FC090: Stevia as an Alternative Osmotic Agent for Peritoneal Dialysis Fluid

Abstract

BACKGROUND AND AIMS Peritoneal dialysis (PD) is a renal replacement technique that requires repeated contact of the peritoneum with hyperosmolar fluids. Unfortunately, it generates metabolic and biomechanical insults that promote structural and functional alterations of the peritoneal membrane (PM). PM is lined by a mesothelial cells (MCs) monolayer that is in direct contact with the PD fluids (PDFs). Repeated PDF exposure causes a mesothelial–mesenchymal transition (MMT) of the MCs, as well as an inflammatory response, angiogenesis and fibrosis of the peritoneal tissue. For decades, glucose (Gl) has been the most commonly used osmotic agent, and Gl itself is responsible for the aforementioned changes. Furthermore, Gl degradation products (GDPs) contribute to deleterious effects on the PM [1]. Therefore, there is an unmet need for more biocompatible osmotic agents able to better maintain the structure and function of the PM. Here we evaluated the biocompatibility of Stevia (STV), a compound of natural origin with a mixture of glycosides. STV has a molecular weight close to the theoretical ideal weight for an osmotic agent in PD [2] and it was used instead of Gl for PDF. METHODS The osmotic capacity of STV-containing fluids was analysed using an in vitro device with artificial membranes and compared to Gl-based fluids. In order to investigate the STV biocompatibility, human omental peritoneal mesothelial cells (HOMCs) were cultured in the presence of different types of Gl-based PDFs (Conventional 4.25% Gl solution with a high GDP level and biocompatible 2.3% Gl solution with low GDPs level), STV-based fluids or TGF-b for 48 h. Mice submitted to surgery of an intraperitoneal catheter insertion were treated for 40 days with STV- or Gl-based PDFs. Peritoneal tissues were collected to determine thickness, MMT, angiogenesis and inflammatory markers, as well as effluents to analyse infiltrating cells. RESULTS Experimentally, using artificial membranes we demonstrated that 1, 0.75 and 0.5% STV-based fluids have a maintained and prolonged osmotic capacity similar to commercial PDFs containing Gl concentrations between 1.5 and 2.3%. Moreover, 1.5% STV-based fluid showed an osmotic capacity higher than a PDF with Gl 2.3% and reached values similar to Gl 4.25% at longer time points. In addition, HOMCs exposed to STV-based fluids showed higher viability and did not induce MMT. Gl-based fluids induced upregulation of mesenchymal markers (VEGF, fibronectin, snail) and also downregulation of the epithelial marker E-cadherin in the case of the highest concentration of Gl (4.25%). None of these markers were significantly affected by STV treatment. In vivo, STV-based fluids preserved MCs monolayer and induced less PM thickness (Figure 1), angiogenesis, leukocyte infiltration and MMT than Gl-based fluids. Furthermore, STV solutions did not promote the release of pro-inflammatory cytokines by HOMCs nor by cells present in mice parietal peritoneal tissue in contrast to Gl-based PDFs. CONCLUSIONS STV-based fluids showed better biocompatibility than Gl-based PDFs in vivo and ex vivo since STV did not induce PM thickness, MMT, angiogenesis or the release of pro-inflammatory cytokines.