Despite NCI support, some cancer trials fail to publish health-related quality of life data
Click Here to Manage Email Alerts
NCI has supported assessment of health-related quality of life in cancer clinical trials for approximately 30 years.
Yet, according to an NCI retrospective review published in Cancer, only about 62% of trials with a parent trial publication that received additional NCI funding to assess health-related quality of life (HRQOL) published the results of these endpoints.
“We were finding that the trials may have had these HRQOL endpoints, but they were never published or they may not have been analyzed,” Lori Minasian, MD, FACP, deputy director for NCI’s division of cancer prevention, a board-certified medical oncologist and a senior author of the study, told Healio. “Patients fill out all of these instruments, they sometimes answer questions multiple times, and they want to know they’re doing it in good faith. They’re participating in the research because they want to benefit, but they’re also participating because they want to give back. So, it’s important that we not just analyze HRQOL endpoints, but that they be published and the findings get back to patients and clinicians.”
Minasian spoke with Healio about the importance of HRQOL data in NCI-sponsored clinical trials.
Question: What prompted you to conduct this study?
Answer: About 15 years ago, co-authors Ann O’Mara, PhD, RN, Joseph Kelaghan, MD, and I started looking specifically at HRQOL endpoints in symptom management studies. We wanted to also explore the HRQOL for those treatment trials for which our program provided additional funding, but we realized that most of those trials were not completed. Recently, we decided to go back and see if we could determine which of those trials had published their HRQOL endpoints. We looked at between 1999 and 2015. Before 1999, NCI did not have the tracking system for the additional funding for HRQOL endpoints and we did not have the data to analyze. So, we took this 15-year time frame and looked at which trials were given additional funding for the HRQOL endpoints and the extent to which the HRQOL endpoints were published.
Q: How did you conduct the study?
A: First, we identified cancer treatment trials for which our program had provided financial support. There were more than 125 treatment trials supported by NCI between 1999 and 2015. Three were still actively accruing and 14 had terminated early. That left us with 108 trials for analysis.
We analyzed these studies in 5-year groups: those conducted from 1999 to 2004, 2005 to 2009 and 2010 to 2015. We are aware that cancer treatment trials may take years to accrue patients and then years of follow up before the primary treatment endpoint is ready for analysis and publication. Thus, we looked at the data in three timeframes. For those trials started between 1999 and 2014, 85% of the primary treatment results were published and of those trials, 62% had a HRQOL publication.
However, for the full 15 years, 54% of the treatment trials were published and 62% of those trials had published their HRQOL endpoints. For the full dataset of 108 trials, 58 had the treatment publication and of those 36 had the HRQOL publication. As a secondary endpoint, HRQOL would not be analyzed and published until the primary treatment endpoint was published. For the HRQOL endpoints, 20 of the trials had an independent HRQOL publication. The treatment trial was published in one journal, and then, the HRQOL findings were published in a different journal. Sixteen of the studies had some mention of HRQOL in the treatment publication.
We also looked at the trials by disease type. Breast cancer trials did extremely well. For the eleven published treatment trials, eight (72%) had published the HRQOL endpoints However, for some of other disease areas, it is a bit concerning. Colorectal cancer had a reasonable showing in terms of publication in general, but only one of the seven trials with HRQOL endpoints were published in terms of publishing HRQOL data.
Q: How did you respond to the results?
A: We wanted to make sure that the groups conducting the trials were aware of the study. We provided each group with their data and also the results of the analysis as a whole. The groups confirmed the data and provided input. Our overall goal was to understand the extent to which these endpoints were published and then begin a dialogue to improve the publication of the HRQOL endpoints.
Another piece to the study was the time frame of publishing HRQOL results. We looked at both the parent trial publication and the HRQOL publication. The median time from parent trial closure to publication was 4 years. Many HRQOL results were published within the first year. However, we were concerned about the studies that took 3 years or 5 years after closure to publish HRQOL data.
One of the challenges is that journals have a limited amount of space per paper. The ideal scenario — and it’s happening in a few circumstances — is to have the treatment results and HRQOL results published together. Typically, though, the treatment trials are published in one journal, and then a year or two later, the quality of life endpoints are published in a different journal. You must pull that data independently. That makes it difficult, because clinicians are not going to go searching for it.
Q: What can be done to improve the way HRQOL data is published?
A: We want the HRQOL endpoints that are captured in NCI studies analyzed and published, because patients have spent the time to participate in that aspect of studies. Those data deserve to be appropriately reported. By the same token, we need to determine the best way to report the HRQOL data. Often, it’s harder for clinicians to understand how to use this type of data and in that regard, it is perceived as less urgent to get that information.
We need to identify the best ways for clinicians and patients to use the data. The FDA has been working on Project Patient Voice to try to create graphic displays of some of this information, so it is more accessible to patients and clinicians. Getting that data out there is the first step. Figuring out how best to use it is the next step.– by Jennifer Byrne
Reference:
St Germain D, et al. Cancer. 2020;doi:10.1002/cncr.32765.
For more information:
Lori Minasian, MD, FACP, can be reached at National Cancer Institute, 9609 Medical Center Drive, Bethesda, MD 20892; email: minasilo@mail.nih.gov.
Disclosure: Minasian reports no relevant financial disclosures.
Collapse
Click Here to Manage Email Alerts