Project Patient Voice a focus of ASCO-FDA workshop

Regulatory NewsRegulatory News
| 17 July 2020 | By Kari Oakes 

A day-long workshop co-sponsored by the US Food and Drug Administration (FDA) and the American Society for Clinical Oncology (ASCO) focused on clinical outcome assessments in cancer clinical trials. A key topic of the day’s work was the pilot version of Project Patient Voice, a new attempt to create a flexible, searchable, accurate, and understandable website for members of the public who wish to learn about patient experience data. The website is meant to give patients and their caregivers realistic information and to be a basis for shared decision-making between patients and healthcare providers.
The session was moderated by Janice Kim, PharmD, of FDA’s Oncology Center of Excellence. She set the stage by noting that Project Patient Voice’s guiding principle is to adhere to the spirit of the 21st Century Cures Act. The pilot program provides a web-based source of patient-reported outcome (PRO) data, designed to give consumers information about the experiences of patients participating in clinical trials. A precedent for public-facing, searchable data of this sort exists in the FDA’s Drug Trials Snapshots, said Kim.
Project Patient Voice has been a partnership between FDA’s drug development team and commercial sponsors, and “came with challenges,” said Kim. From the agency perspective, these included website design and deployment, negotiating with the commercial sponsor to reach agreement on tables and figures, and ensuring that the website was fully accessible to those with hearing- or vision-related disabilities in order to comply with Section 508 of the Americans with Disabilities Act.
The Project Patient Voice website pilot currently features just one drug trial, AstraZeneca’s AURA3 trial of Tagrisso (osimertinib) for non-small cell lung cancer. Katerina Halling, Global Head of Patient Reported Outcomes for AstraZeneca, participated in the workshop. She noted that the AURA strategy planned for including PROs as early as phase 1 of the AURA trials, a tactic “aimed at ensuring the patient experience is fully captured.”
Over 2 dozen symptoms are included in the PRO version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE), noted Halling. This list was put together using a “conservative approach,” with an eye to being over- rather than under-inclusive, she said.
Resources used in developing the tool included other oncology trial questionnaires, documented data about adverse events in those taking epidermal growth factor receptor tyrosine kinase inhibitors, emerging phase one clinical data from AURA, as well as expert input and a targeted literature review. Patients with non-small cell lung cancer were also interviewed and their input was incorporated.
“This has been a journey full of learnings, and I guess we are still learning together,” said Halling. Data from PRO-CTCAE should be used in the same way as other patient reported outcome data, she said, adding that the same best practices for use of the data should be applied during the clinical trial process as for other patient reported outcome data: “The important point is that the now-standardized approach will allow for easier interpretation and comparison.”
Paul Kluetz, MD, Deputy Director of the Oncology Center of Excellence, noted that the data visualizations in Project Patient Voice include both symptoms present at baseline and those that were not present at baseline but developed during treatment. This distinction is a nuance that patients find very valuable, he said. “What does this table tell you that a standard CTCAE safety table would not? It’s an excellent way to bring the patient’s perspective in a complementary fashion to clinician safety tables.”
The visualizations chosen for the pilot run the risk of missing some important data, said Mallorie Fiero, PhD, a mathematical statistician supporting the Office of Oncologic Diseases at FDA’s Center for Drug Evaluation and Research. The visualizations depict PRO responses for both study arms. Fiero pointed out that only 80 of the patients who took at least one dose of Tagrisso were included in building the data visualizations, in part because PRO-CTCAE responses were not available for these patients. “Not all randomized patients were evaluated,” she noted.
The real risk with the Tagrisso PRO data is that there might be different reasons for dropout between the two study arms, she said. If more patients in one treatment arm stopped responding or dropped out because of adverse events or even death, those data wouldn’t be captured when only PRO-CTCAE data are reported. “Missing data interferes with ability to compare effects.”
However, Fiero did say, “We know that PRO data shows higher incidence and higher severity” of symptoms compared with clinician-recorded data, with fatigue being a prominent example. “It’s a real challenging decision” to decide how to gather and portray the totality of the data in a way that is both accurate but accessible and meaningful to patients, she said.
Going forward, Halling said she could envision the data visualizations within Project Patient Voice including a broader snapshot of the patient experience than just PRO side effects. Other patient reported instruments that could be included, she said, could include those capturing health-related quality of life, physical functioning, outcomes in key lung cancer symptoms, as well as patient-priority symptoms that improved and remained stable.
Astra Zeneca’s Halling wrapped up by saying, “To make it short, I think this is a really great milestone in patient- focused drug development,” adding that the opportunity to test a variety of visualizations on the Project Patient Voice website will give industry and FDA more information about how to make data visualizations both accurate and usable.
“The visualizations are about right, but there's room for improvement,” said Halling. The aim, she said, is to help set patient expectations and to facilitate discussions between patients and their health care providers.
Regarding the 28-item questionnaire that was used in the AURA study, she said, “so actually we knew that we were more inclusive than exclusive…. There is always the need to look at the burden as well, but I think it's a delicate question. We clearly think we could reduce the number and still represent the experience in a good way.”
Kluetz added that including a free text question in a patient-reported outcomes questionnaire can “give us a little bit of a mitigation strategy” and an opportunity to capture any patient reported symptoms that were not included in an abbreviated questionnaire.


© 2022 Regulatory Affairs Professionals Society.

Discover more of what matters to you

No taxonomy