Loop diuretics enhance the secretion of prostacyclin in vitro, in healthy persons, and in patients with chronic heart failure

Background: Previous studies suggest that the acute haemodynamic effects of loop diuretics are due to a direct dilation of blood vessels and are not r elated to diuretic properties, but possibly to prostaglandin secretion. Objectives: We investigated whether in vitro human endothelial and renal ep ithelial cells responded to torasemide or furosemide with enhanced secretio n of the vasodilator prostaglandin prostacyclin (PGI(2)). We also investiga ted the effects of loop diuretics on plasma concentrations of PGI(2) and it s physiological antagonist thromboxane after 25 min of administration of dr ugs in 44 patients with congestive heart failure (CHF) and 44 healthy volun teers. Methods: The PGI(2) levels were measured after extraction in ethyl acetate by RIA as levels of 6-KetoPGF(1 alpha), a stable metabolite from a non-enzy matic degradation. TxB2 concentration, the stable hydrolysis product of TxA 2, was also measured by RIA. Results: In human endothelial and renal epithelial cells, both loop diureti cs induced an increase of 6-KetoPGF(1 alpha) secretion that reached a peak after about 5 min and remained stable for 30 min of exposure to the drugs. The magnitude of the phenomenon was lesser in epithelial than in endothelia l cells. Moreover, in both cell lines, there was a significantly higher sec retion of 6-KetoPGF(1 alpha) to torasemide than furosemide (P < 0.05). Conc entrations of 6-KetoPGF(1 alpha) at baseline were similar between the group s of CHF patients receiving the two different drugs. After 25 min of both d rugs, 6-Keto-PGF(1 alpha) significantly increased (P < 0.01), and this was significantly higher in patients treated with 10 mg of torasemide (P < 0.05 vs furosemide). Levels of PGI(2) at baseline were lower in healthy control s than those reached by CHF patients and similar between groups. After 25 m in of both drugs, PGI(2) plasma levels were significantly increased (P < 0. 01). Baseline values of TxB2 were significantly higher in CHF patients comp ared with controls (P < 0.01 vs respective groups), and, more importantly, furosemide but not torasemide increased TxB2 levels in patients and control s (P < 0.05 vs baseline). Conclusions: Our study is the first demonstration in human tissue of increa sed secretion of PGI(2) both in vitro and in vivo, after torasemide or furo semide administration. This phenomenon, which may explain in part the vasod ilatory effects of these drugs, was more evident with torasemide and was re ached at lower concentrations of the drug. Accordingly, we also found that furosemide but not torasemide stimulated the release of the PGI(2) physiolo gical antagonist thromboxane in CHF patients and healthy controls.