Eight DNA Signatures May Help Identify Aggressive Prostate Cancer

A large international prostate-cancer sequencing study has identified eight recurring patterns of DNA damage that together explain about 85% of cases examined. Rather than focusing only on individual cancer genes, the work looks across the genome for the molecular “fingerprints” left by processes that damage DNA or fail to maintain it. Four of the eight patterns were associated with aggressive tumors more likely to spread. 1011

What the eight signatures represent

The researchers sequenced the complete genomes of 959 prostate tumors from men in seven countries. Their analysis combined several kinds of genomic alteration, including point mutations, small insertions and deletions, structural rearrangements, and copy-number changes. 10

This integrated approach identified eight major mutational processes. These are not eight mutually exclusive prostate-cancer subtypes: a tumor can contain a mixture of processes, with their relative contribution varying from patient to patient. Mutational signatures are genomic patterns that can reflect DNA damage, DNA repair, and DNA replication mechanisms. 16

In this study, much of the observed damage was linked to malfunctioning cellular systems that copy DNA, repair it, or respond to androgen signaling in prostate tissue. 10

Four signatures were linked to metastatic risk

Four of the eight integrated signatures were associated with more aggressive disease and a greater likelihood that cancer would spread beyond the prostate. 911

That matters because prostate cancer has a wide range of outcomes. Some tumors remain slow-growing and may be suitable for monitoring, while others need earlier or more intensive treatment. A genome-wide signature profile could eventually add biological information to established clinical assessment, helping distinguish cancers that are less likely to cause harm from those with a higher metastatic risk.

The evidence supports a risk-assessment opportunity, not a replacement for existing diagnosis or staging. The signatures must first prove reliable in additional patient groups and in real clinical decision-making.

Replication errors were a major finding

One of the study’s notable findings was the substantial contribution of DNA-copying errors. Reports on the research indicate that replication-related processes contributed to roughly one-third of tumors—an unexpectedly large role compared with more familiar external mutagens. 10

This supports a broader view of prostate-cancer development: much of its genomic damage may arise from internal failures in the machinery that replicates and preserves DNA, rather than from a single external exposure. The general principle is well established across cancer genomics: each mutational process can leave a characteristic pattern in the tumor genome. 16

Age, ancestry, and individual risk

The study also reported variation in signature patterns by patient characteristics, including age and ancestry. Such findings may help researchers understand biological differences across populations and design more representative genomic studies.

They should not be interpreted as deterministic predictions for an individual patient. Earlier work has described prostate-cancer molecular subtypes enriched in African populations and linked some subtypes to poorer outcomes, but those findings are distinct from this study’s eight-signature framework.

The immediate implication is the need for diverse validation cohorts, particularly before any signature-based tool is used to assess risk across different populations.

Could signatures help choose treatment?

Genome-wide signatures may eventually do more than estimate prognosis. The study generated preliminary evidence that one active mutational process could be associated with differing responses to chemotherapy and androgen-directed hormone therapy.

That possibility is clinically important: if a tumor’s DNA-damage profile can predict which treatment is more likely to work, clinicians could potentially avoid ineffective therapy and better tailor care. But the treatment-response observation was retrospective and based on a small group, not a prospective trial. It is therefore a research lead—not evidence that patients should choose taxane chemotherapy or hormone treatment based on these signatures today.

Why whole-genome sequencing is central

Conventional tumor testing may look for selected mutations or a limited set of biomarkers. The new approach uses whole-genome sequencing to capture multiple forms of genomic damage at once, then interprets their combined patterns.

Whole-genome analyses have shown that mutational processes can be decomposed into signatures associated with distinct DNA damage, repair, and replication mechanisms. 16 Applying that principle to prostate cancer could make existing sequencing data more informative, provided the analysis can be made standardized, reproducible, and practical for clinical laboratories.

What needs to happen before clinical adoption

The findings are promising, but they do not establish a ready-to-use diagnostic or treatment-selection test. Several steps remain:

• Independent replication: The eight signatures and their links to metastatic risk need confirmation in additional, diverse cohorts.
• A standardized assay: Researchers need a reproducible method that can generate comparable results across laboratories and health systems.
• Prospective clinical trials: The critical test is whether acting on signature results improves treatment choices, reduces unnecessary treatment, or improves patient outcomes.
• Careful population validation: Studies must be sufficiently representative to assess how age, ancestry, and other patient factors affect performance.

For now, the study offers a more detailed biological map of prostate cancer. Its key contribution is showing that the balance of DNA-damage processes across a tumor genome may carry information about future behavior—information that could eventually help separate lower-risk disease from cancers requiring earlier, more personalized intervention. 1011