Physician and oncology practice factors have greater influence over who gets enrolled in advanced cancer clinical trials than measured patient characteristics such as age, cancer type, and physical function, according to a recent study. The findings, published by Jiang et al in JAMA Network Open, point to practical ways to broaden access to promising new treatments.
“Oncologists across the country have long sensed that a patient’s chance of joining a clinical trial depends on which doctor and which clinic they see, and now we have data from more than 200,000 patients to show it. That is good news. Unlike a patient’s age or cancer type, a doctor’s trial habits and a clinic’s research support are things a health system can change,” said lead author Changchuan “Charles” Jiang, MD, MPH, Assistant Professor of Internal Medicine in the Division of Hematology and Oncology and of Health Economics, Systems and Policy in the Peter O’Donnell Jr. School of Public Health at UT Southwestern Medical Center (UTSW). He is also a member of the Population Science and Cancer Control Research Program in the Harold C. Simmons Comprehensive Cancer Center.
Study Methodology and Findings
In the study, researchers analyzed de-identified electronic health records from the Flatiron Health research database on adults treated for a broad range of advanced or metastatic solid tumors. The data came from 1,664 physicians at 107 oncology practices across the United States, most of which were in community settings, not National Cancer Institute–Designated Cancer Centers. Clinical trial participation was defined as receiving an investigational drug.
Although clinical trial participation increased during the study period, just over 5% of patients were enrolled in a trial. The study found that 389 of 1,664 physicians (23%) accounted for 84% of all enrollments, and 21 of 107 practices (20%) accounted for 81%, while nearly a third of physicians enrolled no patients.
The research team determined that characteristics of who treated a patient and where care was delivered explained more variation in participation than differences among patients. Patients had higher odds of participating when their physicians specialized in fewer types of cancer, had recently enrolled other patients, or worked at research-active practices. Patients whose physicians focused on one or two cancer types had about four times the odds of joining a trial.
The study also found disparities in trial enrollment. Black patients were 26% less likely to participate in a study than White patients. At practices serving the largest share of patients from low-income areas, the odds of participation were less than half of those at other practices.
“For many patients with advanced cancer, a trial is a way for patients to access promising new treatment. Our results suggest the chance of being offered one depends on where the patient is treated,” said senior study author Tian Zhang, MD, MHS, FASCO, Professor of Internal Medicine in the Division of Hematology and Oncology at UTSW and Associate Director of Clinical Research in the Simmons Cancer Center.
Potential Solutions
Researchers in the study identified potential solutions. Health systems could pair physicians new to trial enrollment with experienced mentors, assign each clinic a dedicated research contact, provide staff support for trial-related testing, and use tools to match patients with appropriate trials. Funders and trial sponsors could help build research capacity in clinics serving low-income communities and design simpler trials that smaller sites can run.
“We have seen this work in our own safety-net clinics, where patient navigation and physician engagement doubled enrollment of underrepresented patients,” Dr. Jiang said.
The authors noted that the database lacked information on whether individual patients were eligible for specific trials at a given time. Participation may also have been underestimated because the analysis counted patients only when they received a clinical study drug.
DISCLOSURE: This study was supported by a Conquer Cancer Career Development Award from ASCO, a Robert Winn Excellence in Clinical Trials Career Development Award, a UT Southwestern Dean’s Scholar in Clinical Research award, a Cancer Prevention and Research Institute of Texas grant (RR210079), and a National Cancer Institute Cancer Center Support Grant (P30CA142543). For full disclosures of the study authors, visit jamanetwork.com.
