Clinical trials ultimately exist to improve patient care through development of new therapeutic agents and regimens that allow patients to live longer, healthier lives. However, clinical trials are not always designed with the patient participant in mind. Traditional trial design often does not include endpoints that capture patient goals or meaningful real-time participant feedback. In addition, traditional trial protocols often lack flexibility regarding eligibility requirements and collection of participant data.
The Clinical Research Workshop, held Tuesday, December 8, highlighted new ways to conceptualize trials—from adoption of Bayesian statistical designs and newer technologies for detection of genomic mutations to incorporation of proactive cultural responsiveness and patient-relevant endpoints—that could raise the bar for the way breast cancer research is performed and reported.
Starting with Statistics
Algorithms, such as 3+3, are the mainstay for phase I clinical trial design but, according to Thomas M. Braun, PhD, Professor of Statistics at the University of Michigan School of Public Health, algorithms often have poor statistical properties and do not generate meaningful data regarding maximum-tolerated dose. In addition, cohort data is often siloed and not well used across the trial to make determinations about dose. Instead of algorithms, Dr. Braun recommends a Bayesian approach because it is generalizable to a complex landscape.
For example, he said that immunotherapy (alone or with chemotherapy) and combinations of agents can be studied in early-phase Bayesian trials, and efficacy endpoints can be used to determine later-phase dosages. Participants are enrolled as they are eligible in a Bayesian-based clinical trial versus after completed observation of a cohort, as in the 3+3 model. Although most researchers assume that Bayesian models are too complicated, Dr. Braun highlighted two cloud-based programs (one from the University of Virginia and the other from MD Anderson Cancer Center) that alleviate the need for a dedicated statistician to generate statistical results.
For phase II and phase III studies, Dr. Braun made recommendations as to how researchers can deviate from the norm in ways that are both useful and efficient. For phase II trials, he asked researchers to consider using more than one dose from the phase I trial. Using historical data, researchers can provide interim rules regarding determination of efficacy as the data accrues. This could allow elimination of the less effective dose and reassignment of patients to the more effective dose.
“This approach of incorporating historical data and using more than one dose is vastly better than the common practice of using an expansion cohort pasted to the end of the 3+3 design,” Dr. Braun said.
Careful planning of subgroups ahead of time and stratification of the subgroups to all arms of the phase III trial is important. Also, incorporation of co-data—historical data from outside of a clinical trial—and use of power priors, which allow for statistical weighting of clinical trial data differently from that of historical data, are more strategic ways to effectively use data and resources.
Incorporating Real-World Knowledge, Real-Time Feedback
There is more to a successful clinical trial than just statistical planning and identification of subgroups, however. Avoiding terms that dehumanize participants and using clear, everyday language in the informed consent to promote health literacy across racial and ethnic populations can go a long way to bolster trial recruitment. Likewise, patients might be more likely to stay on a clinical trial if there are patient-reported outcomes (PROs) from earlier trials that help set participant expectations and provide researchers with heightened cultural sensitivity—a way to see through the patients’ eyes.
“Patients are going through an endurance test when they’re going through treatment,” said Deborah Collyar, president of Patient Advocates in Research (PAIR). “If we start to think about it [this way], the whole tenor of the clinical trial can change, the whole way we approach the clinical trial and talk about clinical trials to people can change.”
Because patients have very different experiences, recruitment plans also should attempt to avoid assumptions based on labels and approach each patient as an individual. For example, older patients may be excluded from a recruitment plan or communicated with differently solely based on age, but an older patient might be healthier and/or have a higher level of health literacy than a younger patient. Ms. Collyar advocated for the creation of “healthy aging” programs that provide different language and therapeutic options for clinical trial participants based not just on age but also on health status.
“COVID-19 has resulted in a silver lining regarding clinical trial participation,” said Ms. Collyar. Changes in the forms of “telehealth visits, local labs and procedures, home nursing care, and home delivery of some agents … will help people be more receptive to clinical trials and help them participate.”
Assumption of shared experience among patients also leads to racial and ethnic disparities among trial participants. PROs can be used to improve disparities in trial participation, however, by creating clear expectations about the type of side effects that interfere with efficacy and adherence. PROs also provide researchers with culturally sensitive insights for vulnerable and at-risk populations.
Oluwadamilola “Lola” Fayanju, MD, MA, MPHS, FACS, is Associate Professor of Surgery and Population Health Sciences at Duke University Medical Center and Associate Director for Disparities & Value in Healthcare at Duke Forge. She discussed a study that found Black patients to have lower baseline distress scores after breast cancer diagnosis than White patients; however, Black patients experienced less improvement in distress over time, indicating that stressors and distress levels might have been underreported at baseline. Being Black also was associated with a longer time to treatment (47 days vs 45 days for White patients), higher disease stage at diagnosis, and presence of practical stressors. Hispanic participants had the longest median length from diagnosis to treatment (71 days), which was associated with existence of practical stressors and higher baseline distress scores (compared with White patients).
“We really need to develop anticipatory and culturally responsive approaches in order to make sure that we are helping people during this critical period between evaluation and treatment and where we know disparities exist,” Dr. Fayanju said. “PROs in this case can allow us … to not only capture areas where patients need help but potentially also decrease disparities.”
New Technologies, Increased Flexibility
Although researchers have known for years that patients with metastatic breast cancer and a high number of circulating tumor cells have a poor prognosis, this has yet to translate into clinical benefit. With the addition of circulating tumor (ct)DNA assessment via blood-based or liquid biopsy into clinical trials, even retrospectively, clinical identification of genomic mutations for better-directed targeted therapy (as in SOLAR1) or to identify resistance mutations (as in PALOMA3) has become a reality.
The key question at this point, according to Nicholas Turner, PhD, Consultant Medical Oncologist and Head of the Ralph Lauren Centre for Breast Cancer Research, is whether clinical trials can be used to assess whether treatment in early breast cancer, based on identification of genetic mutations shortly after diagnosis to predict relapse, can improve outcomes.
Use of CRISPR genome editing is also just beginning to enter the clinic and is, according to Dr. Turner, a “massively powerful clinical tool” that allows for “genome-wide screening to identify cancer targets.”
Although new technologies are not without their challenges, they can potentially provide increased benefit for patients without increased hassle. Liquid biopsy can be performed at local labs, if clinical trials continue to allow the type of flexibility that has been become necessary during the COVID-19 pandemic.
SABCS registrants have exclusive on-demand access to this and other virtual SABCS programming until March 13, 2021.