PUK28 EMULATING EQ-5D DATA FROM AVAILABLE PROS TO OPTIMIZE CLINICAL TRIAL PLANNING IN CHRONIC KIDNEY DISEASE-ASSOCIATED PRURITUS

Abstract

We aimed to assess the expected sensitivity of EQ-5D-5L in a multi-morbid patient population on hemodialysis (HD) suffering from moderate or severe chronic kidney disease-associated pruritus (CKD-aP) to inform design and sample size calculation for a phase-3-study intended to collect payer-relevant quality of life (QoL) data. Using available patient-level data from a phase-2-study (CR845-CLIN2101) in 174 CKD-aP patients treated with difelikefalin (DFK) 0.5, 1.0, and 1.5 mcg/kg or placebo that collected itch intensity, Skindex-10 and 5D-Itch Scale, we emulated EQ-5D-5L health states by matching related questions from available patient-reported outcomes (PRO) instruments to the relevant EQ-5D dimension and applied the United Kingdom (UK) valuation set. To account for differences in the questions and rating scales used in each instrument, different mapping scenarios were assessed. Treatment with DFK resulted in a statistically significant 8.2 point higher improvement on the Skindex-10 and of 2.5 points on the 5D-Itch scale (both p<0.001 versus placebo). Depending on the mapping applied to emulate EQ-5D-5L health states, this translated into improved QoL scores of 0.1 to 0.21 based on the UK tariff. Emulated QoL scores are similar to those directly measured with EQ-5D-5L in HD patients in other studies, supporting the validity of the results. This emulation increased the confidence in the ability of EQ-5D-5L to detect relevant differences in patients’ utilities when treating CKD-aP, despite potential confounding by the underlying morbidity. While these results do not allow to claim any actual effect on patients’ utility, the quantification of the magnitude of the expected effect helped to appropriately power the future studies in the absence of any analogue cases.