adjusted p=0.009) and AKR1B10 (9-fold; adjusted p=0.003) in high grade versus low grade tumors. Lower levels of STAR might lead to decreased de-novo steroid hormone synthesis and tumor differentiation, and lower levels of AKR1B10 to diminished elimination of toxic electrophilic carbonyl compounds in high-grade endometrial cancer. These data thus reveal the potential of STAR and AKR1B10 as prognostic biomarkers, which calls for further validation at the protein level.