Adaptive clinical trials in oncology: Meeting the challenge of Project Optimus

Many organizations are adopting adaptive trial designs that use accumulating data to refine dose selection to meet these expectations. Integrating safety, efficacy, and exposure data in real time allows sponsors to identify the best balance of risk and benefit earlier, improving study efficiency while maintaining statistical and ethical rigor.

Leveraging adaptive clinical trial designs in oncology allows researchers to adjust predefined study elements based on accumulating data, such as dose levels, enrollment ratios, or stopping rules. These approaches embed learning directly into the study, improving efficiency and decision-making confidence.

In dose-finding oncology trials, adaptive methods support real-time assessment of pharmacokinetic (PK), pharmacodynamic (PD), safety, and efficacy data. This allows sponsors to make data-driven decisions during the trial’s progression, instead of waiting until a trial ends to interpret results. As evidence builds, sponsors can refine hypotheses and focus on the most informative doses.

Unlike traditional 3 + 3 designs, which focus only on acute toxicity, adaptive and model-driven approaches capture the broader exposure–response relationship, revealing where benefit plateaus and toxicity begins to rise. This delivers a clearer view of the therapeutic window and a faster path to the optimal dose.

One of the most effective adaptive strategies is the family of Bayesian Logistic Regression Model (BLRM) designs, which evaluates both dose–toxicity and dose–response relationships using all available data, not just results from the latest cohort.

Rather than escalating until toxicity occurs, this approach identifies a minimum safe and biologically effective dose (MSBED) that achieves the desired pharmacodynamic effect while maintaining safety. This objective is more meaningful than simply finding the Maximum Tolerated Dose (MTD).

Model-based methods reduce exposure to high-risk doses and generate stronger evidence for later-stage development. Regulators increasingly endorse these designs under Project Optimus because they enable more comprehensive dose evaluation, while ethics committees value their ability to minimize patients’ unnecessary exposure to overly toxic doses and maximize what can be learned from every patient.

Adaptive methods continue to add value in Phase 2/expansion cohort dose-finding studies through a technique known as response-adaptive randomization.

Early in the study, cancer patients are typically randomized equally across active dose arms. As interim analyses reveal new safety and efficacy data, randomization ratios are updated, directing more patients to doses with the most favorable profiles. An independent Data Monitoring Committee (DMC) typically reviews these interim analyses.

This adaptive process ensures that more participants receive potentially effective treatment as the trial progresses, while maintaining statistical control. By study completion, most patients have contributed to the most relevant dose-response data, accelerating learning and confirmation.

In oncology programs, where every patient’s experience matters, this approach improves operational efficiency and ethical value, aligning research progress with patient benefit.

Adopting adaptive design begins with careful preparation. Before entering first-in-human studies, sponsors need to understand their compound’s pharmacology, its mechanism, receptor targets, and preclinical exposure data. They also need to consider what measurable biomarkers are available to act as indicators of pharmacological activity.

Early exposure–response or PK/PD modeling helps establish the quantitative goals that will guide dose selection as data accumulate. Once initial clinical results are available, clinical trial simulations can compare adaptive and traditional designs to evaluate gains in efficiency, safety, and probability of success.

These simulations provide the evidence needed to refine design choices and build confidence with regulators. By planning early and integrating modeling throughout development, sponsors can streamline decision-making, improve patient safety, and accelerate progress toward the optimal dose.

Adaptive design is now a core element of modern oncology development. By integrating modeling, simulation, and flexible trial structures, sponsors can generate stronger evidence with fewer patients and greater efficiency.

As therapies become more targeted and regulatory expectations evolve, adaptive methods offer a practical path forward, balancing innovation, ethics, and speed. They represent the next step in evidence-based oncology development: smarter trials, better data, and faster access to effective treatments.