Xanthine oxidase (XO) expression is increased in the failing heart, and animal studies in rodents and dogs showed that XO inhibition with allopurinol can improve left ventricular (LV) function and myocardial oxygen efficiency in the failing heart. The purpose of this study was to determine whether chronic XO inhibition by allopurinol or febuxostat, an investigational, potent non-purine, selective inhibitor of XO, could prevent or treat the progression of congestive heart failure (CHF) induced by coronary artery ligation in rabbits, a species that exhibits low intrinsic XO activity similar to humans. One day after coronary ligation, rabbits were assigned to one of four groups (n = 7-8/group): control group (vehicle for 49 days), early treatment (prevention) group (febuxostat for 49 days), and two delayed-treatment groups (vehicle for 21 days followed by either febuxostat or allopurinol for 28 days). An echocardiogram of the LV was obtained on Days 0 (prior to surgery), 21, and 49. Control rabbits developed CHF by Day 21 (significant reduction in LV shortening fraction and ejection fraction, thinning of the LV posterior wall, and increases in LV internal dimensions and end-diastolic volume). Early preventive treatment with febuxostat significantly lessened the reduction of LV function when compared to vehicle on both Days 21 and 49. These cardiac functional improvements were accompanied by moderately less severe changes in LV dimensional parameters relative to vehicle controls. In contrast, when treatments with XO inhibitors were started after the establishment of CHF, no significant relative improvements in cardiac functional or dimensional parameters were observed. These results suggest that chronic preventive treatment with an XO inhibitor initiated shortly after myocardial infarction can delay or prevent the onset of CHF, and that XO inhibition initiated after establishment of the disease does not offer cardiac protection. In contrast to previous rodent studies which do suggest a cardiovascular (CV) benefit of delayed XO inhibition, the results of this rabbit study are in keeping with those of recently completed studies in severe CHF patients treated with oxypurinol, the active metabolite of allopurinol, in which no clinical benefit was observed. This may be due to the fact that rodents have relatively high levels of XO activity, while the levels in rabbits and humans are intrinsically low, suggesting that the rabbit may be the preferred model for investigating the role of XO in CV diseases. (c) 2007 Elsevier Inc. All rights reserved.