In the complex world of drug development, dose optimization plays a crucial role. Project Optimus, a groundbreaking initiative, focuses on refining this process for the development of oncology drugs.

During our recent webinar, we received dozens of questions about this new paradigm in oncology drug development. This blog post highlights some of those questions and answers from our expert speakers.

How critical is the biomarker-clinical efficacy relationship in choosing the minimum effective dose?

While the biomarker-clinical efficacy relationship is vital, it’s important to acknowledge that most biomarkers do not act as surrogate endpoints, meaning they don’t necessarily predict clinical efficacy. Instead, a more pragmatic approach is to treat biomarkers as indicators of target engagement. By doing this, we can use the biomarker response as support for the clinical response observed in the trial. Ideally, the biomarkers measured in Phase 1 should measure this target engagement as directly as possible.
Additionally, pharmacodynamic (PD) endpoints are very useful (in conjunction with clinical response data) in defining the dose range to be studied in Phase 2. Typically, two doses are chosen — one that maximizes the biomarker response and a higher dose to address any uncertainties about the drug’s mechanism. Although this strategy isn’t flawless, it is currently one of the best approaches available.

What are your recommendations for measuring and analyzing biomarkers to select an appropriate dose? What sampling timings are recommended?

The ideal biomarkers directly measure target engagement and should have some correlation with a positive response in animal models. Choosing appropriate biomarkers is challenging and should involve collaboration among clinical pharmacology, translational medicine, and clinical teams.

Sampling timing is crucial. Many biomarkers lose their utility due to insufficient sampling. For instance, measuring biomarkers only after the first dose can miss long-term changes. Early involvement of clinical pharmacology or pharmacometrics teams can ensure that biomarkers and sampling schemes are optimally selected.

How should investigational therapies with a myriad of biomarkers still being studied in Phase 1a/b be evaluated?

This situation is complex and depends a great deal on the target of the investigational therapy. Ideally, biomarkers should directly measure target engagement and be chosen thoughtfully, involving input from pharmacometrics, translational medicine, and clinical teams. It’s better to focus on one or two well-selected biomarkers rather than several chosen without careful consideration.

Could you explain the concept of “short half-life”?

The term “short half-life” is relative to the dosing interval. For instance, a drug with a 30-day half-life would not be considered short in most cases. However, if the drug is dosed annually, a 30-day half-life would be relatively short, suggesting no accumulation of the drug.

A more concrete definition of “short” could be a half-life that results in no accumulation for a specific dosing regimen. For example, a drug with a 4-hour half-life administered weekly would generally not accumulate, and pre-dose values would be zero. In such cases, collecting pre-dose samples would be redundant and post-dose samples would be more informative.

What are the advantages of conducting studies in healthy subjects for small molecule targeted therapies versus starting directly in patients?

Conducting small studies in healthy volunteers (HVs) can provide critical information on the drug’s exposure and half-life in humans, aiding in the design of patient studies. Additionally, safety profiles can be more clearly distinguished in HVs.

However, investors might prefer patient data over HV data. A compromise could be a two-part study, beginning with HVs and followed by patients, allowing for initial pharmacokinetics (PK) knowledge while ultimately providing patient data, which is more compelling to stakeholders.

Most often though, early oncology studies are conducted in patients, so as not to expose volunteers with healthy body systems to potentially toxic treatments.

What does minimal overlap in PK exposure translate into, especially in accelerated titrated designs when there might be insufficient information on PK variability?

The key factor here is the inter-patient variability in clearance. If there is uncertainty about the variability estimate, simulating various estimates can help determine the minimal overlap. For example, if variability ranges between 25% and 45%, simulations using these estimates can guide dose selection to minimize overlap, considering safety and other factors.

By addressing these questions, Project Optimus aims to refine dose optimization strategies, ultimately improving the effectiveness and safety of new therapies.

Interested to learn more? Join our office hours on Project Optimus with James Matcham and Michael Fossler on September 18 — sign up today!

Share Via