New, Inexpensive DNA Methylation Tool Maps Cancer Evolution and Predicts Outcomes

A new study published in Nature introduced a low-cost method that researchers say can track how cancers like chronic lymphocytic leukemia (CLL) evolve over time, offering insight into tumor growth, treatment resistance, and prognosis.¹

The method, known as EVOFLUx, uses natural fluctuations in DNA methylation as a kind of molecular “barcode” to reconstruct a tumor’s evolutionary history. By applying EVOFLUx to nearly 2000 samples of lymphoid cancers, researchers demonstrated that the approach can uncover critical insights into tumor dynamics and even predict patient outcomes

The technique focuses on fluctuating CpG sites (fCpGs), which are locations in the genome where DNA methylation switches on and off over years. These fluctuations leave patterns that reflect a tumor’s ancestry. By analyzing bulk methylation profiles, EVOFLUx can estimate how fast a cancer grew, when it began, and whether subclones emerged along the way.

EVOFLUx requires only bulk methylation data, already widely available in research and diagnostics, making it more accessible than genome sequencing. The study’s researchers believe this opens the door to large-scale clinical application.

“Usually, we study the evolution of cells from normal function to cancer using DNA mutations. These new methylation markers provide more information for a fraction of the cost since they accumulate faster,” said Diego Mallo, a researcher at Biodesign Center for Biocomputing, Security and Society at Arizona State University discussed in a press release.² “The fact that evolutionary parameters estimated using this method are strong predictors of the cancer’s outcome shows their power to improve both cancer management and monitoring patients at risk.”

The study analyzed samples from patients with acute lymphoblastic leukemia (ALL), CLL, mantle cell lymphoma (MCL), diffuse large B-cell lymphoma (DLBCL), and multiple myeloma.

Results revealed striking differences in tumor evolution. Pediatric ALL showed extremely rapid growth rates, shorter evolutionary timelines, and faster epigenetic switching compared to adult lymphoid cancers. Meanwhile, CLL displayed slower, more variable growth, with unmutated subtypes (U-CLL) expanding faster than mutated forms (M-CLL). These growth rates strongly predicted time to treatment.

MCL subtypes differed in aggressiveness, with conventional MCL growing significantly faster than its non-nodal, more indolent counterpart. In DLBCL, large effective tumor sizes were observed despite modest growth rates, underscoring unique evolutionary pressures in this aggressive lymphoma.

Importantly, EVOFLUx-derived growth rates proved to be independent prognostic markers in CLL. Patients with faster-growing cancers required treatment sooner and had worse overall survival, even when accounting for established markers such as TP53 mutations and IGHV status.

In one cohort of 478 patients, those with high inferred growth rates had nearly 4 times the risk of early treatment initiation compared to those with slower-growing disease. These findings were validated in a second independent cohort.

The method also shed light on Richter transformation, an aggressive shift that some CLLs undergo. By analyzing longitudinal samples, EVOFLUx revealed that the seeds of transformed clones were present decades before clinical diagnosis, suggesting new opportunities for early detection and monitoring.

Beyond lymphoid malignancies, the team expects the approach to apply broadly across cancer types. Since methylation profiling is inexpensive and already integrated into many labs, EVOFLUx could become a routine part of cancer workups in the future.

While promising, the study acknowledges limitations. EVOFLUx detects only strongly selected subclones and works best with high-quality methylation data.

References

1. Gabbutt C, Duran-Ferrer M, Grant HE, et al. Fluctuating DNA methylation tracks cancer evolution at clinical scale. Nature. Published online September 10, 2025. doi:10.1038/s41586-025-09374-4

2. Harth R. New cancer test can warn patients up to 10 years before treatment is needed. ASU News. Published online September 10, 2025. https://news.asu.edu/20250910-science-and-technology-new-cancer-test-can-warn-patients-10-years-treatment-needed