Mohammed AA Saleh (1), Katja Fiedler (2), Andreas Ackermann (2), Richard Dambra (3, 4), Joseph Ashour (3), Philippe Slos (2), Eva Germovsek (1)
(1) Translational Medicine and Clinical Pharmacology, Boehringer Ingelheim Pharma GmbH Co. KG., Germany, (2) Cancer Immunology and Immune Modulation, Boehringer Ingelheim Pharma GmbH Co. KG., Germany, (3) Drug Metabolism and Pharmacokinetics, Boehringer Ingelheim Pharmaceuticals, Inc., USA, (4) Drexel University School of Biomedical Engineering, USA
Introduction/Objectives:
Despite recent advancements in cancer treatment, cancer remains a leading cause of death globally [1,2], highlighting the need for new treatment options. Oncolytic viruses, an emerging class of cancer therapeutics, offer a promising alternative with their unique mode of action. Namely, they primarily replicate in tumour cells and induce antitumour immune responses, causing tumour cell death [3]. Although interest in oncolytic viruses as cancer therapies is growing, a better understanding of their underlying pharmacology is warranted to facilitate drug development. One such oncolytic virus is VSV-GP, an RNA virus, based on the vesicular stomatitis virus. To characterise longitudinal tumour viral dynamics, VSV-GP was encoded with the gene for luciferase (VSV-GP-Luc), which produces replication-driven bioluminescence [4]. We aimed to describe VSV-GP(-Luc) preclinical viral dynamics in blood and tumour (following different administration routes), and its effects on tumour growth.
Methods:
We built on our previous work and extended the viral dynamics model, which described the VSV-GP-Luc dynamics in the tumour [5], with more data. We integrated several in vivo preclinical studies in immunocompetent BALB/c mice, healthy or implanted with the CT26.CL25-IFNAR1-KO tumour cell line. The mice received VSV-GP intravenously (IV), intratumourally (IT), or a combination of both administration routes (IVIT), in doses ranging from 105 to 109 median tissue culture infectious dose (TCID50). Vehicle and VSV-GP-Luc were administered IV or IVIT, of which VSV-GP-Luc in doses mentioned above. In addition to tumour volume and bioluminescence observations used previously [5], we included viral load measurements in blood and tumour as TCID50 and genome copies concentrations, representing active and both active and inactive virus, respectively. Nonlinear mixed effects modelling (NONMEM 7.5.1 [6]) was used to simultaneously analyse all data. The previous viral dynamics model [5] was extended to include blood (central and peripheral) compartments and additional compartments for the inactive virus. Different parameterisations [7,8,9] of the model were (re)tested. The models were compared and evaluated by e.g., using various visual diagnostics and checking the precision of model parameter estimates. R 4.3.1 was used for data preparation, visual exploration and reviewing of NONMEM outputs [10]; PsN 5.3.1 was used to execute NONMEM runs and produce the VPCs [11].
Results:
Data from 1060 mice (out of which 256 received vehicle / had no administration) were used in the analysis, with 660 mice receiving IV, 365 mice IVIT and n=35 IT injected drug/vehicle. We excluded observations below the limit of quantification (6% of the post-administration samples). Across all observation types we analysed 3532 observations from mice that received VSV-GP, n=3372 followed VSV-GP-Luc administration, and n=1671 were from the vehicle / no administration group. The viral dynamic model structure that provided the best fit was a combination of models [7,8], with the addition of one compartment for the dying infected tumour cells. Blood distribution was best described with one peripheral compartment. Viral clearance proved different in the tumour and blood, with the clearance being higher in the blood (mean (RSE%), blood: 9.4 mL/day (12.1%), tumour: 2.5 nL/day (5.7%)).
Conclusions:
A viral dynamics model was developed that allowed for a simultaneous analysis of the relationship between VSV-GP(-Luc) dose, viral replication / distribution dynamics and its effects on the tumour growth. Additionally, different markers of viral dynamics (such as, bioluminescence, TCID50 and genome copies concentrations) were linked together in the model. However, although the model described the data adequately, some misspecifications remained. In the future, the model could be further improved by, for example, including the role of the immune system and incorporating more IT data. Still, the developed model is a step towards quantitative description of the complex behaviour of oncolytic viruses and may (once the translation from preclinical species to human has been accounted for) increase efficiency in future trial designs to maximise the benefit for cancer patients.
References:
[1] World Health Organisation, https://www.who.int/news-room/fact-sheets/detail/cancer, last accessed 21 Feb 2024
[2] International Agency for Research on Cancer, World Health Organisation, https://gco.iarc.fr/today/home, last accessed 21 Feb 2024
[3] Shalhout et al, Nat Rev Clin Oncol. 2023 Mar;20(3):160-177
[4] Schreiber et al, Br J Cancer. 2019; 121(8):647-658
[5] Mueller et al, PAGE 31 (2023) Abstr 10539 [www.page-meeting.org/?abstract=10539]
[6] Bauer, CPT Pharmacometrics Syst Pharmacol. 2019 Aug;8(8):525-537
[7] Titze et al, Eur J Pharm Sci. 2017; 97:38-46
[8] Phan and Tian, Comput Math Methods Med. 2017; 2017:6587258
[9] Smith and Perelson, Wiley Interdiscip Rev Syst Biol Med. 2011 Jul-Aug;3(4):429-45
[10] R Core Team, 2023, https://www.R-project.org/
[11] PsN, 2023, https://uupharmacometrics.github.io/PsN/
Funding: The preclinical studies were funded by Boehringer Ingelheim.
Conflict of interest: The authors are employees of Boehringer Ingelheim.
Acknowledgements: The authors want to thank everyone else who was involved and contributed to the preclinical studies used in this abstract.
Reference: PAGE 32 (2024) Abstr 11179 [www.page-meeting.org/?abstract=11179]
Poster: Drug/Disease Modelling - Oncology