Titze M. I. (1), Ehrhardt M. (2), Smola S. (3), Graf N. (2), Lehr T. (1)
(1) Clinical Pharmacy, Saarland University, Germany, (2) University Hospital Homburg, Department for Pediatric Oncology / Hematology, Germany, (3) University Hospital Homburg, Department of Virology, Germany
Objectives: Oncolytic viruses (OV) like reovirus are replication-competent viruses, which are able to specifically infect and replicate in tumor cells and thus lead to rapid cancer cell death. Due to the dependency of viral replication on the number of susceptible tumor cells as well as the dependency of the tumor growth on viral destruction, the relationship between reovirus and glioblastoma cells reflects a highly complex interaction.
The objective of this analysis was to describe this bidirectional interaction between reovirus type 3 dearing and U87-glioblastoma cells. It was aimed to develop a semi-mechanistic mathematical model to describe the dynamics of in vitro tumor cell growth and reovirus replication and to depict the impact of reovirus treatment on tumor growth.
Methods: Relative tumor growth of human U87-glioblastoma cells was measured in vitro via cell viability using the MTT assay [1] for control (200 observations) and treatment group (150 observations) up to 144 hours after seeding. The effect of various reovirus doses (1 to 100 plaque forming units (PFUs)/cell) on tumor cells was investigated and viral titer were quantified up to 120 hours after infection (69 observations) via polymerase chain reaction (PCR) [2]. Model development was done sequentially by first investigating several models (e.g. Gompertz, exponential, logistic) to describe tumor growth. In the next step the model most accurately depicting tumor growth was linked to a viral dynamic model (VDM) adopted from literature [3]. Modeling was performed using non-linear mixed-effects modelling (FOCE-I method) in NONMEM V7.2.0 [4].
Results: Untreated U87-glioblastoma cell growth was best described by an exponential growth model. Viral dynamics were implemented in the VDM. In the VDM two different states of tumor cells were defined: uninfected tumor cells, which are growing exponentially and are infected by free virus particles and infected tumor cells, which are assumed to be no longer replication-competent and die at a first-order-rate constant and release new viruses with a first order rate constant. Goodness of fit plots and visual predictive checks showed an adequate performance of the model.
Conclusion: The developed VDM was capable of accurately describing the bidirectional interaction between tumor cell growth and reovirus dynamics. It can serve as a generic tool to systematically characterize and compare other OV and tumor cell lines.
References:
[1] Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods (1983) 65( 1–2): 55–63.
[2] Bartlett JMS, Stirling D, Carroll P, et al. PCR Protocols, Second Edition. Methods Mol Biol 226.
[3] Ette E , Williams PJ. Pharmacometrics. Wiley (2007).
[4] Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA.
Reference: PAGE 23 () Abstr 3188 [www.page-meeting.org/?abstract=3188]
Poster: Drug/Disease modeling - Oncology