Many Uncertainties With New Cancer Drugs Don’t Make it to FDA Label, Study Finds

With 80% of today’s cancer drugs reaching the US market under the FDA’s accelerated approval pathway,1 it’s no surprise that uncertainties remain when new therapies are first used to treat patients beyond a clinical trial.

But just how much do oncologists know about these uncertainties?

Avi Cherla | Image: LinkedIn

According to a July study in Health Affairs, it’s quite a lot, after investigators from the London School of Economics and Political Science (LSE) set out to quantify how much key information doesn’t make it to the FDA drug label.2
Led by Avi Cherla, a senior health policy associate with LSE Health and a PhD candidate in the Department of Health Policy at LSE, the study scrutinized uncertainties revealed during the regulatory process for 52 new cancer drugs that the FDA approved from 2019 through 2022. The team found that the FDA label, described as the agency’s chief way to communicate with physicians, did not include 26% of all uncertainties, and 48% of those were deemed important to the approval decision.2

Reached by email, Cherla told Evidence-Based Oncology (EBO) that “there does not seem to be an adequate mechanism in place to adequately communicate the known evidence and limitations of drugs to physicians, especially for drugs approved through expedited pathways.”

Cherla attributed the findings to the FDA sticking to past practice instead of finding more effective ways to communicate. “One example is the drug facts box,” which he called “a concise and structured summary” of benefits and risks. He cited trial data that show this method is more effective in relaying information to the general population.3

“Given that the drug label already includes a section on clinical studies, and that labels are most commonly read through online platforms, there seems to be an opportunity to discuss the associated uncertainties and limitations with this clinical evidence,” Cherla wrote.

As reported in Health Affairs, the team focused on cancer drugs because of the widespread use of the accelerated pathway in this specialty.2 EBO asked whether concerns about payer coverage might lead to labels that focus on adverse events (AEs). Cherla wrote that although information that makes it to the label “could potentially influence reimbursement decisions,” it was unclear why the FDA focuses on AEs at the expense of other uncertainties.

What Are Uncertainties, and Why Do They Matter?
FDA reviewers routinely uncover uncertainties in the approval process, which are evaluations of benefits vs risks that cannot be fully answered during the clinical trial process; this may be due to the size or makeup of the study population or the length of the trial. It is well known that trial participants tend to be younger and healthier than real-world populations, and the FDA has previously addressed the lack of diversity in trials as well.4

Of note, as of 30 days after the statutory deadline of June 26, 2025, the FDA had not issued a final guidance on a draft policy that would require drug companies to create Diversity Action Plans for each approval.4

Some uncertainties are related to the study population or trial design, but some are not; these include concerns with data analysis, missing outcome data, and selection of the reported result, the authors wrote.2

Not every detail of a drug review will make it onto the label, which is developed by pharmaceutical sponsors and approved by the FDA. It serves as the agency’s official summary of what is known about the drug, including any limitations. Investigators acknowledge that physicians may not look at the label every time they prescribe a drug, but “they are more likely to consult the label when prescribing a drug for the first time.”2

The authors wrote that uncertainties may include whether evidence from the clinical trial is generalizable to a real-world population and, increasingly, what relationship surrogate end points may have with “clinically meaningful outcomes for patients.”

Study Methods and Findings
Investigators systematically combed through review documents for cancer drugs approved during the study period, starting with uncertainties related to the primary outcome of pivotal trials. Any uncertainty mentioned by an FDA reviewer was included in the data set. The team then included uncertainties discussed by the FDA Oncology Drug Advisory Committee; this group of uncertainties was evaluated separately, they wrote. Finally, they analyzed uncertainties that make up the FDA Benefit-Risk Framework, as these are considered crucial to the approval decision.2

The team identified 13 specific types of uncertainties that may emerge during the review process. Besides generalizability and unvalidated surrogate end points, these include issues with randomization, such as unbalanced demographics between arms; issues with data integrity or missing data; and especially a risk-benefit imbalance, which may result from uncertainty over the size of the dose.2 Overall, questions about generalizability (73%), single-arm trials (68%), and long-term results (66%) came up most often in clinical trials. However, uncertainties about generalizability that were crucial for the FDA’s final decision were seen in only 21% of the trials.2

Through this process, investigators identified 213 clinical trial uncertainties and 107 that were crucial to decision-making. Drug labels included 56 of the 213 overall uncertainties, and 51 of 107 highlighted in the FDA Benefit-Risk Framework as important for final deicsions.2

According to Cherla, the issue of balancing risk vs benefit is critical. “The FDA should make more of a concerted effort to communicate those uncertainties it identifies as important to its assessments in the FDA’s Benefit-Risk Framework,” he wrote in the email to EBO.

“These uncertainties are already summarized in FDA assessments as the most important key uncertainties to its approval decisions,” he said. “It is important that these are prioritized for communication to physicians.”

References
1. Itskovich AC. What is the accelerated approval pathway? understanding the latest FDA guidance. WCG. April 14, 2025. Accessed July 27, 2025. https://www.wcgclinical.com/insights/what-is-the-accelerated-approval-pathway-understanding-the-latest-fda-guidance
2. Cherla A, Woloshin S, Wagner AK, et al. New cancer drug approvals: less than half of important clinical trial uncertainties are reported by the FDA to clinicians, 2019-22. Health Aff (Millwood). 2025;44(7):830-838. doi:10.1377/hlthaff.2024.01134
3. Schwartz LM, Woloshin S, Welch HG. Using a drug facts box to communicate drug benefits and harms: two randomized trials. Ann Intern Med. 2009;150(8):516-527. doi:10.7326/0003-4819-150-8-200904210-00106
4. Anthony VX. The upcoming FDA clinical trial diversity mandate: what you need to know. Imperial Clinical Research Services. May 14, 2025. Accessed July 27, 2025. https://www.imperialcrs.com/blog/business-insights/the-upcoming-fda-clinical-trial-diversity-mandate-what-you-need-to-know