Interaction of diuretics and non-steroidal anti-inflammatory drugs in man.

1. The influence of four diuretics on renal prostaglandins was investigated in a study designed in two parts (A and B): A, 24 normal subjects on a constant sodium intake received frusemide (80 mg daily), or hydrochlorothiazide (100 mg), or triamterene (200 mg) or spironolactone (300 mg); B, the same subjects were pretreated for 3 days with indomethacin (150 mg daily), which was continued during the 3 day administration of the respective diuretics and during a 2 day post-diuretic period. 2. In study A, only triamterene provoked a rise in urinary prostaglandins E2 and F2 alpha (+ 474 +/- SEM 92%, P less than 0.01, and + 192 +/- 7%, P less than 0.01). In study B, prostaglandins were significantly inhibited in all subjects. After indomethacin, the natriuretic effect of frusemide and spironolactone was reduced by 80 +/- 12% (P less than 0.01) and 54 +/- 11% (P less than 0.001), whereas the natriuresis induced by hydrochlorothiazide and triamterene was unchanged. No correlation was found between urinary PGE2 and F2 alpha and natriuresis. 3. When triamterene was associated with indomethacin, two subjects developed reversible acute renal failure. 4. Plasma renin activity and urinary aldosterone were stimulated by the four diuretics in study A, but their response was blunted in study B. Urinary antidiuretic hormone was not modified by diuretics but was suppressed by indomethacin. 5. Diflunisal, a structurally unrelated nonsteroidal anti-inflammatory drug, given to 12 of the subjects provoked similar interactions with frusemide, hydrochlorothiazide and spironolactone. 6. The results suggest that prostaglandins contribute to the natriuretic effects of frusemide and spironolactone, but not to those of hydrochlorothiazide and triamterene.