Data Sharing From Cancer Drug Trials Falls Short of Promises, Study Authors Say

The precision medicine era promises treatments for cancer and other diseases will be increasingly tailored to a patient’s individual traits. That may mean finding therapies that target certain genes or proteins or using data to learn which regimens work best given a patient’s sex, ethnicity, age, or other factors.

As study results are reported, some answers seem clear; others, less so. Results of a study published in May found that women in cancer clinical trials are far more likely than men to experience severe adverse events (AEs) from drug toxicity.¹ Meanwhile, investigators of a study published in 2021 asked whether different pharmacokinetic responses were due to ethnic differences or to factors such as age, diet, and other medicines patients were taking.²

One thing is certain: answering these questions requires more data, not less. And that’s why investigators from Australia sounded the alarm with results from an audit published in the September issue of JAMA Oncology. They found that individual participant data (IPD) from 55% of the cancer clinical trials that led to FDA approvals over a 10-year period are not available to independent investigators, despite a 2014 pledge from the pharmaceutical industry to improve data transparency.³

Lead author and PhD candidate Natansh D. Modi, BPharm; and senior author Ashley M. Hopkins, PhD, both of the College of Medicine and Public Health at Flinders University in Adelaide, Australia, told Evidence-Based Oncology™ (EBO) in an interview how the audit was triggered by their individual frustrations in seeking trial data for other projects.

Although the authors say data availability has improved, some meta-analyses are impossible because data for certain top-selling drugs are off limits. Their paper pointed specifically to nivolumab (Opdivo), pembrolizumab (Keytruda), and pomalidomide (Pomalyst). The audit found that 90% of the trial data for these 3 therapies were unavailable. The authors noted that this trio of therapies generated $30 billion in revenue in 2020.⁴

The top reasons given when trial sponsors were asked why IPD were ineligible for sharing was that the trial was ongoing (53%), or that the trial was complete but the IPD were still under embargo (21%). In their paper and during the EBO interview, Modi and Hopkins rejected these reasons, stating that, given the level of activity in cancer drug development, sharing anonymized IPD is essential for robust research on newer anticancer medicines.

“The concerning aspect of what we found in our study was that some of the most-used medicines—those having the most money spent on their use—were found to have the least transparency in the data,” Hopkins said. “It is a little bit concerning that when medicine seems to be a high-profit kind of medicine, it hasn’t been set up in a way to share that data.”

This fact is of concern to philanthropic groups that Hopkins works with because when drugs are new it may be too early for these entities to find real-world, postmarketing evidence to guide decision-making. But the greater concern with IPD sharing is that evidence be available for secondary analyses and to design new trials, he said.

Support for the audit came from author fellowships or grants from Cancer Council Australia and the National Health and Medical Research Foundation. Two coauthors on the audit reported unrelated research support from Pfizer.

Multiple Requests for Data Status
The audit stemmed from Hopkins’ 2018 research on data availability.⁵ Attempts to gain access were met with statements that IPD were ineligible for release, which Hopkins noted was in direct conflict with the companies’ stated policies.

Modi’s research involved the use of artificial intelligence in treating patients with breast cancer—another project that required requests for clinical trial data. “What we were getting in responses was, ‘These data are not eligible to share.’ And then we said, ‘Hold on, this doesn’t really match your data-sharing policies,” he said.

In the interview, Modi described audit methods, which first required identifying all anticancer medicines approved by the FDA between January 1, 2011, and June 30, 2021. Using product labels accessed through the National Institutes of Health DailyMed database, the authors compiled a list of all clinical trials with results that were used in the approval process for the therapies. This resulted in a list of 304 trials for 115 different therapies. All trials were registered on ClinicalTrials.gov.

Trial start and completion dates were of interest because the authors noted those that started after January 1, 2014. This was the effective date of an agreement on transparency for IPD, which was developed jointly by industry trade groups, the Pharmaceutical Research and Manufacturers of America (PhRMA) and the European Federation of Pharmaceutical Industries and Associations (EFPIA).⁶ Of the 304 trials in the audit, the authors noted that 140 (46%) had a start date before January 1, 2014.

“Less than 3 years had passed since the results summary was added to the product label for 136 (45%) of the trials, 3 to 7 years for 126 (41%) trials, and more than 7 years for 42 (14%) trials,” the authors wrote.³

Among the companies involved, the authors found that 24 belong to 1 or both signatories to the 2014 transparency pledge, 19 have published an IPD policy on their website or a third-party source, and 21 lack a published IPD policy.

Modi then described the painstaking process of contacting all the trial sponsors to learn IPD eligibility. Of the 304 industry-sponsored trials in the audit, IPD sharing eligibility was publicly available for 64 trials (21%). For the remaining trials, Modi contacted the sponsor to determine the status. He created a cadence of reaching out every 30 days until he received a response. In some cases, the drug was no longer owned by the initial trial sponsor, which created delays. In the end, 9 trials (3%) sponsored by 8 different companies offered no response on the eligibility status of the trial data.

The contacts themselves were straightforward, Modi explained. “Our email is just asking them, ‘[Are these] data eligible to share or not? And if not, just give us a reason,’” he said. “The median response time was 42 days because they took their time until I followed up, then followed up again.”

Variation Among Companies
Amid some frustration, there was good news. The authors found the IPD were eligible for sharing from 136 trials, involving 60 anticancer drugs and data from more than 70,000 patients. Data were available from more than 50% of the trials for atezolizumab (Tecentriq), abiraterone (Zytiga), enzalutamide (Xtandi), ibrutinib (Imbruvica), osimertinib (Tagrisso), palbociclib (Ibrance), and pertuzumab (Perjeta).

“Furthermore, 5 of the top 20 pharmaceutical companies by revenue (Amgen, AstraZeneca, Boehringer Ingelheim, Pfizer, and Sanofi) indicated that more than 75% of their sampled oncology trials were eligible for IPD sharing,” the authors wrote. These achievements, they wrote, are a direct result of the 2014 data sharing principles “and are a beacon of hope to one of the least trusted industries in the world.”³

In their paper and in the interview, the authors called on regulatory agencies and journal editors to take stronger stands in support of data sharing. Three specific recommendations are:

• Pharmaceutical sponsors should join PhMRA and/or EFPIA and establish a data sharing policy;
• Sponsors should establish an IPD sharing process external to the company, recognizing there may be conflicts of interest; and
• Enact a policy that states “all IPD underlying results presented in a product label will be immediately available for sharing,” This element is designed to remove extended embargo periods and ensure that registration of product triggers availability of clinical data.³

Hopkins said changes are needed to promote research to improve patient care. A major reason for requesting trial data, he said, is to allow systemic evaluation of data in search of signals related race or sex. A second reason is to address variability in the quality of data provided, Hopkins said.

“Some companies put in a lot of care with the data that they provide. And they’re also transparent as to what data [were] collected throughout the entire trial—you can see that in the data dictionary,” Hopkins said. With other companies, “they don’t give the data dictionary and they redact a lot of the clinical data,” he said.

For investigators looking to study signals into differences in rates of AEs by race or sex, the way these data are reported matters a lot, he said. With the current emphasis on reducing disparities in cancer care, it would seem to matter more than ever how these data are shared. Today, some companies do a very good job in this area, Hopkins said, but others do not.

“So that’s the next step—trying to set standards around actual data provision,” he said.

References
1. Unger JM, Vaidya R, Albain KS, et al. Sex differences in risk of severe adverse events in patients receiving immunotherapy, targeted therapy, or chemotherapy in cancer clinical trials. J Clin Oncol. 2022;40(13):1474-1486. doi:10.1200/JCO.21.02377
2. Olafuyi O, Parekh N, Wright J, Koenig J. Inter-ethnic differences in pharmacokinetics—is there more that unites than divides? Pharmacol Res Perspect. 2021;9(6):e00890. doi:10.1002/prp2.890
3. Modi ND, Abuhelwa AY, McKinnon RA, et al. Audit of data sharing by pharmaceutical companies for anticancer medicines approved by the FDA. JAMA Oncol. 2022;8(9):1310-1316. doi:10.1001/jamaoncol.2022.2867
4. Critical oncology trial data remains hidden. News release. Flinders University. August 1, 2022. Accessed November 24, 2022. https://news.flinders.edu.au/blog/2022/08/01/critical-oncology-trial-data-remains-hidden/
5. Hopkins AM, Rowland A, Sorich MJ. Data sharing from pharmaceutical industry sponsored clinical studies: audit of data availability. BMC Med. 2018;16(1):165. doi:10.1186/s12916-018-1154-z
6. Principles for responsible clinical trial data sharing. Pharmaceutical Research and Manufacturers of America. July 18, 2013. Accessed November 24, 2022. http://phrma-docs.phrma.org/sites/default/files/pdf/PhRMAPrinciplesForResponsibleClinicalTrialDataSharing.pdf

COVER ARTICLES PDF