The landscape of early-phase clinical research continues to evolve rapidly. As new therapeutic modalities emerge and regulatory expectations shift, sponsors are rethinking how dose-finding studies are designed, analyzed, and interpreted. In this blog, we explore some of the latest developments and discussion points in early-phase trials, including recommended dose ranges, backfilling, randomization, and the growing focus on patient-reported outcomes.
Defining a Recommended Dose Range
Traditionally, early-phase studies aimed to identify a single “best” dose, typically the Maximum Tolerated Dose (MTD). However, this approach is increasingly being reconsidered.
Groups such as the Methodology for the Development of Innovative Cancer Therapies (MDICT) [1] task force now recommend that sponsors define a recommended dose range rather than a single value, particularly for monotherapy agents. This reflects a more nuanced understanding of drug behavior and aligns with initiatives like Project Optimus, which encourage comprehensive evaluation of dose-response relationships.
A dose range allows researchers to capture both safety and efficacy signals across multiple levels, improving the likelihood of identifying an Optimal Biological Dose (OBD) and supporting more informed decisions for later-phase studies.
Backfilling: Building Efficiency and Depth
Backfilling has become one of the most talked-about strategies in early-phase oncology trials. [2] Although some form of backfilling has been informally used for years, it has gained renewed attention following recommendations from Project Optimus.
Backfilling involves assigning new patients to existing dose levels, either the current dose or previous ones, to gather additional information on pharmacokinetics, pharmacodynamics, or efficacy. This approach strengthens the dose-response assessment by increasing the number of evaluable patients, helping to identify plateaus or nonlinear effects.
Unlike traditional expansion cohorts, backfilling can occur during dose escalation, improving trial efficiency and minimizing recruitment pauses. Sponsors can apply it flexibly depending on study needs, either after early evidence of safety and efficacy or in parallel with ongoing escalation.
Introducing Randomization in Early-Phase Trials
While randomization is standard in later-phase studies, its role in early development is growing. [1] Randomization can reduce bias, promote comparability between patient groups, and enhance confidence in emerging data.
In practice, randomization can be incorporated in two main ways:
- Between expansion cohorts, once multiple dose levels have been deemed safe for further exploration.
- During dose escalation, through randomized backfilling, where participants are assigned either to continue escalation or to a previously tested dose with promising signals.
Although early randomization improves the interpretability of results, it does not replace the need for a formal Phase II component, as early-phase trials are typically underpowered to detect definitive efficacy differences. Instead, it strengthens data integrity and supports more reliable decision-making.
The Growing Role of Patient-Reported Outcomes
Patient-reported outcomes (PROs) are another area of growing interest in early development. Project Optimus highlights their potential to enhance assessments of tolerability by providing direct insight into patient experience beyond standard adverse event reporting.
However, incorporating PROs at this stage presents challenges. Selecting appropriate measures can be difficult when side effects are unpredictable or the therapeutic class is novel. For drugs with prior safety experience, established tools such as PRO-CTCAE may be feasible [3], but for first-in-human compounds, careful consideration is needed to avoid compromising the study’s primary objectives.
As experience grows, PROs may become a more routine and valuable element of early-phase design, helping researchers capture a fuller picture of benefit and risk.
Innovation in Design Approaches
The field continues to innovate beyond traditional dose-escalation frameworks. New methods continue to be developed to address challenges such as combination therapies, where multiple agents interact in complex ways.
For example, Phastar’s statistical experts have explored robust approaches for introducing new doses mid-study, particularly in combination settings where early data reveal a large gap between existing dose levels.
By applying Bayesian estimation methods, it is possible to identify and add intermediate doses that are most likely to achieve the desired safety and efficacy balance.
These innovations reflect the broader trend toward flexibility, efficiency, and data-driven decision-making in early-phase development.
Conclusion
Early-phase clinical research is evolving faster than ever, driven by regulatory initiatives, new therapeutic modalities, and advances in statistical methodology. Designing these studies now requires more than technical expertise, it demands collaboration, adaptability, and clear scientific intent.
At Phastar, our statisticians and data scientists work closely with sponsors to design and implement innovative early-phase trials. From adaptive dose-finding to model-based simulations and exploratory endpoint analysis, we help ensure every decision is backed by solid data and sound reasoning.
References
- Araujo, D., Greystoke, A., Bates, S., Bayle, A., Calvo, E., Castelo-Branco, L., de Bono, J., Drilon, A., Garralda, E., Ivy, P., Kholmanskikh, O., Melero, I., Pentheroudakis, G., Petrie, J., Plummer, R., Ponce, S., Postel-Vinay, S., Siu, L., Spreafico, A., … Seymour, L. (2023). Oncology phase I trial design and conduct: Time for a change – MDICT Guidelines 2022. Annals of Oncology, 34(1), 48–60. https://doi.org/10.1016/j.annonc.2022.09.158
2.Zhao, Y., Yuan, Y., Korn, E. L., & Freidlin, B. (2024). Backfilling patients in phase I dose-escalation trials using Bayesian Optimal Interval (BOIN) design. Clinical Cancer Research, 30(4), 673–679. https://doi.org/10.1158/1078-0432.CCR-23-2585
3.https://healthcaredelivery.cancer.gov/pro-ctcae/overview.html
