Artificial Intelligence and Mechanistic Modeling for Clinical Decision Making in Oncology

Abstract

The amount of “big” data generated in clinical oncology, whether from molecular, imaging, pharmacological, or biological origin, brings novel challenges. To mine efficiently this source of information, mathematical models able to produce predictive algorithms and simulations are required, with applications for diagnosis, prognosis, drug development, or prediction of the response to therapy. Such mathematical and computational constructs can be subdivided into two broad classes: biologically agnostic, statistical models using artificial intelligence techniques, and physiologically based, mechanistic models. In this review, recent advances in the applications of such methods in clinical oncology are outlined. These include machine learning applied to big data (omics, imaging, or electronic health records), pharmacometrics and quantitative systems pharmacology, as well as tumor kinetics and metastasis modeling. Focus is set on studies with high potential of clinical translation, and particular attention is given to cancer immunotherapy. Perspectives are given in terms of combinations of the two approaches: “mechanistic learning.”

Metronomic Chemotherapy Modulates Clonal Interactions to Prevent Drug Resistance in Non-Small Cell Lung Cancer

“…With many effective therapies now available, it has become more and more difficult to determine the optimal combinatorial treatments and schedules, resulting in an unprecedented number of failed clinical trials over the past few years [ 38 , 39 ]. Mathematical modeling has rapidly evolved and could provide relevant tools in oncology [ 40 , 41 ]. Indeed, preclinical and clinical trials in which mathematical models have been used to estimate the mechanism(s) of treatment failure and explore alternative strategies with new drugs and schedules showed their potential to improve therapeutic efficacy [ 42 , 43 , 44 , 45 ].…”

Metronomic therapy prevents emergence of drug resistance by maintaining the dynamic of intratumor heterogeneity

“…With many effective therapies now available, it becomes more and more difficult to determine the optimal combinatorial treatments and schedules resulting in an unprecedented number of failed clinical trials over the past few years (28,29). Mathematical modeling has rapidly evolved and could provide relevant tools in oncology (30,31). Indeed, clinical trials in which mathematical models have be used to estimate the mechanism(s) of treatment failure and explore alternative strategies with new drugs and schedules showed the potential of mathematical modeling to improve patient outcomes (32)(33)(34).…”

Application of Machine Learning in Translational Medicine: Current Status and Future Opportunities

“…Furthermore, as has been discussed in the setting of cancer immunotherapy, personalized disease trajectory models have the potential to bridge to therapeutics as they can enable response-adaptive treatment decisions for individual patients through iterative development and validation of patient-specific models using baseline and on-treatment biomarker and response data collected longitudinally during treatment (55). Given the multi-dimensional nature of such biomarker measurements in oncology as well as other diseases, convergence and synergy across methodologies ranging from mechanism-based (e.g., QSP) models to biologically agnostic (e.g., ML) models will be necessary (56).…”