Johns Hopkins scientists have developed a blood test that looks for epigenetic instability in DNA, rather than fixed genetic changes, to flag early cancers. The approach could one day complement existing screening tools and help doctors decide who truly needs invasive follow-up tests.
A new blood test from Johns Hopkins researchers may offer a powerful way to catch cancers earlier by looking not at fixed genetic mutations, but at how chaotic the chemical marks on DNA have become.
The experimental test focuses on what the team calls epigenetic instability — random variation in DNA methylation patterns — and uses that signal to distinguish people with early-stage cancers from those without the disease.
In a proof-of-concept study, the approach accurately separated patients with early lung and breast cancers from healthy individuals, and also showed promise in several other cancer types. The findings were published Jan. 27 in Clinical Cancer Research and presented at the 2024 American Association for Cancer Research meeting.
The work marks a turning point in how scientists think about using DNA methylation as a cancer signal, according to lead author Hariharan Easwaran, an associate professor of oncology at the Johns Hopkins University School of Medicine.
“This is the first study where we are trying to really implement measuring that variation, or stochasticity, into a diagnostic tool,” Easwaran said in a news release. “We immediately found that measuring DNA methylation variation performs better than just measuring DNA methylation by itself.”
DNA methylation is a chemical tag that cells place on DNA to help control which genes are turned on or off. Many current liquid biopsy tests look for specific, consistent changes in methylation at particular spots in the genome that are associated with cancer.
Those tests, however, are usually built by studying relatively narrow groups of people — similar in age, race or disease stage — and often do not perform as well when applied to broader, more diverse populations.
The Johns Hopkins team took a different tack. Rather than asking whether a specific site is more or less methylated in cancer, they asked how unpredictable the methylation patterns are overall.
Cancer cells are known to have highly disordered epigenomes. The researchers reasoned that this disorder, captured as variation from one DNA fragment to another, might be a more universal hallmark of cancer than any single methylation change.
To test that idea, first author Sara-Jayne Thursby, a postdoctoral researcher in Easwaran’s lab, analyzed publicly available DNA methylation data from 2,084 samples across multiple cancer types. She searched for genomic regions, called CpG islands, that showed the most variability in methylation in cancers.
“We identified specific genomic regions that tend to be the most variable in DNA methylation marks during cancer,” Thursby said in the news release.
From that analysis, the team built a panel of 269 CpG islands that captured much of the methylation variability seen across cancers. Those regions form the backbone of their new metric, called the Epigenetic Instability Index, or EII.
The researchers then trained a machine learning model on this panel to tell apart cancer signals from healthy ones, and tested it using cross-validation techniques to guard against overfitting.
The results were striking. In lung adenocarcinoma, the EII distinguished stage 1A cancers with 81% sensitivity at 95% specificity. That means the test correctly flagged 81% of true cancers while keeping false alarms low. For early-stage breast cancer, the test reached about 68% sensitivity at the same high specificity.
The tool also picked up signals from colon, brain, pancreatic and prostate cancers, suggesting that epigenetic instability may be a common feature across many tumor types.
In healthy people, cell-free DNA — fragments of DNA that circulate in the bloodstream after cells die and break apart — tends to have relatively stable methylation patterns. Thursby explained that is exactly why measuring variability can be so revealing.
“In cell-free DNA in the blood, that variability shouldn’t be high, but if it is, it is indicative of a developing cancerous phenotype,” she said.
The team believes this instability may also be tied to how dangerous a lesion becomes.
“We hypothesize that early-stage tumors and precancerous lesions that exhibit high degrees of methylation variation, or epigenetic instability, may be more resistant to intrinsic cancer-protective mechanisms and progress more rapidly,” added co-lead author Thomas Pisanic, an associate research professor of oncology at the Johns Hopkins Institute for NanoBioTechnology.
Catching those changes early could give doctors a chance to intervene before tumors grow, spread or become harder to treat.
The group’s working model is that methylation patterns begin to drift even at the very start of cancer development.
“Our hypothesis is that during the earliest stages of cancer development, methylation starts shifting,” Easwaran added. “We can try to pick those signals using these stochasticity metrics, even of early cancer stages, as long as the DNA is shed in the blood.”
In practical terms, the researchers envision the Epigenetic Instability Index as a complement to existing screening tools, not a replacement. It could be paired with other liquid biopsy technologies, such as DELFI and DNA mutation-based assays, developed at Johns Hopkins.
One potential use would be as a secondary triaging measure in situations where standard screening tests often produce false positives. For example, a man with an elevated prostate-specific antigen (PSA) level might face an uncertain decision about whether to undergo an invasive biopsy. An additional blood test that measures epigenetic instability could help clarify whether a biopsy is truly warranted.
Beyond individual cases, a broadly applicable blood test that works across diverse populations could help expand access to early cancer detection. Many people skip or lack access to imaging-based screenings like mammograms or CT scans. A simple blood draw that can be repeated over time could fit more easily into routine care.
At the same time, the researchers caution that their work is still in the early stages. The current study is a proof of concept, and the EII will need to be validated in larger, long-term clinical cohorts before it can be used in practice.
The team is now working to refine the method, improve its performance and test it in more real-world settings.
Pisanic noted that as scientists learn to read these subtle shifts in the epigenome, they may gain a powerful new tool for prevention.
“By leveraging these metrics, we may be able to better identify and intercept tissues in the early stages of carcinogenesis,” he said.
If that promise holds up, a measure of chaos in our DNA’s chemical marks could one day bring more order — and more options — to cancer screening and care.
Source: Johns Hopkins Medicine