Predicting Patient Outcomes With Mathematical Modeling in Virtual Clinical Trials

Preclinical studies are limited by time constraints and cell model systems that do not portray all of the genetic variability observed in different tumors and individual patients. Now, researchers from Moffitt Cancer Center’s Integrated Mathematical Oncology (IMO) Department are overcoming some of the limitations of common preclinical experiments and clinical trials by studying cancer through mathematical modeling. 

Writing in the November issue of the European Journal of Cancer, the Moffitt researchers report on the potential power of mathematical modeling to predict the responses of melanoma to four different common approaches (no treatment, chemotherapy alone, AKT inhibitors, and AKT inhibitors plus chemotherapy in sequence and in combination). The investigators report that preclinical studies with tumor cell models cannot accurately measure changes and adaptations in genetic mutations in a context that accurately reflects what occurs in individual patients, but this approach may overcome those hurdles.

“Purely experimental approaches are unpractical given the complexity of interactions and timescales involved in cancer. Mathematical modeling can capture the fine mechanistic details of a process and integrate these components to extract fundamental behaviors of cells and between cells and their environment,” said lead study author led by Alexander Anderson, PhD, chair of IMO at the Moffitt Cancer Center in Tampa, Fla.

The researchers showed in a phase i trial (i for in silico) or virtual clinical trial that it is possible to reproduce patient responses similar to those observed in an actual published clinical trial. With this method, they were able to stratify patient responses and predict a better treatment schedule for AKT inhibitors in melanoma patients in a way that could help improve patient outcomes and reduce toxicities.

In addition, Anderson said this multimodel, multiscale approach may lead to the creation of many novel approaches for the treatment and understanding of not just melanoma, but many different tumor types. Anderson and his team note that one of the biggest issues in oncology clinical trials is the high failure rate. Part of the problem is the intrinsic homogeneity of preclinical model systems compared to the heterogeneity of actual patient responses.

With a virtual clinical trial, the investigators use mathematics for an assessment of stratification factors and treatment optimization. In this current analysis, they were able to predict melanoma treatment response and resistance to monotherapies and combination therapies. The data were validated with in vitro experiments. The validated model and a genetic algorithm were used to generate virtual patients whose tumor volume responses to the combination therapy matched statistically the actual heterogeneous patient responses in the clinical trial.

The researcher’s analyses on simulated cohorts identified key model parameters. These included the tumor volume doubling rate and a therapy induced phenotypic switch rate. The investigators report that these parameters may have clinical correlates. Of clinical significance, the investigators report that this approach may be able to help predict the optimal AKT inhibitor scheduling, thus leading to less toxic treatment strategies and improved outcomes.