The FDA has just released draft guidance on using Quantitative Systems Pharmacology (QSP)-based dose selection for the minimum anticipated biological effect level (MABEL) in first-in-human trials.
This is an interesting development. Historically, starting doses in first-in-human trials have relied heavily on animal toxicology (NOAEL, STD10, HNSTD). This approach has limitations for certain high-risk products, particularly immunostimulatory biologics such as monoclonal antibodies, where the target may be human-specific and animal data are not easily applicable (consider T-cell activation and cytokine release, for example).
QSP offers a mechanistic alternative, integrating in vitro, in vivo, ex vivo, and in silico data into a single model of drug–target–pathway interactions to estimate the dose at the onset of biological activity.
The draft guidance sets out fairly detailed expectations for sponsors, including model risk assessment, rigorous parameterisation (ideally using human-derived data), verification and validation, and sensitivity/uncertainty analyses. There is also a clear bias towards conservative dosing whenever QSP and traditional methods disagree.
This is a positive indication of the direction in which model-informed drug development (MIDD) is heading, and serves as a reminder that computational pharmacology is becoming an increasingly important part of the regulatory conversation.
