Eight genetic ‘fingerprints’ may help predict prostate cancer metastasis

Eight genetic “fingerprints” that may help predict whether prostate cancer will metastasize have been identified. The fingerprints were found to explain 85% of the mutational processes in primary prostate cancer.

According to a study published on the 28th in the international journal Nature, researchers from the University of Copenhagen in Denmark and other institutions analyzed tumor genomes provided by 959 donors in the Pan-Prostate Cancer Group.

Rather than looking only at known cancer genes, the researchers systematically dissected the mutational processes that shaped the entire genome. They defined eight “integrated mutational fingerprints” (IMFs), combining single-nucleotide substitutions, insertions and deletions, and copy-number features with six features of complex structural variation.

The fingerprints tended to occur disproportionately in specific regions of the genome. The most prominent were mutations mediated by testosterone, a male sex hormone, and errors during DNA replication.

The results showed that damage arising during DNA replication was responsible for about one-third of tumors. The researchers said this could offer a biological clue to why prostate cancer is common.

Clinically, four of the eight fingerprints were found in 37% of primary tumors and were significantly associated with a shorter time to metastasis. These four included mutations induced by reactive oxygen species and classical and nonclassical homologous recombination defects.

Nonclassical homologous recombination defects were more common in patients of African ancestry. The researchers said the fingerprints could reveal the causes of tumors and also be used to classify risk.

The researchers extended their analysis to metastatic prostate cancer and also found that the fingerprints could predict patients’ responses to androgen receptor pathway inhibitors. This means that the more active a particular error process is, the more likely the cancer is to spread.

However, the findings must be validated in other patient populations before they can be applied in clinical practice. The researchers expect the approach could be used for risk assessment within a few years, since the data and sequencing technologies it uses are already widely used in cancer diagnosis.