One-third of the clinical trials that led to new cancer drugs approved between 2008 and 2018 didn’t report on the race of trial participants — and even studies that did report on race often had far fewer black and Hispanic cancer patients than might be expected, given the makeup of the cancer patient population.

That’s according to a new study, published in JAMA Oncology, that looked at 230 clinical trials that supported oncology drugs approved by the Food and Drug Administration. Of those studies, just 145 reported on at least one race of trial participants. Just 18 broke the data down by the four major racial groups — white, Asian, black, and Hispanic — in the U.S. The study’s authors say their findings highlight the clear need for better reporting and representation in cancer trials sponsored by the drug industry.

“It’s important to recognize that this problem is there and this problem is persisting over the years,” said Dr. Kanwal Raghav, an oncologist at MD Anderson Cancer Center and an author of the study.

For trials that did report on race, there were notable disparities in the makeup of participants. White patients accounted for 76% of study participants, while Asian patients accounted for 18%. But black patients made up just 3% of participants in clinical trials for approved cancer drugs during that time, and Hispanic patients accounted for just 6%. The proportion of black and Hispanic patients in cancer trials did improve somewhat during the decade the study examined, though not considerably.

Black and Hispanic patients, in particular, were underrepresented in trials that led to cancer drug approvals. There were far fewer black and Hispanic patients in cancer clinical trials than would be expected, given the share of cancer patients who are black or Hispanic. Raghav and his colleagues say that’s problematic, particularly for trials that play a pivotal role in patient care.

“When you come across clinical trials that establish FDA approval or standard of care [for a new drug], they should definitely be representative of the population it’s used to treat,” Raghav said.

Another striking finding: Different racial groups were more likely to enroll in different types of trials. White patients were more likely to enter larger, randomized, later-stage trials with multiple arms. Minority groups, on the other hand, were more likely to enter smaller, non-randomized trials with just one arm. It’s not clear why, exactly, that’s the case. But taken together, Raghav said the findings raise questions about what the disparities might suggest about how researchers recruit for and run clinical trials.

“You do a trial of cancer patients across the country, you should see that kind of proportional representation. If it’s not being done, why is it not?” he asked.

The National Institutes of Health requires studies it funds to have racial representation that’s proportional to the patient population. But most human trials that contribute to drug approvals are funded by the drug industry. The Food and Drug Administration does recommend that trial researchers collect and report data on the race and ethnicity of participants and has also said study sponsors should enroll patients who reflect the patient populations that would benefit from the drug. But the findings show there’s room for improvement when it comes to representation in industry-sponsored research, Raghav said.

“The issue here is not where the funds are coming from,” Raghav said. “The issue here is doing the right thing. … There should be more efforts made to increase participation of all racial subgroups.”