Anyue Yin (1,2), Dirk Jan A.R. Moes (1,2), Johan G.C. van Hasselt (3), Jesse J. Swen (1,2), Henk-Jan Guchelaar (1,2)*, Lena Friberg (4)* *these authors share senior authorship
(1) Department of Clinical Pharmacy and Toxicology, Leiden University Medical Center, Leiden, the Netherlands, (2) Leiden Network for Personalized Therapeutics, Leiden University Medical Center, Leiden, the Netherlands, (3) Division of Systems Biomedicine and Pharmacology, Leiden Academic Centre for Drug Research (LACDR), Leiden University, Leiden, The Netherlands, (4) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden
Introduction: Evolutionary mechanisms driven by intra-tumor heterogeneity and molecular adaptations in tumor cells to anti-cancer agents are increasingly acknowledged as a key driver for the development of resistance in anti-cancer treatment[1-3].
Real-time monitoring of evolving heterogeneous tumors can thus provide important insights into the development of resistance. Circulating tumor DNA (ctDNA) is increasingly considered as a clinical biomarker to quantify molecular adaptations to treatments and is highly correlated with tumor burden and therapeutic response in many cancer types[4-8].
Quantitative characterization of evolutionary dynamics of tumor under anti-cancer treatment could support identification of novel strategies to optimize treatment schedules. For example, an adaptive intermittent treatment, driven by competition between different clonal sub-populations, has been demonstrated to outperform the continuous treatment in metastatic prostate cancer patients[9]. Another commonly suggested adaptive strategy is to switch to a treatment that specifically targets newly emerged driver mutations that cause resistance[4, 5].
Objectives:1) Build a model structure to characterize treatment resistance evolution based on dynamics of tumor burden and mutant KRAS levels in ctDNA in patients with metastatic colorectal cancer (mCRC) as a proof-of-concept; 2) Evaluate novel anti-cancer treatment schedules using simulation to suggest regimens to improve clinical outcome.
Methods: A dataset including tumor burden and mutant KRAS levels was collected from a published study in mCRC patients treated with anti-EGFR antibody[8]. Data from patients (N=9) with wild type KRAS and developed KRAS mutation during therapy were included.
A mathematical model was developed with ordinary differential equations (ODEs). Tumor tissue was assumed to consist of three clonal populations: TS which is sensitive to treatment 1 (D1), TR1 which is derived from TS during D1 and is resistant to D1 but sensitive to treatment 2 (D2), and TR2 which is derived from TR1 during D2 and is resistant to both D1 and D2. The growth rate of drug resistance tissue was assumed to be limited by the size of drug sensitive tissue. Mutant genes were assumed to be released from TR1 and TR2 and can be detected from ctDNA. Releasing rates of ctDNA were assumed to depend on the size of TR1 and TR2. The delay in the emergence of detectable mutations representing resistance, as is shown in the data[8], was characterized with Hills equation. Tumor growth rate was set to previously reported doubling time[10] while other parameters were set by visually matching the available data. The treatment effect was described with a log-kill pattern and drug exposure was not considered[11].
Regimens with continuous D1, intermittent D1, and switching D1 and D2 adaptively according to mutant gene level in ctDNA (adaptive regimen) were evaluated by simulation. The disease progression time (tDP), as defined in RECIST version 1.1[12], and the time when at least 10 unit mutant gene was detected were evaluated.
Simulations were performed with package RxODE (version 0.6-1) implemented in R (version 3.4.2). Inter-individual variability was not considered.
Results: The developed model structure well captured the reported time curves of tumor burden and mutant KRAS level in ctDNA obtained from mCRC patients.
Simulation results showed that the intermittent regimen with D1 on for 4 weeks and off for 4 weeks prolonged tDP from 29 weeks to 40 weeks compared with continuous regimen. The mutant gene reflecting the emergence of TR1 was detected 12 weeks ahead of the disease progression. Increasing the on or off period of D1 shortened the tDP.
The adaptive regimen showed to further prolong tDP to 60 weeks under a 12-week monitoring frequency, although the mutant gene that represents the emergence of TR2 or regrowth of TR1 was not detectable before disease progression. Adjusting the threshold of mutant gene level where treatment switched between D1 and D2 did not affect the predicted tDP.
Conclusion: The developed model structure well captured the reported data. The alternative anti-cancer treatment schedules, i.e. intermittent and adaptive regimen, inhibited the evolving resistance of cancer and suggested a potential improvement of clinical outcome. An external dataset is warranted to validate the results and ultimately a prospective clinical study is required to prove the added value of these alternative schedules.
References:
[1] Kim, C., et al., Chemoresistance Evolution in Triple-Negative Breast Cancer Delineated by Single-Cell Sequencing. Cell, 2018. 173(4): p. 879-893 e13.
[2] Beksac, A.T., et al., Heterogeneity in renal cell carcinoma. Urol Oncol, 2017. 35(8): p. 507-515.
[3] Gerlinger, M. and C. Swanton, How Darwinian models inform therapeutic failure initiated by clonal heterogeneity in cancer medicine. Br J Cancer, 2010. 103(8): p. 1139-43.
[4] Wan, J.C.M., et al., Liquid biopsies come of age: towards implementation of circulating tumour DNA. Nat Rev Cancer, 2017. 17(4): p. 223-238.
[5] Nangalia, J. and P.J. Campbell, Genome Sequencing during a Patient’s Journey through Cancer. N Engl J Med, 2019. 381(22): p. 2145-2156.
[6] Xiong, L., et al., Dynamics of EGFR mutations in plasma recapitulates the clinical response to EGFR-TKIs in NSCLC patients. Oncotarget, 2017. 8(38): p. 63846-63856.
[7] Dawson, S.J., et al., Analysis of circulating tumor DNA to monitor metastatic breast cancer. N Engl J Med, 2013. 368(13): p. 1199-209.
[8] Diaz, L.A., Jr., et al., The molecular evolution of acquired resistance to targeted EGFR blockade in colorectal cancers. Nature, 2012. 486(7404): p. 537-40.
[9] Zhang, J., et al., Integrating evolutionary dynamics into treatment of metastatic castrate-resistant prostate cancer. Nature Communications, 2017. 8(1).
[10] Blagoev, K.B., et al., Neutral evolution of drug resistant colorectal cancer cell populations is independent of their KRAS status. PLoS One, 2017. 12(10): p. e0175484.
[11] Yin, A., et al., A Review of Mathematical Models for Tumor Dynamics and Treatment Resistance Evolution of Solid Tumors. CPT Pharmacometrics Syst Pharmacol, 2019. 8(10): p. 720-737.
[12] Eisenhauer, E.A., et al., New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer, 2009. 45(2): p. 228-47.
Reference: PAGE () Abstr 9548 [www.page-meeting.org/?abstract=9548]
Poster: Drug/Disease Modelling - Oncology