Indirect treatment comparison of odevixibat and maralixibat for the treatment of cholestatic pruritus in patients with Alagille syndrome

The ASSERT/ASSERT-EXT and ICONIC trials provide supportive evidence of benefit for the use of IBAT inhibitors in patients with ALGS, which has ultimately led to the approval of ODX and MRX for cholestatic pruritus [11–16]. There are no head-to-head data or robust ITC analyses comparing efficacy, safety, and HRQoL outcomes in patients with ALGS treated with ODX versus MRX. Therefore, this is the first analysis applying regression-adjusted matching; an established, published and recommended methodology in current ITC guidelines [19,20], which may help to inform clinical decision-making between ODX and MRX in patients with ALGS.

Population-adjusted ITC methods are used when trials evaluating different treatments include clinically heterogeneous populations, rendering naive comparisons biased and unreliable [19,20]. By adjusting for differences in key effect-modifying baseline characteristics, these methods aim to account for cross-trial heterogeneity and improve the validity of comparative effectiveness estimates. In doing so, they help to ensure that any observed differences in outcomes are more likely attributable to the treatments themselves rather than to underlying differences in patient populations. This strengthens the credibility of evidence for informed clinical decision-making. As it was not possible to adjust for all clinically relevant prognostic factors and effect modifiers using a conventional MAIC approach, an established method, also referred to as a regression-adjusted matching approach, was employed. This approach helps to mitigate residual bias arising from incomplete adjustment by combining two complementary adjustment strategies: weighting and outcome modeling regression techniques. Applying this method enabled adjustment of the ASSERT/ASSERT-EXT population to better align with the ICONIC population, with respect to key prognostic factors and effect modifiers, thus supporting the overall validity of this ODX and MRX comparison and its use to inform treatment decision-making in ALGS.

Overall, ODX demonstrated consistently greater improvements in AM pruritus scores, pruritus response rates and sBA reductions at week 48 versus MRX in patients with ALGS. However, results were not statistically significant and were associated with wide CIs, indicating uncertainty in a treatment effect difference between MRX and ODX. The analysis of PedsQL total score demonstrated that ODX also significantly enhanced patient HRQoL versus MRX, suggesting that clinical improvements with ODX are associated with better HRQoL from the patient perspective. Finally, safety results showed that ODX was overall better tolerated than MRX, with patients experiencing significantly lower rates of GI AEs and abdominal pain; there was also a trend toward fewer SAEs and less vomiting and diarrhea. Taken together, the present data build on the robust, pooled long-term ASSERT/ASSERT-EXT data, which demonstrate that ODX was associated with rapid and significant improvements in pruritus, significant reductions in sBA levels, and a well-tolerated safety profile from baseline up to 96 weeks of treatment in patients with ALGS [21].

GI AEs are the most common side effect with the IBAT inhibitor class, likely due to the increased delivery of bile acids to the colon following inhibition of the apical sodium-dependent bile acid transporter [12,16]. In this ITC analysis, there were significantly lower rates of GI AEs and abdominal pain with ODX versus MRX, and a trend for fewer SAEs with ODX. In ASSERT, no patients had an AE leading to discontinuation whereas [12] in ICONIC, one patient (3%) had AEs leading to discontinuation during the stable dosing period [16]. These findings are consistent with a lower incidence of GI AEs in patients treated with ODX than those treated with MRX, which is clinically relevant, particularly to treatment persistence, overall tolerability and daily disease burden. The GI AEs observed with ODX were generally manageable and rarely led to treatment discontinuation. Thus, despite the limited data on dose-target engagement relationships, ODX (120 μg/kg) appears to achieve greater efficacy than MRX (380 μg/kg), with improved pruritus relief and reductions in sBA levels, yet fewer GI side effects. One mechanistic hypothesis for the observed differences in GI tolerability involves potential differences in bile acid metabolism and microbiome effects; however, these mechanisms were not investigated in the present analysis. In the colon, bacterial conversion of primary to secondary bile acids strongly stimulates TGR5 receptors, thereby promoting intestinal motility, fluid secretion, and faster GI transit [22,23]. MRX has been associated with increased secondary bile acids, whereas these remain largely undetectable with ODX [24,25]. The resulting lower TGR5 stimulation may contribute to the more favorable GI tolerability observed with ODX. This, in turn, may be driven by the distinct physicochemical properties of the compounds, with ODX being a reversible acid with pH dependency and MRX being a permanent cationic base [10,14]; however, these differences were not investigated during this analysis and further research is needed to confirm this potential mechanism.

In this ITC analysis, it was demonstrated that patients had significantly improved HRQoL with ODX versus MRX (per PedsQL total score, which assesses physical, emotional, social and school functioning domains). This significantly improved HRQoL is likely due to improvements in pruritus severity, reduced sleep disturbance and a lower GI AE burden. Pruritus is often cited as the most debilitating symptom of ALGS [3,4] and together with GI tolerability, appears to play a key role in shaping patients' daily functioning and overall HRQoL. In line with this, EASL guidelines recommend the use of IBAT inhibitors (ODX or MRX) for the treatment of cholestatic pruritus in patients with ALGS [7]. Taken together, these ITC data provide supportive evidence of the comparative efficacy and safety of the two IBAT inhibitors available for the treatment of ALGS, and provide guidance for clinicians to make informed treatment decisions together with their patients.

Despite the methodologic advantages of population-adjusted ITCs in reducing bias, several inherent limitations warrant careful consideration when interpreting these findings. First, the differences in study designs of ASSERT/ASSERT-EXT and ICONIC are important to note. ASSERT was the only Phase III, randomized, double-blind, placebo-controlled trial in ALGS, while ICONIC was a placebo-controlled, randomized drug withdrawal period, Phase IIb study [11,16,21], meaning that the anchoring to placebo for MRX was not possible. The lack of anchor requires strong assumptions, including that absolute outcomes can be predicted from observed covariates, and that all relevant prognostic factors and effect modifiers are accounted for. Second, the absence of individual patient data from ICONIC introduces the potential for residual confounding due to unmeasured variables. Third, only two studies were included in the analysis, limiting the ability to reliably assess heterogeneity and, by extension, the consistency between sources of evidence (i.e., agreement across studies), while increasing the likelihood that results are driven by a single study. In addition, the indirect method used for comparison of pruritus scores is acknowledged as a limitation. Finally, both studies had small sample sizes; although this is expected in rare diseases, it further reduces the statistical power to detect meaningful differences. To mitigate these limitations in this analysis, all key prognostic factors available at baseline (including age, sex, pruritus, sBA levels and bilirubin levels) were included to minimize imbalance between studies. While this approach cannot replace a head-to-head randomized trial and some residual uncertainty may remain, ITCs are a well-established and widely accepted method to generate comparative evidence when no direct trials are available. When appropriately adjusted and cautiously interpreted, we believe this approach remains supportive of the comparison between the two studies, despite its acknowledged limitations.

A further consideration is whether observed differences in this ITC translate into clinically meaningful benefits. The difference in improvement in HRQoL with ODX versus MRX exceeded established thresholds for clinical relevance in pediatric populations (minimal clinically important difference of ∼4.5 points on the PedsQL total score). This may translate into observable improvements in day-to-day functioning across physical, emotional and social domains and reduced sleep disruption and caregiver burden. In real-world care, these aspects often drive treatment satisfaction, treatment persistence, and shared decision-making, particularly when efficacy end points (pruritus and sBA) appear directionally similar within statistical uncertainty. Further research, including real-world analyses of patients with ALGS who switch between IBAT inhibitors, may help clarify the clinical relevance of these findings in routine practice.

Enrolling participants in studies of a rare disease is challenging; real-world comparative studies may be more feasible than head-to-head trials and, to date, no head-to-head clinical data have evaluated different IBAT inhibitors in patients with ALGS or other cholestatic liver diseases. This is the first analysis to apply a regression-adjusted matching methodology to indirectly compare efficacy, safety and HRQoL outcomes of ODX versus MRX in patients with ALGS. ODX demonstrated a differentiated clinical profile versus MRX, characterized by significant improvements in GI tolerability and patient HRQoL, alongside consistent trends favoring improved efficacy. Although exploratory and hypothesis-generating, this unanchored population-adjusted ITC provides comparative insights between two available IBAT inhibitors in ALGS, and may support informed clinical discussion in the absence of head-to-head evidence.

E Kaltenbach, C Artaud: Conception and design of the study. E Kaltenbach, AG-Diawara, J Schwarzbard, C Artaud: Data acquisition, analysis, or interpretation. E Kaltenbach, AG-Diawara, J Schwarzbard, C Artaud: Drafting or revising the manuscript. E Kaltenbach, AG-Diawara, J Schwarzbard, C Artaud: Providing resources or materials. E Kaltenbach, C Artaud: Supervising the research.

This study was sponsored by Ipsen (Paris, France).

E Kaltenbach and C Artaud are employees and stockholders of Ipsen Pharma. AG-Diawara and J Schwarzbard are employees and stockholders of Ipsen Innovation (SAS). The authors have no other competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript apart from those disclosed.

Medical writing and editorial support were provided by Christopher Lum, PhD of Nucleus Global, an Inizio company, and were funded by Ipsen.

Qualified researchers with a valid research question may request anonymized patient-level data by contacting an Ipsen representative. Further information on Ipsen’s Data Sharing policy is available here (https://www.ipsen.com/clinical-data-transparency/).

This work is licensed under the Creative Commons Attribution 4.0 License. To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/