Cancer: Signatures of structural genomic variation in cancer

Compilations of signatures of chromosomal instability — a hallmark of cancer — in thousands of human tumours are presented in two Nature papers. These resources could help future investigations of genomic alterations in cancer and may guide treatment options.

Chromosomal instability, which leads to losses, gains and re-arrangements of DNA, is a typical feature of cancer. However, a reference framework to define signatures of such genomic variability and the potential associations with multiple types of cancer has yet to be developed. Isolated genetic alterations have been previously used to develop mutational signatures that correlate with cancer origin, exposure and progression; however, there was not a framework to interpret patterns of structural variation in cancer genomes in a similar way.

Florian Markowetz, Geoffrey Macintyre and colleagues investigated patterns of chromosomal instability across 7,880 tumours, representing 33 types of cancer, such as liver and lung cancer, from The Cancer Genome Atlas (TCGA). Using these data, the authors characterized 17 types of chromosomal instability, based on variations in the copy number of DNA (the number of repetitions of given sequences or regions of DNA). These chromosomal instability signatures were able to both predict how tumours might respond to drugs but also help in the identification of future therapeutic targets.

In an independent study using the TCGA dataset, Nischalan Pillay and colleagues reveal 21 signatures of chromosomal instability in human cancers, after examination of 9,873 cancers of 33 different types, such as breast and ovarian cancer. These signatures were capable of explaining patterns in copy number in 97% of the samples studied. The authors reveal that these identified signatures could be clinically relevant with regard to assessing patient prognosis, which suggests that their consideration might improve the accuracy of existing tests. The authors found no associations between the chromosomal instability signatures that they identified and many known external risk factors for cancer, such as smoking or drinking alcohol. Thus, these structural signatures may reveal different, complementary etiological processes related to tumorigenesis.

The authors conclude that these results may advance our understanding of the origin and diversity of the chromosomal variations that underpin many human cancers. As such, these data could represent a systematic framework to guide future research.