Alicja Puszkiel (1), Benoit You (1,2), Léa Payen (3), Jonathan Lopez (4), Jérôme Guitton (5), Pascal Rousset (6), Juliette Fontaine (7), Julien Péron (2), Denis Maillet (2), Sophie Tartas (2), Nathalie Bonnin (2), Veronique Trillet-Lenoir (2), Olivier Colomban (1), Diane Augu-Denechere (2), Gilles Freyer (1,2), Michel Tod (1,8)
(1) Université Claude Bernard Lyon 1, Faculté de Médecine Lyon Sud, EMR UCBL/HCL 3738 « Ciblage thérapeutique en Oncologie », Lyon, France (2) Medical Oncology, Institut de Cancérologie des Hospices Civils de Lyon (IC-HCL), CITOHL, Centre Hospitalier Lyon Sud, Lyon, France (3) Laboratoire d'Oncologie Moléculaire et Transfert, Centre de Biologie Lyon Sud, Hospices Civils de Lyon, Lyon, France (4) Centre de Recherche en Cancérologie de Lyon - INSERM U1052 CNRS U5286, Service de Biochimie et Biologie moléculaire, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Lyon, France (5) Laboratoire de Biochimie-Toxicologie, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Lyon, France (6) Department of Radiology Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Lyon, France (7) Department of Pathology, Centre Hospitalier Lyon Sud, Hospices Civils de Lyon, Lyon, France (8) Pharmacie de l’Hôpital de la Croix Rousse, Hospices Civils de Lyon, Lyon, France
Objectives: Everolimus (EVE) and sorafenib (SOR) are oral targeted anti-cancer therapies blocking PI3K-AKT-mTOR and RAS-RAF-ERK (MAPK) pathways, respectively, and are currently approved as single agents in several malignancies. Since most cancers will eventually develop resistance to these treatments, a combination therapy of SOR and EVE providing a dual inhibition of MAPK and PI3K-AKT-mTOR pathways may be more effective. However, defining the optimal doses and schedules of combined targeted drugs is challenging. The objective of this study was to develop a pharmacokinetic-pharmacodynamic (PKPD) model for EVE, SOR and biomarkers linked with a time-to-event (TTE) analysis based on data from EVESOR phase I trial (NCT01932177) and propose an optimised dosing schedule of EVE and SOR combination therapy in solid tumor patients.
Methods: Patients were assigned into four treatment arms: continuous SOR 200 mg BID and EVE 5 mg QD (arms A and B), SOR 200-400 mg BID alternating every other week with EVE 5-10 mg QD (arm C), SOR 200-400 mg BID 3 days-on, 4 days-off and EVE 5-10 mg QD continuously (arm D). Each treatment cycle lasted for 28 days. Rich PK sampling was performed on day 1 and 15 (arms A, B and C) and on day 15 and 22 (arm D) of cycle 1 and pre-dose concentrations were measured on day 1 of cycles 2 to 4. The PD biomarkers included expression levels of total AKT, pAKT, pS6K1, total ERK and pERK in peripheral blood mononuclear cells and sVEGF, sVEGFR1 and sVEGFR2 concentrations in serum. Biomarker data was available at day 0, 3 (arm C and D), 8, 15, 22 and 28 of cycle 1. Baseline activation of MAPK pathway was assessed by quantification of mRNA of MAPK specific gene panel in tumor biopsies (Nanostring®) [1]. Progression-free survival (PFS) was defined as time since treatment start to progression or death from any cause.
SOR and EVE PK data were described by a one- and two-compartment models, respectively, with linear absorption and elimination, as described previously [2]. sVEGFR2 kinetics was modeled according to an indirect model. Different parametric TTE models were tested in NONMEM 7.5 using Laplacian estimation method to describe the baseline hazard of disease progression. Treatment arm, age, sex, baseline MAPK activation, individual SOR and EVE PK parameters (AUC) and individual predicted change in sVEGFR2 levels obtained from the PKPD model were evaluated as predictors of PFS. Finally, simulations of sVEGFR2 kinetics under different SOR doses and schedules were performed (n=1000 PD profiles for each) and median PFS (CI95%) were calculated using TTE model to select dose and schedule associated with improved PFS.
Results: Fourty-three patients were included in the study. SOR acted by decreasing sVEGFR2 levels in an Imax manner with IC50 of 1060 ng/mL (RSE=33%) and Imax fixed to 1. EVE PK was not associated with biomarker kinetics. The baseline hazard of disease progression was best described by a Weibull distribution with λ=0.0035 day-1 (RSE=13%) and α=1.54 (RSE=12%). A second Weibull distribution was used to describe the baseline hazard of dropout with λ=0.0126 day-1 (RSE=13%) and α=0.691 (RSE=16%). In the final model, individual model-predicted change in sVEGFR2 at day 21 from day 0 (βsVEGFR2=4.58, RSE=37%) and baseline activation of RAS-RAF-ERK pathway (βMAPK=-0.352, RSE=50%) were significantly associated with baseline hazard of disease progression. Patients exhibiting higher decrease in sVEGFR2 and those with higher baseline activation of MAPK pathway had longer PFS. The final model was validated by a Kaplan Meier VPC (200 simulations) including the dropout model. According to simulations, SOR 200 mg BID 5 days-on 2 days-off (+EVE 5 mg QD continuous) was associated with improved PFS (median=4.1 months, CI95%=2.0-12.3) compared to the median PFS in the EVESOR trial (3.6 months, CI95%=2.7-4.2, n=43).
Conclusions: This study showed that sVEGFR2 kinetics and baseline tumor activation of MAPK pathway were associated with PFS in patients with solid tumors treated with EVE and SOR combination therapy. Since the sVEGFR2 time course was impacted by SOR PK, simulations with the PKPD-TTE model allowed evaluating alternative doses and schedules of SOR to improve PFS in these patients. On that basis, SOR 200 mg BID 5 days-on 2 days-off (+EVE 5 mg QD continuous) was selected for clinical evaluation in the extension of EVESOR trial to confirm its superiority over the previously tested treatment arms.
References:
[1] Wagle et al. Precision Oncology (2018) 2:7
[2] El Madani et al. Future Oncol. (2017) 13(8), 679-693
Reference: PAGE 29 (2021) Abstr 9638 [www.page-meeting.org/?abstract=9638]
Poster: Drug/Disease Modelling - Oncology