Toxicity in phase 3 cancer trials often unreported, minimized
Key takeaways:
- Less than half of phase 3 trials reported toxicity in a guideline-concordant manner.
- Nearly half of trials used what researchers characterized as toxicity-minimizing language.
Toxicity information from phase 3 oncology clinical trials often is incomplete and interpreted in ways that minimize its impact, according to findings published in JCO Oncology Practice.
“We publish a lot of studies that don’t necessarily consider the patients’ perspective, but they are the people who are ultimately impacted by the results of our studies,” Avital M. Miller, BA, research assistant at The University of Texas MD Anderson Cancer Center, told Healio. “It’s important when we’re publishing that we’re representing information that may not be seen as clinically relevant by certain standards, but that definitely would be relevant from the patients’ point of view.”
Increased attention
Lack of complete toxicity reporting and use of subjective minimizing language to describe adverse events are gaining increased attention in oncology.
In 2023, Healio reported results of a systematic review that showed abstracts of randomized phase 3 clinical trials in gastrointestinal oncology frequently use subjective minimizing language to characterize serious toxicities. Findings showed 24% of abstracts used “any subjective minimizing language," and that 95% used subjective or objective toxicity minimization. Information about serious adverse events also was frequently absent or incomplete.
Another study showed more than 15% of published abstracts from ASCO, ASH and European Society for Medical Oncology annual meetings used subjective minimizing language — including words such as “safe” or “tolerable” — when discussing treatment-related toxicities. Nearly 9% of abstracts that used that type of language included grade 5 adverse events.
‘Comprehensive assessment’
Miller — along with Ethan B. Ludmir, MD, assistant professor in the department of gastrointestinal radiation oncology at MD Anderson — and colleagues aimed to add to this evidence base.
“We endeavored to provide a comprehensive assessment of every phase 3 cancer clinical trial we could find from the past 20 years or so,” Ludmir told Healio. “We wanted to assess compliance with toxicity reporting and, in that process, developed additional guidance that we think might be relevant as we look ahead.”
Researchers used ClinicalTrials.gov to identify completed two-arm, superiority-design phase 3 oncology trials. Their analysis focused on 407 trials (n = 322,645) published between 2002 and 2020.
They assessed complete toxicity reporting, defined as reporting all adverse events, all serious adverse events, total deaths and discontinuation of study treatment due to toxicity. They also evaluated use of toxicity minimizing language, defined as “a set of terms that subjectively downplay the harm of therapies.”
More than half (55%) of trials reported total adverse events, 51% reported serious adverse events, 88% reported total deaths and 84% reported study treatment cessation resulting from toxicity.
Less than half (44%) of trials reported toxicity in a manner concordant with guidelines published in BMJ related to reporting of adverse events, deaths and study discontinuation due to toxicity.
“We expanded on that by developing augmented guidance in terms of toxicity reporting that emphasized what we described as complete toxicity reporting, with an emphasis on lower-grade toxicities,” Ludmir said. “Only 32% of trials were compliant with such augmented toxicity reporting criteria.”
A higher percentage of industry-sponsored trials than cooperative group-sponsored trials adhered to complete toxicity reporting (37% vs. 4%).
“This raises the question of why we are seeing better completion of toxicity reporting among industry-supported trials than trials supported by the federal government,” Ludmir said.
In addition to examining how “objective data” were presented, Miller, Ludmir and colleagues evaluated “the subjective side” by examining how researchers were talking about their findings.
“We looked at wording that states that a drug is relatively safe, or not harmful,” Miller said. “These are words that other researchers and patients are noticing. If a patient sees that a drug doesn’t seem dangerous or have too many side effects, it might sound like a good idea. However, that might not be an accurate representation of what the drug really is. So, we were curious about this use of subjective language.”
Their analysis identified toxicity minimizing language in 46% of trials (95% credible interval, 41%-51%).
However, they identified no association between type of trial sponsorship and use of toxicity minimizing language.
“Everyone seems to do this to some extent — minimizing or obfuscating how toxicity is defined,” Ludmir said.
Implications
The findings of this study — and the work of other research teams who have examined this issue — have considerable implications, Ludmir said.
“We’ve learned from recent publications that lower-grade toxicities — which hadn’t been emphasized in older iterations of toxicity reporting guidelines — probably matter a lot more to patients,” Ludmir said. “Grade 3 toxicities are bad, and they carry a lot of significance in their own right. However, we often put aside grade 1 or grade 2 toxicities, when those can be exceedingly debilitating to quality of life.”
For example, a patient may develop a grade 3 gastrointestinal bleed but recover from that adverse event.
In contrast, a person taking a kinase inhibitor may experience grade 2 nausea and grade 2 fatigue “day in and day out,” Ludmir said.
“Their quality of life has very likely suffered,” he said.
Improving transparency and objectivity
The increasing volume of evidence about toxicity underreporting and use of minimizing language to describe adverse events highlights the need for coordinated solutions, researchers concluded.
"There are a plethora of guidelines out there, including for different aspects of clinical trial design and reporting,” Ludmir told Healio. “What we are hoping to do is not just identify areas of improvement and outline a set of criteria that may address those deficits, but actively look for ways to partner with editorial boards, regulatory agencies, trialists, sponsors and cooperative groups to improve transparency and objectivity in toxicity reporting.
“It is easy to be an armchair quarterback,” Ludmir added. “The challenge is in building partnerships to effect change without increasing the burden on trialists."
Reference:
Ethan B. Ludmir, MD, can be reached at ebludmir@mdanderson.org.
Avital M. Miller, BA, can be reached at amiller7@mdanderson.org.
Collapse
Healio Interviews