Anna-Kristina Oma 1,2, Erwin Dreesen 3, Erik Ingmar Hallin 1, Susanna Röblitz 4, Trond Trætteberg Serkland 1,2, Marianne Aanerud, Rune Hørgård Tilseth 6, Silje Skrede 1,2
1 Department of Medical Biochemistry and Pharmacology, Haukeland University Hospital (Bergen, Norway), 2 Department of Clinical Science, University of Bergen (Bergen, Norway), 3 Department of Pharmaceutical and Pharmacological Sciences, KU Leuven (Leuven, Belgium), 4 Department of Informatics, University of Bergen (Bergen, Norway), 5 Department of Thoracic Medicine, Haukeland University Hospital (Bergen, Norway), 6 Department of Medicine, Førde Hospital (Førde, Norway)
Objectives: Immune checkpoint inhibitors (ICI) such as pembrolizumab have transformed outcomes in advanced non-small cell lung cancer (NSCLC) [1]. Still, exposure–response plateaus support the investigation of alternative dosing strategies that could improve resource use and cost-effectiveness of ICI treatment and potentially decrease the risk and severity of immune-related adverse events[2,3].
In this study, we aimed to assess through population pharmacokinetic (popPK) modelling and simulation whether alternative regimens maintain putative minimum effective concentrations (MECs).
Methods: We performed a prospective, observational study in patients with NSCLC receiving standard pembrolizumab therapy. Serum concentrations, dosing histories and clinical covariates (sex, age, body weight, height, Eastern Cooperative Oncology Group (ECOG) performance status, tumour proportion score, smoking status etc.) from patients were used to develop a popPK model (NONMEM 7.5). Model structures with time-stationary and time-dependent clearance were explored. Interindividual variability and residual error were quantified. The base popPK model was selected considering objective function value reductions (difference ≥3.84 points; p ≤0.050), precision of parameter estimates, and goodness-of-fit diagnostics. Covariate effects were investigated through stepwise covariate modelling (αforward=0.01, αbackward=0.001). The tested covariates were body weight, body surface area, body mass index, sex, ECOG performance status, serum albumin and lactate dehydrogenase. Both median imputation and next observation carried backwards were tested for handling covariate missingness.
The final popPK model informed stochastic simulations in 1,000 virtual patients implemented in mrgsolve (R package version 1.6.1). We simulated standard regimens (200 mg Q3W and 400 mg Q6W) and alternative regimens (100 mg Q3W, 200 mg Q6W, and 400 mg Q9W). Attainment of two previously suggested MEC-thresholds (1.5 µg/mL and 15 µg/mL) was investigated both after first dose and at steady state [4,5].
Results: Twenty-eight pembrolizumab-treated patients with NSCLC (female: 43%; body weight: 76.4 ±15.2 kg) contributed a total of 91 serum concentrations (median: 4 per patient, range: 1–6). Trough concentrations (n=46) had a median and range of 35.9 (14.3–80) µg/mL, with one concentration <15 µg/mL. The majority of patients received 200 mg Q3W (n=20). A one-compartment model with time-dependent clearance (sigmoidal decrease) best described the data. Although stepwise covariate modelling identified body surface area as a covariate on Imax, allometric scaling (body weight) was instead included on clearance and volume of distribution to aid interpretability. The estimated typical value of the initial clearance was 0.187 L/day (4% RSE), decreasing by ~60 % to 0.072 L/day by ~7 months after first dose. The estimated volume of distribution was 2.86 L (5%). Estimated Imax and TI50 were -0.614 (12%) and 61.7 days (12%), respectively, while the Hill coefficient was fixed to 3.07[6]. Interindividual variability was 20.6% (coefficient of variation; CV) on volume of distribution and 29.9% on Imax. A proportional error model (10.6%) best described the residual unexplained variability. In simulations, 5th percentile minimum concentrations (95% of concentrations above) after the first dose were 6.5 µg/mL, 2.3 µg/mL and 1.0 µg/mL for 100 mg Q3W, 200 mg Q6W, and 400 mg Q9W, respectively. Corresponding steady-state values (at day 252) were 14.0 µg/mL, 5.9 µg/mL, and 3.1 µg/mL, respectively. Conclusion: We identified two dose-de-escalation/extended interval regimens of pembrolizumab (100 mg Q3W, 200 mg Q6W) that are predicted to meet the 1.5 µg/mL MEC, with 200 mg Q6W aligning with previous in silico findings [4]. While this threshold also could be achieved with a maintenance regimen of 400 mg Q9W, more frequent infusions are needed early in treatment to achieve sufficient concentrations. These findings motivate prospective clinical evaluation, pharmacoeconomic assessment, and consensus on the clinically relevant exposure targets and the role of therapeutic drug monitoring for immunotherapy in NSCLC. References: References 1. Zhou F, Meng Q, Caicun Z. The cutting-edge progress of immune-checkpoint blockade in lung cancer. Cell Mol Immunol. 2021;18(2):279–93. 2. Peer CJ, Goldstein DA, Goodell JC, Nguyen R, Figg WD, Ratain MJ. Opportunities for using in silico-based extended dosing regimens for monoclonal antibody immune checkpoint inhibitors. Br J Clin Pharmacol. 2020;86(9):1769–77. 3. Wesevich A, Goldstein DA, Paydary K, Peer CJ, Figg WD, Ratain MJ. Interventional pharmacoeconomics for immune checkpoint inhibitors through alternative dosing strategies. Br J Cancer. 2023;129(9):1389–96. 4. Peer CJ, Heiss BL, Goldstein DA, Goodell JC, Figg WD, Ratain MJ. Pharmacokinetic Simulation Analysis of Less Frequent Nivolumab and Pembrolizumab Dosing: Pharmacoeconomic Rationale for Dose Deescalation. The Journal of Clinical Pharmacology. 2022;62(4):532–40. 5. Wang N, Zheng L, Li M, Hou X, Zhang B, Chen J, mfl. Clinical efficacy and safety of individualized pembrolizumab administration based on pharmacokinetic in advanced non-small cell lung cancer: A prospective exploratory clinical trial. Lung Cancer. 2023;178:183–90. 6. Li H, Yu J, Liu C, Liu J, Subramaniam S, Zhao H, mfl. Time dependent pharmacokinetics of pembrolizumab in patients with solid tumor and its correlation with best overall response. J Pharmacokinet Pharmacodyn. 2017;44(5):403–14.
Reference: PAGE 34 (2026) Abstr 12135 [www.page-meeting.org/?abstract=12135]
Poster: Drug/Disease Modelling - Oncology