Sex disparities ‘pervasive’ in cancer clinical trial participation
Sex-based disparities remain pervasive in cancer clinical trial participation, according to results of a retrospective meta-analysis published in Journal of the National Cancer Institute.
Researchers observed underrepresentation of women primarily in noncooperative-group-supported trials, which account for the majority of randomized controlled trials.
“In totality, we tried to mirror how other studies have been modeled to give an update on what disparities in clinical trials look like in the broad scope of phase 3 clinical trials in cancer,” Ethan B. Ludmir, MD, resident radiation oncologist at The University of Texas MD Anderson Cancer Center, told HemOnc Today. “It’s hard to exactly assess the impact of these disparities on clinical research. When we run these clinical trials and establish the standard of care, if you have underrepresentation of a specific demographic, you end up getting questionable interpretability of your trial results. Do you really know that your trial results are going to work in a subgroup if they are dramatically underrepresented in your trial? That’s the big concern with these disparities.”
Underrepresentation can make it difficult for people in the community to have confidence when addressing individual patients, C. David Fuller, MD, PhD, associate professor and associate director of MR programmatic development in the department of radiation oncology at University of Texas MD Anderson Cancer Center, told HemOnc Today.
“One problem [in the medical community] we see is there is actually overconfidence when we see a phase 3 clinical study,” he said. “Medical professionals are inclined to think these studies must demonstrate what’s best for everybody, but we just don’t know that.”
Ludmir, Fuller and colleagues searched clinicaltrials.gov to identify 147 randomized phase 3 clinical trials involving a therapeutic intervention for colorectal and lung cancer that listed the proportion of female patients in each study. Researchers chose colorectal and lung cancer because they are the two most common nonsex-specific disease sites.
The trials enrolled a total of 100,907 participants between 1996 and 2014. Investigators calculated the difference in proportion of female patients (DPF) for each trial by subtracting the proportion of female patients in the trial from the proportion of female patients in the population by disease site.
Results showed an average DPF among all trials of 6.8% (95% CI, 8.8 to 4.9; P < .001), with more pronounced differences among the 121 noncooperative-group-sponsored trials than among the 26 cooperative group trials (8% vs. 1.1).
Researchers then examined 16 trials (10.9%) restricted to patients with molecular subtypes associated with a larger proportion of female patients, including ALK-rearranged or EGFR-mutated non-small cell lung cancer. Results showed these trials had a greater proportion of female patients than unrestricted trials (DPF = 2.6 vs. 8%; P = .006).
Linear regression modeling showed no significant improvement in underrepresentation of women over time, with an estimated annual change in trial DPF of 0.1% (95% CI, 0.5% to
Researchers noted the need for further studies of sex-based disparities in trials involving other disease sites.
“These sorts of problems with disparities came to the forefront about 20 years ago and really spurred a lot of interest because it also includes minorities and older patients on phase 3 clinical trials,” Ludmir said. “What they ended up doing at the NCI was creating programs to promote enrollment. These initiatives have worked reasonably well, and the question now is how to do extrapolate that to noncooperative group setting.” – by John DeRosier
For more information:
Ethan Ludmir, MD, can be reached at The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 12345; Twitter: @ebludmir.
C. David Fuller, MD, PhD, can be reached at The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 12345; email: firstname.lastname@example.org.
Disclosures: The authors report no relevant financial disclosures.