T h e authors have previously proposed that the in vitro displacement of urate from its albumin binding site by an acidic anion was highly indicative of potential uricosuric activity by that anion when administered to humans (1). T h e original experiments on which this hypothesis was based utilized an equilibrium dialysis method to measure urate binding and a fluorescence-quenching technique to detect displacement of 5-dimethylamino-naphthalene-1-sulfonamide (DNSA) from binding sites on human serum albumin T h e recent development of a continuous ultrafiltration technique for measuring urate binding under more physiologic conditions has permitted a reexamination of urate displacement from serum albumin in vitro by certain drugs (3). T h e displacement of sodium (2). urate by uricosuric agents from binding sites on human serum albumin (HSA) and normal serum was thereby demonstrated under in vitro physiologic conditions (pH 7.4, ionic strength 0.16). At 37OC, 0.5 mM sodium salicylate reduced urate bound to 5 g per 100 ml HSA from 19.6% to 9.7%. At room temperature (22.5OC) 22.6y0 urate was bound to 5 g per 100 ml HSA; this was reduced to 12.3y0 by 0.5 mM salicylate, to 9.2y0 by 0.5 mM phenylbutazone, to 14.67, by 0.2 mM diflumidone, and to 17.9a/, by 0.2 mM sulfaethyladiazole (Figure 1). At 22.5"C pooled normal human serum containing 4.8 g per 100 ml albumin bound
23.5% urate, and comparable displacement by drugs was demonstrated (4).During the screening of numerous novel compounds for in vitro urate displacing properties, four drugs or their metabolites were identified as potentially useful new uricosurics (5). In short-term clinical trials these four agents (sulfaethylthiadiazole, diflumidone sodium, W-2354, and halofenate) were all clearly uricosuric, lowering serum urate concentrations and significantly increasing the renal clearance of urate (6).Halofenate was particularly interesting. It is now widely recognized that many patients with gout have Type IV hyperlipoproteinemia characterized predominantly by elevated serum triglyceride concentrations (7). Moreover this association is only partly explained by concomitant obesity or heavy alcohol