Multi-cancer early detection screening tests from a sample of blood (liquid biopsy) can detect cancer signals at very early stages, leading to the hope of earlier treatments that may be more effective and even curative. Currently, standard-of-care screening tests exist for only a limited number of cancers, and each is designed to detect only a single cancer type.
To consider a new path for earlier detection of cancers through screening, the Friends of Cancer Research (FOCR) released a new white paper, Multi-Cancer Early Detection Screening Tests: Considerations for Use of Real-World Data. The organization also held an online briefing to introduce concepts in the publication, which grew out of discussions among government officials, screening test developers in industry, academic researchers, and patient advocacy groups.
Currently, there is legislation in Congress which would cover multi-cancer early detection (MCED) screening tests for Medicare patients if they are approved by the FDA. The Medicare Multi-Cancer Early Detection Screening Coverage Act is a bill that would authorize the Centers for Medicare & Medicaid Services to evaluate and cover these blood-based multi-cancer screening tests, which are used to screen for cancer across many cancer types.
A policy insight article published in the March 2022 issue of the journal Health Affairs also highlights the use of systematically collected real-world data in evaluating the clinical validity, utility, and economic value of multi-cancer screening tests (2022; https://doi.org/10.1377/hlthaff.2021.01316).
“Everything in cancer is moving so much faster,” said Norman E. “Ned” Sharpless, MD, Director of the National Cancer Institute (NCI), during the briefing. He pointed out that the last 10 years have produced remarkable strides in progress against cancer based on new technologies. Because of the rapid pace of discovery, he noted that “while we may never eradicate cancer for all people,” he believes President Biden’s recent effort to relaunch and supercharge the Cancer Moonshot will reach its goal of reducing U.S. cancer mortality by at least 50 percent over the next 25 years.
But, Sharpless noted, “The truth is we still find many cancers too late.” One way to improve that outlook is to complement currently used single-organ screening tests with MCED screening tests, he said.
The NCI director praised FOCR for sponsoring the white paper and the online briefing, stating that it is vital to assess the benefits and risks of MCEDs accurately before they are used in clinical practice. While they hold great promise for earlier detection, Sharpless noted that MCEDs could also lead to overdiagnosis and false positives.
If used correctly and carefully, MCEDs could very well revolutionize cancer detection, said Ernest Hawk, MD, MPH, Vice President and Division Head for Cancer Prevention and Population Sciences at the University of Texas MD Anderson Cancer Center.
Data from randomized controlled trials (RCTs) are the gold standard for cancer survival, noted Ruth B. Etzioni, PhD, a member of the Biostatistics Program in the Public Health Sciences Division at the Fred Hutchinson Cancer Research Center and an affiliate professor at the University of Washington. The white paper discusses the limitations of trial surrogate endpoints. “Remember that screening can only be beneficial as part of a continuum of care,” said Etzioni.
The new white paper noted that the NCI is “evaluating the landscape of study designs and seeking to potentially launch a multi-arm, multi-stage pivotal RCT to evaluate multiple MCED screening tests in the years ahead.” The use of real-world evidence gathered from real-world data (RWD) generated in the intended use population could “help supplement data generated in non-RWD clinical screening studies and may be used to inform regulatory decisions,” the paper stated.
Powering studies for each cancer type in an RCT, especially rare cancers, is logistically challenging because it requires large enrollment numbers and one or more decades of follow-up to demonstrate a cancer-specific mortality benefit for individual cancer types. The FOCR white paper includes a figure on types of study designs for screening tests that incorporate real-world data. For example, RWD could serve as an external control arm, a case-control study, or a cohort study.
RWD that rests outside the trial site offers “a chance to cast a wider net,” said Sam Roosz, MBA, the CEO and co-founder of Crescendo Health. He said it is important to continue to plan for interoperability of data and “handling the messiness of data” in the development of MCEDs.
For example, how do test developers determine whether a negative result on an MCED is a true negative? The result from an MCED screening test should show a positive result in people who have cancer (sensitivity) while providing a negative result in people who do not have cancer (specificity).
Key questions for MCED development posed by the white paper include the following:
- Performance characteristics: How well does the MCED screening test detect cancer? How early does it detect cancer?
- Safety: What are the health burdens/harms of MCED screening tests, including the diagnostic confirmation process?
- Clinical outcomes and utility: How does an MCED screening test impact cancer outcomes?
The FDA is currently examining issues raised by MCEDs, said Wendy Rubinstein, MD, PhD, Director of Personalized Medicine at the FDA’s Center for Devices and Radiological Health. She noted that, in the short term, test developers will bring their MCEDs to FDA for assessment and approval, and there is a strong need to determine the tests’ benefits before they are rolled out for use in the general public.
“Unless we plan now, we won’t even have the answers 5 or 10 years from now,” Rubinstein said. She stressed the need for harmonized data collection standards for MCEDs as their development progresses.
She also noted that assessing the potential harms of MCEDs is just as important as assessing their benefits. For example, she said that a positive result from an MCED screening test could lead to an invasive diagnostic workup, causing a harmful outcome for the patient, such as pancreatitis.
“We need to think about MCEDs in a more nuanced manner,” said Seema Singh Bhan, JD, Vice President for Public Policy & External Affairs at Exact Sciences Corp. “There’s no clear pathway for MCEDs.”
She noted that today more than 70 percent of cancers are diagnosed at late stages, and single-organ screening tests are the norm, stressing that looking to future MCEDs will be most useful when they complement current screening tests. “It’s about a future where everyone can get an early diagnosis,” Bhan said.
MCEDs must have both clinical validity and clinical utility. And they must have something more: patient acceptance. From the patient’s point of view, MCEDs may be used in the future to identify at-risk people who could benefit from lifestyle and behavioral changes that reduce their risk of cancer, said Jody Hoyos, MHA, President and Chief Operating Officer of the Prevent Cancer Foundation. But she noted that how well patients accept MCEDs will be critical to how widely they are used in clinical practice.
FOCR President and founder Ellen Sigal, PhD, said she knows MCEDs are a complex topic, and the new white paper is just the start of the discussion on their potential value in diagnosing cancer earlier.
Talks with professionals contributing to the white paper highlighted the need for alignment on terminology used in MCED screening test development, validation, and evaluation. Another area that needs attention, highlighted in the white paper, “is the use of machine learning and artificial intelligence by many of the MCED screening tests to determine cancer status.”
The paper points out that real-world evidence “can enable real-world learning and evaluation of these technologies as they enter clinical practice, helping to achieve the goal of a learning health care system.” In that regard, there may be a role for real-world evidence in “periodic (post-market) re-evaluation of MCED screening tests utilizing machine learning to assess the real-world performance of initial and future versions of these tests,” the white paper stated.