OR25-3 Toward a Circulating Marker of Hepato-Visceral Fat Excess: S100A4 in Adolescent Girls with Polycystic Ovary Syndrome

S100A4 has recently been identified as an adipokine that can serve as a marker of "pernicious" subcutaneous adipose tissue (Arner et al, Int J Ob 2018). However, S100A4 can also be released by an inflammatory subset of macrophages in the liver (Osterreicher et al, Proc Natl Acad Sci USA 2011). Development of polycystic ovary syndrome (PCOS) in adolescent girls is often driven by an early excess of fat in subcutaneous adipose tissue, accompanied by an excess of fat in ectopic locations, notably in viscera and liver (Cree-Green et al, Obesity 2016; de Zegher et al, Trends Endocrinol Metab 2018). Adolescent PCOS phenotypes are normalized more by preferential reduction of hepato-visceral adiposity with a low-dose combination of spironolactone, pioglitazone and metformin (SPIOMET) than by treatment with an oral contraceptive (OC) (Ibanez et al, J Adolesc Health 2017). Nowadays, the assessment of hepato-visceral adiposity (by magnetic resonance imaging) is rather cumbersome. In an attempt to identify a circulating marker of hepato-visceral fat excess, we studied whether S100A4 concentrations are elevated in adolescent girls with PCOS and, if so, whether they tend to normalize with OC or SPIOMET treatment. In 12 healthy controls and in 51 adolescent girls with PCOS (by International Consortium Update criteria; Ibanez et al, Horm Res Pediatr 2017), we assessed circulating S100A4 along with endocrine markers (including fasting insulin, SHBG, testosterone [by LC-MS/MS], high-molecular-weight adiponectin), body composition (by dual X-ray absorptiometry), and abdominal fat partitioning (by magnetic resonance imaging). Girls with PCOS were randomized for treatment with an OC (n=27) or SPIOMET (n=24) for 1 year, and the changes in their study endpoints were analysed. S100A4 concentrations were on average 71% (or 2.1 Z-score) higher in girls with PCOS than in controls (P<0.001), and they associated with hepato-visceral adiposity (r=0.47; p=0.001). Treatment with OC failed to lower the elevated S100A4 concentrations, whereas treatment with SPIOMET lowered circulating S100A4 toward normal (P<0.001 for within-group changes, and P<0.01 for between-group changes); on-treatment reduction of circulating S100A4 associated closely with reduction of hepato-visceral adiposity (r=0.50; p<0.0001). If the present findings are corroborated in other study cohorts with PCOS, then S100A4 is a first candidate-member of a future set of circulating markers that may simplify the diagnosis and follow-up of patients with PCOS, and perhaps also of patients with other phenotypes of central obesity.