A phase II, randomized, open-label, 52-week study of seladelpar in patients with primary biliary cholangitis

, Metacrine, NGM Biopharmaceuticals, Northsea Therapeutics B

Treatment with seladelpar up to 10 mg QD through 1 year resulted in robust, dose-dependent, and clinically significant improvements in biochemical markers of cholestasis and pruritus. Charles McWherter is an employee of and owns stock in CymaBay Therapeutics and holds patents broadly relevant to this work.
Financial Support Statement: Funding for this study and manuscript preparation was provided by CymaBay Therapeutics, which had a role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.

Author Contributions
Study

Data Availability
Data is available from the Study Sponsor CymaBay Therapeutics. The full trial protocol can be accessed at https://www.journal-of-hepatology.eu.
J o u r n a l P r e -p r o o f 9 at risk for disease progression. Seladelpar appeared safe and well tolerated with no increase in pruritus.

LAY SUMMARY
Current treatment options for patients living with primary biliary cholangitis (PBC) are not optimal due to inadequate effectiveness or undesirable side effects. Patients with PBC who took seladelpar, a new treatment being developed for PBC, at increasing doses (2, 5, or 10 mg/day) for 1 year had clinically significant, dose-dependent improvements in key liver tests. Treatment appeared safe and was not associated with any worsening in patient self-reported itch scores.
J o u r n a l P r e -p r o o f INTRODUCTION Primary biliary cholangitis (PBC) is a rare autoimmune liver disease predominantly afflicting middle-aged women, with approximately 1 in 1000 women over 40 years old diagnosed globally. 1,2 This complex cholangiopathy is characterized by immune-mediated destruction of small intrahepatic bile ducts [3][4][5][6] leading to cholestasis, with accumulation of toxic bile acids that are believed to perpetuate chronic, progressive inflammation and fibrosis, which can progress to biliary cirrhosis and liver-related death. 7 Elevated alkaline phosphatase (ALP) and total bilirubin are independent risk factors for decreased transplant-free survival in patients with PBC treated with first-line approved therapy, ursodeoxycholic acid (UDCA). [8][9][10] Normalization of ALP and bilirubin levels have been proposed as ideal treatment goals. 11 The only other approved PBC treatment is obeticholic acid (OCA), a bile acid analog and farnesoid X receptor (FXR) agonist used as a second-line add-on therapy for patients with inadequate response to UDCA or as monotherapy for those intolerant to UDCA. 2,5,6,12 However, less than half of the patients treated with OCA, alone or in combination with UDCA, achieved the composite biochemical response used for its approval (ALP <1.67×upper limit of normal [ULN] and ≥15% decrease from baseline and normal total bilirubin levels). 13 Furthermore, OCA therapy can induce or worsen pruritus in a dose-dependent manner. 13,14 Peroxisome proliferator-activated receptor (PPAR) agonists have gained attention as potential therapy for PBC and other cholestatic liver diseases. 15 PPARδ is a broadly expressed, fatty acidactivated transcription factor involved in fatty acid metabolism and inflammation. [16][17][18] In the liver, PPARδ-regulated genes are expressed in hepatocytes, Kupffer cells, and hepatic stellate J o u r n a l P r e -p r o o f cells. 15,19 Importantly, PPARδ plays a critical role in bile acid homeostasis and has antifibrotic effects. 1,[16][17][18]20 Seladelpar is a novel, potent, selective, first-in-class PPARδ agonist being developed for the treatment of PBC in patients who do not respond to or are intolerant to UDCA. 21 In a previous study, seladelpar at doses of 50 and 200 mg/day reduced ALP >60% in patients with PBC; however, 3 cases of rapid, reversible, asymptomatic elevations in hepatic aminotransferases were reported. 21 Herein, we report the results of a phase 2, dose-ranging, open-label study to establish the safety and efficacy of seladelpar using 5-to 100-fold lower doses. The initial objective of this study was to assess the efficacy, safety, and tolerability of seladelpar doses up to 10 mg once daily (QD) in patients with PBC after 8 weeks of treatment. The study design was subsequently amended to examine the effects of seladelpar at 2, 5, and 10 mg QD through 12 weeks and 1 year of treatment. Efficacy endpoints focused on changes from baseline in ALP and bilirubin and other relevant liver biochemistries, including gamma-glutamyl transferase (GGT), 5'nucleotidase, alanine aminotransferase (ALT), and aspartate aminotransferase (AST). Patients received seladelpar (CymaBay Therapeutics, Newark, California) along with UDCA at their prestudy dose levels unless they were UDCA intolerant.

Initial study design
The initial study design included an 8-week dose-ranging period followed by an 18-week extension period (Supplementary Fig. S1A). After a 2-week screening period, patients were centrally randomized (1:1 using an interactive voice/web response system) to treatment with J o u r n a l P r e -p r o o f seladelpar 10 or 5 mg QD. A planned interim analysis was conducted after the enrollment of the 24th patient in the 10-mg and 5-mg cohorts, initially to determine whether a 25-mg cohort should be enrolled.

Amended study design
At the initial 8-week interim analysis, patients who had completed the dose-ranging treatment period had significant declines in ALP without elevations in aminotransferases. The sponsor subsequently amended the protocol to adjust the study design to add a 2-mg QD cohort (up to 18 patients), increase the dose-ranging period from 8 to 12 weeks, increase the sample size to 116, and increase the total treatment period to 52 weeks (Supplementary Fig. S1B). Since the decreases in ALP in the 10-mg cohort at that time were approaching those of the 50-mg cohort in the prior study, 21 it was also decided not to enroll the seladelpar 25-mg cohort.
Randomization to the 5 or 10 mg QD doses continued, and after the amendment, patients at sites in the United Kingdom were registered in chronological order to treatment with seladelpar 2 mg QD for 12 weeks until the cohort was enrolled. Among patients in the 2-mg and 5-mg cohorts, beginning at Week 12 (Week 26 for patients with cirrhosis), the seladelpar dose could be titrated up to 10 mg QD based on investigator judgment for patients with an inadequate biochemical response. The seladelpar dose could be titrated down at any time during the study for safety reasons.

Study endpoints and assessments
The primary efficacy endpoint was the mean percent change in ALP from baseline at Week 8.
Secondary efficacy endpoints included mean absolute and percent changes from baseline at

Statistical analysis
The final planned sample size was 49 patients each in the 10-mg and 5-mg cohorts (increased from 12), and up to 18 in the 2-mg cohort, allowing for detection of at least a 10% mean difference in ALP percent change between the 5-mg and 10-mg cohorts with a 15% standard deviation at 90% power using a 2-sided, 2-sample t-test at α=0.05 significance level. Efficacy analyses were conducted using data from the modified intent-to-treat (mITT) population (any patient diagnosed with PBC who received ≥1 dose of seladelpar with ≥1 post-baseline ALP measurement). A sensitivity analysis on the intent-to-treat (ITT) population (any patient who received ≥1 dose of seladelpar with ≥1 post-baseline ALP measurement) was also conducted.
Safety analyses were conducted for all patients who received ≥1 dose of seladelpar (safety population). Efficacy and safety were analyzed by initial dose assignment. Select efficacy endpoints were also analyzed for patients who were randomized to seladelpar 5 mg QD and remained on this dose (5/5-mg cohort) or uptitrated to 10 mg QD (5/10-mg cohort).
Demographic and baseline characteristics were summarized using descriptive statistics for continuous variables and frequency distributions for discrete variables. Where specified, the last observation carried forward was used for missing laboratory data; other missing data were not imputed. Primary and secondary efficacy analyses were carried out using 2-sided tests at the α=0.05 significance level. For biochemistry measures, within-group comparisons with baseline using a paired t-test were performed at Weeks 12 and 52, and pairwise comparisons of least squares (LS) means between treatment cohorts using an analysis of covariance model were performed at Weeks 8, 12, and 52. Since the seladelpar dose was uptitrated in the majority of patients in the 2-mg and 5-mg cohorts after Week 12, the differences among treatment cohorts were not reported after that time point. Responder rates were compared between treatment cohorts using Fisher's exact test.

Patient disposition and baseline characteristics
Among 192 patients screened, 121 were randomized to daily oral seladelpar at doses of 2 mg   2-mg (26%), 5-mg (33%), and 10-mg (41%) ( Fig. 2A-B). At the end of the dose-ranging period  Table S1). Overall, ALP decreased from baseline in the majority of patients at Week 52 (Fig. 2C). ALP normalization was observed in 31% of patients in the 10-mg cohort as early as Week 12 and was maintained at 33% through Week 52 ( Fig. 2D). In contrast, in the 2-mg and 5-mg cohorts, normalization occurred in 0% and 11% of patients, respectively, at Week 12 and in 9% and 13%, respectively, at Week 52. Similar results were observed for the ITT population (data not shown).

Composite and published criteria response rates
At Weeks 12 and 52, the composite biochemical response endpoint was achieved by 46% and 64% of patients, respectively, in the 2-mg cohort; 49% and 53%, respectively, in the 5-mg cohort; and 67% in the 10-mg cohort at both time points (Fig. 2E). Responder rates for published PBC response criteria, including the Barcelona, Paris, and Toronto criteria (59 to 79% for 10 J o u r n a l P r e -p r o o f mg), were consistent with rates for the composite endpoint (Supplementary Table S2).

Additionally, Global PBC Study Group (GLOBE) and United Kingdom-Primary Biliary
Cirrhosis (UK-PBC) 5-, 10-, and 15-year risk scores decreased from baseline in a dose-dependent manner at Weeks 12 and 52 (Supplementary Table S3). Main efficacy outcomes for the initial treatment cohorts are summarized in Supplementary Table S4.
Small decreases from baseline in mean AST levels (3 U/L [9%] at Week 12 and 6 U/L [14%] at Week 52) were observed in the 10-mg cohort (Supplementary Fig. S3A-B). Mean AST levels were slightly decreased from baseline in the 2-mg and 5-mg cohorts at Week 12 ( Fig. S3C-D).

Lipids
Mean changes from baseline in lipids are summarized in Supplementary Table S2 Fig. S4A-B). Mean low-density lipoprotein cholesterol (LDL-C) levels were also decreased from baseline in the 2-mg, 5-mg, and 10-mg cohorts at  Fig. S4C-D).

J o u r n a l P r e -p r o o f
At Week 12, median levels of the bile acid precursor, C4, were decreased from baseline by 4.5 ng/mL in the 5-mg cohort, and 7.2 ng/mL in the 10-mg cohorts, and increased by 2.5 ng/mL in the 2-mg cohort (Fig. 5A). At Week 52, median levels of FGF19 were decreased from baseline by 17%, 6%, and 15% in the 2-mg, 5-mg, and 10-mg cohorts, respectively (Fig. 5B).

Pruritus VAS score
Mean pruritus VAS scores were decreased from baseline in all cohorts at Week 12: by 4 mm in the 2-mg cohort, 6 mm in the 5-mg cohort, and 12 mm in the 10-mg cohort (Fig. 6). At Week 52, mean VAS scores decreased further from baseline by 10 mm in the 5-mg cohort and 17 mm in the 10-mg cohort, but the initial decrease (3 mm) in the 2-mg cohort remained stable.

Efficacy and Safety in Patients With Cirrhosis
In a prespecified subgroup analysis of patients with and without clinically documented cirrhosis at baseline, reductions in ALP, total bilirubin, and ALT were generally similar between groups at Weeks 12 and 52 (Supplementary Table S6). In the 10-mg cohort, at Week 52, ALP was reduced by 48.5% in patients with (n=9) and 43.2% in patients without cirrhosis (n=40). The J o u r n a l P r e -p r o o f reduction in ALP was smaller in patients with cirrhosis in the 5-mg cohort (n=13) at both time points, although mean baseline ALP was also lower in this subgroup. Safety results were similar in patients with and without cirrhosis (Supplementary Table S7).

DISCUSSION
This study demonstrates the significant anticholestatic effects of seladelpar in patients with PBC who had suboptimally responded to UDCA or were UDCA intolerant. Evidence of seladelpar's dose-dependent efficacy, safety, and improvement in patient-reported pruritus was substantial and durable. Biochemical improvement and safety were generally similar in patients with and without cirrhosis. Significant reductions in mean ALP levels were observed in all cohorts as early as 3 months, with a 43% reduction from baseline and normalization of ALP in 31% of patients in the 10-mg cohort. ALP reductions, an evidence-based surrogate for long-term transplant-free outcomes 9-11 , were maintained through 1 year in patients in the 10-mg cohort. In Synthesis of the bile acid precursor, C4, was dose-dependently reduced from baseline, indicating that seladelpar may protect against cholestatic liver injury by reducing production of bile acids through downregulation of bile acid synthesis. 21 In addition, median reductions from baseline in FGF19 levels of approximately 10% were observed in all cohorts through Week 52. Since FGF19 transcription is activated in response to bile acid binding to FXR, these results indicate that seladelpar's effect on bile acid synthesis is distinct from those of FXR agonists like OCA. 13 Seladelpar was safe and well tolerated at doses up to 10 mg QD, and the safety profile was similar among treatment cohorts. There were no deaths during the study, and no other serious safety signal was identified for seladelpar at doses up to 10 mg QD. Reversible elevations in aminotransferases were observed in 2 patients, although both patients were cirrhotic at baseline and one was able to continue the study without further issue. Pruritus is among the most common symptom of PBC and significantly reduces quality of life. 25  This study has several limitations that should be kept in mind when interpreting results. First, there was an imbalance in baseline ALP levels among cohorts; however, ALP reductions were compared with baseline, and several biochemical endpoints were evaluated. Second, the criteria for dose uptitration were not standardized, except that uptitration could only be done at or after  J o u r n a l P r e -p r o o f c Rotterdam score categories were early (normal total bilirubin and normal albumin), moderately advanced (abnormal albumin OR abnormal total bilirubin), and advanced (abnormal albumin AND abnormal total bilirubin). ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; GGT, gamma-glutamyl transferase; INR, international normalized ratio; MELD, model for end-stage liver disease; mITT, modified intent-to-treat; NC, not calculated; OCA, obeticholic acid; PBC, primary biliary cholangitis; SD, standard deviation; SI, International System of Units; UDCA, ursodeoxycholic acid; VAS, visual analog scale.