Can we predict inadequate response to allopurinol dose escalation? Analysis of a randomised controlled trial

Abstract

Objectives. To determine factors that predict inadequate serum urate (SU) lowering response in a randomized controlled trial of allopurinol dose escalation (DE) in gout. Methods. Participants undergoing allopurinol DE were classified as: complete response (CR)—reached target SU at month 9 and 12 of the DE phase or if still dose escalating at month 9 reached target SU by month 12; partial response (PR)—reached target at some stage but not fulfilling criteria for CR; or inadequate response (IR)—did not reach target SU at any time. Results. IR was uncommon, occurring in 13/150 (8.7%), compared with 82 (54.7%) CR, and 55 (36.6%) PR. Mean (S.E.M.) SU was higher at the end of the 12-month DE in IR compared with both CR and PR groups; 7.6 (0.31) vs 5.01 (0.06) and 5.97 (0.17) mg/dl respectively (P < 0.001). In univariate analysis, compete responders tended to be older, be receiving less allopurinol, have longer gout duration and were more likely to be New Zealand (NZ) European ethnicity, compared with IR+PR. Using multi-variate logistic regression analysis, only longer duration of gout and NZ European ethnicity remained significant independent predictors of CR. Baseline SU ≥ 8 mg/dl had a sensitivity of 69.2% and specificity of 85.1% in predicting IR. The odds ratio for an IR if baseline SU was ≥ 8 mg/dl was 11.7 (95% CI 3.3, 41.2). Conclusion. A minority of people with gout never reach target SU when allopurinol dose is increased in a treat-to-target manner. Approximately one-third of those with SU ≥ 8mg/dl despite allopurinol ≥300mg/d have an IR to DE.