Pharmacokinetic/Pharmacodynamic framework to support the subcutaneous dosing regimen of nemolizumab in subjects with moderate-to-severe atopic dermatitis.

Floris Fauchet (1), Antonio Goncalves (1), Wannee Kantasiripitak (1), Eline van Maanen (1), Anna Largajolli (1), Vincent Duval (1), Liliana Ulianov (2), Michael Graeber (3), Patricia Fleuranceau Morel (3), Christophe Piketty (4), Luca Loprete (4), Nathalie Wagner (2)

(1) Certara - CDDS (2) Galderma R&D, Basel, Switzerland (3) Galderma R&D, Dallas, USA (4) Galderma R&D, Zug, Switzerland

Introduction: Nemolizumab (NEMO), a humanized anti-IL-31 receptor A monoclonal antibody was developed in atopic dermatitis (AD). Previous PK/PD models showed that the 30-mg dose given every four weeks (Q4W) with a 60-mg loading dose (LD) was the optimal subcutaneous (SC) dose to achieve therapeutic benefit and was selected for the Phase 3 pivotal studies [1-3]. These models have been updated with Phase 3 data to support the proposed dosing regimen in AD.

Objectives: 1) To characterize the PK/PD relationships between NEMO concentrations and the clinical efficacy; the eczema area and severity index (EASI: 0-72 scale), the investigator global assessment (IGA IGA: 0-4 scale) and the weekly average peak pruritus numerical rating scale (PP NRS: 0-10 scale) 2) To support the proposed dosing regimen in adult and adolescent patients with moderate to severe AD not adequately controlled by topical therapies.

Methods: Data were available from 2209 subjects with AD from six randomized placebo-controlled clinical studies (1 Phase 1, 3 Phase 2 and 2 Phase 3 studies).

Population pharmacokinetic (popPK) model described NEMO concentrations with a 1-compartment distribution model, a linear elimination, a 1st order absorption with lag time and a dose effect on bioavailability. Apparent clearance and volume increased with increasing bodyweight (BW), according to a power function 0.662 and 0.802, respectively.

The PKPD relationships between NEMO, EASI and weekly PP NRS assessments were described by an indirect response model (IRM) with inhibiting drug effect on the rate of production of PD signal using IPP approach [2, 4]. For IGA, a continuous time Markov model (CTMM) was applied. A previously developed CTMM [1,3] was extended to describe IGA score. Models were implemented in NONMEM software version 7.4 [5].

As NEMO exposures decreased with increasing body weight, PK/PD simulations were conducted to assess the impact of BW on clinical efficacy endpoints of interest of the Phase 3 studies. The proportion of EASI, IGA and PP NRS responders were evaluated in 3 different arms (N=1000 virtual patients) receiving the dosing regimen of the phase 3 studies and resampled from subjects included in Phase 3 studies:

Arm 1: Overall subject population (BW range observed in the clinical studies: 31 to 171 kg)
Arm 2: Low BW subject population (BW range: 31 to 89 kg)
Arm 3: High BW subject population (BW range: 90 to 171 kg)

Results: An IRM with a maximum inhibitory effect of NEMO and a constant placebo effect on the production of the EASI and PP NRS response was found adequate to describe both endpoints. The amount in the turnover compartment was constrained by defining a logit-transformed score of both endpoints. No significant covariates were identified in both models.

A 4-compartment CTMM was used to describe IGA observations with a common stimulating drug effect Edesc (linear relationship with concentrations) on all descending transitions, λdesc. An effect of baseline IGA score on all λasc was characterized (higher IGA baseline were associated with higher λasc), suggesting a lower response rate for patients with higher baseline values. Descending transition rate constants λdesc were found to be higher in Female (i.e., higher IGA responder rate) compared to Male.

Table 1 – Parameter estimates for the popPK and PK/PD models

Parameter	Point Estimate	%RSE	IIV	%RSE
popPK
CL/F (L/day)	0.263	1.3	0.109	2.2
V/F (L)	7.670	1.3	0.110	6.2
Ka (1/day)	0.476	7.7	0.344	10.0
lag-time(day)	0.043	23.0
Weight effect on CL/F	0.662	6.0
Weight effect on V/F	0.802	5.3
Covariance (CL/F and V/F)	0.068	9.97
EASI
K_OUT (1/day)	0.0328	4.1	0.946	3.7
P₁	1.43	4.5	2.18	2.7
Imax	1.40	20.8
IC50 (µg/mL)	4.27	43.1
IGA
λ₁₂ (day⁻¹)	0.00794	7.8
λ₂₁ (day⁻¹)	0.00708	6.4	0.444	5.3
λ₂₃ (day⁻¹)	0.00909	5.8
λ₃₂ (day⁻¹)	0.01200	5.1	0.444	5.3
λ₃₄ (day⁻¹)	0.00334	6.9
λ₄₃ (day⁻¹)	0.0266	6.2	0.444	5.3
Slope/1000	0.0964	14.0
Baseline effect (4/5)	2.25	6.9
Sex effect (female)	0.412	16.0
PP NRS
tPP NRS0	1.82	0.76	0.132	1.5
KOUT (1/day)	0.0396	2.5	1.28	2.5
P1	0.545	8.1	2.34	1.8
Imax	2.23	0.60
IC50 (µg/mL)	0.654	2.6

Results of PK/PD simulations showed a similar efficacy across arms, thus, confirming that exposure variability due to BW did not result in a clinically meaningful difference in efficacy.

Conclusion: Current framework was used to support the proposed SC dosing regimen in adult and adolescent subjects with moderate to severe AD not adequately controlled by topical therapies, i.e. 30-mg Q4W dose (60-mg LD) with no dose adjustment based on BW.

References:
[1] Wagner N, Loprete L, Duval V, Jauslin P, Benkali K, Silverberg JI, Wollenberg A, Saito T, Ahmad F, Graeber M, Winkelman W, Piketty C. Selection of Nemolizumab Clinical Dosage for Atopic Dermatitis. J Drugs Dermatol. 2023 Oct 1;22(10):1017-1020. doi: 10.36849/JDD.7437R1. PMID: 37801534.
[2] Jauslin et al. PAGE 28 (2019) Abstr 9088 [www.page-meeting.org/?abstract=9088
[3] Shindler et al. PAGE 28 (2019) Abstr 9178 [www.page-meeting.org/?abstract=9178
[4] Zhang L, Beal SL, Sheiner LB. Simultaneous vs. sequential analysis for population PK/PD data I: best-case performance. J Pharmacokinet Pharmacodyn. 2003 Dec;30(6):387-404.
[5] Beal SL et al. NONMEM 7.3.0 users guides. (1989–2013). Hanover: ICON Development Solutions.

Reference: PAGE 32 (2024) Abstr 10968 [www.page-meeting.org/?abstract=10968]

Poster: Drug/Disease Modelling - Other Topics

PDF poster / presentation (click to open)