Chiara Zecchin (1), Lia Liefaard (1)
(1) GlaxoSmithKline
Introduction: GSK2831781 is a first in class monoclonal antibody (mAb) depleting cells expressing Lymphocyte Activation Gene-3 (LAG3) on the cell surface. The proposed indication is ulcerative colitis (UC). Clinical experience with GSK2831781 included in this analysis comes from a single IV dose escalation in healthy volunteers and patients with psoriasis (NCT02195349).
Objectives:
- To develop a PKPD model to describe systemic PK and LAG3+ cells depletion.
- To simulate PK and cell depletion in plasma and at site of action for UC (colon) with repeat dosing.
- To inform the design of the phase 2 study and the doses to be tested.
Methods: Data collected during the Phase I single IV dose escalation study 200630 (GSK; NCT02195349) were available. Doses ranging from 0.003 to 0.15 mg/kg were evaluated in healthy male volunteers and doses of 0.5, 1.5 and 5 mg/kg were evaluated in psoriasis patients. Longitudinal PK, soluble target (sLAG3) and number of LAG3+ T-cells samples were measured for all cohorts. This information was used to develop a mPBPK model [1] to describe total drug and soluble target concentration values in plasma. The number of LAG3+ T-cells was described using an indirect effect model. As LAG3+ T-cell counts in blood are the same in healthy volunteers and patients with UC, this mPBPK-PD model was used to simulate PK and cell depletion in plasma after repeat doses of GSK2831781 at different dose levels. To predict cell depletion in colon, the PK-PD relationship was extrapolated to colon assuming distribution into colon based on literature data for “typical mAbs” [2]. Baseline colon LAG3+ T-cell count was obtained from colon biopsies of patients with active UC [3]. Simulations of PK and target cells depletion in plasma and colon guided the selection of doses and dosing schedule for the ph2A/B study, aiming for a wide range of exposure (approximately 10-fold) and target cells depletion (between 20 and >95%).
Results: The mPBPK model [1] with target mediated drug disposition (TMDD) in plasma and both linear and Michaelis-Menten clearance best described drug and soluble target concentration in plasma. The drug/soluble target affinity equilibrium and dissociation constants were fixed to the values determined in vitro (0.223 nM and 15.5 1/h respectively). The estimated (95%CI) values of PK parameters were: linear clearance 0.0119 (0.0090-0.0156) L/h, Michaelis-Menten parameters characterising the non-linear clearance Vmax=0.0282 (0.0240-0.0330) mg/L/h and Km=0.2254 (0.165-0.307) mg/L; degradation rate of drug + soluble target complex 0.0046 (0.00341-0.00628) 1/hr, soluble target zero order synthesis rate constant 0.143 (0.108-0.189) nM/hr and first order degradation rate constant 1.10 (0.966-1.25) 1/hr. An indirect effect model, with drug exposure increasing target cells elimination with a Hill relationship described target cells depletion in plasma. The estimated values of PD parameters were: Emax 22, EC50 9509 ng/mL and Hill coefficient 1.13. Target cells zero order production and first order elimination rate constant were, respectively, 4.34 cells-count/hr and 0.0046 1/hr. PK and PD model parameters were estimated sequentially in NONMEM 7.3 [4]. PK model parameters were estimated with the SAEM followed by IMP estimation method. PD model parameters were estimated with the FOCE-I estimation method, assuming individual PK parameters were known and fixed to the previously estimated values.
Population PK and cell depletion were simulated in plasma and colon using parameter estimated values and baseline LAG3+ T-cell count of 2500 cell/mm3, using the software R [5]. Doses between 45 mg and 450 mg where selected to be tested in the induction of the phase 2A/B study.
Conclusions: Choosing the right dose(s) is key to design an informative Phase 2A/B study. In this analysis a mPBPK model with TMDD and linear and non-linear clearance was developed to describe plasma PK and soluble target. Target cell depletion data were described by an indirect response model, with drug exposure increasing target cells elimination. The mPBPK + PD model was used to simulate drug exposure and cell depletion in colon (target tissue in UC) and plasma and guide selection of doses for the Ph2A/B planned study.
References:
[1] Cao Y and Jusko WJ. “Incorporating target-mediated drug disposition in a minimal physiologically-based pharmacokinetic model for monoclonal antibodies.” J pharmacokinet pharmacodyn 41.4 (2014): 375-387.
[2] Shah DK, Betts AM. (2012) Towards a platform PBPK model to characterize the plasma
and tissue disposition of monoclonal antibodies in preclinical species and human. J
Pharmacokinet Pharmacodyn; 39(1): 67–86.
[3] Slevin S, Tan M, Lahiff C, Williamson K, Geremia A, Hughes S, Leavens K, Nevin K,
Marks DJB, Tarzi R, Srinivasan N, Arancibia C, Keshav S. (2018). Intestinal expression
of LAG-3 correlates with inflammatory activity and response to biological therapy in
ulcerative colitis. Journal of Crohn’s and Colitis, Volume 12, Issue supplement_1; S125.
[4] Beal SL., et al. “NONMEM 7.3. 0 Users Guides. (1989–2013).” ICON Development Solutions, Hanover, MD.
[5] The R Foundation for Statistical Computing, version 3.3.2 (2016-10-31).
Reference: PAGE 29 (2021) Abstr 9764 [www.page-meeting.org/?abstract=9764]
Poster: Drug/Disease Modelling - Other Topics