ServicesVirtual Tumour Preclinical

Physiomics Virtual Tumour technology

The Virtual Tumour (figure 1) is a sophisticated computer model that simulates tumour cell division and the effect of antineoplastic drugs, taking into consideration the differences between proliferative cells and those that are part of the necrotic core. The complexity of the model is deliberately constrained so that it can be parameterised with data that is usually produced during drug development:

  • drug concentration profile
  • biomarkers measuring the cell population response to the drug effect
  • xenograft growth measurements showing how tumour growth is affected

This technology provides a rationale for designing an appropriate schedule, and allows our partners to prioritise the most effective drug combinations.

Physiomics Virtual Tumour
Fig. 1 Physiomics Virtual Tumour simulation platform. The Virtual Tumour is a computer simulation of a growing tumour, which integrates the cell division dynamic with the effect of antineoplastic agents. The platform is composed of PK models of the drugs of interest, as well as a pharmacodynamic model of cell cycle progression. Drug effect can be calibrated by using various data sources: in vivo target inhibition (IVTI), xenograft growth timecourses, flow cytometry and public literature data.

The Physiomics Virtual Tumour service

With the Virtual Tumor, simulations of different drug administrations can be quickly made and the best schedule regimen chosen for in-vivo verification. This technology can be used to:

  • design new regimens with proprietary compounds as well as standard of care, small molecules or biotherapeutical agents;
  • help test possible schedules for combinations of different drugs that would be effectively impossible to investigate experimentally;
  • allow prioritisation of the most effective drug combinations;
  • reduce the number of animals used in studies.

Using data that can be gathered during the development of a drug, this could lead to dramatic improvement in its usage, efficacy and prevent loss of synergy due to timing issues.

Case study 1: Optimising the Gemcitabine-Docetaxel combination

The combination of fixed-dose-rate gemcitabine and docetaxel has become an established first- or second-line treatment option for many types of cancer. Several clinical studies have attempted to improve efficacy of these two drugs by designing innovative gemcitabine-docetaxel sequences and schedules in various cancer types. By producing xenograft and biomarker data of each drug in isolation, we have built a Virtual Tumour capable of predicting the outcome of various regimens using this combination, and demonstrated how optimal administration schemas can be determined in silico.

We identified 3 schedules that were of particular interest:

  • one schedule expected to result in the worst synergy: Gemcitabine then Docetaxel 4 hours later;
  • two schedules expected to result in better synergy: Gemcitabine then Docetaxel 12 hours apart and Docetaxel then Gemcitabine 10 hours apart.

Three cycles of drug administration were made for each regimen. Total drug dose was constant in all regimens (each dose: Gemcitabine 60 mg/kg, Docetaxel 7.5 mg/kg). These three proposed schedules were experimentally tested in xenograft models, along with no-drug or single-drug treatment controls (Figure 2).

Gem-Dtx xenograft growth time courses
Fig. 2 Efficacy study of predicted combination regimens. Vehicle (dark blue), Gemcitabine alone (red),
Docetaxel alone (green), Gem->Dtx 4h (brown), Dtx->Gem 10h (yellow), Gem->Dtx 12h (light blue)

The Virtual Tumor allowed us to predict a combination regimen ~50% more efficient than another predicted regimen that was lacking synergy. Furthermore, toxicity was not higher in the more efficient combinations, as shown in mean body weight plots. This shows how critical timing can be when administrating drugs having different mechanisms of action, and how predictive models could be used for optimization.

Arrow More information can be found about this case study in our poster: "Modeling the sequence-sensitive gemcitabine-docetaxel combination using the Virtual Tumor."

Case study 2: Single-blind validation study with Lilly

As another example, the major pharmaceutical company provided us with data for two drugs (compounds are undisclosed for confidentiality reasons). The data consisted of xenograft growth and biomarker information for the drugs taken individually. We were asked to predict xenograft growth when the two separate drugs were used in two different combinations. Our predictions were then compared against experimental data in a single-blind test.

Virtual Tumour predicitions
Fig. 3 Tumour growth predictions for simultaneous (top panel) and sequential (bottom panel) schedules.
The green lines show our predictions, along with estimated upper and lower bounds. The black lines show
the actual average xenograft growth, along with 5 and 95 percentile error bounds. Schedules for the two
drugs are indicated in red and blue on the bottom axis.

Simulations such as those shown here allow our partners to avoid costly trial-and-error approaches to determining the best administration schedules. Furthermore, it can be used to simulate thousands of possible schedules for combinations of different drugs that would be effectively impossible to investigate experimentally, and allow prioritization of the most effective drug combinations.

Arrow More information can be found about this case study in our poster: "Predicting the effect of combination schedules on xenograft tumor using the Virtual Tumor" and our article: "Using Predictive Mathematical Models to Optimise the Scheduling of Anti-Cancer Drugs".