Supplementary Table 1. Summary of Sonidegib Studies Included in Population Pharmacokinetic
1
Supplementary Table 1.Summary of sonidegib studies included in population pharmacokinetic analysis
Study / N / Description / Sonidegib Doses (food intake) / PK Collection Time PointsA1102 / 36 / Phase 1, open-label, dose-escalation study to assess the PK of a single dose of sonidegib in healthy Japanese subjects / Single dose of 200, 400, 800 mg
(after an overnight fast) / Predose and 0.5, 1, 2, 4, 6, 8, 10, 12, 24, 48, 72, 96, 120, 168, 336, 504, 672, 1008, 1344, 1680, and approx. 2016 h (approx. 12 weeks) postdose starting on day 1
A2114 / 49 / A randomized, open-label study to evaluate the relative bioavailability of three FMI formulations of sonidegib compared with the CSF capsulea formulation and the effect of food in healthy subjects / Single dose of 200, 800, or 1200 mg
(after an overnight fast)
Single dose of 800 mg (after a high-fat meal) / Predose and 0.5, 1, 2, 4, 6, 8, 10, 12, 24, 48, 72, 96, 120, 168, and 336 h postdose starting on day 1
Postdose at 504, 672, 1008, 1344, 1680, and approx. 2016 h (approx. 12 weeks) starting on day 1 (stages 1 and 2 only)
X2101 / 103 / Phase 1, open-label, dose-escalation
study of sonidegib in patients with advanced solid tumours / QD doses: 100, 200, 400, 800, 1000, 1500, and 3000 mg
BID doses: 250, 450, and 750 mg
(2 h after a light meal) / PK run-in: Predose and 0.5, 1, 2, 4, 6, 8, 24, 48, 72, and 96 h postdose starting on day 1
Predose on day 1, 8, 15, 16, and 22 of cycles 1 and 2
For day 15 of cycle 1, samples were collected at 0.5, 1, 2, 4, 6, and 8 h postdose
Predose on day 1 of all subsequent cycles
X1101 / 21 / Phase 1, open-label, dose-escalation study of sonidegib in patients of East Asian descent with advanced solid tumoursb / 400, 600 mg QD
(2 h after a light meal) / PK run-in: Predose and 0.5, 1, 2, 4, 6, 8, 24, 48, 72, and 96 h postdose starting on day 1
Predose on day 1, 8, 15, 16, and 22 of cycle 1; Postdose at 0.5, 1, 2, 4, 6, and 8 h on day 15 of cycle 1
Predose on day 1, 8, 15, and 22 of cycle 2
Predose on day 1 of all subsequent cycles
A2201c / 227 / Phase 2, randomized double-blind study of efficacy and safety of two dosages of sonidegib in patients with laBCC or mBCC / 200, 800 mg QD
(2 h after a light meal) / All patients: Predose weeks 1, 3, 5, 9, 13, 17, 21, 33, 45, 57, and 69
Subset (86): up to 4 additional samples at week 17
BID, twice daily; CSF, clinical service formulation; FMI, final market image; laBCC, locally advanced basal cell carcinoma; mBCC, metastatic basal cell carcinoma; PPI, proton-pump inhibitors; QD, once daily.
aIncludes participants taking the capsule formulation (FMI or CSF) only.
bIncludes Japanese cohort only; Chinese cohort was not completed at the time of the analysis.
c Only 227 of 230 patients enrolled in this study were included in the analysis.
Supplementary Table 2.Description of covariate groups
Covariate / Covariate Groups and Reference CategoryBody weighta / 41.9 – 69.9 kg, ≥ 69.9 – 84.4 kg (ref)b, ≥ 84.4 – 181 kg
Age / < 65 years (ref)b, ≥ 65 years
Gender / Male (ref)b, Female
Ethnicity / Subjects enrolled in Western countries (ref)b, Japanese
Coadministration of H2 receptor antagonist or PPI / No (ref)b, Yes
Baseline liver function / Normal, mild, moderate, and severe hepatic impairmentc
Baseline renal function / Normal, mild, moderate, and severe renal impairmentd
AST, aspartate aminotransferase; PPI, proton pump inhibitor; ULN, upper limit of normal.
a Cut-off values for body weight correspond to the approximate 1st and 3rd quartiles of the corresponding covariate range in the study.
bReference group.
c Normal: total bilirubin ≤ ULN and AST ≤ ULN; Mild: total bilirubin ≤ ULN and AST ≥ ULN or total bilirubin > 1 to ≤ 1.5 ×ULN and AST = any value; Moderate: total bilirubin > 1.5 to ≤ 3.0 × ULN and AST = any value; Severe: total bilirubin > 3 × ULN and AST = any value [20].
d Normal: creatinine clearance ≥ 90 ml/min; Mild: creatinine clearance 60 to < 90 ml/min; Moderate: creatinine clearance 30 to < 60 ml/min; Severe: creatinine clearance 15 to 30 ml/min.
Supplementary Table 3.Covariate relationships in the full model
Covariate / Parameters / Reason for investigationDose (Ca) / F / There was evidence of dose-dependent bioavailability.
Healthy subject vs. Cancer Patient (D) / CL/F, F / CL/F differences may occur between healthy subjects and cancer patients. In addition, the meal protocols for healthy subjects and cancer patients were different, which may affect F.
High fat meal vs 2 hours post light meal (D) / F, Ka / High fat meal conditions may alter bioavailability.
Age (C) / CL/F / Clearance generally decreases with age.
Body weight (C) / CL/F, Vc/F,
Vp/F / Clearance and volume are often correlated with size. The impact of body weight was assumed to be same on Vp/F and Vc/F.
Sex (D) / CL/F / Sex can be correlated with differences in clearance.
Ethnicity (Western vs. Japanese) (D) / CL/F / Different ethnic populations may have differences in clearance.
Estimated creatinine clearance by Cockcroft-Gault formula (C) / CL/F / Renal function may affect clearance.
Albumin (C) / CL/F, Vc/F,
Vp/F / Protein binding may affect clearance and volume
ALT (C) / CL/F / Measures of liver function may be correlated with differences in clearance
Normalized -bilirubin (C) / CL/F / Measures of liver function may be correlated with differences in clearance
Multiple dosing in patients vs single dose (D) / F / Variable compliance with food restriction during the multiple dose phase may affect the bioavailability relative to the first dose.
Concurrent H2 receptor antagonists (D) / F / As sonidegib solubility is pH dependent, medications that modify the gastric pH could affect the bioavailability.
Concurrent proton pump inhibitors (D) / F / As sonidegib solubility is pH dependent, medications that modify the gastric pH could affect the bioavailability
aC=Continuous, D=Dichotomous
Supplementary Table 4.Effects of covariates on steady-state AUC and Cmax in patients
Cov / Cat:Ref / Ratio of steady-state AUC0-24h / Ratio of steady state CmaxGeo mean / 5% / 95% / Geo mean / 5% / 95%
Ethnicity / Japanese:Western / 0.96 / 0.50 / 2.12 / 0.97 / 0.52 / 2.01
PPI / Yes:No / 0.66 / 0.34 / 1.31 / 0.66 / 0.36 / 1.21
Gender / Female:Male / 1.15 / 0.51 / 2.53 / 1.14 / 0.53 / 2.42
Body weight / [41.9,69.9]:(69.9,84.4] / 0.99 / 0.44 / 2.07 / 0.99 / 0.46 / 2.01
(84.4,181]:(69.9,84.4] / 1.02 / 0.54 / 2.09 / 1.02 / 0.56 / 2.01
Baseline renal function / Moderate:Normal / 1.04 / 0.49 / 2.10 / 1.03 / 0.50 / 2.00
Mild:Normal / 1.10 / 0.56 / 2.19 / 1.09 / 0.58 / 2.08
Baseline liver function / Mild:Normal / 0.72 / 0.33 / 1.56 / 0.74 / 0.34 / 1.50
Age / >=65:<65 / 1.13 / 0.56 / 2.40 / 1.12 / 0.56 / 2.27
Note: Cov refers to covariate category; Cat:Ref is the ratio of category to a reference category within a covariate; 5th and 95th refers to the percentile of the distribution, Geo mean is the geometric mean; Western refers to patients enrolled in Western countries
Additional details on covariate categories are provided in Supplementary Table 2.
Supplementary Figure 1. The time course of LGE899 to sonidegib ratio over time after repeated dosing in the A2201 trial
Supplementary Figure 2. Goodness-of-fit plots for full model for all data (A) in healthy subjects after a single dose (B) and in patients after a single dose (C) or after multiple dosing during the first 20 weeks of treatment (D)
CWRES, conditional weighted residuals; HV, healthy volunteers; IWRE, individual weighted residuals.
Supplementary Figure 3. Visual predictive checks in healthy subjects (A) and patients (B) after a single dose of sonidegib
A: Observed sonidegib concentrations (dots) plotted vs time, with the predicted (solid line) and observed (dashed line) medians and the 95% PI (shaded area) overlaid; data are shown for fasting conditions.
B. Observed sonidegib concentrations (dots) plotted vs time, with the predicted (solid line) and observed (dashed line) medians and the 95% PI (shaded area) overlaid.
Supplementary Figure 4.Histograms of estimated random effects from the full model