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, 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 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 RAS-RAF-ERK and PI3K-AKT-mTOR pathways may be more effective. However, defining the optimal doses and dosing schedules of combined targeted drugs is challenging. The multi-parameter EVESOR phase I trial (NCT01932177) was designed to assess alternative doses and schedules of SOR and EVE combination therapy associated with optimal efficacy/toxicity ratio using a pharmacokinetic-pharmacodynamic (PKPD) model linked with a time-to-event (TTE) analysis. The objective of this study was to develop a TTE model and identify predictive factors of progression-free survival in solid tumor patients treated with EVE and SOR combination therapy.
Methods: Patients were assigned into four treatment arms:
- arm A: continuous EVE 5 mg once daily (QD) alone during a 2-week run-in period before adding continuous SOR 200 mg twice daily (BID) on day 15;
- arm B: continuous SOR 400 mg BID alone during a 2-week-run-in period before adding continuous EVE 5 mg QD on day 15;
- arm C: SOR 200-400 mg BID alternating every other week with EVE 5-10 mg QD,
- arm D: SOR 200-400 mg BID for 3 days-on, 4 days-off and EVE 5-10 mg QD on a continuous basis.
In all arms, each treatment cycle lasted for 28 days. Rich pharmacokinetic (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 the expression levels of total AKT, pAKT, pS6K1, total ERK and pERK in the peripheral blood mononuclear cells and VEGF, VEGFR1 and VEGFR2 concentrations in serum. Progression-free survival 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 [1]. Different parametric TTE models (exponential and Weibull) were tested in NONMEM 7.4.2 using Laplacian estimation method to describe the baseline hazard of disease progression. Treatment arm, cancer type (gynecological cancers, cholangiocarcinomas or other types of cancer), age, sex, area under the concentration-time curve of SOR and EVE cumulated over 28 days (AUCSOR,Cum28days and AUCEVE,Cum28days, respectively) obtained from the PK models were evaluated as predictors of progression-free survival.
Results: Fourty-three patients were included in the study (n = 7, n = 7, n = 16 and n = 13 for arm A, B, C and D, respectively) and most of them had gynecological cancers (25.6%) and cholangiocarcinomas (23.2%). The baseline hazard of disease progression was best described by a Weibull distribution with λ = 0.00022 day-1 (RSE = 11.6%) and α = 1.96 (RSE = 10.4%). A second Weibull distribution was used to describe the baseline hazard of drop out with λ = 0.0126 day-1 (RSE = 13.3%) and α = 0.691 (RSE = 16.1%). In the univariate analyses, cancer type (p < 0.00001), treatment arm (p = 0.02) and AUCSOR,Cum28days > 1295 mcg/mL.h (i.e. median AUCSOR,Cum28days, p = 0.02) had significant impact on baseline hazard of disease progression. Cancer type and AUCSOR,Cum28days were retained in the final model. Cholangiocarcinomas and gynecological cancers had a decreased risk of disease progression compared to other types of cancer with estimated effects (β) of -3.31 (RSE = 24.4%) and -1.66 (RSE = 28.5%), respectively. AUCSOR,Cum28days > median (1295 mcg/mL.h) was associated with a decreased risk of disease progression (β = -0.718 [RSE = 49.4%]). The final model was validated by a Kaplan Meier Visual Predictive Check of 100 simulations including the drop out model.
Conclusions: The TTE model shows that the efficacy of EVE and SOR combination therapy might be higher in patients with cholangiocarcinomas and gynecological cancers than in patients with other types of cancer evaluated in EVESOR trial. The current analysis suggests that the doses and dosing schedules of SOR and EVE combination therapy could be optimized on the basis of SOR plasma exposure. Further PKPD analyses including evaluation of biomarkers and adverse events aiming to determine the SOR and EVE dosing regimens associated with the optimal efficacy/toxicity ratio will be presented.
References:
[1] El Madani et al. Future Oncol. (2017) 13(8), 679-693
Reference: PAGE () Abstr 9372 [www.page-meeting.org/?abstract=9372]
Poster: Drug/Disease Modelling - Oncology