Generalized dilation of the visceral microvasculature by peritoneal dialysis solutions.

OBJECTIVES: Conventional peritoneal dialysis solutions are vasoactive. This vasoactivity is attributed to hyperosmolality and lactate buffer system. This study was conducted to determine if the vasodilator property of commercial peritoneal dialysis solutions is a global phenomenon across microvascular levels, or if this vasodilation property is localized to certain vessel types in the small intestine. DESIGN: Experimental study in a standard laboratory facility. INTERVENTIONS: Hemodynamics of anesthetized rats were monitored while the terminal ileum was prepared for in vivo intravital microscopy. Vascular reactivity of inflow arterioles (A1), branching (A2), and arcade, as well as pre-mucosal (A3) arterioles was assessed after suffusion of the terminal ileum with a non-vasoactive solution or a commercial 4.25% glucose-based solution (Delflex; Fresenius USA, Ogden, Utah, USA). Vascular reactivity of three different level venules was also assessed. Maximum dilation response was obtained from sequential applications of the endothelial-dependent dilator, acetylcholine (10(-5) mol/L), and the endothelial-independent nitric oxide donor, sodium nitroprusside (NTP; 10(-4) mol/L). RESULTS: Delflex induced an instant and sustained vasodilation that averaged 28.2% +/- 2.4% of baseline diameter in five different-level arterioles, ranging in size between 7 mu and 100 mu. No significant vascular reactivity was observed in three different-level venules. Delflex increased intestinal A1 blood flow from baseline 568 +/- 31 nL/ second to 1,049 +/- 46 nL/sec (F= 24.7, p< 0.001). Similarly, intestinal venous outflow increased to 435 +/- 17 nL/sec from a baseline outflow of 253 +/- 59 nL/sec (F= 4.7, p < 0.05). Adjustment of the initial pH of Delflex from 5.5 to 7.4 resulted in similar microvascular responses before pH adjustment. CONCLUSIONS: Ex vivo exposure of intestinal arterioles to conventional peritoneal dialysis solutions produces a sustained and generalized vasodilation. This vasoactivity is independent of arteriolar level and the pH of the solution. Dialysis solution-mediated vasodilation is associated with doubling of A1 intestinal arteriolar blood flow. Addition of NTP at an apparent clinical dose does not appear to produce any further significant arteriolar dilation than that induced by dialysis solution alone. Experimental data that estimate the exchange vessel surface area per unit volume of tissue will be required to make a correlation with permeability in order to extrapolate our findings to clinical in vivo conditions.