Ana M Novakovic (1), Justin Wilkins (2), Haiqing Dai (3), Satjit Brar (4), Carlo L. Bello (5), Berend Neuteboom (3), Pascal Girard (6), Akash Khandelwal (1)
(1) Merck KGaA, Darmstadt, Germany (2) Occams Coöperatie U.A., Amstelveen, The Netherlands (3) EMD Serono Research & Development Institute, Inc., Billerica, MA, USA (4) Pfizer, La Jolla, USA (5) Pfizer, San Francisco, USA (6) Merck Institute for Pharmacometrics, Lausanne, Switzerland
Objectives: Avelumab is a human anti–PD-L1 IgG1 antibody that has demonstrated meaningful efficacy and clinical activity across various tumor types with a manageable safety profile. Avelumab 10 mg/kg IV every two weeks (Q2W) is approved in the US, EU and Switzerland for treatment of metastatic Merkel cell carcinoma (MCC), in Japan for curatively unresectable MCC, and for advanced or metastatic urothelial carcinoma (UC) in the US (1), and is in clinical development for other cancer types. In the case of monoclonal antibodies, a flat dose regimen (independent of body weight) is likely to minimize drug wastage, facilitate preparation and administration, and reduce pharmacy errors in comparison to body weight dosing. In addition, a flat dosing regimen is expected to result in more consistent exposure across patient populations, when the power exponent describing the effect of body weight on clearance (CL), using population PK models, is estimated to be less than 0.5 (2). The choice of the flat dose to be considered (800 mg) was based on median body weights of the advanced or metastatic UC and mMCC patients treated with avelumab and general populations of patients with cancer. The objectives of this analysis were to compare avelumab exposure between weight-based (10 mg/kg Q2W) and flat (800 mg Q2W) dosing regimens using simulations based on two population PK models (PK CYCLE and PK SS) describing avelumab PK after single (3) and multiple dose infusions (4). Time-varying CL was introduced to adequately describe the data across all cycles of avelumab treatment. Body weight was found to be a significant covariate on clearance and central volume of distribution. The estimated magnitudes of weight exponents on CL are 0.335 (95% CI: 0.277-0.393) for the PK CYCLE model and 0.324 (95% CI: 0.260-0.388) for the PK SS model.
Methods: Data from 1827 patients across three clinical trials were used for modeling, all given the weight based dosing regimen (dose of 1, 3 or 10 mg/kg Q2W). PK CYCLE and PK SS models were used to simulate exposures for weight-based dosing and flat dosing regimens. A total of 10,000 subjects were simulated taking into account both parameter uncertainty and inter-individual variability. Sets of covariates (including body weight) were resampled from the original data set, with all covariates present in the final model retained at the subject level. Various PK exposure metrics, including maximum serum concentrations (Cmax), minimum serum concentrations (Ctrough), and area under the serum concentration-time curve (AUC) after a single dose and at steady state, were simulated for both dosing scenarios and summarized across the entire weight range and by quartiles of weight.
Results: Simulations of exposure comparing weight-based (10 mg/kg Q2W) and flat dosing (800 mg Q2W) after the first cycle (PK CYCLE) and at steady state (PK SS) suggested that the overall variability in exposure was lower for the flat dosing group (coefficient of variation (CV) = 27.1% vs 29.0% for AUC0-336h; 38.6% vs 41.2% for AUCss). The flat 800 mg dose resulted in slightly higher exposures for the simulated population with the median increasing by approximately 12% and 11% for AUC0-336h and AUCss, respectively. For weight-based 10 mg/kg dosing, the lowest weight quartile is associated with the lowest exposure. The opposite is true for the flat 800 mg dose where the highest weight quartile is associated with the lowest exposures. Simulations based on other exposure metrics confirm the results presented here with AUC. Overall, the simulated exposure for the 800 mg flat dosing regimen fell within the exposure range of avelumab known to be clinically effective and safe.
Conclusions: Modeling and simulation-based analyses were performed to simulate PK exposure for the currently approved 10 mg/kg Q2W and the proposed flat 800 mg Q2W dosing regimens. Predicted PK exposures were similar. Lower variability was observed for flat dosing compared to weight-based dosing, as expected for monoclonal antibodies with weight exponent on CL estimated to be less than 0.5, according to findings from Wang et al. (2). Given the results of these simulations, 800 mg flat dose of avelumab should result in similar exposures and thus maintain the positive benefit risk profile of avelumab while simplifying administration. A flat dose regimen is currently being implemented in several trials.
References:
[1] Bavencio (avelumab) [package insert]. Darmstadt, Germany: Merck KGaA; 2017.
[2] Wang D, et al. J Clin Pharmacol. 2009;49:1012-1024.
[3] Wilkins J, et al. ACOP8 2017: Abstract W-079.
[4] Wilkins J, et al. ACOP8 2017: Abstract W-080.
Reference: PAGE 27 (2018) Abstr 8678 [www.page-meeting.org/?abstract=8678]
Poster: Drug/Disease Modelling - Oncology