A pilot study has revealed promising early results indicating that real-world external control arms can be constructed alongside patient enrollment in phase 2 clinical trials for HER2-positive breast cancer. This approach, presented by Jessica Paulus, ScD, during the 2025 ASCO Annual Meeting, could significantly expedite the drug development process for this specific cancer type.
The study focused on a single-arm phase 2 trial (NCT05748834) assessing the combination of tucatinib (Tukysa) and liposomal doxorubicin in patients with locally advanced or metastatic HER2-positive breast cancer. At the time of the interim analysis, the trial had enrolled eight patients. Researchers employed multiple imputation techniques to create a simulated dataset of 40 patients based on baseline characteristics. This simulated cohort was then compared with a real-world data external control arm comprising 77 patients.
Following propensity score matching, 82% of the simulated cohort aligned successfully with the external control group. The study demonstrated that covariate matching could be achieved across several baseline characteristics, including mean age, number of prior treatments, and prior exposure to tucatinib. As the phase 2 trial aims to enroll 36 patients, researchers anticipate that the balance between the cohorts will enhance as recruitment progresses.
The primary endpoint of the trial is the overall response rate, while key secondary endpoints include safety and progression-free survival. Paulus emphasized the significance of this research, stating, “The big-picture reason we’re doing this is in the hopes of demonstrating that this approach can be valid for phase 2 studies and that assembling these kinds of real-world cohorts would offer a way to speed up the clinical development pipeline.”
Paulus, who serves as vice president of Real World Research at Ontada in Boston, Massachusetts, discussed the increasing interest in leveraging real-world data to support external control arms. These arms aid in evaluating the efficacy and safety of experimental therapies, especially in single-arm trials that often lack comparator groups.
Traditionally, most efforts to establish external control arms have focused on phase 3 trials. By extending these methods to phase 2 studies, the research team faced unique challenges due to the typically smaller sample sizes and the critical necessity of comparator arms. The current study aims to create a real-world data-derived external control arm for a phase 2 clinical trial sponsored by the Sarah Cannon Research Institute in Nashville, Tennessee.
The interim findings presented at ASCO 2025 indicate that the researchers successfully developed the external control arm in parallel with the ongoing trial. By anchoring their real-world cohort around the first eight trial patients, they were able to match 36 real-world patients on approximately six key baseline characteristics. This approach aims to demonstrate the comparability of trial and real-world patients, thus validating the external control arm as a reliable reference point.
Efforts to assemble the correct medical oncology and epidemiology expertise were crucial to ensure accurate matching of real-world patients with trial participants. Paulus noted that while real-world data can often be compiled more rapidly than trial data, the process still demands substantial effort. The initial results showed that four out of six baseline characteristics exhibited great balance, with only two variables slightly out of alignment.
As the trial progresses toward its target enrollment of 36 patients, the researchers expect to enhance the balance between the cohorts. They aim for a ratio of approximately three real-world patients for every trial participant, anticipating that a larger sample size will improve comparability.
Ultimately, this innovative approach could facilitate faster, evidence-driven decisions within the life sciences sector, particularly in advancing to phase 3 clinical trials. By integrating real-world data into the clinical development pipeline, researchers hope to streamline processes and ensure more robust evaluation of therapeutic efficacy.