Anson K. Abraham, Tommy Li*, Dinesh de Alwis, Kapil Mayawala
Quantitative Pharmacology & Pharmacometrics, PPDM (West Point, PA and Rahway, NJ), MRL, Merck & Co. Inc., Kenilworth, NJ, USA. *Currently affiliated with Genmab US, Inc.
Introduction: For immune-oncology (IO) targets, selection of a biologically effective dose (BED) is a key decision in early drug development [1]. BED is expected to maximize efficacy based on understanding of target pharmacology. This work outlines an approach for BED determination for IO targets expressed on immune cells, e.g., myeloid-derived suppressor cells, macrophages, whose distribution is not restricted to the tumor and blood alone [2,3].
Objectives: To predict a BED that ensures adequate target engagement in the tumor for a monoclonal antibody (mAb) in early oncology clinical development.
Methods: For targets with high expression on tissue resident immune cells, a minimal physiologically based pharmacokinetic (mPBPK) model [4] framework was used as the starting point. A central and peripheral TMDD model, parameters of which characterize mAb PK and target engagement in both central and peripheral compartments, was added to the mPBPK model. A spatial tumor PBPK model was also incorporated to characterize spatial variations in target engagement in the tumor. For purposes of the simulation, the mAb dose selected to achieve target saturation in blood was 300 mg. Simulations were then performed to predict target engagement in different types and sizes of tumor, ranging from small vascular tumor to large avascular tumor, representing variations across human cancer types.
Results: Target-mediated drug disposition (TMDD) binding models have been used to describe drug-target binding dynamics for prediction of target saturation in blood and a relevant “site-of-action” [5]. A key assumption in these models is that target expression and binding in peripheral tissues can be ignored. While this assumption generally holds true, incorporating drug-target binding in tumor as well as non-tumor tissues is critical for BED calculation for targets with widespread expression across tissues.
Exploration of model simulations demonstrate that conventional TMDD model, i.e., without including drug-target binding across tissues, can significantly overpredict BED. This counterintuitive behavior is a result of misinterpretation of blood target engagement data by the conventional central TMDD model.
Typically, target saturation in blood can be clinically characterized by experimental measurement of target engagement. In the scenario presented here, the dose selected to achieve target saturation in blood was 300 mg. The conventional TMDD model within the mPBPK framework predicted BED as 2000-3000 mg based on prediction of target engagement in tumor. However, the proposed mPBPK model, including peripheral TMDD and spatial variations within tumor, predicted BED as 800 mg.
This work shows that selection of BED must consider drug-target binding across tissues. Without such a consideration, this work shows that blood saturation data alone underpredicts BED (as 300 mg), while the use of conventional TMDD model within the mPBPK framework overpredicts BED (as 2000-3000 mg).
Conclusions: A novel spatial tumor mPBPK model including central and peripheral TMDD was developed to predict target engagement in blood, tissue and tumor enabling accurate estimation of BED. The simulations demonstrate that inclusion of peripheral TMDD is critical for targets with high expression on tissue resident immune cells. Ignoring this aspect would lead to a ~3-fold overprediction of BED in the clinic.
References:
[1] Mayawala K, Tse A, Rubin EH, Jain L, Alwis DD. Dose Finding Versus Speed in Seamless Immuno-Oncology Drug Development. The Journal of Clinical Pharmacology. 2017; 57:S143-S145.
[2] Solito S, Falisi E, Diaz-Montero CM, Doni A, Pinton L, Rosato A, et al. A human promyelocytic-like population is responsible for the immune suppression mediated by myeloid-derived suppressor cells. Blood. 2011; 118:2254–2265
[3] Condamine T, Gabrilovich DI. Molecular mechanisms regulating myeloid-derived suppressor cell differentiation and function. Trends Immunol. 2011; 32:19–25
[4] Cao Y, Balthasar JP, Jusko WJ. Second-generation minimal physiologically-based pharmacokinetic model for monoclonal antibodies. Journal of Pharmacokinetics and Pharmacodynamics. 2013; 40 597–607.
[5] Chudasama VL, Zutshi A, Singh P, Abraham AK, Mager DE, Harrold JM. Simulations of site-specific target-mediated pharmacokinetic models for guiding the development of bispecific antibodies. Journal of pharmacokinetics and pharmacodynamics. 2015; 42 (1), 1-18
Reference: PAGE () Abstr 9310 [www.page-meeting.org/?abstract=9310]
Poster: Drug/Disease Modelling - Oncology