Alix Démaris (1,2)*, Ana-Marija Grišić (1,2)*, Wilhelm Huisinga (3), Walter Reinisch (4), Charlotte Kloft (1)
(1) Dept. of Clinical Pharmacy and Biochemistry, Institute of Pharmacy, Freie Universitaet Berlin, Germany, (2) Graduate Research Training program PharMetrX, Germany, (3) Institute of Mathematics, University of Potsdam, Germany, (4) Dept. for Gastroenterology and Hepatology, Medical University of Vienna, Austria, * Shared first authorship
Introduction: Anti-tumour necrosis factor α (TNFα) monoclonal antibodies (mAbs) are a well-established treatment for patients with Crohn’s Disease (CD). However, up to 30% of the treated patients do not respond to treatment (primary non-response) and another 50% demonstrate a loss of response (LOR) over time [1]. Subtherapeutic concentrations of anti-TNFα mAbs have been associated with LOR. Several factors such as immunogenicity and a decrease in body weight have been related to an increased clearance of those mAbs and thus decreased drug concentrations. An appropriate dosing strategy is a crucial aspect to be taken into account to prevent underexposure in those patients. Different dosing strategies (fixed or body size-adapted) are recommended for anti-TNFα mAbs, with no clear conclusion on the consequences of one or the other on pharmacokinetics (PK) and therapeutic outcome. The dosing strategy of several mAbs in other indications have already been called into question, leaning toward the use of fixed dosing rather than body size-adjusted dosing, thus avoiding dosing mistake [2-8].
Objectives: The aim of our study was to (1) assess the impact of different dosing strategies (fixed dose or body size-adapted) on drug exposure and target attainment for an anti-TNFα mAb certolizumab pegol (CZP, currently recommended fixed dosing), and (2) compare it to previously reported results for infliximab (IFX, currently recommended body weight (BW)-based dosing) [9], by performing a comprehensive PK simulation study.
Methods: Based on the distribution of CD patient characteristics from an in-house clinical database (n=116), a virtual population of 1000 clinically representative CD patients was generated in R. Seven dosing regimens were investigated: (i) fixed dose (mg) and dose adjusted to either (ii) BW (mg/kg), (iii) lean BW (LBW, mg/kg LBW), (iv) body surface area (BSA, mg/m²), (v) height (HT, mg/m), (vi) body mass index (BMI, mg/kg/m²) or (vii) fat-free mass (FFM, mg/kg FFM). The individual body size-adjusted doses were calculated from patient generated body size descriptor values. Using previously published PK model for CZP in CD patients [10], 1000 stochastic simulations were performed using NONMEM. For each dosing strategy, the between patients variability in Cmax, Cmin, average mAb concentration (Cavg), AUC and the percent of patients reaching target concentration (Cavg=17 µg/mL [11]) were assessed at steady state, during maintenance therapy. The analysis was performed using NONMEM (version 7.4.3), PsN (version 4.8.1), R (version 3.6.0), and RStudio (version 1.2.1335).
Results: In total, 7000 patients were simulated (seven dosing strategies each including 1000 patients). The model used for the CZP simulation was a one-compartment model with a linear elimination taking into account the effect of immunogenicity, albumin concentration, BSA and C-reactive protein concentration on clearance as well as BSA and race on volume of distribution. The virtual population included patients with BW, HT and BMI ranging from 47.2 to 111 kg, 171 to 190 cm and 14.7 to 41.0 kg/m² respectively. While the previous analysis for IFX reported similar median individual doses across all investigated dosing strategies (~365 mg), in this analysis the median individual doses for CZP ranged from 342 mg to 422 mg (for HT and LBW-adjusted doses, respectively). Similar to IFX, Cmin showed the highest variability between patients (CV~80%) with a similar extent across all dosing strategies and similar value as that found for IFX (CV~110%). In contrast to IFX, for which the percent of patients reaching the target concentration (Cmin=5 µg/mL) was similarly low across all dosing strategies (~15%), for CZP, the percent of patients reaching the target average concentration ranged from 52% to 71% for HT and LBW-adjusted dosing, respectively.
Conclusions: By using a PK simulation approach, different dosing regimens of anti-TNFα mAbs revealed the highest interindividual variability for Cmin, the most commonly used PK parameter guiding treatment decisions, independent upon dosing regimen. For IFX, fixed dosing led to similar target concentration attainment as currently recommended BW-based dosing, calling into question the need for this adjusted dosing regimen. For CZP, the current fixed dosing strategy led to comparable percentage of patients reaching target as the best performing body size-adjusted dosing, LBW-adjusted dosing (66% vs 71%, respectively).
References:
[1] Adegbola S.O. et al. Int. J. Mol. Sci. 19: 2244 (2018).
[2] Wang D.D. et al. J. Clin. Pharmacol. 49: 1012–1024 (2009).
[3] Stroh M. et al. Clin. Pharmacol. Ther. 102: 305–312 (2017).
[4] Novakovic A.M. et al. Clin. Pharmacol. Ther. 0: 1–9 (2019).
[5] Baverel P.G. et al. Clin. Pharmacol. Ther. 103: 631–642 (2018).
[6] Zhao X. et al. Ann. Oncol. 28: 2002–2008 (2017).
[7] Freshwater T. et al. J. Immunother. Cancer. 5: 1–9 (2017).
[8] Wilkins J.J. et al. Adv. Ther. 36: 2414–2433 (2019).
[9] Grisic A. M. et al. J. Crohn’s Colitis. 11 (Suppl. S1): S326–S325 (2017).
[10] Wade J.R. et al. J. Clin. Pharmacol. 55: 866–874 (2015).
[11] U.S. Food and Drug Administration. Cimzia [Label]. (2008).
Reference: PAGE 29 (2021) Abstr 9768 [www.page-meeting.org/?abstract=9768]
Poster: Clinical Applications