Simulation of Disease Progression for Pharmaceutical Forecasting

Simulation of Disease Progression for Pharmaceutical Forecasting

Acute myeloid leukemia (AML) is a type of cancer in which abnormal cells in bone marrow and blood interfere with normal blood cell production. This is a very serious cancer and can progress rapidly, resulting in death within weeks or months if left untreated. It is, therefore, essential to diagnose it early and ensure treatment is given as soon as possible.


Treatments are available but depend on a number of factors such as the subtype, morphology, patients’ preferences, access to care, and numerous others. AML is very complex in how it can be diagnosed, how it progresses, and how it is treated. This complexity provides an opportunity for an epidemiology model based on systems science to be developed.


System science based models can create complete systems of disease and treatment pathways to help decision makers understand how health conditions develop and their consequences. This allows the patient to be treated, understand the progress of the patient, and finally how the patient is interacting with the system.

The benefit of systems science methodologies is that there can be integration of data and evidence from many different sources at many levels of analysis. The model below shows this in the model breadth and at a lower level – the interventions. This creates a deeper understanding of the patient and the entire market overall.

System science based epidemiology model

System science based epidemiology model (click to enlarge)

Astellas used an agent-based model to simulate how patients progressed through screening, diagnosis, disease progression, and treatment. External factors that could impact this model were also considered.

Agent-based model to follow the journey of patients

Agent-based model to follow the journey of patients

The model used publicly available data to be able to ensure accuracy and adjust for the real world. This methodology could help optimize forecasting and product planning by identifying risk.


The model matched the published literature at a very high level by using different data sources. By aggregating the results into one model, the researchers could see how patients would progress from diagnosis to palliative care or a relapse group, or other alternative progressions.

The assumptions were visualized using the private cloud feature of AnyLogic. Internal decision making such as forecasting, sensitivity analysis, and running Monte Carlo simulations could also be done here.

Astellas could then understand how changes in the market could impact patients, and then simulate how to better treat them in the future.

The case study was presented by Alexander Chettiath of Astellas, at the AnyLogic 2021 Conference.

Similar case studies

More case studies

Get a brochure with industry case studies