BACKGROUND: Hyperuricemia is common in chronic heart failure (CHF), and it is a strong independent marker of prognosis. Upregulated xanthine oxidase (XO) activity and impaired renal excretion have been shown to account for increased serum uric acid (UA) levels in CHF. Therapeutic interventions with allopurinol to reduce UA levels by XO inhibition have been shown to be beneficial. Discussions are ongoing whether UA itself is actively involved or it is a mere marker of upregulated XO activity within CHF pathophysiology. Therefore, the aim of this study was to test the effect of lowering UA by uricosuric treatment without XO inhibition on hemodynamic and metabolic characteristics of CHF. Impaired renal excretion of UA was taken into account. METHODS AND RESULTS: Serum UA (SUA), urinary UA (uUA) excretion, and renal clearance test for UA (Cl(UA)) were measured in 82 patients with CHF. SUA was significantly increased compared with controls of similar age (control, 5.45+/-0.70 mg/dL; New York Heart Association I, 6.48+/-1.70 mg/dL; New York Heart Association II, 7.34+/-1.94 mg/dL; New York Heart Association III, 7.61+/-2.11 mg/dL; P<0.01). Patients with CHF showed lower uUA excretion and Cl(UA). On multivariate analysis, insulin, brain natriuretic peptide (P<0.01), and creatinine levels (P=0.05) showed independent correlation with SUA. The treatment effect of the uricosuric agent benzbromarone was tested in 14 patients with CHF with hyperuricemia in a double-blind, placebo-controlled, randomized crossover study design. Benzbromarone significantly decreased SUA (P<0.01). Brain natriuretic peptide, left ventricular ejection fraction, and dimensions in echocardiographic assessment did not change after benzbromarone therapy. In contrast, fasting insulin (placebo, 18.8+/-8.9 microU/mL; benzbromarone, 11.0+/-6.2 microU/mL; P<0.05), homeostasis model assessment of insulin resistance index (placebo, 5.4+/-2.6; benzbromarone, 3.0+/-1.7; P<0.05), and tumor necrosis factor-alpha (placebo, 2.59+/-0.63 pg/mL; benzbromarone, 2.14+/-0.51 pg/mL; P<0.05) improved after benzbromarone, and the changes in tumor necrosis factor-alpha levels were correlated with reduction of SUA (P<0.05). CONCLUSIONS: These results show that UA lowering without XO inhibition may not have an effect on hemodynamic impairment in CHF pathophysiology. To the extent that these data are correct, this finding suggests that upregulated XO activity rather than UA itself is actively involved in hemodynamic impairment in CHF. Clinical Trial Registration- clinical trials.gov. Identifier: NCT00422318.