Progression Free Survival In Humans Cancer Therapy: Can It Be PredictedFrom Mouse Models?
The primary clinical endpoint used for measuring the efficacy of an anti-cancer therapy is Progression Free Survival (PFS) or Overall Survival (OS).
In preclinical (Nonclinical) pharmacology studies the mouse models of cancer form the basis for evaluating efficacy of anticancer therapies. The models include, with increasing degree of complexity, the Xenograft, Orthotopic, Syngeneic, Genetically Engineered Models, and the Patient Derived Xenograft (PDX). The primary
measures of efficacy in these models are tumor growth inhibition, tumor regression and survival is a secondary measure in some cases.
Since PFS is the clinically relevant endpoint, can it also be used as a primary parameter in evaluating clinical drug candidates in preclinical models? Can PFS be predicted in humans using data from mice?
Allometry is the study of relationship between physiological parameters (Lifespan, Heart beats, Metabolic rates, Volumes, Blood flows, Oxygen consumption) and body mass(1). Allometry is used to predict Pharmacokinetic parameters (Clearance and Volume of Distribution) in humans based on data from mice, rats, dogs, and non-human primates (2).
Based on Allometric principles Physiological time is defined as a species dependent unit of chronological time required to complete a species independent physiological event.
For example, the average lifespan of a mouse is two years and the average lifespan of a human is 75 years. It takes one year for the mouse to complete 50% of its life and the 37.5 years for a human to complete 50% of life. In other words 1 year of mouse is equivalent to 37.5 years in humans. The time taken to complete 50% life is the species dependent physiological event and the lifespan is the species independent physiological event.
Applying this concept in the mouse models of cancer, a PFS of 1 month is equivalent to a PFS of 37.5 months in humans.
The challenges in applying this concept in mouse models are a) differences in PK between mice and humans; b) the relevance of the model (orthotopic models are physiologically similar to humans for metastasis than xenograft); c) the genomic heterogeneity in tumors (cell line derived versus patient derived);d) the stage of
tumors, and e) the immune system differences (Immunocompromised versus immunocompetent).
For a reliable prediction of PFS in humans, models that mimic the development of cancers in humans in terms of tumor stage, tumor heterogeneity, metastasis, and immune status are required. Since most cancer deaths occur by metastasis of stage 3 and 4 tumours, the orthotopic models in humanized mice using patient derived xenografts could be the appropriate model to evaluate PFS in mice.
Cancer is a highly heterogenous disease. For a therapy to be truly meaningful it has to either cure or extend the lives of patients by years than a few months. Many approved therapies at best extend survival be a few months to years.
Data for drugs where PFS and OS were measured in mice and humans can be used to evaluate the relationships. Such relationships can be invaluable to assess the potential of a clinical drug candidate to bring significant clinical benefit in humans- which is longer PFS and OS than the standard of care.
References
Boxenbaum H. 1984. Interspecies pharmacokinetic scaling and the evolutionary-comparative paradigm. Drug Metab Rev.;15(5-6):1071-121
Boxenbaum. H. 1982. Interspecies Scaling, Allometry, Physiological Time, and the Ground Plan of Pharmacokinetics. Journal of Pharmacokinetics and Biopharmaceutics, Vol. 10, No. 2.