Application of an immune-oncology framework to explore the role of regulatory T cells in antitumor response.
Serrano-Alcaide, Alejandro1; Zalba, Sara1; Sancho Araiz, Aymara1; Casares, Noelia2; Lasarte, Juan José2; Trocóniz, Iñaki F1; Garrido, María J1.
1) Pharmacometrics & Systems Pharmacology Group, Department of Pharmaceutical Technology and Chemistry, School of Pharmacy and Nutrition, University of Navarra, Pamplona, Spain. 2) Immunology & Immunotherapy Research Program, Applied Medical Research Centre, University of Navarra, Pamplona, Spain
Introduction and Objective: Regulatory T cell (Treg), characterized by the expression of Foxp3 intracellularly and CD25 at the surface, is a key element in immune homeostasis and self-tolerance. Blockade of both biomarkers by using a Foxp3 inhibitor peptide P60 and an anti-CD25 mAb, combined in a targeted nanoliposome, induced tumor regression in a syngeneic murine tumor model.
In this work, we expanded a computational model previously developed [1] that described the combination of vaccines, Toll-like receptor agonist and anti-PD1 mono-clonal antibodies, aiming to characterize Treg effects on tumor resistance after administering a combination treatment of P60 and two mAbs (an anti-CD25, and anti-PD1). The peptide was given either free or encapsulated into non- or CD25 targeted liposomes (TgLP) [2].
Methods: Longitudinal tumor size data were obtained from 80 mice that received a subcutaneous (s.c) injection of 105 MC38 murine cancer cells. Animals were split into even groups: control (n=16); free P60 peptide (n=16); non-targeted P60 liposomes (n=8); CD25-targeted empty liposomes (n=8); TgLP (n=16); anti-PD1 (n=8); and combination of TgLP with anti-PD1 (n=8).
Data were log-transformed and analysed using Monolix version 2023R1 (https://lixoft.com/products/monolix/). Parameters were estimated by using the stochastic approximation expectation maximization algorithm (SAEM), assuming a lognormal distribution of the individual parameters and a constant error model for the residual variability of the observations. Observations below the limit of quantification (BLQs) were established according to our experimental data and coded as censored observations.
Model selection and evaluation was assessed by checking the precision of parameter estimates and goodness-of-fit plots; and Visual Predictive Checks (VPCs) that included the 5th, 50th and 95th percentiles of the data, respectively. Data pre-processing and plots were performed using tidyverse R package.
Results: Control group was characterised first using the Simeoni model. The estimated parameters were: initial tumor size (TS0 = 97.35 mm3; RSE = 8.00 %), exponential (Kge = 0.18 day-1; RSE = 5.90) and linear growth rates (Kgl = 162.9 mm3/day; RSE = 11.2 %). Lack of pharmacokinetic data led to consider a K-PD approach, where kinetics was inferred from the dynamics of the observed response [3]. Administration of P60 increased the anti-PD1 tumor response, which was greater in the case of targeted anti-CD25 liposomes. In the original model, Treg effects were implicitly included reducing the anti-exhaustion CD8 shield built by anti-PD1 mAbs. The additional data were described in the model by a reduced activity rate of Treg induced by P60. The fact that the P60 response was increased in the case of non- and targeted liposomes allowed to include an intracellular compartment accounting for the enhanced tumor bioavailability of P60 of the targeted delivery device. This bioavailability was included in the model as a correction factor affecting P60 elimination rate (F = 0.22) and extracted from in-vitro release studies. As a result, inclusion of TgLP favoured the reduction of Treg activity compared to free P60 administration even at doses that were 20 times lower than the original, free one, achieving 50% tumor rejection in mice. Anti-PD1 treatment, with a more stable kinetics profile compared to P60 administration (kPD1 = 0.75 day-1; kP60 = 103.8 day-1), resulted in a 62.5% tumor rejection, achieving total response when combination therapy was selected.
Conclusions: Application of a previously developed computational platform for describing antitumor effects of immune combination therapies in the pre-clinical arena successfully described the activity of Treg immune modulators, identifying the role of this subset of T cells in the tumor response and progression. These results set a starting point for future combinational therapies where more antitumor agents could be evaluated, in order to expand the mechanisms of antitumor resistance in oncology.
- Sancho-Araiz A, Zalba S, Garrido MJ et al. Semi-mechanistic model for the antitumor response of a combination cocktail of immuno-modulators in non-inflamed (cold) tumors. Cancers 2021; 13(20):5049. doi: 10.3390/cancers13205049
- Casares N, Rudilla F, Arribillaga L et al. A Peptide Inhibitor of FOXP3 Impairs Regulatory T Cell . Journal of Immunology, 2010, 185: 5150–5159.
- Jacqmin P, Snoeck E, Van Schaick EA, et al. Modelling response time profiles in the absence of drug concentrations: Definition and performance evaluation of the K-PD model. J Pharmacokinet Pharmacodyn. 2007;34(1):57-85. doi:10.1007/s10928-006-9035-z