Meredith Salisbury
August 2025—Detection of extrachromosomal DNA in a tumor was found in a study published late last year to be associated with tumor stage, more prevalent after targeted and cytotoxic therapy, and associated with metastases and shorter overall survival (Bailey C, et al. Nature.2024;635[8037]:193–200).
Paul Mischel, MD, coauthor of the study, reported the results in April at the American Association for Cancer Research meeting and spoke with CAP TODAY in June. He is professor and vice chair for research in the Department of Pathology at Stanford University and an institute scholar in Sarafan ChEM-H, which is Stanford’s innovation hub for chemistry, engineering, and medicine for human health.
Dr. Mischel started his own laboratory in 1998, “just at the time of the sequencing of the human genome,” he notes, where he was studying why some cancer patients were not improving with therapy. “What were we missing? It turns out what we were missing was the extrachromosomal DNA,” he says. The tumors seemed to be able to change their oncogenes quickly, he adds, “and the inheritance pattern didn’t seem like classic Mendelian genetics.”
“That’s the work that took us to extrachromosomal DNA. Once you see it,” he says, “it’s so remarkably simple. It explains these mysteries . . . about how cancer can change its genome so quickly.”
He and colleagues have now been studying ecDNA for over a decade.
Individual extrachromosomal DNAs are large, circular DNA particles that can be found in the nuclei of many cancer cells. They were first observed in the 1960s, when the analytical tools were limited. Next-generation sequencing technologies provide “a completely different view of how to study cancer,” he notes.
When Dr. Mischel started his work on ecDNA in 2014, tumors containing ecDNA were thought to be rare—“1.4 percent of all cancers, something that had been seen but not really understood,” he says. But it’s closer to 20 percent of cancer patients who are affected. “Those are the patients who do worse than all other patients,” he says, calling it “an area of huge unmet need.”
In the study published in Nature last November, he and a large team of collaborators analyzed data from 14,778 patients with 39 tumor types from the United Kingdom’s 100,000 Genomes Project. They found that 17.1 percent of tumor samples contain ecDNA.
“ecDNA presents a complex challenge,” Dr. Mischel and coauthors write. “Its non-chromosomal inheritance drives intratumoural genetic heterogeneity fuelling accelerated evolution, thereby enabling tumours to resist treatment.”
“We’ve learned,” Dr. Mischel says, “that it can arise very early or very late,” and its presence may prove to be the tipping point between precancer and cancer.
In trying to understand when ecDNA can arise during tumor formation, Dr. Mischel and colleagues Vineet Bafna, PhD, and Jens Luebeck, PhD, studied patients with Barrett’s esophagus who did or did not develop esophageal adenocarcinoma. They and Rebecca Fitzgerald, MD, of Cambridge University and Thomas Paulson, PhD, and Patricia Galipeau of Fred Hutchinson Cancer Center examined whole genome sequencing data from biopsies from various locations in the esophagus, and at various time points in the patients who had serial biopsies. They found that ecDNA could arise during high-grade dysplasia, always in the context of p53 loss, and that it was strongly associated with the transition to cancer (Luebeck J, et al. Nature. 2023;616[7958]:798–805).
“We think this [ecDNA] is going to be important for diagnosis and intervention and hopefully even prevention, and that might be relevant to other cancer types,” Dr. Mischel says.
In the more recently published study, 39 solid and hematological tumor subtypes from the 14,778 patients (15,832 samples) were classified, and focal amplifications were quantified. Focal amplifications were defined as regions of the genome between 50 kb and 20 Mb in size, with a minimum copy number of 4.5 and twice the estimated ploidy of the tumor. A total of 4,716 unique ecDNAs were identified from 2,532 ecDNA-positive tumors.
The amplifications detected in the 17.1 percent of tumor samples varied in frequency by cancer type. ecDNA was detected in 54.9 percent of liposarcomas, 49.1 percent of glioblastomas, 46.4 percent of HER2-positive breast cancer, 37.9 percent of upper gastrointestinal adenocarcinomas, 22.4 percent of lung squamous cell carcinomas, 24.6 percent of bladder cancers, and 20.4 percent of ovarian cancers. Some tumor types had a low prevalence of ecDNA, among them oligodendrogliomas, in which no ecDNA was detected in 57 samples. Prostate cancers had low rates of ecDNA before treatment but higher rates as treatment progressed. Thyroid cancers maintained low frequencies of ecDNA.
Dr. Mischel and coauthors report that the amplification of specific oncogenes varied greatly by tissue type.
The authors say one of the “unanticipated results” is the high level of ecDNA found in HER2-positive breast cancer (39.3–53.7 percent), including amplification of ERBB2 on ecDNA in 26 percent of HER2-positive breast cancers. They write, “Given the known role of ecDNA in driving intercellular heterogeneity, the demonstration that increased HER2+ copy-number heterogeneity is associated with shorter disease-free survival is notable.”
The work in this study was done on the cancer samples of adult patients. Dr. Mischel and his team and other researchers in the field are working now to map ecDNA frequencies in pediatric cancers. “We already know it affects children too,” he says.
Dr. Mischel and his team in 2022 were awarded $25 million to investigate ecDNA, part of the Cancer Grand Challenges initiative of the National Cancer Institute and Cancer Research UK. Two weeks before he spoke with CAP TODAY, Dr. Mischel attended a Cancer Grand Challenges ecDNA conference of the Royal Society in London. “It was packed,” he says. “Brilliant scientists from all across the world, all people with tools, ideas, concepts to add. We’re all working to reduce any barriers to entry to the field.”
Ultimately, it’s pathologists who will be detecting ecDNA and by whom many advances will be made, he notes. “This is where the action is going to happen that makes it possible to identify patients.” For now, he says, an awareness of ecDNA is important.
“It’s been hiding inside the nucleus of cancer cells in plain sight,” he says.
Meredith Salisbury is a writer in the New York City area.