QSP: Review, Use Case for Hyperkalemia Risk Assessment

Quantitative systems pharmacology (QSP) is widely used to address questions in drug discovery and development, including identification of the mechanism of action of a therapeutic agent, patient stratification, and the mechanistic understanding of the progression of disease.

Ryuta Saito and Tomohisa Nakada assessed the current landscape of QSP modeling by reviewing QSP publications over 10 years (2013-2022). They also developed a use case for QSP in the assessment of hyperkalemia risk in patients with diabetic nephropathy (DN) treated with mineralocorticoid receptor antagonists (MRAs). Their report was published in Drug Metabolism and Pharmacokinetics.

QSP is the core technology of model-informed drug discovery and development (MID3), a quantitative framework for prediction and extrapolation. Because results of nonclinical studies are often insufficient to quantitatively predict clinical outcomes and the related signatures in terms of drug efficacy and adverse effects, MID3 is often used to boost quality and productivity of research and development. QSP is the integrated mathematical model of MID3 and its analysis.

The use of QSP in drug development has steadily increased since 2013. Prior to 2018, QSP models were mostly applied to certain central nervous system diseases, such as schizophrenia and Alzheimer disease, and cardiotoxicities, such as cardiac arrhythmias. However, QSP modeling has spread to an array of therapeutic areas since 2019. Breakthroughs in artificial intelligence (AI) and machine learning (ML) are likely to further accelerate the application of QSP modeling to drug discovery and development.

Questions that QSP modeling has been used to address include identification or verification of drug target molecules, prediction of clinical efficacy and safety from nonclinical data, decision support on clinical trials using biomarkers, differentiation from competitive drugs, patient stratification for clinical trial design, and pathophysiological interpretation of clinical data.

In a recent survey by the IQ Consortium, QSP models were found to be used mainly to support internal decision-making in preclinical and early phases of clinical development. However, the review also found examples of QSP models being applied retrospectively, for example to clarify mechanisms of action of anticoagulants and endocrine disruptors in nonclinical studies, predict drug-induced liver injury, and model and simulate sodium-glucose transport protein 2 inhibitors.

With this in mind, the researchers also presented their use case for the risk assessment of hyperkalemia (increase in serum potassium concentrations beyond 5.5 mEq/L) in patients with DN treated with MRAs. A QSP model to generate virtual patients with DN was used to quantitatively determine that the nonsteroidal MRAs finerenone and apararenone are associated with a lower risk of hyperkalemia than the steroidal MRA eplerenone. Such simulation studies using a QSP model can help prioritize pharmaceutical candidates in clinical development and validate pharmacological concepts related to the risk-benefit ratio before conducting large-scale clinical trials and incurring large costs.

Looking forward, the authors wrote, “Recent breakthroughs of AI/ML technologies are expected to accelerate the application of QSP models to drug discovery and development, with the hybrid models of QSP and AI/ML-encoded clinical and experimental data and mechanistic knowledge having good predictive power. We would like to improve the low productivity and success rates in drug discovery and development as much as possible by the applications of QSP models to prospective prediction, such as the use case presented in this review article.”

Source: Drug Metabolism and Pharmacokinetics