CADTH Reimbursement Review

Maralixibat (Livmarli)

Sponsor: Mirum Pharmaceuticals Inc.

Therapeutic area: Alagille syndrome

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Ethics Review

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Table 1: Background Information of Application Submitted for Review

NOC = Notice of Compliance.

Introduction

Alagille syndrome (ALGS) is a rare, life-threatening, genetic, complex, multisystem disorder that presents with a range of clinical features, including cholestatic liver disease, failure to thrive, cardiovascular disease, skeletal abnormalities, ocular abnormalities, renal and vascular abnormalities, and distinct facial features.^1,2 In most cases, the liver dysfunction associated with ALGS is an early and a serious feature of this genetic condition and typically presents in the first 3 months of life. Elevated levels of serum bile acids (sBAs) and jaundice (elevated bilirubin) are hallmarks of ALGS and indicate the presence of impaired bile flow.³ Clinically important manifestations of cholestasis in ALGS include debilitating and intractable pruritus, disfiguring xanthomas, sleep disturbances, chronic debilitating fatigue, and failure to thrive (i.e., insufficient growth).^4,5 ALGS is predominantly caused by mutations in the Jagged-1 gene in greater than 90% of cases⁶ and in the NOTCH2 gene in approximately 4% of cases, resulting in bile duct paucity, bile flow obstruction, and bile accumulation in the liver. ALGS is inherited in an autosomal-dominant pattern. Cholestasis is defined as a decrease in bile flow due to impaired secretion by hepatocytes or to obstruction of bile flow through intrahepatic or extrahepatic bile ducts.⁷ Cholestasis leads to hepatic and systemic accumulation of substances normally excreted via the biliary tract, such as bile acids and conjugated bilirubin. Elevated levels of bile acids are hepatotoxic and contribute to disease progression. Bile acids have been shown to induce damage and necrosis in hepatocytes and cholangiocytes and are associated with increased morbidity and mortality in chronic cholestatic diseases.⁷ The systemic and hepatic accumulation of bile acids and other toxins leads to incapacitating and chronic cholestatic symptoms such as pruritus, and clinical sequelae.

The clinical manifestations of cholestasis associated with ALGS are severe, even in the absence of liver disease, with cholestatic pruritus being the leading cause of liver transplant in patients.^8-10 These clinical manifestations present in the first few years of life and as early as 3 months of age. The symptoms include severe and unremitting pruritus (74%),¹⁰ xanthomas (disfiguring and sometimes disabling subcutaneous lipid deposits, 40%), chronic fatigue (between 65% and 85%), and growth failure (between 50% and 87%¹¹). In addition, fat-soluble vitamin (FSV) malabsorption and increased risk of bone fractures because of trabeculae malformation can also be present. A second wave of portal hypertension and associated complications occurs later in adolescence.⁶ Collectively, all cholestasis-related symptoms result in poor health-related quality of life (HRQoL). As cholestasis progresses and symptoms worsen, as described previously, a majority of patients will either receive a liver transplant (50.4%) or die (9.3%) by 18 years of age, with only 40.3% of patients reaching adulthood with their native liver.^6,10 The estimated 20-year life expectancy is 75% for patients diagnosed with ALGS, 80% for patients with ALGS who do not require liver transplant, and 60% for patients with ALGS who require liver transplant.¹² For patients with ALGS who undergo liver transplant, the estimated 1-year survival rate is 87%.¹³ The majority of early liver transplantations occur because of complications associated with cholestasis, including pruritus. The reported incidence of ALGS is 1 in 30,000 to 50,000 births.^3,11 In the absence of Canadian statistics, the sponsor estimated prevalence (based on 1 in 30,000 live births) to be a total of 1,032 patients with ALGS in 2023.

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of maralixibat (Livmarli), 9.5 mg/mL, oral solution, for the treatment of cholestatic pruritus in patients with ALGS aged 2 months and older.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient and clinician groups that responded to the CADTH call for input and from clinical expert(s) consulted by CADTH for the purpose of this review.

Patient Input

Two patient groups, the Canadian Liver Foundation (CLF) and the Alagille Syndrome Alliance (ALGSA), provided input. The CLF is the only national health charity committed to supporting people in Canada affected by liver diseases. Based in the US, the ALGSA is a nonprofit organization dedicated to supporting families affected by ALGS globally. The CLF submission included phone and virtual interviews conducted in May 2023 with 8 patients and caregivers in Canada. Of those, 4 respondents had experiences with maralixibat through clinical trials. The ALGSA gathered data online through family surveys (2020), personal conversations, and topic-specific discussions among support or focus groups, including at least 76 members in Canada. Both groups stated that the itchiness (pruritus) is the most bothersome symptom affecting patients’ and caregivers’ lives. For example, the itchiness interrupts patients and families’ sleep, making those affected fatigued, anxious, depressed, irritable, and worried. Patients said they feel isolated at school and that it is challenging to maintain employment. Also, patients and families have difficulty finding a specialist who could recognize and make a proper diagnosis of ALGS and manage disease treatment. Patients and families from both groups want a new therapy that can provide significant relief of itchiness with long-term effects without high risks such as liver transplant and immunosuppression. Patients who have taken maralixibat during clinical trials said that their itchiness has been resolved with minor side effects, such as upset stomach and diarrhea; that they felt more like their true self and were able to engage in normal day-to-day activities; and that their households were also positively affected.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical expert panel stated that cholestatic pruritus remains a very significant management problem for patients with ALGS and their families, due to partial, incomplete, or null response to currently available treatments. Current treatments are used off-label and are supportive in nature. The experts noted that surgical options such as an external or internal biliary diversion can be offered to patients with ALGS with cholestatic pruritus refractory to medical therapies; however, these are not very effective and seldomly used in clinical practice. Finally, the experts stated that between 50% and 75% of patients with cholestatic liver disease will require a liver transplant and that cholestatic pruritus is a leading indication for a transplant. Liver transplant is associated with increased morbidity, mortality, and lifelong immune suppression. As such, the experts noted that there is an unmet need for effective symptomatic and curative treatment for cholestatic pruritus in the indicated patient population.

The clinical experts stated that maralixibat would likely be used in combination with current off-label treatments in patients experiencing ongoing pruritus, and that it is possible some patients could discontinue some of the off-label treatments once they are established on maralixibat and their pruritus is under control. The experts noted that, if easily accessible, maralixibat may be used as an initial therapy for new patients presenting with severe pruritus. The clinical experts stated that the estimated incidence of ALGS in Canada is about 1 in 30,000 to 50,000 live births, with about 200 new cases each year. The experts noted that pediatric patients with ALGS most suited for treatment with maralixibat are those who present with cholestatic pruritus that is persistent with current off-label treatments, which makes up about a third of patients in a clinical expert’s practice. Patients least suited to treatment with maralixibat are those who have minimal liver involvement (i.e., minimal liver enzyme abnormalities and no FSV deficits) and patients who do not experience cholestatic pruritus.

According to the expert panel, a clinically meaningful response to treatment would include a reduction in the frequency and severity of pruritus, a reduction in sleep deprivation among patients and their caregivers, the ability for patients and their caregivers to attend school or work, reduced damage to the patients’ skin, and improved patient weight and growth. The clinical experts consulted on this review noted that response to therapy would likely be evaluated via subjective family reporting of symptoms including itching and sleep disturbances as well as by visual assessments of excoriations on the patient’s skin, which are often indicative of severe cholestatic pruritus. Standard scratch scales are not commonly used in clinical practice, according to the experts. Measurements of sBA could be used to assess response to therapy; however, the experts noted that this is not common in clinical practice due to the high cost and limited availability of such testing in some practice settings. The clinical experts would initially assess patients monthly for approximately 3 months, at which time the frequency of visits would be reduced to every 3 to 6 months if a response to treatment is evident. The clinical experts stated that treatment with maralixibat will likely be lifelong for most patients. The panel noted that treatment discontinuation may be considered if there is no effect on cholestatic pruritus after approximately 6 months of treatment initiation, if a patient’s liver disease progresses and they undergo liver transplant, or due to serious adverse events (SAEs); however, the experts stated that AEs associated with maralixibat are likely self-limited and may be addressed by titrating the dose of maralixibat. The clinical experts noted that a pediatric or adult liver or gastrointestinal specialist would be the preferred specialist to prescribe and monitor treatment with maralixibat.

Clinician Group Input

One clinician from the Canadian Association for the Study of the Liver provided input. The clinician group and 2 clinical experts consulted by CADTH agree on the unmet need, which is a lack of effective therapy specifically indicated for cholestatic pruritus associated with ALGS refractory to current off-label treatments. They also agree that all the existing therapies are not effective at reducing cholestatic pruritus associated with ALGS and that there are no guidelines for treating cholestatic pruritus in patients with ALGS. In alignment with clinical experts, the clinician group stated that treatment goals are mainly improvement in pruritus, improvement in quality of life (i.e., sleep duration), and optimizing nutritional goals (i.e., treating FSV deficiency). Also, both groups agree that patients with ALGS and cholestatic pruritus that is persistent on standard-of-care medical treatment would be an eligible population. The clinician group stated that if a patient’s liver disease progresses and they undergo liver transplant, discontinuation is considered and the clinical experts stated that if there is no effect on itch as measured clinically, then discontinuation is considered after adequate trial — i.e., 6 months. Otherwise, both groups agree that adverse events (AEs) would be an unlikely reason to discontinue since maralixibat is well tolerated. Lastly, all the clinician group and clinical experts agree that maralixibat should be prescribed by a pediatric gastroenterologist or hepatologist. None of the clinician group or clinical experts consulted by CADTH had declared experience with maralixibat.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH Reimbursement Review processes by identifying issues that may affect their ability to implement a recommendation. The drug plans identified implementation issues related to considerations for initiation of therapy and continuation or renewal of therapy. The clinical expert consulted by CADTH weighed evidence from the pivotal trial and other clinical considerations to provide responses to the drug program’s implementation questions. Refer to Table 6 for more details.

Clinical Evidence

Systematic Review

Description of Studies

The pivotal LUM001-304 (ICONIC) trial was an open-label, phase IIb study to evaluate the safety and efficacy of maralixibat in children with ALGS between the ages of 1 and 18 years. A total of 31 patients enrolled into the study, which was conducted at 10 clinical sites in Australia, Europe, and the UK between November 25, 2014, and September 11, 2015. The study comprised an 18-week, open-label run-in period during which all patients received maralixibat, up to 380 mcg/kg/day (0 weeks to 18 weeks); a 4-week, randomized, double-blind, placebo-controlled drug-withdrawal phase (weeks 18 to 22), during which 13 patients continued receiving active treatment while 16 patients shifted to placebo; followed by a 26-week stable-dosing period (weeks 23 to 48), during which all patients received active treatment at doses up to 380 mcg/kg/day; and an optional long-term treatment period. It should be noted that during the long-term extension (LTE) phase (as of week 103) eligible patients could have received a dose of maralixibat of up to 760 mcg/kg/day (given as twice-daily doses of 380 mcg/kg) which is outside of the proposed Health Canada indication of 380 mcg/kg/day. As such, efficacy and safety data after this period is not aligned with the recommended dose. Assessed efficacy outcomes included change in sBA, change in pruritus assessed using the Itch Reported Outcome (ItchRO) observer (Obs) and patient (Pt) tools, change in liver biomarkers and enzymes (alanine transaminase [ALT], alkaline phosphatase [ALP], total and direct bilirubin), change in body height and weight z scores, and HRQoL as measured by the Pediatric Quality of Life Inventory (PedsQL) total score (parent) and the PedsQL Multidimensional Fatigue Scale score (parent). Assessed harms included AEs such as diarrhea, abdominal pain, and FSV deficiency; and SAEs.

In the overall study population (N = 31), there were more males (19 of 31 [61.3%]) than females (12 of 31 [38.7%]) at baseline and in the maralixibat (9 of 13 [69.2%]) and placebo groups (n = 10 of 16 [62.5%]) during the randomized withdrawal (RWD) period. The mean age in the overall study population was 5.4 years (range, 1 to 15 years) and was similar between the maralixibat and placebo groups. Most patients were from Australia and France (9 of 31 [29.0%] each) in the overall study population. ||| |||| |||| ||||| ||| |||||||| ||||||||| || |||| ||| |||| |||||| || ||| ||||||| ||||| ||||||||||| |||| |||| |||||| || ||||||||||| ||||| ||| |||| |||||| || ||| || ||||||| ||||| |||||| ||| |||||||||| |||||||||| |||||| In the overall study population, 8 of 31 (25.8%) of patients had a family history of ALGS (1 of 13 [7.7%] and 7 of 16 [43.8%] in the maralixibat and placebo groups, respectively). All enrolled patients had the Jagged-1 mutation present. Race and ethnicity data were not collected in the ICONIC trial.

Efficacy Results

In the ICONIC trial, the primary efficacy end point was the change in sBA during the 4-week RWD phase in the modified intent-to-treat (mITT) population (patients with sBA reduction ≥ 50% at week 12 or 18). A total of 15 participants were in the mITT population and were analyzed in the primary end point (5 randomized to maralixibat; 10 to placebo). The least squares (LS) mean difference in change from weeks 18 to 22 in sBA between the maralixibat and placebo groups was –117.28 (95% confidence interval [CI], –211.699 to –23.103; P = 0.0464) μmol/L, in favour of maralixibat. A consistent difference was observed in the overall randomized intent-to-treat (ITT) population.

In the ICONIC pivotal trial, the change from weeks 18 to 22 in ItchRO(Obs) weekly average morning severity score was a secondary end point. The LS mean difference between the maralixibat and placebo groups was –1.48 (95% CI, –2.12 to –0.84; P < 0.0001), in favour of maralixibat. In the overall population, there was a decrease (improvement) in ItchRO(Obs) weekly average morning severity score from baseline to week 18 (secondary end point) with a mean change of –1.70 (95% CI, –2.05 to –1.36; P < 0.0001) and from baseline to week 48 (additional end point) with a mean change of –1.62 (95% CI, –2.12 to –1.12; P < 0.0001). The prespecified sensitivity analyses for ItchRO(Obs) weekly average morning severity score was consistent with the results of the ItchRO(Obs). A total of 14 patients met the age cut-off for completion of the ItchRO(Pt) (≥ 9 years of age or ≥ 5 years of age with the assistance of their caregiver) in the pivotal trial. The LS mean difference between the maralixibat and placebo groups from weeks 18 to 22 for the change in ItchRO(Pt) weekly average morning severity score was –1.98 (–3.01 to –0.97; P = 0.0013), in favour of maralixibat. In the overall population, there was a decrease (improvement) in ItchRO(Pt) weekly average morning severity score from baseline to week 18 (secondary end point) with a mean change of –2.07 (95% CI, –2.65 to –1.50; P < 0.0001) and from baseline to week 48 (additional end point) with a mean change of –2.25 (95% CI, –2.84 to –1.67; P < 0.0001).

From weeks 18 to 22, the LS mean difference between the maralixibat and placebo groups for ALP was 10 (95% CI, –52.6 to 72.6; P = 0.7455) U/L compared with placebo. From weeks 18 to 22, the LS mean difference between treatment groups for ALT was 15.1 (95% CI, –25.1 to 55.2; P = 0.4472) U/L. From weeks 18 to 22, the LS mean difference between the maralixibat and placebo groups for total bilirubin was –0.14 (–0.88 to 0.60; P = 0.7000) mg/dL. From weeks 18 to 22, the LS mean difference between the maralixibat and placebo groups for direct bilirubin was –0.02 (95% CI, –0.56 to 0.53; P = 0.9517) mg/dL.

In the overall study population, there was an increase from baseline to week 100 (last observation carried forward [LOCF]) in mean height z scores with a mean change of 0.25 (95% CI, –0.86 to 2.04; P = 0.0216). In the overall study population, there were no major changes from baseline in mean weight z scores at any time point with a mean change from baseline to week 100 (LOCF) of –0.05 (95% CI, –0.12 to 0.23; P = 0.5306).

The pivotal trial assessed HRQoL using the PedsQL as additional efficacy end points, and the LS mean difference from weeks 18 to 22 in the PedsQL total score (parent) between the maralixibat and placebo groups was 2.33 (95% CI, –10.08 to 14.75; P = 0.7018). In the overall population, the mean change in the PedsQL total score (parent) from baseline to week 18 was 10.73 (95% CI, 4.43 to 17.03; P = 0.0016). The LS mean difference for the PedsQL Multidimensional Fatigue Scale score (parent) from weeks 18 to 22 between the maralixibat and placebo groups was 14.03 (95% CI, –2.78 to 30.84; P = 0.0966). In the overall population the mean change in the PedsQL Multidimensional Fatigue Scale score (parent) from baseline to week 18 was 20.30 (95% CI, 8.98 to 31.63; P = 0.0013).

Harms Results

The incidence of AEs was similar during the open-label phase, after RWD phase, and the LTE phase, with at least 25 of 29 patients (86.2%) experiencing any AEs in these treatment periods. During the RWD phase, patients who stayed on maralixibat had a lower incidence of AEs (7 of 13 patients [38%]) compared with patients receiving placebo (12 of 16 patients [75%]). The most frequently reported AEs (> 30% in at least 1 phase) were abdominal pain, pyrexia, diarrhea, nasopharyngitis, vomiting, cough, and pruritus. During the RWD phase, SAEs were reported for 1 of 13 patients (7.7%) receiving maralixibat and 1 of 16 patients (6.3%) receiving placebo. None of the SAEs were considered related to study medication. A total of 6 patients (2 each in the open-label phase, after RWD phase, and the LTE phase) experienced AEs leading to study drug discontinuation. No deaths were noted during the study. During the RWD phase, patients who stayed on maralixibat had a similar incidence of diarrhea and abdominal pain (1 of 13 patients [7.7%]) compared with those on placebo (1 of 16 patients [6.3%]). No patients experienced events associated with FSV deficiency during the RWD phase.

Critical Appraisal

During the open-label phases of the pivotal trial, patients’ and/or caregivers’ knowledge of treatment assignment may have biased subjective outcomes such as ItchRO(Obs), ItchRO(Pt), and PedsQL in favour of maralixibat. Reporting of harms could also have been biased, potentially in favour of maralixibat. Discontinuation was low, with 3 of 31 patients (9.7%) discontinuing due to an AE through to week 48. Regarding differences in baseline characteristics between patients in the maralixibat and placebo groups, the clinical experts noted that patients in the maralixibat group may have had a higher degree of disease severity than those in the placebo group as indicated by the higher sBA, ALT, and bilirubin values, which may have biased results in favour of placebo. Descriptive post hoc data from the ICONIC pivotal trial found that reductions in sBA from baseline to week 48 were associated with reductions in mean ItchRO(Obs) weekly average morning severity scores (Appendix 1). The data may show an association between sBA and ItchRO in some patients, but as the data were descriptive in nature and the assessment was conducted posthoc on a small number of patients (n = 28), it is unclear the extent to which sBA levels may be associated with pruritus in patients with cholestatic liver diseases.

The clinical experts consulted on this review noted that a minimal important difference (MID) of 1 for the ItchRO tool is clinically meaningful; however, the experts noted that such tools are not commonly used in clinical practice. HRQoL was assessed using the PedsQL as an additional efficacy outcome in the pivotal trial and MID estimates of 4 to 5 points align with the clinical experts’ expectations of a clinically meaningful change. It should be noted that the number of patients assessed for the PedsQL Multidimensional Fatigue Scale score was low during the RWD phase, with 9 of 13 patients (69.2%) in the maralixibat group and 12 of 16 patients (75.0%) in the placebo group contributing to the analysis of mean change from weeks 18 to 22. The impact of missing data on this outcome is unclear in the absence of sensitivity analyses.

The clinical experts consulted on this review stated that patients included in the ICONIC trial generally align with the selection criteria for candidates for maralixibat, although patients with mild cholestatic pruritus would not necessarily be excluded from treatment in clinical practice. Nonetheless, the clinical experts did not expect the exclusion of these patients to significantly affect the generalizability of the patient population in this study. The clinical trial only enrolled patients aged 12 months or older with a Jagged-1 mutation; however, the clinical experts note that the trial results would be applicable to patients younger than 12 months as well as patients with a NOTCH2 mutation, respectively. Although race and ethnicity data were not assessed in the pivotal trial, the clinical experts stated that the results would be applicable to the patient population in Canada. The efficacy outcomes measured in the study were of clinical importance to patients and clinicians, including change in sBA. However, the clinical experts noted that the change in sBA is not often assessed in clinical practice due to high costs and logistical limitations, as sBA testing is often sent to specialized laboratories and is not readily available in all gastroenterology practice settings. The clinical experts consulted for this review indicated that although tools such as PedsQL are frequently used in clinical trials, they are not typically used in clinical practice. Furthermore, the double-blind phase in the pivotal ICONIC trial was 4 weeks in length, limiting the ability to assess the long-term efficacy and safety of maralixibat compared with placebo for the indicated dose of 380 mcg/kg/day. While maralixibat has been approved by Health Canada for use in patients for the treatment of cholestatic pruritus in patients with ALGS, aged 2 months and older, the ICONIC trial only enrolled patients aged 12 months or older. As such, there is an absence of comparative efficacy and safety data assessing maralixibat versus placebo among patients aged younger than 12 months in the ICONIC trial due to the challenges of conducting a controlled clinical trial in this age group. However, the trial results are expected to be applicable to patients younger than 12 months based on clinical experts’ feedback. Furthermore, during the LTE phase of the ICONIC pivotal trial (as of week 103) eligible patients could have received a dose of maralixibat of up to 760 mcg/kg/day (given as twice-daily doses of 380 mcg/kg), which is outside of the proposed Health Canada indication of 380 mcg/kg /day. As such, efficacy and safety data after this period is are aligned with the recommended dose.

GRADE Summary of Findings and Certainty of the Evidence

Methods for Assessing the Certainty of the Evidence

For the pivotal study (ICONIC) identified in the sponsor’s systematic review, Grading of Recommendations Assessment, Development and Evaluations (GRADE) was used to assess the certainty of the evidence for outcomes considered most relevant to inform CADTH expert committee deliberations, and a final certainty rating was determined as outlined by the GRADE Working Group.^14,15 Following the GRADE approach, evidence from the pivotal study started as high-certainty evidence and could be rated down for concerns related to study limitations (i.e., internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, and publication bias.

The selection of outcomes for GRADE assessment was based on the sponsor’s Summary of Clinical Evidence, consultation with clinical experts, and input received from patient and clinician groups and public drug plans. The following list of outcomes was finalized in consultation with expert committee members: change in fasting sBA levels, change in pruritus as measured by ItchRO(Obs) and ItchRO(Pt) weekly average morning severity scores, change in liver biomarkers and enzymes (ALT, ALP, total, and direct bilirubin), change in body height and weight z scores, HRQoL as measured by the PedsQL total score (parent) and the PedsQL Multidimensional Fatigue Scale score (parent), and AEs including SAEs, diarrhea, abdominal pain, and FSV deficiency.

When possible, certainty was rated in the context of the presence of an important (nontrivial) treatment effect; if this was not possible, certainty was rated in the context of the presence of any treatment effect (i.e., the clinical importance is unclear). In all cases, the target of the certainty of evidence assessment was based on the point estimate and where it was located relative to the threshold for a clinically important effect (when a threshold was available) or to the null. The target of the certainty of evidence assessment was the presence or absence of any (non-null) effect for all outcomes except the ItchRO and PedsQL due to the lack of a formal MID estimate.

Results of GRADE Assessments

Table 2 presents the GRADE summary of findings for maralixibat versus placebo for the treatment of cholestatic pruritus in pediatric patients with ALGS.

Table 2: Summary of Findings for Maralixibat Versus Placebo for the Treatment of Cholestatic Pruritus in Patients With ALGS

ALGS = Alagille syndrome; ALP = alkaline phosphatase; ALT = alanine transaminase; CI = confidence interval; FSV = fat-soluble vitamin; HRQoL = health-related quality of life; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); LOCF = last observation carried forward; MID = minimal important difference; NA = not applicable; NR = not reported; PedsQL = Pediatric Quality of Life Inventory; RCT = randomized controlled trial; SAE = serious adverse event; sBA = serum bile acid.

Note: Study limitations (i.e., internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, and publication bias were considered when assessing the certainty of the evidence. All serious concerns in these domains that led to the rating down of the level of certainty are documented in the table footnotes.

^aStatistical testing for all outcomes was not adjusted for multiplicity. The potential for type I error (false-positive results) is increased.

^bRated down 2 levels for very serious imprecision; evidence from 1 trial with small sample size. The small sample size raises concerns about the potential for prognostic imbalance and potential overestimation of the true effect. No known MID so target of certainty appraisal was any effect; 95% CI did not cross the null.

^cRated down 2 levels for very serious imprecision; evidence from 1 trial with small sample size. The small sample size raises concerns about the potential for prognostic imbalance and potential overestimation of the true effect. The 95% CI did not considerably cross the threshold of importance (based on an MID of 1).

^dRated down 2 levels for very serious imprecision; evidence from 1 trial with small sample size. There is no known MID and the clinical experts consulted by CADTH could not provide a threshold of important difference; however, the CADTH review team judged that the effect estimate was likely to correspond with no important difference, and CI was unlikely to include both important benefit and harm.

^eIn the absence of a comparator group at the assessed time point, conclusions about efficacy relative to any comparator cannot be drawn and certainty of evidence started at very low. Rated down 2 levels for very serious imprecision, evidence from 1 arm of 1 trial with small sample size.

^fRated down 2 levels for very serious imprecision. The 95% CI for difference between groups included possible important benefit and important harm (based on MID of 4 to 5 points).

^gRated down 1 level for serious study limitations. Risk of bias due to missing outcome data, results of analysis available for 9 of 13 patients (69.2%) in the maralixibat group and 12 of 16 patients (75.0%) in the placebo group. Rated down 1 level for serious imprecision; the 95% CI for difference between groups included the potential for little-to-no difference (based on MID of 4 to 5 points).

^hRated down 1 level for serious indirectness. The clinical experts noted that the 4-week randomized withdrawal period was not sufficient to fully assess the comparative safety of maralixibat compared with placebo for this outcome. Rated down 2 levels for serious imprecision; the sample size is small and the results are based on very few or no events in each group.

Source: ICONIC Clinical Study Report.¹⁶ (Note: details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)¹⁷

LTE Studies

The pivotal ICONIC trial included an LTE phase described in the Systematic Review section of this report. No other LTE studies were submitted.

Indirect Comparisons

No indirect comparisons were conducted comparing maralixibat with other comparators for this submission.

Studies Addressing Gaps in the Evidence From the Systematic Review

Description of Studies

The sponsor submitted a natural-history comparison study, which is presented in this report, comparing disease outcomes among patients with ALGS treated with maralixibat (N = 84) with an external controls cohort from the Global ALagille Alliance (GALA) clinical research database (n = 469), with follow-up data up to 6 years. Outcomes assessed included event-free survival (EFS), a composite end point of first event of liver decompensation (ascites, variceal bleeding, surgical biliary diversion [SBD], liver transplant, and death), and transplant-free survival (TFS). Of note, the natural-history comparisons were conducted independent of the sponsor (Mirum Pharma).

Results from patient-level data from 3 long-term studies of patients with ALGS treated with maralixibat, including the LUM001-303 (IMAGINE) trial, the ICONIC pivotal trial (LUM001-304), and the IMAGINE-II (LUM001-305) trial to identify predictors of EFS and TFS were submitted by the sponsor and presented in this report.

Efficacy Results

Results from the natural-history comparison study reported a 70% improvement in EFS with maralixibat treatment compared with the GALA control group (hazard ratio [HR] = 0.305; 95% CI, 0.189 to 0.491; P < 0.0001) and a 67% improvement in TFS with maralixibat treatment compared with the GALA control group (HR = 0.332; 95% CI, 0.197 to 0.559; P < 0.0001). Additional relevant evidence assessing patient-level data (n = 76) from 3 ALGS clinical trials (IMAGINE, IMAGINE-II, and ICONIC) stated that clinical parameters (sBA levels, total serum bilirubin, and change in pruritus from baseline as measured by the ItchRO[Obs]) obtained after 48 weeks of maralixibat treatment were potential predictors of subsequent TFS and EFS.

Critical Appraisal

Concerns regarding the natural-history comparison include the potential residual confounding, incomparability in disease severity, and the lack of sBA data available among patients in the GALA registry. Although the study showed statistically and clinically significant reduction in liver transplant, death and other associated events in patients who received maralixibat treatment compared with patients who received standard of care, there is uncertainty in the results and they should therefore be interpreted with caution. Results from the 3 ALGS clinical trials (IMAGINE, IMAGINE-II, and ICONIC) are subject to uncertainty due to various limitations including the limited sample size, a lack of control group, and uncertainty if the improvements in EFS and TFS observed in this analysis are solely the result of improvements in pruritus.

Conclusions

There is an unmet need for symptomatic and curative treatment options for cholestatic pruritus in pediatric patients with ALGS. Patients and clinicians highlighted the need for treatments that reduce the frequency and severity of pruritus and reduce patient and caregiver fatigue. The pivotal phase II, double-blind, placebo-controlled, randomized drug withdrawal trial (ICONIC) was included in this review, which assessed the treatment of cholestatic pruritus in pediatric patients with ALGS (aged 12 months to 18 years). The trial was an exploratory study.

The study demonstrated that maralixibat may result in a decrease in sBA levels and results in a clinically meaningful improvement in pruritus as assessed by the ItchRO(Obs) and ItchRO(Pt) weekly average morning severity scores compared with placebo. It is important to note that the “low” certainty of evidence as assessed by GRADE for sBA and pruritus outcomes is due to the imprecision observed, and not from bias due to study limitations. This imprecision due to small sample size is clearly connected to the nature of the rarity of the disease. Improvements in the PedsQL total score (parent) and the PedsQL Multidimensional Fatigue Scale score (parent) were uncertain. It also remained uncertain whether maralixibat may have increased ALT and ALP levels or resulted in any difference in total and direct bilirubin compared with placebo. Moreover, although the sponsor provided some evidence to support sBA as a predictor of long-term outcomes such as EFS and TFS, the relationship between sBA and severity of cholestatic pruritus still remains uncertain. Despite certain limitations inherent with observational study design, it is likely that there is a significant treatment effect of maralixibat on long-term outcome such as EFS and TFS.

Maralixibat was generally well-tolerated in the ICONIC trial, with limited grade 3 AEs or SAEs. Of note, however, due to the rare nature of the disease, the severity of the condition, and the lack of approved or effective treatments, the study sample size was small and the study was short, in a 4-week randomized duration. Longer-term evidence to support the overall benefit and safety of maralixibat include LTE of the ICONIC trial and the natural-history comparison study. Real-world data currently being collected through the established GALA registry provide the opportunity to continuously monitor the efficacy and safety of maralixibat.

Introduction

The objective of this report is to review and critically appraise the evidence submitted by the sponsor on the beneficial and harmful effects of maralixibat (Livmarli), 9.5 mg/mL, oral solution, for the treatment of cholestatic pruritus in patients with ALGS, 2 months of age and older.

Disease Background

Contents within this section have been informed by materials submitted by the sponsor and by clinical expert input. The following has been summarized and validated by the CADTH review team.

ALGS is a rare, life-threatening, genetic, complex, multisystem disorder that presents with a range of clinical features including cholestatic liver disease, failure to thrive, cardiovascular disease, skeletal abnormalities, ocular abnormalities, renal and vascular abnormalities, and distinct facial features.^1,2 In most cases, the liver dysfunction associated with ALGS is an early and serious feature of this genetic condition and typically presents in the first 3 months of life. Elevated levels of sBAs and jaundice (elevated bilirubin) are hallmarks of ALGS and indicate the presence of impaired bile flow.³ Clinically important manifestations of cholestasis in ALGS include debilitating and intractable pruritus, disfiguring xanthomas, sleep disturbances, chronic debilitating fatigue, and failure to thrive (i.e., insufficient growth).^4,5

ALGS is predominantly caused by mutations in the Jagged-1 gene in greater than 90% of cases⁶ and in the NOTCH2 gene in approximately 4% of cases, resulting in bile duct paucity, bile flow obstruction, and bile accumulation in the liver. ALGS is inherited in an autosomal-dominant pattern. Children of an individual with ALGS have a 50% chance of inheriting the causative gene mutation in the Jagged-1 or NOTCH2 gene.² In 50% to 70% of affected individuals, however, the mutation is de novo.² Due to the diverse role of Notch signalling, disruption of either gene results in a broad range of clinical manifestations, as described previously. Cholestasis is defined as a decrease in bile flow due to impaired secretion by hepatocytes or to obstruction of bile flow through intrahepatic or extrahepatic bile ducts.⁷ Cholestasis leads to hepatic and systemic accumulation of substances normally excreted via the biliary tract, such as bile acids and conjugated bilirubin. Elevations of sBA up to 100 times the upper limit of normal (ULN) and serum bilirubin up to 30 times the ULN are not uncommon.⁶ Elevated levels of bile acids are hepatotoxic and contribute to disease progression. Bile acids have been shown to induce damage and necrosis in hepatocytes and cholangiocytes and are associated with increased morbidity and mortality in chronic cholestatic diseases.⁷ The systemic and hepatic accumulation of bile acids and other toxins leads to incapacitating and chronic cholestatic symptoms such as pruritus, and clinical sequelae.

The clinical manifestations of cholestasis associated with ALGS are severe, even in the absence of liver disease, with cholestatic pruritus being the leading cause of liver transplant in patients.^8-10 These clinical manifestations present in the first few years of life and as early as age 3 months. The symptoms include severe and unremitting pruritus (74%¹⁰), xanthomas (disfiguring and sometimes disabling subcutaneous lipid deposits, 40%), chronic fatigue (between 65% and 85%), and growth failure (between 50% and 87%¹¹). In addition, FSV malabsorption and increased risk of bone fractures because of trabeculae malformation can also be present. A second wave of portal hypertension and associated complications occurs later in adolescence.⁶ Collectively, all cholestasis-related symptoms result in poor HRQoL. As cholestasis progresses and symptoms worsen, as described previously, a majority of patients will either receive a liver transplant (50.4%) or die (9.3%) by age 18 years, with only 40.3% of patients reaching adulthood with their native liver.^6,10 The estimated 20-year life expectancy is 75% for patients diagnosed with ALGS, 80% for patients with ALGS who do not require liver transplant, and 60% for patients with ALGS who require liver transplant.¹² For patients with ALGS who undergo liver transplant, the estimated 1-year survival rate is 87%.¹³ The majority of early liver transplantations occur because of complications associated with cholestasis, including pruritus.

The reported incidence of ALGS is 1 in 30,000 to 50,000 live births.^3,11 Table 3 presents sponsor-submitted estimates of disease incidence based on the patient population in CADTH-participating jurisdictions. In the absence of Canadian statistics, the sponsor estimated prevalence (based on 1 in 30,000 live births) to be a total of 1,032 patients with ALGS in 2023.

Table 3: Estimated Incidence in Each Region

Source: Sponsor’s Clinical Summary Report.¹⁷

Diagnosis of ALGS can be challenging due to variable presentation of the clinical manifestations. The phenotypic presentation may vary even among individuals from the same family sharing the same genetic mutation.² Additionally, there is a lack of a strong correlation between mutation type, clinical manifestation, and disease severity.¹⁸ The clinical diagnostic criteria for ALGS can be made if 3 of the following 7 major clinical features are present: cholestasis, ophthalmologic abnormalities, characteristic facial features, cardiac defects, skeletal abnormalities, kidney abnormalities, and vasculature. Liver histology showing bile duct paucity and genetic testing is sometimes conducted to diagnose or confirm the diagnosis of ALGS.^2,3,6,11 ALGS can be diagnosed through clinical criteria alone; however, molecular genetic testing can also be used for diagnosis or can provide valuable confirmation to clinical diagnosis, especially in milder cases.¹¹ Genetic testing for ALGS is currently available in Canada and covered by the ministries of health.¹⁹ Additionally, Prevention (formerly Emory Medical Laboratories, US) provides free genetic panel for patients who have cholestasis. This service is made available to patients in Canada as well.¹⁹

In Canada, patients with ALGS are primarily followed in large specialized tertiary care centres covering large catchment areas.¹⁹ These centres receive referrals from specialists from smaller community hospitals, general practitioners or pediatricians, and neonatal intensive care units. The main treating team in these tertiary care centres comprises hepatologists or gastroenterologists and other allied professionals. Hepatologists or gastroenterologists are responsible for the primary management of the disease and patient follow-up, as well as coordination with local hepatologists or gastroenterologists for patients with ALGS who reside far from the tertiary care centres. Other allied professionals involved in the care of patients with ALGS are dieticians or nutritionists, social workers, and psychologists. Other care departments are also involved as needed (e.g., cardiology, radiology, genetics, nephrology, neurology, endocrinology).

ALGS pediatric patients who have progressed in their disease and may require liver transplants are referred to 1 of 3 pediatric transplant centres for transplant assessment. These pediatric transplant centres are located at The Hospital for Sick Children (SickKids) in Toronto, Ontario; the University of Alberta Hospital in Edmonton, Alberta; and the Centre hospitalier universitaire Sainte-Justine in Montreal, Quebec. For patients eligible to receive a liver transplant, the centres are also responsible for preparing patients (i.e., administrative, clinical work-up, and so forth) leading up to the procedure. Ultimately, the liver transplant will take place at these centres.

Standards of Therapy

Contents within this section have been informed by materials submitted by the sponsor and by clinical expert input. The following has been summarized and validated by the CADTH review team.

Currently in Canada, there are no medical treatments approved by Health Canada to treat ALGS or the cholestatic pruritus associated with it. According to the clinical experts consulted, there are no clinical practice guidelines for the treatment of ALGS and current treatment approaches focus on improving symptoms. Certain drugs are prescribed off-label based on limited clinical data, many of which have limited or transient efficacy and may have undesirable adverse effects.²⁰ According to the clinical experts consulted, current treatments for cholestatic pruritus include antihistamines, which are considered for mild cases and for their sedative effect. Ursodeoxycholic acid promotes bile excretion and, due to its attractive safety profile, it is typically used early in the management of cholestasis; however, its effectiveness for pruritus is not certain. Rifampicin has been reported to improve but not completely resolve pruritus in approximately 50% of patients with ALGS. Cholestyramine, a bile-acid binding resin, may be considered, but its poor palatability and interference with absorption of other drugs limits its use in clinical practice. Sertraline and naltrexone provide marginal additional benefit; however, according to the clinical experts, their use in pediatric patients is limited and their efficacy to treat pruritus in pediatric patients with ALGS is unclear.

As existing pharmacological therapies often fail in patients with ALGS, surgical alternatives are often required. Surgical interventions to treat cholestasis in ALGS include SBD procedures and liver transplant. SBD procedures (partial internal biliary diversion and ileal exclusion) have been used in ALGS to ameliorate cholestasis with variable results and short-term and long-term surgical and medical complications^6,8,10,11,21 and are seldomly used in Canada, according to the clinical experts consulted.

Intractable pruritus or disfiguring xanthomas can be severe enough to warrant liver transplant, even in the absence of liver failure.^10,11,22 Most patients with ALGS will undergo a liver transplant in the first 2 decades of life, with only 40.3% of patients reaching adulthood with their original liver. ALGS-associated pruritus is a leading cause of transplant in these patients.^6,10 Liver transplant in ALGS has increased the risk of complications, with studies reporting 1-year survival rates below 80%.¹² Patients with ALGS who survive liver transplant have the burden of lifelong immunosuppression and other long-term morbidity.¹¹ In particular, children who require transplant at a young age have poorer outcomes. Infants younger than 1 year have higher rates of mortality on the liver transplant waitlist and higher rates of rejection. Children younger than 4 years have lower rates of graft survival.²³ Of note, extrahepatic manifestations of ALGS such as significant cardiac disease can often be a contraindication for a major surgery such as liver transplant, leaving these patients with even further limited treatment options.²⁴

In alignment with clinical experts consulted by CADTH, the sponsor stated that the management of cholestasis and related pruritus in patients with ALGS remains largely supportive. Alternatively, surgical treatment may be required for patients, continuing to experience extensive morbidity. The clinical experts consulted by CADTH said that improving symptoms (such as very debilitating pruritus of ALGS that has a significant impact on patient quality of life) is the current treatment focus. Also, the clinical experts noted that there is a lack of curative and symptomatic treatments, and even liver transplant is not necessarily curative for liver disease due to high morbidity.

Drug Under Review

Key characteristics of maralixibat are summarized in Table 5, along with other treatments available for pruritus associated with ALGS.

Maralixibat (Livmarli), 9.5 mg/mL, oral solution, is approved by Health Canada for the treatment of cholestatic pruritus in patients with ALGS. The reimbursement request is the same as the indication to Health Canada. Maralixibat is the first drug approved by the FDA for cholestatic pruritus in patients with ALGS who are aged 3 months and older.²⁵ Maralixibat is indicated for the treatment of cholestatic pruritus in patients with ALGS, aged 2 months and older in the European Union.²⁶ Maralixibat does not have any other indications and has not been previously reviewed by CADTH. The recommended maintenance dose is 380 mcg/kg once daily in the morning after 1 week of a starting dose of 190 mcg/kg orally once daily. The maximum daily dose in volume for patients above 70 kg is 3 mL (Table 4).

Table 4: Individual Dose Volume by Patient Weight

Source: Sponsor’s Clinical Summary Report.¹⁷

As an inhibitor of the ileal bile acid transporter (IBAT), maralixibat interrupts the enterohepatic circulation of bile acids, leading to statistically significant decreases in sBA levels and increased fecal bile acid secretion. Bile acids are synthesized in the liver and are the major lipid components of bile, making up approximately two-thirds of the solute mass of normal human bile.²⁷ The enterohepatic circulation of bile acids acts as a feedback mechanism to maintain bile acid homeostasis and control bile acid production.²⁸ Circulation of abnormal bile acid levels is associated with a variety of illnesses, such as cholestatic liver disease. IBAT is responsible for the active reabsorption of about 90% to 95% of intestinal bile acids in the terminal ileum.^27,29,30 Due to its key role in bile acid re-uptake, IBAT is a target for pharmacologic regulation of bile acid reabsorption.³¹ The bile duct paucity associated with ALGS leads to impaired bile flow, accumulation of bile acids, and cholestatic liver injury. Through the reduction of sBA, maralixibat provides an improvement in cholestatic pruritus in patients with ALGS.

Table 5: Key Characteristics of Maralixibat, UDCA, Rifampin, and Antihistamines

ALGS = Alagille syndrome; ALT = alanine transaminase; AST = aspartate transaminase; FSV = fat-soluble vitamin; IBAT = ileal bile acid transporter; NA = not applicable; QTc = corrected QT interval; RNA = ribonucleic acid; sBA = serum bile acid; UDCA = ursodeoxycholic acid.

^aAntihistamines include certrizine hydrochloride, hydroxyzine hydrochloride, diphenhydramine, and so forth.

^bHealth Canada–approved indication.

Source: Sponsor’s Clinical Summary Report.¹⁷

Stakeholder Perspectives

Patient Group Input

This section was prepared by the CADTH review team based on the input provided by patient groups. The full original patient input received by CADTH has been included in the Stakeholder Perspectives section of this report.

Two patient groups, CLF and ALGSA, provided input. CLF is the only national health charity committed to supporting people in Canada affected by the liver diseases. Based in the US, ALGSA is a nonprofit organization dedicated to supporting families affected by ALGS globally. The CLF submission included phone and virtual interviews conducted in May 2023 with 8 patients and caregivers in Canada. Of those, 4 respondents had experiences with maralixibat through clinical trials. The ALGSA gathered data online through family surveys (2020), personal conversations, and topic-specific discussions among support or focus groups, including at least 76 members in Canada. Both groups stated that itchiness (pruritus) is the most bothersome symptom affecting patients’ and caregivers’ lives. For example, itchiness interrupts patients and families’ sleep, making those affected fatigued, anxious, depressed, irritable, and worried. Patients said they feel isolated in school and that it is challenging to maintain employment. Also, patients and families have difficulty finding a specialist who could recognize and make a proper diagnosis of ALGS and manage disease treatment. Patients and families from both groups want a new therapy that can provide significant relief of itchiness with long-term effects without high risks such as liver transplant and immunosuppression. Patients who have taken maralixibat during clinical trials said that their itchiness has been resolved with minor side effects, such as upset stomach and diarrhea, that they could become more of themselves and engage in normal day-to-day activities, and that their households were also positively affected.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). In addition, as part of the review of maralixibat, a panel of 3 clinical experts from across Canada was convened to characterize unmet therapeutic needs, assist in identifying and communicating situations where there are gaps in the evidence that could be addressed through the collection of additional data, promote the early identification of potential implementation challenges, gain further insight into the clinical management of patients living with a condition, and explore the potential place in therapy of the drug (e.g., potential reimbursement conditions). A summary of this panel discussion follows.

Unmet Needs

The clinical expert panel stated that cholestatic pruritus remains a very significant management problem for patients with ALGS and their families due to partial, incomplete, or null response to currently available treatments. Current treatments are used off-label and are supportive in nature. These treatments include antihistamines, which can be effective for mild cholestatic pruritus and provide a sedative effect to augment sleep; however, the experts noted that antihistamines are rarely effective for treating cholestatic pruritus. Ursodeoxycholic acid promotes bile excretion and, due to its attractive safety profile, it is typically used early in the management of cholestasis; however, its effectiveness for cholestatic pruritus is not certain. Cholestyramine is often unpalatable and rarely used in pediatric patients. Rifampin does provide some symptomatic relief; however, it is usually ineffective in substantially ameliorating or eradicating cholestatic pruritus. Naltrexone, an opioid antagonist, seldomly used in pediatric patients, has been associated with minimal improvement in some children with ALGS but may include symptoms of opioid withdrawal syndrome and gastrointestinal intolerance. Sertraline, a selective serotonin re-uptake inhibitor, has been used in refractory cases in adults; however, there is limited pediatric data available to support its use as an additional therapy for pruritus. The experts noted that surgical options such as an external or internal biliary diversion can be offered to patients with ALGS with cholestatic pruritus who are refractory to medical therapies; however, these are not very effective and seldomly used in clinical practice. Finally, the experts stated that between 50% and 75% of patients with cholestatic liver disease will require a liver transplant and that cholestatic pruritus is a leading indication for a transplant. Liver transplant is associated with increased morbidity, mortality, and lifelong immune suppression. As such, the experts noted that there is an unmet need for effective symptomatic and curative treatment for cholestatic pruritus in the indicated patient population.

Place in Therapy

The clinical expert panel noted that there are currently no established clinical practice guidelines for ALGS and that none of the currently available off-label treatments target the underlying disease mechanism of bile duct paucity. Many patients with ALGS and cholestatic pruritus have inadequately treated pruritus with currently available off-label medical therapy. The clinical experts stated that maralixibat would likely be used in combination with current off-label treatments in patients experiencing ongoing pruritus, and that it is possible some patients could discontinue some of the off-label treatments once they are established on maralixibat and their pruritus is under control. The experts noted that, if easily accessible, maralixibat may be used as an initial therapy for new patients presenting with severe pruritus.

Patient Population

The clinical experts stated that the estimated incidence of ALGS in Canada is about 1 in 30,000 to 50,000 live births with about 200 new cases each year. A clinical diagnosis can be made by the presence of 3 of 7 clinical features (chronic cholestasis, cardiovascular abnormalities, butterfly vertebrae, posterior embryotoxon, renal anomalies, vascular abnormalities, or characteristic facies). ALGS is an autosomal-dominant condition, and a genetic diagnosis can be confirmed in approximately 95% of patients with clinical features. The experts noted that with the advent of accessible molecular testing, most pediatric patients receive genetic testing to confirm their diagnosis. The experts noted that pediatric patients with ALGS most suited for treatment with maralixibat are those who present with cholestatic pruritus that is persistent with current off-label treatments, which is about a third of patients in a clinical expert’s practice. Patients least suited to treatment with maralixibat are those who have minimal liver involvement (i.e., minimal liver enzyme abnormalities and no FSV deficits) and those who do not experience cholestatic pruritus.

Assessing the Response to Treatment

According to the expert panel, a clinically meaningful response to treatment would include a reduction in the frequency and severity of pruritus, a reduction in sleep deprivation among patients and their caregivers, the ability for patients and their caregivers to attend school or work, reduced damage to the patients’ skin, and improved patient weight and growth. The clinical experts consulted on this review noted that response to therapy would likely be evaluated via subjective family reporting of symptoms including itching and sleep disturbances as well as by visual assessments of excoriations on the patient’s skin, which are often indicative of severe cholestatic pruritus. Standard-reporting itch scales such as the Visual Analogue Scale may be used to assess response to treatment, although such scales are not commonly used in clinical practice. according to the experts. Improvements in patient weight and growth are most often assessed using standardized parametrizes including midarm circumference and skin-fold thickness. Measurements of sBA could also be used to assess response to therapy; however, the experts noted that this is not common in clinical practice due to the high cost and the limited availability of such testing in some practice settings. The clinical experts would initially assess patients monthly for approximately 3 months, at which time the frequency of visits would be reduced to every 3 to 6 months if a response to treatment was evident.

Discontinuing Treatment

The clinical experts stated that treatment with maralixibat will likely be lifelong for most patients. The panel noted that treatment discontinuation may be considered if there is no effect on cholestatic pruritus after approximately 6 months of treatment initiation. Treatment may be discontinued if a patient’s liver disease progresses, and they undergo liver transplant. Treatment may be discontinued due to severe AEs; however, the experts stated that AEs associated with maralixibat are likely self-limited and may be addressed by titrating the dose of maralixibat.

Prescribing Considerations

The clinical experts noted that a pediatric or adult liver or GI specialist would be the preferred specialist to prescribe and monitor treatment with maralixibat.

Clinician Group Input

This section was prepared by the CADTH review team based on the input provided by clinician groups. The full original clinician group input received by CADTH has been included in the Stakeholder Perspectives section of this report.

One clinician from the Canadian Association for the Study of the Liver provided input. The clinician group and 2 clinical experts consulted by CADTH agree on the unmet need, which is a lack of effective therapy specifically indicated for cholestatic pruritus associated with ALGS refractory to current off-label treatments. They also agree that all the existing therapies are not effective at reducing cholestatic pruritus associated with ALGS in most patients and that there are no guidelines for treating cholestatic pruritus in patients with ALGS. In alignment with clinical experts, the clinician group stated that treatment goals are mainly improvement in pruritus, improvement in quality of life (i.e., sleep duration), and optimizing nutritional goals (e.g., treating FSV deficiency). Also, both groups agree that patients with ALGS and cholestatic pruritus that is persistent on standard-of-care medical treatment would be an eligible population. The clinician group stated that if a patient’s liver disease progresses and they undergo liver transplant, discontinuation is considered, and the clinical experts stated that if there is no effect on itch as measured clinically then discontinuation is considered after adequate trial — i.e., 6 months. Otherwise, both groups agree that AEs would be an unlikely reason to discontinue since maralixibat is well-tolerated. Lastly, all the clinician group and clinical experts agree that maralixibat should be prescribed by a pediatric gastroenterologist or hepatologist. None of the clinician group or clinical experts consulted by CADTH had declared experience with maralixibat.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH Reimbursement Review processes by identifying issues that may affect their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 6.

Table 6: Summary of Drug Plan Input and Clinical Expert Response

ItchRO = Itch Reported Outcome; UDCA = ursodeoxycholic acid.

Clinical Evidence

The objective of CADTH Clinical Review report is to review and critically appraise the clinical evidence submitted by the sponsor on the beneficial and harmful effects of maralixibat, 9.5 mg per mL, oral solution in the treatment of cholestatic pruritus in patients with ALGS. The focus will be placed on comparing maralixibat with relevant comparators and identifying gaps in the current evidence.

A summary of the clinical evidence included by the sponsor in the review of maralixibat is presented in 2 sections, with CADTH critical appraisal of the evidence included at the end of each section. The first section, the Systematic Review, includes pivotal studies and randomized controlled trials (RCTs) that were selected according to the sponsor’s systematic review protocol. CADTH assessment of the certainty of the evidence in this first section using the GRADE approach follows the critical appraisal of the evidence. The second section includes additional studies that were considered by the sponsor to address important gaps in the systematic review evidence. There was no indirect evidence submitted for this review.

Included Studies

Clinical evidence from the following are included in the CADTH review and appraised in this document:

• One pivotal study identified in systematic review

• Two additional studies addressing gaps in evidence.

Systematic Review

Contents within this section have been informed by materials submitted by the sponsor. The following has been summarized and validated by the CADTH review team.

Description of Studies

Characteristics of the included pivotal study are summarized in Table 7.

Table 7: Details of the Study Included in the Systematic Review

AE = adverse event; ALGS = Alagille syndrome; ALP = alkaline phosphatase; ALT = alanine transaminase; b.i.d. = twice a day; CCS = Clinician Scratch Scale; FSV = fat-soluble vitamin; GGT = gamma-glutamyl transferase; GI = gastrointestinal; INR = international normalized ratio; ItchRO = Itch Reported Outcome; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); LDL-C = low-density lipoprotein cholesterol; LOCF = last observation carried forward; mITT = modified intent to treat; sBA = serum bile acid; PedsQL = Pediatric Quality of Life Inventory; ULN = upper limit of normal.

Source: ICONIC Clinical Study Report.¹⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)¹⁷

The ICONIC pivotal trial was an open-label, phase IIb study with a double-blind, placebo-controlled, randomized drug withdrawal period and a long-term open-label extension (Figure 1). The study was designed to evaluate the safety and efficacy of maralixibat in children with ALGS aged between 12 months and 18 years. A total of 31 patients enrolled into the study which was conducted at 10 clinical sites in Australia, Europe, and the UK. Patients were enrolled between November 25, 2014, and September 11, 2015.

The periods of the study are grouped as follows:

• screening period: study eligibility assessments (up to 4 weeks)

• open-label phase: 18-week open-label run-in period

• RWD phase: 4-week randomized, double-blind, placebo-controlled drug-withdrawal period

• after RWD phase: 26-week stable-dosing period

• LTE phase: optional long-term treatment periods consisting of a 52-week follow-up treatment period followed by a long-term follow-up treatment period.

Randomization and Masking

All participants were randomly assigned (1:1) in a blinded fashion to continue receiving the same dose of maralixibat or receive placebo for a period of 4 weeks (RWD; Figure 1). Randomization used a permuted block algorithm stratified by predefined response criteria (≥ 50% sBA reduction from baseline to weeks 12 or 18) and with entire blocks (size 4) assigned by study site using SAS software (version 9.4) by an unblinded statistician not involved in the conduct of the trial or analysis of the data. The randomization code was assigned to each participant in sequence in the order of enrolment, and then the participants received the investigational products labelled with the same code. Both maralixibat and placebo were identical in appearance. All participants, investigators, and laboratory staff were masked to treatment allocation.

The Clinical Study Report for the ICONIC pivotal trial presented maralixibat treatment doses as maralixibat chloride. The dosage of maralixibat described in this report are of maralixibat free base as in the proposed Health Canada indication, unless otherwise specified. A dose of 380 mcg/kg maralixibat free base is equivalent to 400 mcg/kg maralixibat chloride.

Upon completion of the 52-week optional follow-up treatment period, patients were considered for the optional long-term follow-up treatment period. If eligible, based on efficacy and safety assessments, patients received up to 760 mcg/kg/day of maralixibat (given as twice-daily doses of 380 mcg/kg), which is not aligned with the approved Health Canada dosage.

Figure 1: Study Design for the ICONIC Pivotal Trial

ALGS = Alagille syndrome; BID = twice a day; ItchRO(Obs) = Itch Reported Outcome (observer); QD = every day; R = randomization; sBA = serum bile acid.

Populations

Inclusion and Exclusion Criteria

The ICONIC trial included pediatric patients aged 12 months to 18 years with a diagnosis of ALGS based on diagnostic criteria as shown in Table 8, evidence of cholestasis (described in Table 7), and significant pruritus, defined as an average daily score of greater than 2 on the ItchRO questionnaire for 2 consecutive weeks in the screening period. Key exclusion criteria included chronic diarrhea requiring specific IV fluid or nutritional intervention, surgical disruption of the enterohepatic circulation, liver transplant, decompensated cirrhosis, or history or presence of other concomitant liver disease.

Patients were considered eligible for the initial 52-week optional follow-up treatment period if they had completed the protocol through the week 48 visit with no safety concerns. Patients who had undergone a surgical disruption of the enterohepatic circulation were not eligible to enter into the follow-up treatment period. Patients were considered eligible for the optional long-term follow-up treatment period if they completed the protocol through the week 48 visit with no major safety concerns or discontinued due to safety reasons judged unrelated to maralixibat, and laboratory results had returned to levels acceptable for this patient population or individual and the patient did not meet any of the protocol’s stopping rules at the time of entry into the follow-up period. Patients who were discontinued for other reasons were considered for the follow-up periods on an individual basis. All exclusion criteria mentioned for the core study applied upon entry into the optional long-term follow-up treatment period, with the exception of patients weighing more than 50 kg at screening.

Table 8: ALGS Diagnostic Criteria Used in the ICONIC Pivotal Trial

ALGS = Alagille syndrome.

^aFamily history = ALGS present in a first-degree relative.

^bMajor clinical criteria or features for ALGS include: cholestasis consistent cardiac, renal, vascular, ocular, or skeletal involvement; or characteristic Alagille facies.

^cIdentified = Jagged-1 or NOTCH2 mutation identified in clinical laboratory.

^dNot identified = not identified on screening, or not screened for.

Source: Sponsor-submitted internal package.³⁷

Interventions

Open-Label Phase: Dose-Escalation Period (Day 0 to Week 6)

During the first 6 weeks of the study, the dose-escalation period, the dose of maralixibat was increased at weekly intervals up to 380 mcg/kg/day once daily as follows:

• Week 1 dose: 13.3 mcg/kg/day once daily

• Week 2 dose: 33.3 mcg/kg/day once daily

• Week 3 dose: 66.5 mcg/kg/day once daily

• Week 4 dose: 133 mcg/kg/day once daily

• Week 5 dose: 266 mcg/kg/day once daily

• Week 6 dose: 380 mcg/kg/day once daily

Patients received maralixibat daily for at least 7 days at each dose level during the dose-escalation period.

The dosage and dosing regimen of concomitant drug therapy other than that specified by the protocol was not changed during the first 22 weeks of study, except for weight-based dose adjustments and vitamin supplementation. No new medications used to treat pruritus could be added during the first 22 weeks of the study.

Open-Label Phase: Stable-Dosing Period (Weeks 7 to 18)

Following the 6-week dose-escalation period, patients continued dosing for another 12 weeks at the dose administered at week 6, which may have been 380 mcg/kg/day or the highest tolerated dose below 380 mcg/kg/day.

RWD Phase (Weeks 19 to 22)

At the week 18 visit, patients were randomized 1:1 to continue to receive maralixibat or a corresponding placebo for 4 weeks. Randomization was stratified by whether the patient achieved a greater than 50% reduction in sBA between baseline and weeks 12 or 18.

Long-Term Exposure Period (Weeks 23 to 48)

Following the 4-week study drug RWD period, patients who received placebo returned to the maralixibat dose received during the initial escalation. Patients who were randomized to receive maralixibat during this period continued to receive the same dose of maralixibat and, following week 22, a simulated dose escalation occurred to maintain the blind in the RWD period. Dosing with maralixibat continued in a 26-week, long-term exposure period to complete 48 weeks of study. Dose adjustments were made as needed due to intolerance or change of greater than or equal to 10% in a patient’s body weight since the screening visit.

Fifty-Two–Week Optional Follow-Up Treatment Period

At week 48, the investigator assessed all patients to determine their willingness and eligibility to move into the 52-week optional follow-up treatment period. The following 3 possible scenarios may have occurred:

• Patients who were eligible to roll over into the optional follow-up treatment period with no maralixibat dosing interruption or an interruption of less than 7 days were maintained at the same dose level that they had been taking at week 48.

• Patients who were eligible to roll over into the optional follow-up treatment period with a maralixibat dosing interruption of 7 days or more were dose escalated beginning at 35 mcg/kg/day up to a maximum of 380 mcg/kg/day or highest tolerated dose.

• Patients who did not wish to enter the follow-up treatment period were contacted via telephone by the study site approximately 30 days after the last dose of maralixibat.

Long-Term Optional Follow-Up Treatment Period

Upon completion of the 52-week optional follow-up treatment period, patients eligible for the optional follow-up treatment period either continued receiving the same dose level of maralixibat that they had been taking at the completion of the follow-up period with morning dosing only or started twice-daily dosing. Eligibility for twice-daily dosing was determined based on efficacy as measured by sBA level and ItchRO score. Patients without dosing interruption or interruption of less than 7 continuous days who met sBA and ItchRO eligibility criteria started twice-daily dosing (afternoon dose escalation) up to their maximum tolerated dose or at most a maximum daily dose of 380 mcg/kg twice daily (i.e., 760 mcg/kg/day). Patients with a maralixibat dosing interruption of greater than or equal to 7 days who met sBA and ItchRO eligibility criteria underwent dose escalation starting with 4-times-per-day dosing (at most a maximum daily dose of 380 mcg/kg), followed by an afternoon dose escalation up to their maximum tolerated dose or to a maximum daily dose of 380 mcg/kg twice daily (i.e., 760 mcg/kg/day).

Patients could continue to receive maralixibat until they were eligible to enter another maralixibat study, until maralixibat was available commercially, or until the sponsor stopped the program or development in this indication, whichever occurred first.

Concomitant Medications

Patients were required to discontinue bile acid resins for at least 28 days before screening and for the duration of the study. The dosage and dosing regimen of concomitant drug therapy other than that specified by the protocol should not change during the first 22 weeks of study, with the exception of weight-based dose adjustments and vitamin supplementation. No new medications used to treat pruritus were to be added during the first 22 weeks of the study.

Outcomes

Summarized outcomes are based on outcomes included in the sponsor’s Summary of Clinical Evidence as well as any outcomes identified as important to this review according to the clinical expert(s) consulted by CADTH and stakeholder input from patient and clinician groups and public drug plans. Using the same considerations, the CADTH review team selected outcomes that were most relevant to inform CADTH expert committee deliberations and finalized this list of outcomes in consultation with members of the expert committee. All summarized efficacy end points identified in Table 9 were assessed using GRADE. Select notable harms outcomes considered important for informing CADTH expert committee deliberations were also assessed using GRADE. A description of outcome measures and their measurement properties is provided in Table 10. Details regarding additional efficacy outcomes not assessed by GRADE can be found in Appendix 1.

Serum Bile Acid

In other pediatric cholestatic diseases (progressive familial intrahepatic cholestasis and biliary atresia), sBA was shown to have prognostic implications, with patients with higher sBA levels being more likely to progress to end-stage liver disease or to undergo liver transplant.^38,39 It has not yet been defined what sBA level or reduction can be associated with improved long-term outcome in patients with ALGS. In the pivotal trial, an sBA responder was defined as a patient with a reduction in sBA greater than or equal to 50% from baseline to weeks 12 or 18. An sBA responder at week 48 was defined as a patient with a reduction in sBA greater than or equal to 50% from baseline to week 48. According to the clinical experts consulted by CADTH for this review, there is insufficient evidence on the association between sBA change and pruritis in patients with ALGS.

Itch Reported Outcome

Pruritus was assessed using the ItchRO, which comprises 2 clinical outcome assessment instruments — namely, ItchRO(Obs), the caregiver-reported version, and the ItchRO(Pt), the patient-reported version for patients aged 9 years or older.⁴⁰ The caregiver-rated ItchRO(Obs) instrument was selected as a tool for assessing pruritus given the expectation, that a large proportion of study patients would be aged younger than age 9 years. The ItchRO(Pt) instrument was included in the ICONIC trial to collect supportive data in children who were old enough to assess their own pruritus (≥ 9 years of age or ≥ 5 years of age with the assistance of their caregiver).

The ItchRO(Obs) consists of a morning and an evening diary entry that asks the caregiver about the severity of the child’s itch-related symptoms using a single item (item 1). The item pertaining to nighttime itching (from bedtime until the time the child woke up) is answered in the morning diary entry, and the item pertaining to daytime itching (from the time the child woke up in the morning until bedtime) is answered in the evening diary entry. Item 1, also called the ItchRO severity score, has a range from 0 to 4, with a higher score indicating increasing itch severity (i.e., 0 = none, 1 = mild, 2 = moderate, 3 = severe, 4 = very severe). The daily score was taken as the highest score of the morning and evening scores, representing the greatest severity of itching over the 24-hour period. The weekly average severity score, defined as the 7-day average of the daily maximum of morning and evening ratings before a study visit, was used in the psychometric validation analysis.

A description of the ItchRO instrument, as well as its psychometric properties and MID estimates, are summarized in Table 10.

Liver Biomarkers and Enzymes

Efficacy end points in the pivotal trial included liver chemistry markers for cholestasis (total and direct bilirubin, ALP) and hepatocellular involvement (ALT). For these parameters, the prespecified secondary end point analyses included 1) change from baseline to weeks 18 and 2) change from weeks 18 to 22. Furthermore, change from baseline to weeks 22, 48, and then every 12 weeks were analyzed as additional efficacy end points.

Health-Related Quality of Life: PedsQL

The PedsQL Generic Core Scales are composed of 23 items to assess pediatric HRQoL measurements across 6 subscales: physical functioning, physical symptoms (only applicable for infants aged 1 to 24 months), emotional functioning, social functioning, cognitive functioning (only applicable for infants aged 1 to 24 months), and school functioning (only applicable for children aged 2 to 18 years). The total scale score encompasses all 23 items. Each item consists of 5-level verbal descriptor response options (0 to 4 points). Items are reverse scored and linearly transformed to a 0 to 100 scale so that higher scores indicate better HRQoL (less negative impact). The PedsQL Multidimensional Fatigue Scale is composed of 18 items across 3 subscales: general fatigue, sleep/rest fatigue, and mental fatigue. Respondents use the scale to indicate how frequently certain fatigue-related symptoms and complaints trouble them. The Multidimensional Fatigue Scale score is computed from all items of the PedsQL Multidimensional Fatigue Scale. Items are reverse scored and linearly transformed to a 0 to 100 scale so that higher scores indicate better HRQoL (fewer symptoms of fatigue). In the ICONIC trial, PedsQL questionnaires were prospectively collected via a caregiver proxy-report during the ICONIC study and analyzed retrospectively. Measurements from baseline and week 48 were included in this analysis.

A description of the PedsQL, its psychometric properties, and MID estimates are summarized in Table 10.

Change in Body Height and Weight

To compare growth parameters to the normal population, z scores were calculated in the ICONIC trial. Changes greater than 0 in height and weight z scores indicate that patients grow faster than their healthy peers (catch-up growth), whereas changes less than 0 indicate a deepening growth deficit. Height and weight z scores are based on a patient’s sex and age at each scheduled visit. For patients younger than 24 months, the WHO growth charts are recommended by the Centers for Disease Control and were used to derive z scores. For patients aged at least 24 months, the Centers for Disease Control growth charts were used to derive z scores.

Harms Outcomes

The harms outcomes assessed in the pivotal trial included the proportion of patients with AEs, SAEs, and notable harms. AEs were any unfavourable and unintended sign, symptom, or disease temporally associated with the study or use of an investigational drug product, which may or may not be related to the study drug. The criteria for SAEs included death; events that were life-threatening, required hospitalization, or prolongation of hospitalization, or resulted in persistent or significant disability or incapacity; was a congenital anomaly or birth defect; or an event that may have required medical or surgical intervention to prevent 1 of the other outcomes listed.

Table 9: Outcomes Summarized From the ICONIC Trial

AE = adverse event; ALP = alanine phosphatase; ALT = alanine transaminase; FSV = fat-soluble vitamin; HRQoL = health-related quality of life; ItchRO = Itch Reported Outcome; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); LOCF = last observation carried forward; PedsQL = Pediatric Quality of Life Inventory; SAE = serious adverse event; sBA = serum bile acid.

^aIdentified as an end point of importance by the clinical experts consulted but not measured in the ICONIC pivotal trial.

Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

Source: ICONIC Clinical Study Report.¹⁶

Considerations that went into the selection of efficacy outcomes to be summarized and assessed using GRADE include the following:

• The primary outcome in the pivotal trial; change in fasting sBA levels (μmol/L) in patients who previously responded to treatment with maralixibat, defined by a reduction in sBA of greater than or equal to 50% from baseline to weeks 12 or 18 was defined as important by the clinical experts consulted and is an outcome that provides the source for key input in the sponsor’s pharmacoeconomic model.

• Change in pruritus was defined as a very important outcome by the clinical experts consulted and noted by the patient and clinician group input received, resulting in the selection of the outcome of change in pruritus as measured by ItchRO(Obs) and ItchRO(Pt) weekly average morning severity score in patients who previously responded to maralixibat treatment, defined by a reduction in ItchRO scale of greater than 1 point from baseline to weeks 12 or 18.

• Other liver-related outcomes defined as important by the clinical experts consulted included change in liver biomarkers and enzymes (ALT, ALP, total and direct bilirubin).

• Change in body height and weight was described as important by the clinical experts consulted and the patient input received.

• HRQoL outcomes were defined as important by the clinical experts and patient input received. The PedsQL total score (parent) was selected as it encompasses all 23 items of the PedsQL, and the PedsQL Multidimensional Fatigue Scale score (parent) was selected given the prevalence and severity of fatigue in the indicated patient population as noted by the clinical experts consulted and patient input received.

• Safety outcomes defined as important to the clinical experts consulted included SAEs, diarrhea, abdominal pain, and FSV deficiency.

Statistical Analysis

Sample Size and Power Calculation

The planned sample size of 30 evaluable patients with ALGS was based on practical considerations, rather than a desired power for a prespecified difference.

Table 10: Summary of Outcome Measures and Their Measurement Properties

ALGS = Alagille syndrome; CIC = Caregiver Impression of Change; CSS = Clinician Scratch Scale; HRQoL = health-related quality of life; ICC = intraclass correlation coefficient; ItchRO = Itch Reported Outcome; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); MDFM = Multidimensional Fatigue Module; MID = minimal important difference; PedsQL = Pediatric Quality of Life Inventory; SEM = standard error of mean.

Statistical Testing

For the primary efficacy end point, the difference between treatment groups from weeks 18 to 22 in sBA was evaluated using an analysis of covariance (ANCOVA) model with treatment group as a factor, and week 18 sBA as a covariate, assessed in the mITT population. An ANCOVA model that includes the stratification variable sBA responder indicator as an additional covariate was also performed on the ITT population. This model also included the sBA responder covariate by treatment sequence interaction term. The LS mean difference between treatment groups with standard error, 95% CI for the LS mean difference, and P value for testing if the treatment group LS means were equal, was calculated to determine if the change in sBA levels between the treatment groups were statistically significant.

Secondary and other efficacy variables that are continuous measures were analyzed similarly to the primary efficacy analyses in the ITT population using summary statistics, and the majority were assessed by ANCOVA. If there appeared to be a relative significant difference among treatment sequences with respect to baseline characteristics, that baseline variable may have been added to the statistical model as a blocking factor or covariate to determine the effect on treatment. For this analysis, an interaction term for the covariate by treatment sequence was not included. ANCOVA models that only included the baseline value (of the variable of interest) as a single covariate were also included.

Efficacy measures that are categorical binary responder outcomes were analyzed using the chi-square or Fisher’s exact test, as appropriate based on sample sizes. Fisher’s exact test was chosen over the chi-square test if any of the cell counts were less than 5.

Multiple Comparisons

No adjustments were made for multiple comparisons.

As the definition of the pruritus end point used to assess efficacy was not prespecified in the protocol and the type I error was not specifically controlled for, P values for the analyses were treated as nominal.

Handling of Drop-Outs or Missing Data

For ItchRO(Obs) and ItchRO(Pt) weekly average scores, an average of daily scores was computed if at least 4 of the 7 daily ItchRO scores for a 7-day period were reported. If fewer than 4 ItchRO scores were reported, then the weekly average from the previous compliant week was used in an LOCF format. ItchRO(Obs) and ItchRO(Pt) weekly average morning severity scores derived for each week over the 4-week RWD phase were analyzed using ANCOVA and a restricted maximum likelihood (REML)–based repeated-measures approach as the principal sensitivity analysis method. The mixed model for repeated measures (MMRM) analysis model was used for handling missing data, with change from baseline (week 18) as the dependent variable; it included the fixed, categorical effects of treatment group, visit (time point), and treatment-group-by-visit interaction as well as the continuous covariates of baseline and baseline-by-visit interaction. The protocol stated that missing data imputation was not done for other efficacy end points.

In addition to the time points specified in the protocol, efficacy variables analyzed by time point were analyzed (as a sensitivity analysis) at the following LOCF time points: week 18 (LOCF), week 22 (LOCF), week 48 (LOCF), and week 100 (LOCF) time points, where appropriate.

For patients who terminated early from the study before week 100 or were otherwise missing weeks 18, 22, 48, and/or 100, data were imputed in an LOCF approach as follows:

• open-label phase: last observation on or before week 18 imputed as week 18 (LOCF)

• RWD phase: last observation between week 18 and the end of week 22 imputed as week 22 (LOCF) (only applicable for ItchRO[Obs] weekly average score variables)

• after RWD phase: last observation between week 22 and the end of week 48 imputed as week 48 (LOCF) and last observation after week 22 and end of week 100 imputed as week 100 (LOCF).

For PedsQL scale scores, if more than 50% of the items in the scale were missing, the scale score was not computed.

If a patient had a missing sBA, ItchRO, or CSS value at a week required in determining responder status, then the missing change from baseline value would be considered as not meeting the criteria for a responder.

Subgroup Analyses

For ItchRO(Obs) weekly average morning severity scores and sBA levels, summary statistics by analysis visit were conducted for the following subgroups:

• Age group (up to 24 months, 2 to 12 years, greater than 12 years)

• Baseline sBA (less than 275 μmol/L, greater than or equal to 275 μmol/L)

• Baseline total bilirubin (less than 3.8 mg/dL, greater than or equal to 3.8 mg/dL)

• Baseline ALT (less than 90 U/L, greater than or equal to 90 U/L)

• Baseline ItchRO(Obs) weekly average morning severity score (less than 3 points, greater than or equal to 3 points)

For the subgroup analyses, observed and change from baseline (day 0) values were presented for the open-label and after RWD phases. For the RWD phase, observed and change from week 18 values were presented for the maralixibat and placebo treatment groups.

Detailed statistical analysis of efficacy end points including adjustment factors and sensitivity analyses are presented in Table 11.

Table 11: Statistical Analysis of Efficacy End Points in the ICONIC Trial

ALP = alkaline phosphatase; ALT = alanine transaminase; ANCOVA = analysis of covariance; ItchRO = Itch Reported Outcome; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); LOCF = last observation carried forward; mITT = modified intent to treat; MMRM = mixed model for repeated measures; NA = not applicable; PedsQL = Pediatric Quality of Life Inventory; REML = restricted maximum likelihood; sBA = serum bile acid.

Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

Source: ICONIC Clinical Study Report.¹⁶

Analysis Populations

Table 12 summarizes the analysis populations in the ICONIC pivotal trial.

Table 12: Analysis Populations of ICONIC (LUM001-304) Study

ITT = intent to treat; mITT = modified intent to treat; sBA = serum bile acid.

Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

Source: ICONIC Clinical Study Report.¹⁶

Results

Patient Disposition

Thirty-six children with ALGS were screened for the ICONIC pivotal trial, 5 (13.9%) of whom were excluded due to screening failure. In total, 31 patients were enrolled into the open-label period of the study, during which 2 (6.5%) discontinued due to AEs. In total, 29 patients entered the RWD period and were randomized to continue maralixibat treatment (n = 13) or administered placebo (n = 16). A total of 28 patients (90.3%) completed the 48-week core study, with 1 (3.4%) patient discontinued due to an AE. After finalization of the core study, 23 patients (79.3%) consented to enter the LTE period at week 48 and 14 (60.9%) remained on study at the time of the interim report. A summary of patient disposition is presented in Table 13.

Baseline Characteristics

In the overall study population (N = 31), there were more males (19 of 31 [61.3%]) than females (12 of 31 [38.7%]) at baseline and in the maralixibat (9 of 13 [69.2%]) and placebo groups (10 of 16 [62.5%]) during the RWD period. The mean age in the overall study population was 5.4 years (range, 1 to 15 years) and was similar between the maralixibat and placebo groups. The majority of the patients were from Australia and France (9 of 31 [29.0%] each in the overall study population). The mean time since the original diagnosis of ALGS was 66.2 months in the overall study population, with 64.5 months in the maralixibat group and 73.2 months in the placebo group during the RWD phase. In the overall study population, 8 of 31 (25.8%) of patients had a family history of ALGS (1 of 13 [7.7%] and 7 of 16 [43.8%] in the maralixibat and placebo group, respectively). All enrolled patients had the Jagged-1 mutation present. During the RWD, 4 of 13 (30.8%) of patients in the maralixibat group had documented bile duct paucity (46.2% unknown), compared with 12 of 16 (75.0%) (18.8% unknown) in the placebo group. The majority of patients in the overall study population (29 of 31 [93.5%]) had used previous antipruritic treatment, with a similar distribution in the maralixibat and placebo groups.

Of note, the mean baseline sBA (μmol/L) was higher in the maralixibat group (317.97) than in the placebo group (249.56). ItchRO(Obs) weekly average morning severity score was similar between the 2 arms (2.88 versus 2.93). All liver biomarkers and enzymes (ALT, ALP, total and direct bilirubin) were consistently higher in the maralixibat group than in the placebo group. Mean baseline 7alpha-hydroxy-4-cholesten-3-one levels were higher in the maralixibat group (14.77 [standard deviation (SD) = 19.874] ng/mL) than in the placebo group (6.53 [SD = 8.728] ng/mL).

The baseline characteristics outlined in Table 14 are limited to those that are most relevant to this review or that were felt to affect the outcomes or interpretation of the study results.

Table 13: Patient Disposition in the ICONIC Trial

ALT = alanine transaminase; ARW = after randomized withdrawal; INR = international normalized ratio; ITT = intent to treat; LTE = long-term extension; mITT = modified intent to treat; MRX = maralixibat; NA = not applicable; PA = protocol amendment; sBA = serum bile acid; ULN = upper limit of normal.

Notes: Percentages are based on the number of patients in the safety population within each treatment period and treatment group. Assigned treatment groups are presented.

Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

^aAverage daily score greater than 2 on the ItchRO questionnaire (maximum possible daily score of 4) for 2 consecutive weeks in the screening period, before dosing. A daily score is the higher of the scores for the morning and evening ItchRO. The average daily score is the sum of all daily scores divided by the number of days the ItchRO was completed.

^bDecompensated cirrhosis (ALT 15 times greater than the ULN, INR greater than 1.5 [unresponsive to vitamin K therapy], albumin < 3.0 g/dL, history or presence of clinically significant ascites, variceal hemorrhage, and/or encephalopathy).

^cThe safety population includes all patients who were enrolled and received at least 1 dose of study drug.

^dThe ITT population includes all patients who were enrolled and received at least 1 dose of study drug.

^eThe mITT population includes all patients who were enrolled, received study drug through week 18, and had a reduction from baseline in sBA of greater than or equal to 50% at the weeks 12 or 18 measurement (sBA responder).

^fPatients are considered to have completed the treatment period if they were in the study at the time of the data cut-off.

^gPatients who did not consent to the optional long-term treatment extensions (PA3 or PA4) are not considered to have completed treatment.

Source: ICONIC Clinical Study Report.¹⁶

Table 14: Summary of Baseline Characteristics in the ICONIC Pivotal Trial

ALGS = Alagille syndrome; ALP = alkaline phosphatase; ALT = alanine transaminase; ItchRO(Obs) = Itch Reported Outcome (observer); sBA = serum bile acid; SD = standard deviation; UDCA = ursodeoxycholic acid.

Source: ICONIC Clinical Study Report.¹⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)¹⁷

Exposure to Study Treatments

Table 15 summarizes study drug exposure and compliance by study phase and by overall study population. Maralixibat doses are shown using maralixibat chloride. In the overall study population, the mean treatment duration was 944 days (SD = 587.07), with a mean average daily dose of 439.8 mcg/kg (SD = 133.47) once daily at the time of the data cut-off. During the RWD phase, the mean duration of study drug exposure was 29.9 days (SD = 4.09) in the maralixibat group and 29.1 days (SD = 2.94) in the placebo group. Study drug compliance was high (≥ 98.0%) throughout all phases of the study.

Concomitant Medications

Table 16 summarizes concomitant medications that treat pruritus by phase for the safety population. In all study phases, most patients (> 82% in all study phases) were taking at least 1 concomitant medication to treat pruritus. At study entry, 74.2% and 80.6% of patients were receiving stable doses of rifampicin and ursodeoxycholic acid, respectively. The most common medications used (> 10% in the open-label phase) included ursodeoxycholic acid, rifampicin, and phenobarbital. No additional concomitant medications with a frequency of greater than 10% were taken in any other study phase.

Efficacy

Efficacy end points in the ICONIC pivotal trial assessing changes from weeks 18 to 22 are shown in Table 17, and changes from baseline to weeks 18 or 48 are shown in Table 18.

Serum Bile Acid

In the ICONIC trial, the primary efficacy end point was the change in sBA during the 4-week RWD phase in the mITT population (patients with sBA reduction ≥ 50% at weeks 12 or 18). A total of 15 patients met this criterion and were analyzed in the primary end point (5 randomized to maralixibat; 10 to placebo by following a prespecified block randomization procedure). The LS mean difference in change from weeks 18 to 22 in sBA between the maralixibat and placebo groups was –117.28 μmol/L (95% CI, –211.699 to –23.103; P = 0.0464), in favour of maralixibat.

Change in sBA during the 4-week RWD phase was also assessed in the ITT population (Figure 2). The LS mean difference in change from weeks 18 to 22 in sBA between the maralixibat (n = 13) and placebo (n = 16) groups was –113.95 μmol/L (95% CI, –212.68 to –15.21; P = 0.0254), in favour of maralixibat.

In the overall ITT population (N = 31 patients), there was a decrease in sBA during the open-label phases from baseline to week 18 (secondary end point) with a mean change of –87.73 μmol/L (95% CI, –133.37 to –42.09; P = 0.0005) and baseline to week 48 (additional end point) with a mean change of –96.44 μmol/L (95% CI, –162.36 to –30.52; P = 0.0058).

The prespecified sensitivity analyses for change in sBA were consistent with the results of the primary efficacy analysis. Ancillary sensitivity analyses that applied 3 different imputation methods (LOCF, MMRM, and multiple imputation) to assesses mean changes in sBA from baseline to week 48 were consistent with the results of the primary efficacy analysis.

Subgroup analyses were generally consistent with the overall results of the pivotal trial.

In a post hoc analysis assessing data from baseline to week 48, sBA reductions of at least 50%, 60%, 70%, 80%, and 90% had associated ItchRO(Obs) reductions of –1.9, –2.1, –2.3, –2.8, and –2.7, respectively (Appendix 1).

Figure 2: Mean (Standard Error) Change From Baseline in sBA Over Time in the ICONIC Trial (ITT Population)

sBA = serum bile acid; SE = standard error; MRX = maralixibat; PBO = placebo.

Table 15: Study Drug Exposure and Compliance in the ICONIC Trial — Safety Population

ARW = after randomized withdrawal; b.i.d. = twice daily; LTE = long-term extension; q.d. = every day; SD = standard deviation.

Notes: Compliance (%) equals 100 times (number of days a dose was taken as prescribed divided by treatment duration in days) is calculated for each treatment period, where treatment duration (days) equals date of last dose during the treatment period minus date of first dose during the treatment plus 1 day. A dose was considered “taken as prescribed” if, during q.d. dosing, the patient took any amount of study drug, and during b.i.d. dosing the patient took any amount of both the morning and evening doses. Drug interruptions due to a patient being off-study, between protocol amendments, are not included in the calculation of compliance.

Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

Source: ICONIC Clinical Study Report.¹⁶

Table 16: Summary of Concomitant Medications That Treat Pruritus in the ICONIC Trial — Safety Population

ATC = Anatomic Therapeutic Chemical; MRX = maralixibat.

Notes: Patients were counted only once for each ATC or preferred term. Concomitant medications with a start date > 14 days after the patient's last dose of study drug are not included in the presentation. For patients with study drug interruptions, any concomitant medication that starts > 14 days after the last dose before the drug interruption and ended before the study drug was re-initiated will not be considered as a concomitant medication. Included medications are listed from a prespecified dictionary search of medications. Any medication taken during the specified analysis phase(s) will be included in each respective analysis phase and treatment group (where applicable). When the use of concomitant medications continues into subsequent analysis period(s), the medications are counted in each applicable analysis period.

Source: ICONIC Clinical Study Report.¹⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)¹⁷

Itch Reported Outcome (Observer)

In the ICONIC pivotal trial, the change from weeks 18 to 22 in ItchRO(Obs) weekly average morning severity score was a secondary end point. The LS mean difference between the maralixibat and placebo groups was –1.48 (95% CI, –2.12 to –0.84; P < 0.0001) in favour of maralixibat (Figure 3).

In the overall population, there was a decrease (improvement) in ItchRO(Obs) weekly average morning severity score from baseline to week 18 (secondary end point), with a mean change of –1.70 (95% CI, –2.05 to –1.36; P < 0.0001) and from baseline to week 48 (additional end point) with a mean change of –1.62 (95% CI, –2.12 to –1.12; P < 0.0001).

The prespecified sensitivity analyses for ItchRO(Obs) weekly average morning severity score were consistent with the results of the primary efficacy analysis. Additional sensitivity analyses that applied different imputation methods (LOCF, MMRM, and multiple imputation) to assesses mean changes in ItchRO(Obs) weekly average morning severity score from baseline to week 48 were consistent with the results of the primary efficacy analysis.

Subgroup analyses were generally consistent with the overall results of the pivotal trial, except for where there was a very small number of patients in a given category (n ≤ 4).

Similar results were observed for the ItchRO(Obs) weekly average evening severity score, 4-week average morning severity score, 4-week average evening severity score, and weekly average severity score (daily average of morning and evening scores) in the ITT population, with use of an ANCOVA model.

Itch Reported Outcome (Patient)

A total of 14 patients met the age cut-off for completion of the ItchRO(Pt) (≥ age 9 years or ≥ age 5 years, with the assistance of their caregiver) in the pivotal trial. The LS mean difference between the maralixibat and placebo groups for the change from weeks 18 to 22 in ItchRO(Pt) weekly average morning severity score was –1.98 (–3.01 to –0.97; P = 0.0013) in favour of maralixibat.

In the overall population, there was a decrease (improvement) in ItchRO(Pt) weekly average morning severity score from baseline to week 18 (secondary end point) with a mean change of –2.07 (95% CI, –2.65 to –1.50; P < 0.0001) and from baseline to week 48 (additional end point) with a mean change of –2.25 (95% CI, –2.84 to –1.67; P < 0.0001).

The sensitivity analysis for change in ItchRO(Pt) weekly average morning severity score was consistent with the results of the primary efficacy analysis.

Liver Biomarkers and Enzymes

Alkaline Phosphatase

In the ICONIC pivotal trial, the change from weeks 18 to 22 in ALP was a secondary end point. The LS mean difference between the maralixibat and placebo groups was 10 (95% CI, –52.6 to 72.6; P = 0.7455) U/L.

In the overall population, the mean change in ALP from baseline to week 18 (secondary end point) was –27.8 (95% CI, –72.8 to 17.2; P = 0.2163) U/L and from baseline to week 48 (additional end point) the mean change was –51.5 (95% CI, –95.7 to –7.2; P = 0.0242) U/L.

The sensitivity analyses for change in ALP were consistent with the results of the primary efficacy analysis.

Alanine Transaminase

In the ICONIC pivotal trial, the change from weeks 18 to 22 in ALT was a secondary end point. The LS mean difference between the maralixibat and placebo groups was 15.1 (95% CI, –25.1 to 55.2; P = 0.4472) U/L.

In the overall population, the mean change in ALT from baseline to week 18 (secondary end point) was –1.3 (95% CI, –33.4 to 30.9; P = 0.9358) U/L and the mean change from baseline to week 48 (additional end point) was 17.5 (95% CI, –15.0 to 50.2; P = 0.2787) U/L.

The sensitivity analysis for change in ALT was consistent with the results of the primary efficacy analysis.

Total Bilirubin

In the ICONIC pivotal trial, the change from weeks 18 to 22 in total bilirubin was a secondary end point. The LS mean difference between the maralixibat and placebo groups was –0.14 (–0.88 to 0.60; P = 0.7000) mg/dL.

In the overall population, the mean change in total bilirubin from baseline to week 18 (secondary end point) was –0.47 (95% CI, –1.01 to 0.08; P = 0.0893) mg/dL and the mean change from baseline to week 48 (additional end point) was 0.03 (95% CI, –0.70 to 0.76; P = 0.9285) mg/dL.

The sensitivity analysis for change in total bilirubin was consistent with the results of the primary efficacy analysis.

Direct Bilirubin

In the ICONIC pivotal trial, the change from weeks 18 to 22 in direct bilirubin was a secondary end point. The LS mean difference between the maralixibat and placebo groups was –0.02 (95% CI, –0.56 to 0.53; P = 0.9517) mg/dL.

In the overall population, the mean change in direct bilirubin from baseline to week 18 (secondary end point) was –0.50 (95% CI, –0.90 to –0.11; P = 0.0139) mg/dL and the mean change from baseline to week 48 (additional end point) was –0.24 (95% CI, –0.56 to 0.09; P = 0.1489) mg/dL.

The sensitivity analysis for change in direct bilirubin was consistent with the results of the primary efficacy analysis analyzing direct bilirubin levels change from weeks 18 to 22 and from baseline to week 18.

Change in Body Height and Weight

In the overall study population, there was an increase from baseline in mean height z scores at all time points after weeks 3 to 252. The mean change from baseline to week 100 (LOCF) in height z scores (additional end point) was 0.25 (95% CI, –0.86 to 2.04; P = 0.0216).

In the overall study population, there were no major changes from baseline in mean weight z scores at any time point. The mean change from baseline to week 100/LOCF in weight z scores (additional end point) was –0.05 (95% CI, –0.12 to 0.23; P = 0.5306).

Health-Related Quality of Life

PedsQL Total Score (Parent)

In the ICONIC pivotal trial, the LS mean difference from weeks 18 to 22 in the PedsQL total score (parent) was an additional end point. The LS mean difference between the maralixibat and placebo groups was 2.33 (95% CI, –10.08 to 14.75; P = 0.7018).

In the overall population, the mean change in the PedsQL total score (parent) from baseline to week 18 (additional end point) was 10.73 (95% CI, 4.43 to 17.03; P = 0.0016) and the mean change from baseline to week 48 (additional end point) was 8.94 (95% CI, 1.53 to 16.35; P = 0.0200).

The sensitivity analysis for change in PedsQL total score was consistent with the results of the primary efficacy analysis.

PedsQL Multidimensional Fatigue Scale Score

In the ICONIC pivotal trial, the change from weeks 18 to 22 in the PedsQL Multidimensional Fatigue Scale score (parent) was an additional end point. The LS mean difference between the maralixibat and placebo groups was 14.03 (95% CI, –2.78 to 30.84; P = 0.0966).

In the overall population, the mean change in the PedsQL Multidimensional Fatigue Scale score (parent) from baseline to week 18 (additional end point) was 20.39 (95% CI, 8.91 to 31.87; P = 0.0013) and the mean change from baseline to week 48 (additional end point) was 20.30 (95% CI, 8.98 to 31.63; P = 0.0013).

The sensitivity analysis for change in the PedsQL Multidimensional Fatigue Scale score was consistent with the results of the primary efficacy analysis.

Figure 3: Mean Change From Baseline in ItchRO(Obs) Weekly Average Morning Severity Score Over Time in the ICONIC Trial (ITT Population)

SE = standard error; MRX = maralixibat; PBO = placebo.

Table 17: Summary of Key Efficacy Results in the ICONIC Trial — Randomized Withdrawal Phase, Weeks 18 to 22

ALT = alanine transaminase; ALP = alkaline phosphatase; ANCOVA = analysis of covariance; CI = confidence interval; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); ITT = intent to treat; LS = least square; mITT = modified intent to treat; PedsQL = Pediatric Quality of Life Inventory; sBA = serum bile acid; SE = standard error.

Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

^aThe primary efficacy end point was conducted on the mITT population, including all patients who were enrolled, received study drug through week 18, and had a reduction from baseline in sBA of ≥ 50% at the week 12 or 18 measurement (sBA responder). Estimates are from a mixed model with treatment group as a fixed effect and baseline value as a covariate. Analysis uses a tabulation of fitted summary statistic from ANCOVA in the mITT population.

^bThe P value was not adjusted for multiple testing (i.e., the type I error rate was not controlled).

^cEstimates are from a mixed model with treatment group as a fixed effect and baseline value as a covariate.

^dAnalyses for the primary efficacy outcome variable were also done on the ITT population.

^eItchRO scores have a range from 0 to 4, a higher score indicating increasing itch severity.

^fPer protocol, only administered to 14 patients at baseline. Children aged at least 9 years complete the ItchRO(Pt). Children aged 5 to 8 years complete the instrument with their caregiver's assistance. There is no ItchRO(Pt) report for patients under age 5 years.

Source: ICONIC Clinical Study Report.¹⁶

Table 18: Summary of Efficacy Results From the ICONIC Trial, Overall Maralixibat — Baseline to Week 18 and Baseline to Week 48 (ITT Population)

ALT = alanine transaminase; ALP = alkaline phosphatase; CI = confidence interval; ITT = intent to treat; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); LOCF = last observation carried forward; PedsQL = Pediatric Quality of Life Inventory; sBA = serum bile acid; SD = standard deviation; SE = standard error.

^aThe P value was not adjusted for multiple testing (i.e., the type I error rate was not controlled).

^bStudent’s t test used to test if mean change is statistically significant.

^cSecondary end point.

Source: ICONIC Clinical Study Report.¹⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)¹⁷

Harms

Refer to Table 19 for harms data.

Adverse Events

The incidence of AEs was similar during the open-label phase, the after RWD phase, and the LTE phase, with at least 25 of 29 patients (86.2%) experiencing any AEs in these treatment periods (Table 17). During the RWD phase, patients who stayed on maralixibat had a lower incidence of AEs (7 of 13 patients [38%]) compared with patients on placebo (12 of 16 patients [75%]). The most frequently reported AEs (> 30% in at least 1 phase) were abdominal pain, pyrexia, diarrhea, nasopharyngitis, vomiting, cough, and pruritus.

Serious Adverse Events

The incidence of SAEs was similar during the open-label phase (4 of 31 patients [12.9%]) and after RWD phase (5 of 29 patients [17.2%]) and was slightly higher during the LTE phase (6 of 23 patients [26.1%]). During the RWD phase, SAEs were reported for 1 of 13 patients who stayed on maralixibat and 1 of 16 patients on placebo. None of the SAEs were considered related to study medication.

Withdrawals Due to AEs

A total of 6 patients (2 each in the open-label, after RWD, and LTE phases) experienced AEs leading to study drug discontinuation.

Mortality

No deaths were noted during the study.

Notable Harms

Diarrhea

AEs of diarrhea were similar during the open-label phase (13 of 31 patients [41.9%]) and the LTE phase (10 of 23 patients [43.5%]) and were lower during the after RWD phase (7 of 29 patients [24.1%]). During the RWD phase, patients who stayed on maralixibat had a similar incidence of events of diarrhea (1 of 13 patients [7.7%]) compared with those receiving placebo (2 of 16 patients [12.5%]).

Fat-Soluble Vitamin Deficiency

AEs associated with FSV deficiency were similar during the open-label phase (7 of 31 patients [22.6%]) and LTE phase (6 of 23 patients [26.1%]) and were lower during the after RWD phase (1 of 29 patients [3.4%]).

Elevated Transaminases

AEs associated with elevated transaminases were only reported during the LTE phase (4 of 23 patients [17.4%]).

Elevated Bilirubin

AEs associated with elevated bilirubin were only reported during the after RWD phase (1 of 29 patients [3.4%]).

Critical Appraisal

Internal Validity

The ICONIC trial was a phase IIb, placebo-controlled study with open-label extension. Overall, the study, including both the RWD phase and the extension phase, showed considerable evidence with moderate certainty that maralixibat improved sBA, and itch-related outcome. The study is an exploratory trial, with small sample size, Uncertainty remains as to whether maralixibat could improve liver function and patients’ HRQoL. The beneficial effect may only exist in a certain proportion of patients with ALGS in the trial. In particular, after open-label treatment through to week 18, patients remaining in the study were block randomized in a 1:1 ratio to receive maralixibat or placebo in the 4-week RWD phase. The primary efficacy analysis on the changes in mean sBA at week 22 was based on 15 of 29 patients with at least 50% reduction of sBA (responder) at week 12 or 18 (n = 5 in maralixibat, n = 10 in placebo). However, the magnitude of mean change was similar between the selected 15 patients (–117) versus the overall 29 patients (–114), indicating that the treatment effect in the overall randomized study population was primarily driven by those responders representing 50% (15 of 29) of the total study population.

The definition of the pruritus end point used to assess efficacy was not prespecified in the protocol; however, analyses performed using different summaries of the daily pruritus scores (i.e., morning scores, evening scores, weekly average of the 2 daily scores) were in line with the primary analysis.

Subgroup analyses were difficult to interpret due to small sample sizes and a lack of control of multiplicity. During the open-label phases of the pivotal trial, patients’ and/or caregivers’ knowledge of treatment assignment may have biased subjective outcomes such as ItchRO(Obs), ItchRO(Pt), and PedsQL in favour of maralixibat. Reporting of harms could also have been biased, potentially in favour of maralixibat. Discontinuation through to week 48 of the pivotal trial was low, with 3 of 31 patients (9.7%) discontinuing due to an AE through to week 48, with 2 patients discontinuing during the initial open-label phase, and 1 patient discontinuing after the RWD phase.

Differences in baseline characteristics between patients in the maralixibat and placebo group were noted in the ICONIC trial. Those in the maralixibat group were slightly younger, had a smaller proportion of patients with family history of ALGS, had a smaller proportion with known bile duct paucity, and had higher sBA, ALT, and total bilirubin values than those in the placebo group. The clinical experts consulted on this review stated that these baseline differences were not likely to lead to biased results. The clinical experts noted that patients in the maralixibat group may have had a higher degree of disease severity than those in the placebo group as indicated by the higher sBA, ALT, and bilirubin values, which may have biased results in favour of placebo.

Change in sBA was assessed as a primary efficacy end point in the clinical trial. Descriptive posthoc data from the ICONIC pivotal trial found that reductions in sBA from baseline to week 48 were associated with reductions in mean ItchRO(Obs) weekly average morning severity scores (Appendix 1).The data may show an association between sBA and ItchRO in some patients, but as the data were descriptive in nature and the assessment was conducted posthoc on a small number of patients (n = 28), it is unclear the extent to which sBA levels may be associated with pruritus in patients with cholestatic liver diseases.

Table 19: Summary of Harms Results in the ICONIC Trial

AE = adverse event; ALT = alanine transaminase; FSV = fat-soluble vitamin; SAE = serious adverse event.

^aFrequency ≥ 10%.

^bReported in 2 or more patients.

Source: ICONIC Clinical Study Report.¹⁶ (Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.)¹⁷

Pruritus was assessed using the ItchRO(Obs) and ItchRO(Pt) tool as a secondary outcome measured by patients and/or caregivers using an electronic diary twice per day (morning and evening). Procedures were in place to ensure proper completion of the electronic diary, including training patients and caregivers in the use of the electronic diary during the screening visit, and requiring caregivers and patients 9 years of age or older to have completed at least 10 electronic diary reports (morning and/or evening) during each of 2 consecutive weeks of the screening period; re-training on the use of the diary was offered as appropriate.

There were minimal missing data for the ItchRO(Obs) weekly average morning severity scores during the RWD phase, with only 1 randomized patient (maralixibat group) having a missing ItchRO(Obs) weekly average morning severity score at week 22. Only 9 patients qualified to complete the ItchRO(Pt) tool, with results that were supportive of those attained by the ItchRO(Obs). Predefined sensitivity analyses were conducted to assess the impact of missing data for ItchRO outcomes, and their results were consistent with the primary analysis.

Limited available evidence on the ItchRO tool as assessed among pediatric patients with cholestatic liver disease suggests that it is likely valid and reliable. The clinical experts consulted on this review noted that an MID of 1 for the ItchRO tool is clinically meaningful; however, the experts noted that such tools are not commonly used in clinical practice.

HRQoL was assessed using the PedsQL as an additional efficacy outcome in the pivotal trial. Psychometric properties of the instrument have not been assessed in pediatric patients with ALGS. MID estimates for the PedsQL have not been established in pediatric patients with ALGS; however, MID estimates for the PedsQL Generic Core Scales ranged from 4 to 5 points among a sample of children with and without chronic health conditions, which aligns with the clinical experts’ expectations of a clinically meaningful change. It should be noted that the number of patients assessed for the PedsQL Multidimensional Fatigue Scale score was low during the RWD phase, with 9 of 13 patients (69.2%) in the maralixibat group and 12 of 16 patients (75.0%) in the placebo group contributing to the analysis of mean change from weeks 18 to 22. The impact of missing data on this outcome is unclear in the absence of sensitivity analyses.

Various efficacy outcomes deemed important to the clinical experts were not assessed in the pivotal trial, including liver transplant or TFS, time to liver event, cirrhosis, or renal disease as measured by the gradual loss of kidney function.

External Validity

The clinical experts consulted on this review stated that patients included in the ICONIC trial generally align with the selection criteria for candidates for maralixibat, although patients with mild cholestatic pruritus would not necessarily be excluded from treatment in clinical practice. Nonetheless, the clinical experts did not expect the exclusion of these patients to significantly affect the generalizability of the patient population in this study. The clinical trial only enrolled patients 12 months of age or older with a Jagged-1 mutation; however, the clinical experts note that the trial results would be applicable to patients less than 12 months of age as well as patients with a NOTCH2 mutation, respectively. The experts noted that patients in clinical practice are likely to also receive confirmatory genetic testing, which was not conducted in the pivotal trial. Although race and ethnicity data were not assessed in the pivotal trial, the clinical experts stated that the results would be applicable to the patient population in Canada. The remaining baseline patient characteristics were similar to the indicated patient population in clinical practice in Canada.

The efficacy outcomes measured in the study were of clinical importance to patients and clinicians, including change in sBA. However, the clinical experts noted that the change in sBA is not often assessed in clinical practice due to high costs and logistical limitations, as sBA testing is often sent to specialized laboratories and is not readily available in all gastroenterology practice settings. Change in pruritus and HRQoL outcomes were described as important outcomes by the clinical experts, as well as in the patient and clinician group input received, which noted the need for a treatment that reduces the severity and frequency of pruritus as well as reducing patient and caregiver fatigue. Such outcomes were captured by the ItchRO and the PedsQL tools in the ICONIC trial. The clinical experts consulted for this review indicated that although tools such PedsQL are frequently used in clinical trials they are not typically used in clinical practice. Furthermore, the double-blind phase in the pivotal ICONIC trial was 4 weeks in length, limiting the ability to assess the long-term efficacy and safety of maralixibat compared with placebo for the indicated dose of 380 mcg/kg/day. The durability of treatment effect and long-term safety still remain somewhat uncertain, although the beneficial effect is likely maintained, as reflected by the long-term curve on changes of sBA and ItchRO(Obs).

While maralixibat has been approved by Health Canada for use in patients for the treatment of cholestatic pruritus in patients with ALGS, the ICONIC trial only enrolled patients aged between 1 and 15 years. As such, there is an absence of efficacy and safety data assessing maralixibat versus placebo among patients younger than 12 months due to the challenges of conducting a controlled clinical trial in this age group. However, the trial results are expected to be applicable to patients less than 12 months of age based on clinical experts’ feedback. Furthermore, during the LTE phase of the ICONIC pivotal trial (week 103) eligible patients could have received a dose of maralixibat of up to 760 mcg/kg/day (given as twice-daily doses of 380 mcg/kg) which is outside of the proposed Health Canada indication of 380 mcg/kg /day. As such, efficacy and safety data after this period is not aligned with the recommended dose.

GRADE Summary of Findings and Certainty of the Evidence

Methods for Assessing the Certainty of the Evidence

For pivotal studies and RCTs identified in the sponsor’s systematic review, GRADE was used to assess the certainty of the evidence for outcomes considered most relevant to inform CADTH expert committee deliberations, and a final certainty rating was determined as outlined by the GRADE Working Group:^14,15

• High certainty: We are very confident that the true effect lies close to that of the estimate of the effect.

• Moderate certainty: We are moderately confident in the effect estimate — the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. We use the word “likely” for evidence of moderate certainty (e.g., “X intervention likely results in Y outcome”).

• Low certainty: Our confidence in the effect estimate is limited — the true effect may be substantially different from the estimate of the effect. We use the word “may” for evidence of low certainty (e.g., “X intervention may result in Y outcome”).

• Very low certainty: We have very little confidence in the effect estimate — the true effect is likely to be substantially different from the estimate of effect. We describe evidence of very low certainty as “very uncertain.”

• For RCTs: Following the GRADE approach, evidence from RCTs started as high-certainty evidence and could be rated down for concerns related to study limitations (i.e., internal validity or risk of bias), inconsistency across studies, indirectness, imprecision of effects, and publication bias.

When possible, certainty was rated in the context of the presence of an important (nontrivial) treatment effect; if this was not possible, certainty was rated in the context of the presence of any treatment effect (i.e., the clinical importance is unclear). In all cases, the target of the certainty of evidence assessment was based on the point estimate and where it was located relative to the threshold for a clinically important effect (when a threshold was available) or to the null.

Results of GRADE Assessments

Table 2 presents the GRADE summary of findings for maralixibat versus placebo.

LTE Studies

The pivotal ICONIC trial included an LTE phase described in the systematic review section of this report. No other LTE studies were submitted.

Indirect Evidence

No indirect comparisons were conducted comparing maralixibat with other comparators for this submission.

Studies Addressing Gaps in the Pivotal and RCT Evidence

Contents within this section have been informed by materials submitted by the sponsor. The following has been summarized and validated by the CADTH review team.

Table 20: Summary of Gaps in the Evidence

ALGS = Alagille syndrome; CI = confidence interval; EFS = event-free survival; GALA = Global ALagille Alliance; HR = hazard ratio; LTE = long-term extension; RCT = randomized controlled trial; SAP = statistical analysis plan; sBA = serum bile acid; TFS = transplant-free survival.

Source: Sponsor’s Summary of Clinical Evidence,¹⁷ GALA SAP.⁴³

Description of Studies

• Patient-level data from the 3 long-term studies of patients with ALGS treated with maralixibat (the LUM001-305, LUM001-303, and LUM001-304 studies), and a natural-history comparison study (GALA registry) have been summarized to provide evidence regarding maralixibat for the treatment of cholestatic pruritus in patients with ALGS, aged 2 months and older.

• The following gaps in the pivotal and RCT evidence have been identified, which the studies included in this section address.

  • Population: Although the LUM001-304 (ICONIC) trial did not enrol children aged younger than 12 months with ALGS, there remains an unmet need for the treatment in infants since ALGS is typically diagnosed within the first months of life,^1,12 and this population suffers from the clinical burden of cholestasis from a very young age.^6,44 In this regard, infants with ALGS may present with the key features of cholestasis: elevations in sBAs, pruritus, hypercholesterolemia, xanthomas, and failure to thrive. The most burdensome clinical feature of cholestasis in ALGS is cholestatic pruritus. Older infants may exhibit typical scratching behaviour and skin mutilations. In younger infants, before development of motor coordination to scratch properly, children may exhibit coarse scratching movement, fussiness, and irritability as signals of pruritus. Therefore, it is clear that pruritus represents a medical need in infancy.⁴⁰ However, given the rarity of the disease and the differences in presentation of pruritus in this age group, it would be challenging to generate new evidence to demonstrate clinical efficacy in the treatment of cholestatic pruritus in the subgroup of patients with ALGS who are less than age 12 months.

  • Outcomes: Clinical end points, such as long-term EFS and TFS in patients with ALGS, were not evaluated in the ICONIC study.

  • Duration: In a life-threatening condition such as ALGS, long-term, randomized, placebo-controlled studies are not possible because the risks of forgoing other treatment options that a patient would otherwise receive outside of a study are so high that it is not feasible or ethical to ask patients to accept those risks.

Patient-Level Data From the LUM001-305 (IMAGINE), LUM001-303 (ICONIC), and LUM001-304 (IMAGINE-II) Studies

To identify predictors of long-term EFS and TFS in patients with ALGS, individual patient data from 3 long-term clinical trials of maralixibat, with up to 6 years of follow-up, were integrated. All trials used ItchRO(Obs) as a primary end point, with a less than 1 point reduction defined as clinically meaningful. Inclusion and exclusion criteria, dosing, and end points were similar for all 3 trials.

Table 21: Details of LTE Studies: LUM001-303 (IMAGINE), LUM001-304 (ICONIC), and LUM001-305 (IMAGINE-II) Studies

AE = adverse event; ALGS = Alagille syndrome; ALP = alkaline phosphatase; ALT = alanine transaminase; AST = aspartate transaminase; FSV = fat-soluble vitamin; GGT = gamma-glutamyl transferase; ItchRO = Itch Reported Outcome; ItchRO(Obs) = Itch Reported Outcome (observer); ItchRO(Pt) = Itch Reported Outcome (patient); LOCF = last observation carried forward; LTE = long-term extension; mITT = modified intent to treat; NA = not applicable; NR = not reported; RWD = randomized withdrawal; SAE = serious adverse event; sBA = serum bile acid; ULN = upper limit of normal.

^aThe IMAGO study (NCT01903460) had a similar study design as the ITCH trial but was conducted in the UK and patients were eligible to enrol in an LTE study (IMAGINE; NCT 02047318) for up to 288 weeks.

^bThe ICONIC study (NCT02160782) was a double-blind, randomized-withdrawal study of maralixibat in children with ALGS. The ICONIC study was a placebo-controlled, phase IIb study with an open-label extension in children with ALGS (1 to 18 years of age) with more than 3 times the ULN of sBA levels, and intractable pruritus. After 18 weeks of maralixibat, patients were randomized 1:1 to continue maralixibat or receive placebo for 4 weeks, after which all patients received open-label maralixibat up to week 48. Patients could then enrol in an LTE study.

^cThe ITCH study (NCT02057692) was a double-blind, RCT conducted in North America of maralixibat for 13 weeks to evaluate the safety and efficacy in the reduction of pruritus and sBA by 3 doses of maralixibat or placebo. Patients enrolled in the ITCH trial were then eligible to enrol in a long-term, open-label extension study (IMAGINE-II) in which all patients received maralixibat (NCT02117713) for up to 220 weeks.

Source: Clinicaltrials.gov — NCT02047318 (LUM001-303),⁴⁶ NCT02160782 (LUM001-304),⁴⁷ and NCT02117713 (LUM001-305) trials.⁴⁸

Populations

Inclusion criteria for this analysis:

• Patients who were on maralixibat 48 weeks from the first dose and had laboratory results at 48 weeks

• No prior clinical event

Outcomes

EFS was defined as first occurrence of any of the following listed events:

1. Liver transplant

2. SBD

3. Liver decompensation (variceal bleeding, ascites requiring therapy)

4. Death

TFS was defined as absence of liver transplant or death.

Patients were followed from the start of maralixibat treatment for a minimum of 53 weeks and a maximum of 380 weeks.

Interventions

• Maralixibat was administered in the LUM001-303 (IMAGINE) and LUM001-305 (IMAGINE-II) studies, up to 266 mcg/kg/day (doses presented as maralixibat free base).

• In the LUM001-304 (ICONIC) study, maralixibat was administered to a final dose of up to 380 mcg/kg/day. Patients were increased to twice-daily dosing starting after week 103.

Statistical Analysis

The statistical analyses conducted are as follows:⁴⁹

• Forty-six predictors were considered, which included, but were not limited to, total bilirubin at week 48, total sBA at week 48, pruritus (as assessed by ItchRO[Obs] 0 to 4 scale) change from baseline to week 48, and age at initiation of maralixibat.

• Harrell’s concordance statistic (C-statistic) was computed for each predictor, with the C-statistic defined as the proportion of observations that the variable can order correctly in terms of survival times indicating the goodness of fit.

• Once variables and time points with the highest C-statistics were identified, a survival analysis using a Cox regression model for each result was carried out to confirm that each predictor (treated as a continuous variable) was a statistically significant predictor of outcomes of interested assessed at alpha = 0.05 confidence level for 2-sided comparisons.

• Variables with a value equal to or greater than 0.7 (indicating a good model) were selected for further analysis.

• Cut-offs for each variable were determined via a grid search across the range of values.

• Statistical comparisons between the cut-off groups were calculated using a log-rank test.

A total of 84 patients were enrolled across all 3 clinical trials and their LTE studies; however, only 76 patients with complete data, including laboratory values at baseline and week 48 (with a window of less than 12 weeks to more than 6 weeks), were included in integrated analyses.

Results

Baseline Characteristics

This analysis included 76 patients treated with maralixibat, with a median follow-up of 266 weeks (range, 53 to 380). Median duration of maralixibat treatment was 4.7 years (interquartile range [IQR], 1.6 to 4.8). The median age (IQR) of 76 patients was 70 months (33 to 126). The median sBA (IQR) at baseline was 184 µmol/L (78 to 361 µmol/L). The median ItchRO(Obs) score at baseline (IQR) was 2.7 (2.1 to 3.1). The median total bilirubin, ALT, and gamma-glutamyl transferase (GGT) at baseline (IQRs) were 0.03 mmol/L (0.01 to 0.09 mmol/L), 134 U/L (95 to 193 U/L), and 392 U/L (188 to 751 U/L), respectively.

Predictors of EFS

Sixty out of 76 patients remained event-free at the time of analysis. Sixteen patients experienced clinical events: liver transplant (n = 10), decompensation (n = 3), death (n = 2), and SBD (n = 1).

Kaplan-Meier analyses demonstrated that at week 48, lower total bilirubin levels (< 0.07 mmol/L versus ≥ 0.07 mmol/L) and lower sBA levels (< 200 µmol/L versus ≥ 200 µmol/L) were associated with increased EFS (90% versus 43%; P < 0.0001, and 85% versus 49%; P = 0.0010, respectively). A larger reduction in ItchRO(Obs) scores from baseline to week 48 (> 1 point reduction versus ≤ 1 point reduction) and a higher age at initiation of maralixibat were also associated with improved EFS (88% versus 57%; P = 0.0046, and 83% versus 57%; P = 0.0059, respectively) (Figure 4).

Week 48 sBA and change from baseline to week 48 in pruritus (ItchRO[Obs]) were identified as predictors of EFS. Specifically, 6-year EFS improved with a clinically meaningful greater than 1-point ItchRO(Obs) reduction from baseline to W48 (88% versus 57%; P = 0.005) and W48 sBA < 200 µmol/L (85% versus 49%; P = 0.001). These 2 variables had high C-statistics over time, indicating that these cut-offs were stable predictors for 2 to 5 additional years after 48 weeks of maralixibat treatment. These variables were similarly predictive for TFS (Table 22).

Table 22: Predictors of EFS and TFS in Patients With ALGS Treated With Maralixibat

ALGS = Alagille syndrome; C-statistic = concordance statistic; EFS = event-free survival; ItchRO(Obs) = Itch Reported Outcome (observer); sBA = serum bile acid; TFS = liver transplant-free survival.

Source: Sponsor’s Clinical Summary.¹⁷

Figure 4: KM Estimates of 6-Year EFS in Children With ALGS Treated With Maralixibat: sBA Levels at Week 48 and Change in Pruritus (ItchRO[Obs]) From Baseline to Week 48

ALGS = Alagille syndrome; EFS = event-free survival; ItchRO = Itch Reported Outcome; ItchRO(Obs) = Itch Reported Outcome (observer); KM = Kaplan-Meier; sBA = serum bile acid.

Source: Sokol (2021),⁵⁰ Sokol (2023).⁴⁹

Discussion

In patients with ALGS, predictors of EFS with maralixibat treatment include sBA (at week 48) and pruritus reduction (from baseline to week 48). These potential markers of disease may help identify disease progression in patients with ALGS treated with maralixibat, and inform long-term outcomes. As pruritus is a leading cause of liver transplant in patients with ALGS, these data also demonstrated that improvements in pruritus with maralixibat are significantly associated with improved EFS and TFS. These data identify potential prognostic markers that may better inform patient and provider discussions of clinical outcomes in patients receiving maralixibat treatment.

One limitation pointed out by the sponsor is the small sample size, although it is considered large for rare diseases. Another limitation the sponsor stated is the absence of liver histology or data from non-invasive techniques to assess hepatic fibrosis. According to the sponsor, it is unclear if improvements in EFS and TFS observed in these analyses were due to improvements in pruritus, which could be the major indication for transplant, or due to improvement or stabilization of hepatic fibrosis and consequent portal hypertension related to the reduction in hepatic retention of toxic bile acids. Lastly, only patients who have shown benefit from maralixibat — i.e., improved pruritus, laboratory parameters, or nutrition — were maintained in the LTE studies with sufficient follow-up that allowed for examination of long-term outcomes. This selection bias may contribute to overestimation of long-term outcomes.

Natural-History Comparison Study: Patients Treated With Maralixibat Versus GALA Natural-History Cohort

A natural-history comparison study was conducted to compare disease outcomes between 2 groups of patients with ALGS: 1) maralixibat-treated study patients and 2) patients who are brown adipose tissue inhibitor–naive from a natural-history global clinical research database (GALA).

The GALA clinical research database, established in 2018, is the only global ALGS database with a robust data collection and appropriate quality-assurance measures in place. GALA is recognized internationally by academic and tertiary referral liver transplant centres and is led by Dr. Binita Kamath at The Hospital for Sick Children in Toronto, with a high level of participation. It is supported by more than 100 physicians, surgeons, scientists, and research coordinators from 35 countries, who contribute more than 1,400 patient records into the clinical database. The GALA clinical research database includes clinical and laboratory data as well as disease outcomes.¹⁰

Study Design and Objectives

The study is a global, open-label, nonrandomized, external-control study to compare time-to-event clinical outcomes in patients with ALGS treated with maralixibat with selected external controls from the international registry of the GALA registry. This natural-history comparison followed a prespecified SAP. The statistical team performing the analysis remained blinded to treatment outcomes before the selection of patients from the GALA registry.

The objective is to evaluate the effect of long-term maralixibat treatment on clinical outcomes in patients with ALGS (duration of follow-up is up to 7 years). The prespecified hypothesis is that the time to a clinical event in patients treated with maralixibat is delayed compared with natural-history controls.

There are 2 cohorts in the study: patients with ALGS treated with maralixibat in the LUM001-301, LUM001-302, LUM001-303, LUM001-304, and LUM001-305 studies, and those contained within the natural history or standard of care cohort (GALA registry). The GALA cohort (standard of care) included patients with ALGS who aligned with maralixibat key study entry criteria (listed in the following section) from the registry.

Populations

The GALA registry includes multiple visits per patient from the date of diagnosis of ALGS disease until the patient has an event. Patients are otherwise in follow-up, and the last follow-up will be the date of analysis. Selection of patients for the GALA cohort (standard of care) followed a stepwise procedure of patients’ or visit start times to align with the maralixibat cohort, as follows.

• Identification of patients and visits within the GALA registry was selected by the following:

  • inclusion criteria:

    • age at inclusion: aged older than 12 months and younger than 18 years

    • cholestasis, defined by 1 or more of the following:

      • total sBA greater than 3 times ULN

      • conjugated or direct bilirubin greater than 0.01 mmol/L

      • total bilirubin greater than 0.02 mmol/L

      • GGT greater than 3 times ULN

  • exclusion criteria:

    • ALT greater than 15 times ULN

    • clinical event, defined as liver transplant, SBD, liver decompensation, or death before inclusion

    • diagnosis of hepatocellular carcinoma by biopsy-proven histopathology

    • born before 1997

    • participation in a maralixibat study at any time

    • patients with only a single visit that do not have any follow-up information available.

• Selection of the start of follow-up: A patient may be eligible with multiple visits at different time points. The best visit to represent baseline (i.e., the start of follow-up aligning with maralixibat studies) was selected by maximum likelihood methods. To be specific, the maximum estimated prediction within a patient, based on a stratified logistic regression analysis of all eligible visits, was compared with the maralixibat cohort (yes or no), including covariates of age and total bilirubin.

• Balance assessment: The balance between the maralixibat cohort and the GALA cohort at baseline was assessed — i.e., the 2 cohorts should show comparable distributions of the variables such as age, sex, bilirubin (total and direct or conjugated), GGT, and ALT before selection was completed and before describing effects of treatment on the outcome events. If the 2 cohorts were not well balanced, the analysis was repeated with adjustment for confounders. In this case, inverse probability of treatment weights methods were applied using the estimated weights from a propensity score method.

The total number of patients in the GALA clinical research database at the cut-off for the GALA cohort selection and analyses was 1,438, with a total of 12,535 visits. Following the prespecified selection process, a total of 469 patients with 3,906 visits were included in the GALA control group (Figure 5). The maralixibat clinical study data included in this natural-history comparison comprised the aggregated data from all patients treated with maralixibat from the long-term maralixibat ALGS program (N = 84), with follow-up data up to 6 years.

Outcomes

EFS: First occurrence of any of the following listed events (maralixibat versus natural history):

1. Liver transplant

2. SBD

3. Liver decompensation (variceal bleeding, ascites requiring therapy)

4. Death

A sensitivity analysis included confirmed hepatocellular carcinoma by biopsy-proven histopathological diagnosis as an additional event.

Figure 5: Selection of the GALA Cohort (Standard-of-Care Control Group)

ALT = alanine transaminase; GALA = Global ALagille Alliance; mo = month; ULN = upper limit of normal; yr = year.

Note: Asterisks indicate that, to avoid immortal time bias, all eligible patients must have had 3, 6, or 12 moths of follow-up.

^a Among the 490 GALA eligible patients, only 469 patients had all covariates needed to perform the maximum likelihood selection.

Source: Sponsor’s Clinical Summary.¹⁷

Statistical Analysis

No formal sample-size calculations were performed. Available patients meeting the selection criteria previously described were analyzed.

The statistical analysis conducted were as follows:

• To select the baseline visit for the GALA cohort, logistic regression analysis was used.

• Kaplan-Meier analyses were performed, and corresponding figures were generated to compare the time-to-event differences between the maralixibat and GALA cohorts.

• Cox proportional hazards models were applied to describe and test differences between the treated and the control cohort, adjusting for potential confounders.

• If inverse probability of treatment weights methods were applied (due to imbalance), the analysis was done using the estimated weights.

• Sensitivity and subgroup analyses performed included the following:

  • Inclusion of hepatocellular carcinoma diagnosed by biopsy-proven histopathology as an additional event.

  • Selection of baseline visit for patients in GALA:

    • Definition of baseline for GALA cohort: Use first eligible visit.

    • Definition of baseline for maralixibat and GALA cohorts: Use date of birth instead of selected visit or treatment start.

  • Subgroup analyses:

    • Analysis was performed by region: North America versus Europe.

    • Analysis was performed using overlapping study centres participating in maralixibat studies and GALA registry.

• Imputation: For patients treated with maralixibat, the date of the clinical event was used in the time-to-event analysis unless it was unavailable, in which case the date the patient discontinued was used. For patients who discontinued a maralixibat study, follow-up data on outcome events were collected through an appropriate institutional review board/ethics committee approval and consent process.

• Missing data: Patients with missing outcome data will be censored at the time of discontinuation.

Results

Baseline Characteristics

Baseline characteristics generally demonstrate balance between the maralixibat cohort and the GALA control group, especially bilirubin, GGT, and ALT (Table 23).

Currently, treatment decisions are based on the management of pruritus and/or disease progression; therefore, measurement of sBA is not typically used for clinical decision-making. For this reason, limited sBA data are available in the GALA cohort (approximately 85% do not have sBA measures), as these are not sampled regularly on a clinical basis, often only at a single time point in some patients (at entry into the registry), and not longitudinally. Compared with patients for whom sBA data were available at baseline in the GALA cohort mean sBA was significantly higher in the maralixibat cohort, suggesting that the patients receiving maralixibat may have had worse disease at baseline; however, that cannot be confirmed given that most patients in the GALA cohort do not have sBA continuously monitored. (Table 23). To address the sBA limitations, a subgroup analysis was performed; it demonstrated results consistent with the primary analysis (Figure 7).

Balance in baseline pruritus levels between the groups could not be assessed as pruritus data are not recorded in the GALA cohort because of the lack of both a validated tool outside of clinical studies and inconsistent data collection across centres, regions, and countries. However, pruritus is a common reason for liver transplant and, as such, is recorded in the GALA cohort as 1 of the options as an indication for liver transplant. In the maralixibat studies, patients were required to have moderate-to-severe pruritus at baseline. The absence of pruritus severity data in the GALA cohort is a limitation of the dataset. Although up to 80% of patients with ALGS have pruritus,¹¹ not all patients with ALGS have moderate-to-severe pruritus. Therefore, the selected GALA control group has likely a larger heterogeneity of pruritus severity, which would include those with a milder pruritus status than those in the maralixibat studies. As such, the presence of pruritus and its severity are unlikely to have influenced results in favour of the maralixibat cohort.

Patients with cardiac disease were not excluded from the maralixibat studies; 82.1% of patients from studies LUM001-301 and LUM001-305 and studies LUM001-302 and LUM001-303 had cardiac disease reported at baseline, and 93.5% had cardiac disease reported at baseline in the LUM001-304 study. Similarly, cardiac disease was not excluded from the GALA cohort selection. It has been previously reported that 91% of patients in the GALA database had cardiac anomalies,¹⁰ so similar rates in the selected GALA control group can be expected in the absence of relevant exclusion criteria.

Table 23: Baseline Characteristics of Maralixibat Cohort and GALA Cohort (Control Group)

ALT = alanine transaminase; GALA = Global ALagille Alliance; GGT = gamma-glutamyl transferase; Q = quartile; sBA = serum bile acid; ULN = upper limit of normal.

^aBaseline sBA was available for 73 patients in the GALA control group. Approximately 85% values were not captured in the GALA cohort.

Source: Sponsor’s Clinical Summary.¹⁷

Efficacy

Event-Free Survival Analysis

The primary analysis was the comparison of the time to the first clinical event between the maralixibat cohort and GALA control group. EFS adjusted for age, sex, bilirubin, and ALT was statistically significantly higher in the maralixibat cohort compared with the GALA control group (HR = 0.305; 95% CI, 0.189 to 0.491; P < 0.0001), indicating a 70% improvement in EFS with maralixibat treatment (Figure 6).

The impact of different clinically meaningful parameters (such as age, sex, serum, levels of total bilirubin, ALT, GGT, region, and year of birth), different baseline definitions, and potential differences in standard of care between regions and centres was also assessed (Figure 7). The immediate events in the GALA control group (Figure 6) may be suggestive of an immortal time bias. Pruning analyses for comparison of EFS that excluded patients who had events in the first 3, 6, or 12 months confirmed significant improvement in EFS with maralixibat treatment (Figure 7).

Liver TFS Analysis

The majority of events in both cohorts were liver transplantations. A separate analysis considered liver TFS, where only events of liver transplant or death were considered. Liver TFS was statistically significantly higher in the maralixibat cohort compared with the GALA control group (HR = 0.332; 95% CI, 0.197 to 0.559, P < 0.0001), indicating a 67% improvement in liver TFS with maralixibat treatment (Figure 8).

For the GALA control group (N = 469), the main reasons for liver transplant were end-stage liver disease and pruritus, followed by growth impairment. In total, there were 24 deaths in the GALA control group. Importantly, there were 2 cardiac-related deaths, with the remaining deaths being most commonly related to liver disease and noncardiac vascular complications. In the maralixibat cohort (N = 84), there were 12 liver transplants reported, and the most common indications were disease progression and pruritus. There were 5 deaths reported in the maralixibat cohort.

Figure 6: Kaplan-Meier Plot for EFS — Maralixibat Cohort Versus GALA Control Group (Primary Analysis)

CI = confidence interval; GALA = Global ALagille Alliance; HR = hazard ratio; ML = maximum likelihood; MRX = maralixibat; SAP = statistical analysis plan; vs. = versus; yr = year.

^a Cox regression models: Primary: Cox regression – effect of MRX vs. GALA log likelihood test adjusted for age, sex, bilirubin, and ALT (according to the SAP). Unadjusted: Crude model.

Source: Sponsor’s Clinical Summary¹⁷

Discussion

The analysis included data from the natural-history clinical research registry and interventional dataset in ALGS. Given the challenge of conducting large, randomized studies in severe, rare pediatric settings, the availability of these datasets has important implications for understanding the effects of maralixibat in ALGS. The breadth of the GALA clinical research database allowed for the selection of a balanced external comparator group to the maralixibat cohort to evaluate EFS and TFS. Important baseline demographic and disease characteristics appeared to be comparable between the 2 groups, except for sBA, which was higher in the maralixibat group than in the control group (200 µmol/L versus 125 µmol/L). Of note, sBA was unavailable in approximately 85% of the patients in the control group. The study showed statistically and clinically significant reduction in liver transplant, death, and other associated events in patients who received maralixibat treatment compared with patients who received standard of care. In conclusion, the natural-history comparison supplements evidence supporting maralixibat as the potential therapeutic option to improve EFS and TFS in patients with ALGS.

Critical Appraisal

Internal Validity

The sponsor attempted to address gaps in evidence by providing long-term outcomes (EFS, TFS) data not available in pivotal trial and a natural-history study to complement the single-arm design of the pivotal trial. Although these attempts partially addressed these gaps, there are still limitations remaining. Primarily, in alignment with the pivotal trial, the relationship between sBA and pruritus has not been fully established.

Moreover, the sponsor applied the external control for maralixibat cohort to address gaps in evidence — i.e., the beneficial effect on long-term clinical outcomes. The sponsor used many methods to minimize selection bias between the comparison groups. First, the statistical team performing the analysis remained blinded to treatment outcomes before the selection of patients from the GALA registry. Second, to balance the baseline characteristics of both cohorts, the sponsor followed a stepwise selection process using a common set of cohort eligibility criteria as the maralixibat trials and a prespecified selection process using logistic regression. The GALA study sites were located in Canada as well as in the US, Europe, and Australia, which share similar health care systems as in Canada. Also, the 2 cohorts were likely recruited from comparable catchment areas to which rare genetic diseases are referred and where genetic testing and specialist care are available. As stated by the sponsor, some of the patients in maralixibat trials are from the same clinical sites as GALA registry clinical sites.

Figure 7: Forest Plot for EFS Analysis

ATT = average treatment effect in the treated; CI = confidence interval; GALA = Global ALagille Alliance; GGT = gamma glutamyltransferase; HR = hazard ratio; IPTW = inverse probability of treatment weights; SAP = statistical analysis plan; sBA = serum bile acid; Std = standardized; vs. = versus.

Notes: SAP specified refers to the Cox regression adjusted for age, sex, total bilirubin, and ALT. Adjusted1 refers to the Cox regression adjusted for age, total bilirubin, and GGT. Adjusted2 refers to the Cox regression adjusted for age, total bilirubin, GGT, ALT, and region. Adjusted3 refers to the Cox regression adjusted for age, total bilirubin, GGT, ALT, sex, and year of birth. Adjusted4 refers to the Cox regression adjusted for age, total bilirubin, GGT, and sBA.

Source: Sponsor’s Clinical Summary.¹⁷

Figure 8: Kaplan-Meier Plot for Liver TFS: Maralixibat Cohort Versus GALA Control Group

CI = confidence interval; GALA = Global ALagille Alliance; HR = hazard ratio; ML = maximum likelihood; MRX = maralixibat.

Source: Sponsor’s Clinical Summary.¹⁷

Although both comparison groups are closely matched with respect to age, sex, genetic mutations, bilirubin (total and direct/conjugated), GGT, and ALT, sBA data were available in only 73 of 469 patients from the GALA registry, with approximately 85% missing information. Moreover, although patients in the GALA registry have confirmed diagnoses of neonatal cholestasis, with the majority having Jagged-1 or NOTCH2 mutations, it is unknown how severe their cholestatic pruritus was in the absence of information on pruritus severity at baseline, which is a significant limitation. Other important prognostic factors besides pruritus severity, such as patients’ history of disease, previous treatments, duration of disease, and comorbidities, are unavailable. Therefore, it is highly likely that many residual confounding factors are still unaccounted for, particularly the disease severity, which could have a significant impact on long-term clinical outcomes such as liver transplant or death. Other factors, such as socioeconomic factors that affect patients’ or physicians’ choice of treatment, as well as differences in clinical settings in different regions, standards of care, and health care systems, could also affect the study results.

Taken together, despite the potential biases due to retrospective cohort design and missing important information on disease characteristics, the cohort study likely showed significant treatment effect of maralixibat on long-term outcome such as EFS and TFS in patients with moderate-to-severe cholestatic pruritus of ALGS.

External Validity

Given the rarity of ALGS and the challenges associated with conducting clinical trials in children, it is extremely difficult to assess the generalizability of long-term effect on clinically outcomes such as liver transplant and death. This study, with external control based on disease registry, may have provided important insights into coping with those evidence gaps. Since most patients treated with maralixibat included in this report have moderate-to-severe cholestatic pruritus (as measured by ItchRO questionnaire), the itch severity in the study may not be representative of that found in the overall ALGS population, which shows a spectrum of itch severity. Missing information on important baseline and disease characteristics, especially information about severity of cholestatic pruritus at baseline, makes it unclear how generalizable the results can be.

Discussion

Summary of Available Evidence

One pivotal trial, the ICONIC trial, met the inclusion criteria for the systematic review conducted by the sponsor. The ICONIC trial (9 sites in 6 countries, N = 31) is a double-blind, placebo-controlled, RWD study with a long-term, open-label extension in children with ALGS (range, 1 to 15 years) designed to evaluate the safety and efficacy of maralixibat. The trial’s primary outcome was the mean change from weeks 18 to 22 of fasting sBA levels in patients who previously responded to maralixibat treatment (n = 15), as defined by a reduction in sBA greater than or equal to 50% from baseline to weeks 12 or 18. Maralixibat was administered according to a dose-escalation schedule over 6 weeks of treatment to a final dose of 380 mcg/kg/day. Patients were eligible to increase to twice-daily dosing for a total of 760 ug/kg/day, starting after week 103, which is outside of the Health Canada recommended dose.

The sponsor submitted a natural-history comparison study, which is presented in this report, comparing disease outcomes among patients with ALGS treated with maralixibat (N = 84) with an external-controls cohort from the GALA disease registry database (n = 469), with follow-up data up to 6 years. Outcomes assessed included EFS (composite end point of first event of liver decompensation [ascites, variceal bleeding], SBD, liver transplant, and death) and TFS (liver transplant and death). Results from patient-level data from 3 long-term studies of patients with ALGS treated with maralixibat, including the LUM001-303 (IMAGINE) trial, the ICONIC pivotal trial (LUM001-304), and the IMAGINE-II (LUM001-305) trial to identify predictors of EFS and TFS was submitted by the sponsor and presented in this report.

Interpretation of Results

Efficacy

The ICONIC pivotal trial included a double-blind phase and appeared to have appropriate methods for blinding, allocation concealment, and randomization during the RWD phase. Although practical constraints associated with sample size are acknowledged, given the rarity of ALGS, small sample size remains a significant limitation, as the level of certainty was rated down across the outcomes. Furthermore, it should be noted that, during open-label phases of the study, patients’ and/or caregivers’ knowledge of treatment assignment may have biased subjective outcomes such as the ItchRO, PedsQL, and report of harms.

Evidence from the pivotal ICONIC trial showed that maralixibat may result in a decrease in sBA from weeks 18 to 22 in the mITT population (n = 15, defined as responders of study patients with an sBA reduction ≥ 50% at weeks 12 or 18), compared with placebo. Subgroup analyses were generally supportive of the findings of the primary analyses. Similar results were demonstrated when change in sBA from weeks 18 to 22 was assessed in the ITT population (N = 31), suggesting that the benefit from maralixibat observed in the trial was primarily from those responders (n = 15), which represented about 50% of the overall study population. Furthermore, trial results found that decreases in sBA levels were maintained over time in the overall study population as demonstrated by change in sBA levels from baseline to week 18 (secondary outcome) and from baseline to week 48 (additional efficacy outcome). The clinical experts consulted for this review indicated that although sBA levels are not measured in clinical practice due to high costs and barriers in access to testing, the results of the primary outcome provide supportive evidence of the mechanism of action of maralixibat. Descriptive posthoc data from the ICONIC pivotal trial suggested that reductions in sBA from baseline to week 48 were associated with reductions in mean ItchRO(Obs) weekly average morning severity scores (Appendix 1). These internal data supported a potential correlation in the improvement between sBA and ItchRO. Yet, according to the clinical expert consulted for this review, there was no substantial and consistent literature on the association between sBA change and itching in patients with ALGS. Moreover, due to the small number of patients (n = 28), it is unclear the extent to which sBA levels may be associated with pruritus in patients with ALGS.

Input from patient and clinician groups and clinical experts highlighted the importance of reduction in pruritus as an important treatment goal for patients. In the ICONC trial, pruritus was assessed using the ItchRO instrument, which has been reported to be generally valid, reliable, and responsive, as assessed either by patients or caregivers (observers). Patients who were dropped off from maralixibat treatment had experienced an increase (worsening) in mean ItchRO(Obs) and ItchRO(Pt) weekly average morning severity scores from weeks 18 to 22, which nearly rebound back to the same levels at baseline, whereas those who continued to receive maralixibat generally maintained lower mean scores during the RWD period. The magnitude of improvement in both measures may be clinically meaningful both in favour of maralixibat according to the MID estimate of 1 point identified in the literature and clinical expert input. Subgroup analyses for the ItchRO(Obs) were generally consistent with the overall results of the pivotal trial, except for where there was a very small number of patients in a given category (n ≤ 4). Similarly, in the overall population, improvements were observed in the ItchRO(Obs) and ItchRO(Pt) weekly average morning severity scores from baseline to week 18 (secondary end point) and baseline to week 48 (additional efficacy end point), providing supportive evidence for maralixibat.

HRQoL was highlighted as an important treatment goal for patients from the input of patient groups and the clinical expert, especially in terms of patient and caregiver fatigue. The pivotal trial assessed HRQoL using the PedsQL, which suggested that maralixibat would likely have contributed to an improvement in the PedsQL total score (parent) and also likely contributed to a considerable reduction of the PedsQL Multidimensional Fatigue Scale score compared with placebo. However, due to a short 4-week randomization phase and, moreover, a small sample size, there was high uncertainty of these findings as result of wide 95% CI in the estimation of the changes, which included null values. Nevertheless, such a trend in improvement of fatigue, in particular, was supported by the LTE phase of the study. In the overall study population, from baseline to week 18 and from baseline to week 48, patients achieved an 8-point improvement in the PedsQL total score and a nearly 20-point improvement in the PedsQL Multidimensional Fatigue Scale score. These changes are larger than the estimated MID of 4 to 5 points for the PedsQL Generic Core Scales in a sample of children with and without chronic health conditions. However, in a single-arm, LTE phase, any improvements over time compared with baseline may not be completely attributable to the treatment itself alone. There might have been other contemporaneous changes that could also contributed to the observed temporal changes overtime. For example, it should be noted that the number of patients assessed for the PedsQL Multidimensional Fatigue Scale score was low (22 out of 31) in the assessment from baseline to week 48. As such, those who completed the questionnaires may be fundamentally different than those who did not (i.e., differences in treatment response). The extent to which the impact of missing data due to selective reporting on the assessment of this outcome is unclear.

Other liver-related outcomes identified as important by the clinical experts consulted included change in liver biomarkers and enzymes. Secondary end points assessed in the RWD phase of the ICONIC trial found that maralixibat may increase ALT and ALP levels and may result in little-to-no difference in total and direct bilirubin compared with placebo. Again, due to the short duration of the RWD phase and the small sample size, the estimation of the differences included a null value across the 95% CIs, indicating the lack of precision, which rendered high uncertainty in these potential improvements. Similarly, no clear and consistent pattern of improvements in ALT and ALP were observed at weeks 18 and 48 in the overall trial population, and no notable changes in total and direct bilirubin during these time periods. The lack of a comparator at these time points and the small sample sizes may have contributed to the high uncertainty of the evidence on improvement of liver biomarkers. The clinical experts noted the importance of change in body height and weight as an important outcome and stated that it is reasonable to assess these outcomes in the indicated patient population in clinical practice. Numerical increases in body height and weight z scores were noted from baseline to week 48 and to week 100; however, the evidence is very uncertain, as conclusions about their efficacy relative to any comparator cannot be drawn.

Results from the natural-history comparison study reported a large improvement in EFS and TFS in the maralixibat cohort compared with standard of care in the control cohort. The study applied an external control based on disease registry, which provided a large sample size with a study design that matched on all known potential confounders including age, sex, genetic mutations, and so forth. However, the disease severity was not well known and therefore, not appropriately controlled, which is a significant concern. Meanwhile, the potential for unknown confounding factors may also have had an impact on the results. Yet it is unlikely that such potential biases would be able to explain all the large differences in EFS and TFS between the comparison groups. In combination with the evidence based on the observed benefit on sBA during the randomized phase of the trial and the potential large improvements over the long-term phase, it could extrapolate that an improvement on EFS and TFS is highly likely. In other words, despite the nonrandomized, retrospective cohort design, this observational study supplements evidence supporting maralixibat as the potential therapeutic option to improve EFS and TFS in patients with ALGS.

Harms

Evidence from the ICONIC trial during the RWD period suggest that maralixibat may result in little-to-no difference in SAEs, diarrhea, abdominal pain, and FSV deficiency compared with placebo. Overall, the most frequently reported AEs (> 30% in at least 1 phase) were abdominal pain, pyrexia, diarrhea, nasopharyngitis, vomiting, cough, and pruritus. A total of 6 patients (2 each in the open-label phase, the after RWD phase, and the LTE phase) experienced AEs leading to study drug discontinuation. According to the clinical experts consulted, the 4-week RWD period was not sufficient to fully assess the comparative safety of maralixibat to placebo. Most AEs were generally of mild-to-moderate severity and no deaths were noted during the study.

Conclusions

There is an unmet need for symptomatic and curative treatment options for cholestatic pruritus in pediatric patients with ALGS. Patients and clinicians highlighted the need for treatments that reduce the frequency and severity of pruritus and reduce patient and caregiver fatigue. The pivotal phase II, double-blind, placebo-controlled, randomized drug withdrawal trial included in this review assessed the treatment of cholestatic pruritus in pediatric patients with ALGS (aged 12 months to 15 years). The trial was an exploratory study.

The study demonstrated that maralixibat may result in a decrease in sBA levels and results in a clinically meaningful improvement in pruritus as assessed by the ItchRO(Obs) and ItchRO(Pt) weekly average morning severity scores compared with placebo. It is important to note that the “low” certainty of evidence as assessed by GRADE for sBA and pruritus outcomes is due to the imprecision observed and not from bias due to study limitations. This imprecision due to small sample size is clearly connected to the nature of the rarity of the disease. Improvements in the PedsQL total score (parent) and the PedsQL Multidimensional Fatigue Scale score (parent) were uncertain. It also remained uncertain whether maralixibat may have increased ALT and ALP levels or resulted in any difference in total and direct bilirubin compared with placebo. Despite certain limitations inherent with observational study design, it is likely that there is a significant treatment effect of maralixibat on long-term outcome such as EFS and TFS.

Maralixibat was generally well-tolerated in the ICONIC trial, with limited grade 3 AEs or SAEs. Of note, however, due to the rare nature of the disease, the study sample size was small and the study was short, in a 4-week randomized duration; therefore, more evidence is needed to support the overall benefit and safety of maralixibat.

References

1.Subramaniam P, Knisely A, Portmann B, et al. Diagnosis of Alagille syndrome-25 years of experience at King's College Hospital. J Pediatr Gastroenterol Nutr. 2011;52(1):84-89. PubMed

2.Saleh M, Kamath B, Chitayat D. Alagille syndrome: clinical perspectives. Appl Clin Genet. 2016;9:75-82. PubMed

3.Turnpenny PD, Ellard S. Alagille syndrome: pathogenesis, diagnosis and management. Eur J Hum Genet. 2012;20(3):251-257. PubMed

4.Osawa Y, Seki E, Adachi M, et al. Role of acid sphingomyelinase of Kupffer cells in cholestatic liver injury in mice. Hepatology. 2010;51(1):237-245. PubMed

5.Mehl A, Bohorquez H, Serrano MS, Galliano G, Reichman TW. Liver transplantation and the management of progressive familial intrahepatic cholestasis in children. World J Transplant. 2016;6(2):278-290. PubMed

6.Kamath BM, Ye W, Goodrich NP, et al. Outcomes of Childhood Cholestasis in Alagille Syndrome: Results of a Multicenter Observational Study. Hepatol Commun. 2020;4(3):387-398. PubMed

7.Monte MJ, Marin JJ, Antelo A, Vazquez-Tato J. Bile acids: chemistry, physiology, and pathophysiology. World J Gastroenterol. 2009;15(7):804-816. PubMed

8.Lykavieris P, Hadchouel M, Chardot C, Bernard O. Outcome of liver disease in children with Alagille syndrome: a study of 163 patients. Gut. 2001;49:431-435. PubMed

9.Englert C, Grabhorn E, Burdelski M, Ganschow R. Liver transplantation in children with Alagille syndrome: indications and outcome. Pediatr Transplant. 2006;10(2):154-158. PubMed

10.Vandriel SM, Li LT, She H, et al. Natural history of liver disease in a large international cohort of children with Alagille syndrome: Results from the GALA study. Hepatology. 2023;77(2):512-529. PubMed

11.Kamath BM, Baker A, Houwen R, Todorova L, Kerkar N. Systematic Review: The Epidemiology, Natural History, and Burden of Alagille Syndrome. J Pediatr Gastroenterol Nutr. 2018;67(2):148-156. PubMed

12.Emerick KM, Rand EB, Goldmuntz E, Krantz ID, Spinner NB, Piccoli DA. Features of Alagille syndrome in 92 patients: Frequency and relation to prognosis. Hepatology. 1999;29:822-829. PubMed

13.Kamath BM, Yin W, Miller H, et al. Outcomes of liver transplantation for patients with Alagille syndrome: the studies of pediatric liver transplantation experience. Liver Transpl. 2012;18(8):940-948. PubMed

14.Balshem H, Helfand M, Schünemann HJ, et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64(4):401-406. PubMed

15.Santesso N, Glenton C, Dahm P, et al. GRADE guidelines 26: informative statements to communicate the findings of systematic reviews of interventions. J Clin Epidemiol. 2020;119:126-135. PubMed

16.Clinical Study Report: LUM001-304. ICONIC: Long-term, open-label study with a double-blind, placebo-controlled, randomized drug withdrawal period of lum001, an apical sodium-dependent bile acid transporter inhibitor (asbti), in patients with Alagille syndrome [internal sponsor's report]. Foster City (CA): Mirum Pharmaceuticals Inc.; 2020 Oct 13.

17.Summary of Clinical Evidence [internal sponsor's report]. In: Drug Reimbursement Review sponsor submission: Livmarli (Maralixibat), oral solution, 9.5 mg/mL [internal sponsor's package]. Foster City (CA): Mirum Pharmaceuticals Inc.; 2023 May 05.

18.Fischetto R, Palmieri VV, Tripaldi ME, et al. Alagille Syndrome: A Novel Mutation in JAG1 Gene. Front Pediatr. 2019;7:199. PubMed

19.Canadian ALGS and Livmarli Landscape Assessment. Data on file. In: Drug Reimbursement Review sponsor submission: Livmarli (maralixibat), oral solution, 9.5 mg/mL [internal sponsor's package]. Foster City (CA): Mirum Pharmaceuticals Inc.; 2023 May 05.

20.Dull MM, Kremer AE. Newer approaches to the management of pruritus in cholestatic liver disease. Curr Hepatol Rep. 2020;19:86-95.

21.Emerick KM, Whitington PF. Partial external biliary diversion for intractable pruritus and xanthomas in Alagille syndrome. Hepatology. 2002;35(6):1501-1506. PubMed

22.Mattei P, von Allmen D, Piccoli D, Rand E. Relief of intractable pruritus in Alagille syndrome by partial external biliary diversion. J Pediatr Surg. 2006;41(1):104-107; discussion 104-107. PubMed

23.Kwong AJ, Ebel NH, Kim WR, et al. OPTN/SRTR 2020 Annual Data Report: Liver. Am J Transplant. 2022;22 Suppl 2:204-309. PubMed

24.Shneider BL. Liver transplantation for Alagille syndrome: the jagged edge. Liver Transpl. 2012;18(8):878-880. PubMed

25.Search Orphan Drug Designations and Approvals: maralixibat Silver Spring (MD): U.S. Food and Drug Administration; 2023: https://www.accessdata.fda.gov/scripts/opdlisting/oopd/detailedIndex.cfm?cfgridkey=401513. Accessed 2023 Jun 1.

26.Product information: Livmarli (maralixibat chloride). (European public assessment report). Amsterdam (NL): European Medicines Agency; 2023: https://www.ema.europa.eu/en/medicines/human/EPAR/livmarli. Accessed 2023 Jun 1.

27.Di Ciaula A, Garruti G, Lunardi Baccetto R, et al. Bile Acid Physiology. Ann Hepatol. 2017;16 Suppl 1:S4-S14. PubMed

28.Reue K, Lee JM, Vergnes L. Regulation of bile acid homeostasis by the intestinal Diet1-FGF15/19 axis. Curr Opin Lipidol. 2014;25(2):140-147. PubMed

29.Al-Dury S, Marschall HU. Ileal Bile Acid Transporter Inhibition for the Treatment of Chronic Constipation, Cholestatic Pruritus, and NASH. Front Pharmacol. 2018;9:931. PubMed

30.Dawson PA. Role of the intestinal bile acid transporters in bile acid and drug disposition. Handb Exp Pharmacol. 2011(201):169-203. PubMed

31.Dawson PA, Lan T, Rao A. Bile acid transporters. J Lipid Res. 2009;50(12):2340-2357. PubMed

32.Product monograph: pms-Ursodiol C (Ursodiol). Montreal (QC): Pharmascience; 2017: https://pdf.hres.ca/dpd_pm/00040988.PDF. Accessed 2023 Jun 1.

33.Product monograph: Rifadin (Rifampin). Laval (QC): sanofi-aventis Canada Inc.; 2017: https://pdf.hres.ca/dpd_pm/00040693.PDF. Accessed 2023 Jun 1.

34.Product monograph: Novohydroxyzin (Hydroxyzine Hydrochloride). Scarborough (ON): Teva Canada Limited; 2014: https://www.tevacanada.com/globalassets/canada-scs/product-monographs/novo-hydroxyzin-capsules-10-mg-25-mg--50-mg_pi-feb19_1987-wis-eng.pdf. Accessed 2023 Jun 1.

35.Donkers JM, Roscam Abbing RLP, van de Graaf SFJ. Developments in bile salt based therapies: A critical overview. Biochem Pharmacol. 2019;161:1-13. PubMed

36.Gonzales E, Hardikar W, Stormon M, et al. Maralixibat Treatment Significantly Reduces Pruritus and Serum Bile Acids in Patients with Alagille Syndrome: Results from a Randomized Phase II Study with 4 Years of Follow-Up. SSRN Electronic Journal. 2020.

37.Drug Reimbursement Review sponsor submission: Livmarli (maralixibat), oral solution, 9.5 mg/mL [internal sponsor's package]. Foster City (CA): Mirum Pharmaceuticals Inc.; 2023 May 05.

38.van Wessel DBE, Thompson RJ, Gonzales E, et al. Genotype correlates with the natural history of severe bile salt export pump deficiency. J Hepatol. 2020;73(1):84-93. PubMed

39.Harpavat S, Hawthorne K, Setchell KDR, et al. Serum bile acids as a prognostic biomarker in biliary atresia following Kasai portoenterostomy. Hepatology. 2023;77(3):862-873. PubMed

40.Kamath BM, Abetz-Webb L, Kennedy C, et al. Development of a Novel Tool to Assess the Impact of Itching in Pediatric Cholestasis. Patient. 2018;11(1):69-82. PubMed

41.Foster B, Gauthier M, Vig P, Jaecklin T, Kamath B, Andrae DA. Itch reported outcome tool for caregivers of pediatric patients with cholestatic liver disease: an analysis of validation and scoring from the ICONIC maralixibat study. ISPOR; May 18-20, 2020; Virtual.

42.Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr. 2003;3(6):329-341. PubMed

43.Clinical Study Report: GALA-MRX-ALGS. Long-term outcomes of long-term maralixibat treatment in patients with Alagille Syndrome compared with a natural history cohort from the GALA registry. Statistical Analysis Plan (SAP). In: Drug Reimbursement Review sponsor submission: Livmarli (maralixibat), oral solution, 9.5 mg/mL [internal sponsor's package]. Foster City (CA): Mirum Pharmaceuticals Inc.; 2023 May 05.

44.Ayoub MD, Kamath BM. Alagille Syndrome: Diagnostic Challenges and Advances in Management. Diagnostics (Basel). 2020;10(11). PubMed

45.Gonzales E, Hardikar W, Stormon M, et al. Efficacy and safety of maralixibat treatment in patients with Alagille syndrome and cholestatic pruritus (ICONIC): a randomised phase 2 study. Lancet. 2021;398(10311):1581-1592. PubMed

46.Shneider BL, Spino CA, Kamath BM, et al. Impact of long-term administration of maralixibat on children with cholestasis secondary to Alagille syndrome. Hepatol Commun. 2022;6(8):1922-1933. PubMed

47.Mirum Pharmaceuticals Inc. NCT02160782: Safety and Efficacy Study of LUM001 (Maralixibat) With a Drug Withdrawal Period in Participants With Alagille Syndrome (ALGS) (ICONIC). ClinicalTrials.gov. Bethesda (MD): U.S. National Library of Medicine; 2021: https://clinicaltrials.gov/ct2/show/NCT02160782. Accessed 2023 Jun 5.

48.Mirum Pharmaceuticals Inc. NCT02117713: An Extension Study to Evaluate the Long-Term Safety and Durability of Effect of LUM001 in the Treatment of Cholestatic Liver Disease in Pediatric Subjects With Alagille Syndrome (IMAGINE-II). ClinicalTrials.gov. Bethesda (MD): U.S. National Library of Medicine; 2021: https://clinicaltrials.gov/ct2/show/NCT02117713. Accessed 2023 Jun 5.

49.Sokol RJ, Gonzales EM, Kamath BM, et al. Predictors of 6-year event-free survival in alagille syndrome patients treated with maralixibat, an ileal bile acid transporter inhibitor. Hepatology. 2023. PubMed

50.Sokol RJ, Gonzales E, Kamath B, et al. Predictors of 6-year event-free survival in patients with ALGS treated with maralixibat, an IBAT inhibitor. American Association for the Study of Liver Diseases (AASLD) The Liver Meeting Digital Experience; November 12-15, 2021; Virtual.

51.Rovner AJ, Schall JI, Jawad AF, et al. Rethinking growth failure in Alagille syndrome: the role of dietary intake and steatorrhea. J Pediatr Gastroenterol Nutr. 2002;35(4):495-502. PubMed

52.Feranchak AP, Sokol R. Medical and nutritional management of cholestasis in infants and children. Liver disease in children. 2007;3:190-231.

53.Wasserman D, Zemel BS, Mulberg AE, et al. Growth, nutritional status, body composition, and energy expenditure in prepubertal children with Alagille syndrome. J Pediatr. 1999;134(2):172-177. PubMed

54.Kronsten V, Fitzpatrick E, Baker A. Management of cholestatic pruritus in paediatric patients with alagille syndrome: the King's College Hospital experience. J Pediatr Gastroenterol Nutr. 2013;57(2):149-154. PubMed

55.Thebaut A, Habes D, Gottrand F, et al. Sertraline as an Additional Treatment for Cholestatic Pruritus in Children. J Pediatr Gastroenterol Nutr. 2017;64(3):431-435. PubMed

56.Wang KS, Tiao G, Bass LM, et al. Analysis of surgical interruption of the enterohepatic circulation as a treatment for pediatric cholestasis. Hepatology. 2017;65(5):1645-1654. PubMed

57.Kriegermeier A, Wehrman A, Kamath BM, Loomes KM. Liver disease in alagille syndrome. Alagille Syndrome. Cham (CH): Springer; 2018:49-65.

58.Sheflin-Findling S, Arnon R, Lee S, et al. Partial internal biliary diversion for Alagille syndrome: case report and review of the literature. J Pediatr Surg. 2012;47(7):1453-1456. PubMed

59.Kamath BM, Ye W, Goodrich NP, et al. Outcomes of Childhood Cholestasis in Alagille Syndrome: Results of a Multicenter Observational Study. Hepatol Commun. 2020;4(3):387-398. PubMed

60.Hori T, Egawa H, Takada Y, et al. Long‐term outcomes after living‐donor liver transplantation for Alagille syndrome: a single center 20‐year experience in Japan. Am J Transplant. 2010;10(8):1951-1952. PubMed

Appendix 1: Detailed Outcome Data

Note this appendix has not been copy-edited.

Table 24: Additional Efficacy Outcomes in the ICONIC Trial

SD = standard deviation; SE = standard error.

Note: Details from the table have been taken from the sponsor’s Summary of Clinical Evidence.¹⁷

Source: ICONIC Clinical Study Report.¹⁶

Table 25: ItchRO(Obs) Weekly Morning Average Severity Score by sBA Response in the ICONIC Trial (Safety Population)

ItchRO(OBs) = Itch Reported Outcome (observer); sBA = serum bile acid; SD = standard deviation.

Source: Sponsor submitted internal package.³⁷

Appendix 2: New Data Submitted by Sponsor for Reconsideration

Note this appendix has not been copy-edited.

Itch-Related Symptoms Based on Individual Patient Data for ItchRO(Obs)

As part of the Request for Reconsideration, the sponsor provided a posthoc analysis of individual patient data on ItchRO among the participants who completed 48-week follow-up in ICONIC trial (additional efficacy end point), which is presented in Table 26.

As described in the CADTH initial clinical report, average values of weekly morning severity score indicate a decrease (improvement) in ItchRO(Obs) from baseline to week 18 with a mean change of −1.70 (95% CI, - 2.0 to –1.4;) and from baseline to week 48 with a mean change of −1.6 (95% CI, - 2.1 to −1.1).

When looking at the individual patient values, 12 patients, of which 7 were randomized to placebo and 5 were randomized to maralixibat during the RWD phase, showed decreases in ItchRO score from an average score of about 3 at baseline to an average score of around zero at 48 weeks. The sponsor indicated that change in ItchRO score from a moderate-to-severe itch, indicated by a score of 3 on the ItchRO, to no itch, indicated by a score of less than or equal to 1, represents a binary result and when observed over a period of 48 weeks, does not imply a biased subjective outcome for pruritus measure.

As summarized, the analysis included in the sponsor’s initial submission supports an improvement in itch severity based on the ItchRO(Obs) from baseline to week 48. As a continuous outcome in the trial, change from baseline in ItchRO(Obs) was assessed through an ANCOVA model, using treatment groups as a fixed effect and baseline value as a covariate. It is important to note that changes from baseline to week 48 in ItchRO(Obs) were considered as an additional efficacy end point in the trial. Moreover, caregivers’ knowledge of treatment assignment during the open-label phases of the ICONIC trial might have introduced bias in measuring this subjective outcome.

CADTH acknowledges the observation of patients reporting a score corresponding to no itch at week 48 that was derived from the table of individual patient data, provided in the sponsor’s reconsideration request; however, this is an inappropriate method of analysis that is extremely limited in its interpretability at a population level. Importantly, this analysis has been conducted as posthoc analysis by the sponsor and as such, offers limited evidence regarding durability of the effect of maralixibat.

Table 26: Individual ItchRO(Obs) Data Within Participants Who Completed the Week 48 Study Period (N = 29)

ItchRO(Obs) = Itch Reported Outcomes (observer); NA = not available; PBO = placebo; RWD = randomized withdrawal.

Notes:(1) denotes patients randomized to maralixibat during RWD, while (2) denotes patients randomized to PBO during RWD.

This analysis was conducted post hoc.

^aThis participant was observed to have an average daily score greater than 2, consistent with study inclusion criteria.

Source: Gonzales, E., et.al. Efficacy and safety of maralixibat treatment in patients with Alagille syndrome and cholestatic pruritus (ICONIC): A randomized phase 2 study. Lancet (2021), issue 10311, 1581 to 92.

Clinician Observed Scratching and/or Physical Evidence of Scratching Based Individual Patient Data on the Clinician Scratch Scale

To support the validation of the changes in ItchRO observed in the ICONIC trial, the sponsor also provided new information based on the Clinician Scratch Scale (CSS) for reconsideration by CDEC. More specifically, individual-level data on clinician-rated severity of cholestatic pruritus measured by the CSS at baseline and week 48 were provided in the reconsideration request (Table 27). CSS contains a 5-point scale, reporting values that range from zero (none) to 4 (cutaneous mutilations, hemorrhage, scarring).

CSS outcome data were predefined as an additional efficacy end point in the ICONIC trial. The individual-level data are considered new information as part of the reconsideration; however, as noted in the CADTH initial Clinical Review report, average CSS baseline values for the ITT population were 3.3, and the average CSS values at week 48 were 1.5.

||| |||||||||| ||||||| |||||||| |||||||| || ||| ||||||||||||||| ||||||| |||||||||||| ||||||| |||| |||||||| || ||| ||| |||||||||| ||| ||||||||| |||| |||||||||||||||||| ||| ||||||||| ||||| |||| || ||||||| ||||| || ||| || ||||||||| ||||| ||||||||||| || |||||||||||||||||| ||||||| |||||||||| || |||| || |||| ||| ||||| ||||||||||| || || |||||||||| || |||||||||| CADTH acknowledges the observed data in certain individuals, but this new information provided by the sponsor is subject to the same limitations noted for the individual-level data for the ItchRO. While the individual-level data may provide some insight or context to the analysis of the ItchRO and CSS reported in the ICONIC trial, this data are very limited in its interpretability. In reference to subjectivity of pruritus measurements, assessed by ItchRO instrument, and its associations with the objective CSS tool, data provided both in sponsor’s initial submission to CADTH and reconsideration request are coming from a published abstract, which offers only limited opportunity to appraise the measurement properties of the 2 instruments (Foster B, Gauthier M, Vig P, Jaecklin T, Kamath B, Andrae DA. Itch reported outcome tool for caregivers of pediatric patients with cholestatic liver disease: an analysis of validation and scoring from the ICONIC maralixibat study. Value Health. 2020;23(5):PIH76).

Table 27: Redacted