Abstract
IntroductionBiochemical response endpoints in primary biliary cholangitis (PBC) utilize reductions in alkaline phosphatase (ALP) and total bilirubin (TB) and may be influenced by the distribution of baseline values. In the phase III ELATIVE® trial (NCT04526665), patients with baseline ALP ≥ 3x the upper limit of normal (ULN) required a ≥ 44% reduction from baseline to meet the ALP threshold component of the primary endpoint (ALP < 1.67 x ULN), compared with a ≥ 17% reduction in those with baseline ALP ≤ 2 x ULN. ALP levels can be highly variable in patients with PBC. Here, we explored the impact of baseline ALP levels on biochemical response and ALP normalization in ELATIVE®.MethodsIn ELATIVE®, patients with unstable ALP levels during the screening period were not eligible to participate. The primary endpoint was biochemical response, defined as ALP < 1.67 x ULN, with a ≥ 15% reduction from baseline, and normal TB. The first key secondary endpoint was ALP normalization. Here, these endpoints were analyzed at Week 52 in subgroups of patients in ELATIVE® with a range of baseline ALP levels (≤ 2, 2 – ≤ 2.5, 2.5 – ≤ 3, 3 – ≤ 4, and > 4 x ULN).ResultsBaseline ALP subgroups comprised (elafibranor, placebo): ≤ 2 x ULN (14%, 28%); 2 – ≤ 2.5 x ULN (23%, 26%); 2.5 – ≤ 3 x ULN (23%, 8%); 3 – ≤ 4 x ULN (25%, 17%); and > 4 x ULN (15%, 21%). The primary endpoint was met in 50.9% of patients receiving elafibranor overall, with 86.7% with baseline ALP ≤ 2 x ULN and 18.8% with baseline ALP > 4 x ULN meeting the endpoint. For placebo, the overall biochemical response rate was 3.8% and only occurred in patients with a baseline ALP ≤ 2 x ULN (figure 1A). ALP normalization occurred in 14.8% receiving elafibranor and 0% receiving placebo overall. The majority of patients receiving elafibranor who had ALP normalization had baseline ALP ≤ 2 x ULN (53.3%). However, with elafibranor, ALP normalization also occurred in two patients (12.5%) with baseline ALP > 4 x ULN (figure 1B). A higher proportion of patients receiving elafibranor met either endpoint compared with placebo, regardless of baseline ALP levels, and lower baseline ALP levels yielded higher response rates.Abstract P250 Figure 1Conclusion(s)Patients on elafibranor with lower baseline ALP levels had higher biochemical response and ALP normalization rates versus those with higher baseline ALP. Though important, dichotomous endpoints may limit comparability of trials with varying baseline levels; a higher proportion of patients with low baseline ALP leads to higher response rates in both investigational and placebo groups.Study FundingThis study is funded by Ipsen.