Predicting Human Infection Risk: Do Rodent Host Resistance Models Add Value?

Area of Science:

• Immunology
• Pharmacology
• Translational Medicine

Background:

• Genetically engineered rodents are used to evaluate safety concerns of pathway inhibition, particularly immunomodulatory risks.
• Infection models in these animals are challenged with pathogens to predict human risks, but model applicability is rarely assessed against human data.

Purpose of the Study:

• To compare infectious pathogen challenge outcomes in mice with genetic deficiencies in TNF-α, IL17, IL23, or JAK pathways against human outcomes from pathway inhibitors.
• To evaluate the predictive value of mouse infection models for identifying human safety risks.

Main Methods:

• Review and comparison of published data on infectious pathogen challenge outcomes in genetically modified mice (TNF-α, IL17, IL23, JAK deficient).
• Comparison of mouse model outcomes with retrospective human data on infections associated with inhibitors of these pathways.

Main Results:

• Mouse infection models demonstrated modest utility for hazard identification, predicting only overall trends in infection risk.
• These models showed limited value in predicting risks for specific prevalent (e.g., rhinoviruses) or concerning (e.g., herpes zoster) pathogens.
• Mouse models tended to overestimate the severity of infection risk compared to human patients.

Conclusions:

• Genetically engineered mouse models have limited predictive value for human infection risks associated with pathway inhibition.
• There is an unmet need for more human-relevant models with improved predictive capabilities for drug safety assessment.
• Large meta-analyses and post-marketing evaluations remain the gold standard for characterizing true patient infection risk.