The Friends of Cancer Research (Friends) held a conference in Washington, DC, to focus on new findings from two of its research projects, ai.RECIST and ctMoniTR.
“Today’s discussion is about one thing above all: patients,” said Ellen V. Sigal, PhD, Friends Founder and Chairperson. She stressed that patients need to know their tests are accurate and are based on rigorous science. She added that rigorous science leads to smart policy.
Keynote speaker Rep. Diana DeGette (D-CO) put the new research findings in context, noting that she worked on the 21st Century Cures Act, which passed in 2016 with the goal of accelerating cures for diseases, including cancer. DeGette said it is vitally important to invest in health agencies to maintain the accelerated pace supported by the Cures Act. Since 1991, 5 million lives have been saved from cancer, she noted. “This is what progress looks like,” DeGette said. On January 21, 2026, she introduced the Follow the Science Act, which aims to protect the National Institutes of Health (NIH) from political interference in biomedical research.
Before the meeting, Friends released a new white paper, “Leveraging AI-Enabled Tumor Assessment Tools on Radiological Images to Evaluate Treatment Effect and Support Clinical Trial Endpoints in Solid Tumors.” The white paper states that for more than 25 years, RECIST (Response Evaluation Criteria in Solid Tumors) has been used successfully for tumor measurement. However, the white paper notes that AI offers opportunities to improve upon RECIST.
Specifically, the paper states that “AI-enabled tumor assessment tools can detect, segment, track, and quantify tumors across the entire body with potentially greater consistency and improved biological detail.” AI-enabled approaches include enhanced RECIST, radiomics, volumetric analysis, and tumor growth kinetics.
“The role of all of us will change,” said panelist Lawrence H. Schwartz, MD, Chair of the Department of Radiology at Memorial Sloan Kettering Cancer Center (MSKCC), a member of the Quantitative Biomarker Alliance and a member of the Oncology Biomarker Qualification Initiative. Calling AI a “disruptive technology,” Schwartz said it will be disruptive in good ways, which will ultimately improve cancer care. He said a major challenge will be how AI-enabled tumor assessment tools are presented to oncologists. “It really will need to be a stepwise approach,” he said, with the goal of a future where validated AI-driven imaging biomarkers and AI-driven tools for tumor assessment are used consistently, transparently, and with trust.
“What we want is AI with humans, not without,” said panelist Lia Ridout, a lung cancer survivor and Friends advocate. Ridout said patients are generally supportive of AI as long as a human expert makes the decisions on cancer care.
Panelist Nathaniel Braman, PhD, Co-Founder and Vice President of AI Research & Development at Picture Health, agreed. All AI should be human-centered, with the human user who will interpret the results, he said. At the end of the day, the goal will be what AI-enabled tumor assessment can do for overall survival, he said. “We’re just at the beginning of thinking about these scans in terms of precision medicine,” he added.
Panelists at the Friends conference were also hopeful about the promise of circulating tumor DNA (ctDNA) to improve cancer care. Aggregate data from the Friends ctMoniTR Project show that decreased ctDNA levels are associated with improved overall survival, said Nevine Zariffa, MMath, Principal and Founder of the NMD Group, LLC. There are lots of data on ctDNA in the metastatic setting but fewer in the early setting, said Minetta Liu, MD, Chief Medical Officer of Oncology at Natura. “How do we know when we can finally say they’re cured?” she asked of patients whose ctDNA test results are excellent.
“Will a fall in ctDNA be sustained at future time endpoints?” asked Paz Vellanki, MD, PhD, Supervisory Associate Director in the Division of Oncology 2, Center for Drug Evaluation and Research (CDER) at the FDA. Although ctDNA is a promising biomarker, Vellanki noted that it is not yet validated for regulatory use. While the FDA is very interested in ctDNA as an early endpoint, “We’re not quite there yet,” she said, stressing that more data are needed.
Vellanki said one promising future use of a ctDNA assay may be for patients who have no radiographic evidence of disease.
In a session on regulation and policy, panelists said that it is difficult for regulators to keep up with the pace of innovation. “We are seeing a tsunami of innovation,” and it is “paradigm-changing,” said Michael Montaldo, PhD, Vice President of Precision Medicine, Global Development at Amgen, Inc. “I think that implies innovation on the policy side,” so that a regulatory bottleneck does not occur, said Montaldo, former Chief Scientific Officer at PathAI and a member of the board of the Journal of AI in Precision Oncology.
There is a need for people with multidisciplinary expertise in the regulatory space, emphasized panelist Tala H. Fakhouri, PhD, MPH. She noted that developers of these new biomedical technologies are often outside the traditional pharmaceutical arena. They may come from academia or from small, innovative technology companies.
In closing the conference, Friends President and CEO Jeff Allen, PhD, said, “These are works in progress.” He said Friends would keep the audience informed on ai.RECIST and ctMoniTR as new results become available.
Peggy Eastman is a contributing writer.