Marc Cerou (1), Sophie Fliscounakis-Huynh (2), Clemence Pouzin (3), Nathalie Fagniez (3), Frano Mihaljevic (4), Emmanuelle Comets (5), Mustapha Chadjaa (6), Hoai-Thu Thai (1)
(1) Translational Disease Modelling Oncology, Sanofi, 91380 Chilly-Mazarin, France. (2) Translational Disease Modelling Oncology, Sanofi, on behalf IT&M Stats, 91380 Chilly-Mazarin, France (3) Pharmacokinetics Dynamics and Metabolism, Sanofi, 91380 Chilly-Mazarin, France (4) Simulations Plus, Lixoft division, Antony, France (5) Université Paris Cité, IAME, Inserm, F-75018, Paris, France ; Univ Rennes, Inserm, EHESP, Irset (Institut de recherche en santé, environnement et travail) - UMR_S 1085, F-35000 Rennes, France (6) Lung Cancer Oncology Dev, Sanofi, 75000 Paris, France
Introduction:
Tusamitamab ravtansine (Tusa Rav) is an antibody-drug conjugate targeted to carcinoembryonic antigen-related cell adhesion molecule 5 (CEACAM5), a cell-surface glycoprotein expressed in tumor cells, and delivering DM4, a cytotoxic agent which can disrupt cell division and lead to cell death. In a first-in-human phase 1 (NCT02187848) study 1 , Tusa Rav monotherapy has shown encouraging antitumor activity and a favorable safety profile in patients with non-squamous (NSq) non-small cell lung cancer (NSCLC). The observed adverse events (AE), in particular corneal events, were manageable by dose modifications (e.g., dose delay/reduction/discontinuation). Evaluation of the benefit-risk balance for dose optimization was needed but challenging considering the impact of AE-related dose modifications.
Objectives:
The aims of this work were to:
- develop a Drug-Disease Model (DDM) characterizing the relationship between exposure, progression-free survival (PFS) and adverse events (AEs) (corneal vs non-corneal AEs) using Tusa rav phase 1 data,
- evaluate the impact of dose modifications and benefit-risk of different dosing regimens of interest by iterative simulations.
Methods:
Data from Tusa rav phase 1 NCT02187848 study [1] was used to build the DDM. 254 patients with different dose levels and tumor types were used to develop a PK/PFS and PK/safety model characterizing the relationship between exposure and dose-altering AEs, taking into account their impact on dose modification. Exposure parameters considered included Cmax, AUC, and Ctrough, which were time-varying due to dose modifications caused by AEs. AEs were categorized as corneal vs non-corneal events. A parametric repeated time-to-event (RTTE) model was used to describe the time to AE, a categorical model with mixed effects described AE grade, with a linear mixed effects model for AE duration.
Covariate were selected for both models using the COSSAC approach [2].
To evaluate the models and simulate clinical trials, an iterative procedure was employed. First, simulated doses were generated using the toxicity models considering the AE-related dose modifications, and these doses were used to estimate time-varying exposure parameters. Next, the simulated doses and/or associated exposure parameters were considered as input into the PFS model.
A simulation study was conducted using the time-varying exposure/safety/PFS models previously developed to compare the effects of different Tusa rav dosing regimens (80, 100 & 120 mg/m²) for patients with Nsq NSCLC and high CEACAM5 expression (≥2+ intensity in ≥50% tumor cells) (N=64).
Model parameters were estimated using the SAEM algorithm implemented in Monolix2023R1, and simulations were performed using SimulX2023R1 and R version 4.2.3.
Results:
The developed DDM model was composed of a PK model [3], a log logistic proportional hazard model for PFS, an exponential model for the risk of both corneal and noncorneal event, a linear mixed effect model without slope for the duration of corneal event, and a linear model without slope for the duration of noncorneal event.
Time-varying Cmax up to the current administered dose was found to be the best predictor of the risk of corneal events, with a beta coefficient of 7.6. Additionally, older age was associated with a lower risk of corneal events, while increased aspartate amino-transferase and BMI were linked to an increased risk of noncorneal events.
In the exposure-response model for PFS, time-varying Ctrough up to the previously administered dose was found to be associated with the risk of disease progression. Other factors that influenced PFS included baseline tumor size and alkaline phosphatase, with larger tumors and higher alkaline phosphatase values corresponding to an increased risk of progression. Furthermore, patients in the Nsq NSCLC cohorts or those who had a higher duration between diagnosis to first dose tended to have a lower risk of progression.
Internal model evaluation showed that this model accurately predicted PFS profile of Tusa rav phase 1 data. All Tusa rav AE endpoints were also well-described by the model.
The simulation study in virtual patients with Nsq NSCLC & with high CEACAM5 expression predicted no significant difference in median PFS between the doses (3.2, 3.3 and 3.6 months for 80, 100 and 120 mg/m² Q2W, respectively). However, the simulation predicted a lower proportion of patients experiencing corneal events inducing dose modification with the 80 mg/m² Q2W dose (4% vs. 12% and 24% for 100 and 120 mg/m² Q2W, respectively).
Conclusions:
This novel drug-disease model accurately predicted both the observed efficacy and safety endpoints of Tusa Rav phase 1. This model considered the AE-related dose modifications and allowed to evaluate the benefit-risk of different dosing regimen of Tusa Rav.
Conflicts of interests:
M.Ce., H-T.T., C.P, N.F. and M.C. are employees of Sanofi and may hold stock/stock options. All other authors declared no competing interests for this work.
References:
[1] Gazzah, A. et al. Safety, pharmacokinetics, and antitumor activity of the anti-CEACAM5-DM4 antibody–drug conjugate tusamitamab ravtansine (SAR408701) in patients with advanced solid tumors: first-in-human dose-escalation study. Ann. Oncol. 33, 416–425 (2022).
[2] Ayral, G., Si Abdallah, J.-F., Magnard, C. & Chauvin, J. A novel method based on unbiased correlations tests for covariate selection in nonlinear mixed effects models: The COSSAC approach. CPT Pharmacomet. Syst. Pharmacol. 10, 318–329 (2021).
[3] Pouzin, C. et al. Integrated multiple analytes and semi-mechanistic population pharmacokinetic model of tusamitamab ravtansine, a DM4 anti-CEACAM5 antibody-drug conjugate. J. Pharmacokinet. Pharmacodyn. 49, 381–394 (2022).
Reference: PAGE 32 (2024) Abstr 11078 [www.page-meeting.org/?abstract=11078]
Poster: Drug/Disease Modelling - Oncology