Personalized medicine is more than matching a drug to an individual. It also means getting the dose right. Yet dose optimization, as a strategy to improve patient care, is under-used.  Testing dose-effect relationships among different patient groups is expensive and takes time, which not all patients have. Drugs for serious diseases are often approved if they show some efficacy, with acceptable toxicity – even if the dose doesn’t hit the optimal effect/side-effect balance. Patients with few alternatives may be reluctant to test a lower dose, if it is perceived to be less potent.

In silico-based approaches – under the umbrella of ‘model-informed drug development’ (MIDD) – help skirt some of these hurdles. Pharmacokinetic analyses, dose-exposure response models and simulations are already used in dose analysis and selection. Integrated platforms such as nova’s jinkō go even further. jinkō uniquely assembles disease models, treatment models and a virtual population, allowing drug developers to test multiple doses of a given drug without risk to patients. Using the jinkō platform, developers can precisely simulate a therapy’s effects – both wanted and unwanted – at a range of doses, or along a (feasible) dosing continuum.  The data generated from these simulations helps identify the best and safest dose range, before or alongside dose-finding clinical trials. 

This solution comes as regulators are pushing for change. FDA’s Project Optimus, for instance, encourages developers to shift away from the standard “maximum tolerated dose” (MTD) approach used in oncology- whose roots lie in decades-old systemic chemotherapy – to one that seeks the best dose to maximize both efficacy and tolerability. For instance, patients may be able to remain on a lower dose for longer, achieving the same (or better) outcomes than those on a high-dose regimen with greater side-effects. Lower doses are also less likely to generate treatment resistance. The most effective dose may not always be the highest dose. 

In-person trials testing a wider range of doses are still necessary. For example, a 2021 white paper from Friends of Cancer, written in collaboration with FDA, industry and patient advocates, calls for randomized studies that evaluate at least two doses in oncology, rather than the default MTD approach. 

Jinkō gives sponsors a head start in identifying those most promising doses for confirmatory trials. (Testing two doses is better than one, but even this welcome advance falls short of a complete dose-response exploration.) The platform extrapolates existing pharmacokinetic, pharmacodynamic and dose-exposure response data and integrates this with patient-level clinical outcome predictions. The result is a robust, data-backed evidence package. 

The jinkō platform can also match dosing regimens to patient subgroups. Virtual patients or patient cohorts can be designed with specific characteristics (co-morbidities, body composition, genetic profiles) that, alongside drug attributes and disease severity, feed into optimized dosing. Regulators are also paying more attention to individuals with obesity, for instance, who may require adjusted dosing due to different drug distribution and metabolism. Oral contraceptives are shown to be less effective in these patients. 

With few limits on the number of simulations, doses or patient groups, in silico modeling with platforms like jinkō provide a powerful tool to help industry and regulators shift to a new drug dosing paradigm.  Exploring a wider range of dosing regimens early on helps identify not just one optimal dose, but potentially several, suited to different populations. This maximizes outcomes, product potential and R&D efficiency.