Mutational signatures

Cancer genomes are highly unstable and, as they evolve, they acquire thousands of somatic (tumour) mutations. Some of these mutations are ‘driver’ mutations, which confer a growth and survival advantage to the cell, thereby promoting the process of oncogenesis. Most somatic variants are ‘passenger’ mutations, which do not confer any growth advantage to the cell but accumulate as a consequence of the inherent genomic instability of cancer cells.

While a cancer genome may appear totally chaotic, in fact it represents an overlay of non-random patterns, characteristic of the underlying mutational processes. With the advent of widespread cancer exome and genome sequencing, there has been much interest in trying to unpick these patterns, to distinguish specific ‘mutational signatures’ associated with particular underlying mutational processes. A mutational signature is the ‘molecular footprint’ of a tumour.

These mutational signatures may be associated with external environmental exposures, for example that typically seen in smokers with lung cancer. Or they may be associated with internal biological processes such as defective DNA repair; for example, the homologous recombination deficiency signature seen in carriers of pathogenic constitutional (germline) variants in BRCA1 or BRCA2. The basis of some signatures has yet to be determined.

Collaborative initiatives such as the Pan-Cancer Analysis of Whole Genomes and The Cancer Genome Atlas (TCGA) have analysed thousands of cancer whole genome and exome sequences, enabling the discovery of characteristic mutational signatures across a wide variety of tumour types. For a full list of mutational signatures and their cancer associations, see COSMIC (Catalogue of Somatic Mutations in Cancer).

Such studies have shown that certain signatures may be shared between tumours of different histological type if they share an underlying aetiology. For example, breast and ovarian cancers may both demonstrate the homologous recombination deficiency signature if they occur in carriers of a constitutional (germline) pathogenic BRCA1 variant.

Conversely, tumours of the same histopathological subtype may demonstrate different mutational signatures if they have different underlying aetiologies.

At present, mutational signatures are largely used in the research setting, and this type of analysis is not readily available in the clinical setting.

Mutational signature analysis of tumour exome/genome sequencing data can be used to provide information about an individual cancer’s mutational mechanisms. There is increasing interest in using this information to guide diagnosis, prognosis and therapeutics.

Identifying mutational signatures from prior exposures from environmental toxins (such as smoking and UV light) is of great interest from a public health perspective, in quantifying the relative cancer risk of potential genotoxins. From an individual patient’s perspective, however, more pertinent is the identification of potentially therapeutically targetable pathways in their cancer.

Watch the video to hear specialist registrar and clinical research fellow Dr Alison Berner talks about how analysis of mutational signatures in tumour DNA is bringing better outcomes for patients.

This film was created in collaboration with the Royal College of Physicians.

Determining causality

Assessment of the underlying mutation signature in a cancer type may be useful in variant interpretation, or in determining the contribution of a particular variant to cancer development. For example, tumours associated with a pathogenic variant in both copies of the NTHL1 gene demonstrate a characteristic signature.

If a tumour is identified in an individual who has been found to have a single heterozygous variant in NTHL1, but the tumour demonstrates the NTHL1-signature, this indicates that the single identified variant has contributed to the patient phenotype – and that a second variant ‘hit’ on the other allele may be, as yet, undetected.

Assessment of homologous recombination repair

Homologous recombination is essential for repairing double-stranded breaks in DNA. If these breaks are not repaired, they can contribute to the development of cancer.

Mutational signature analysis can be used to identify tumours that are homologous recombination (HR) deficient, as a consequence of constitutional (germline) or somatically (tumour) acquired pathogenic variants in BRCA1, BRCA2 or other homologous recombination repair (HRR) genes.

Such tumours have been found to demonstrate increased sensitivity to treatment with poly-ADP-ribose (PARP) inhibitors. A mutational-signature-based algorithm has been successfully used to predict BRCA1 and BRCA2 deficiency in breast cancer, and to identify breast cancers that are HR-deficient which, regardless of BRCA status, may benefit from targeted PARP inhibition.

Individuals with constitutional (germline) variants in HRR genes are also at risk of developing sporadic cancers by chance, which may not demonstrate HR deficiency – and therefore may not demonstrate a response to targeted agents. Demonstrating a lack of the HR-deficient signature in a breast cancer occurring in a constitutional BRCA carrier favours a sporadic event, and may spare the patient the potential side-effects of a targeted agent from which they are unlikely to derive benefit.

Assessment of MMR deficiency

Mismatch repair (MMR) is another DNA repair mechanism: it repairs small base changes, insertions and deletions. Many of these changes include the repetition of a small number of bases. These repeats are called microsatellites. Not all microsatellites are pathogenic, but when they become unstable with many additional repeats added, this can become problematic.

Mutational signature analysis can be used in cancer to identify tumours that are MMR-deficient, either owing to an underlying constitutional (germline) or somatically (tumour) acquired pathogenic variant in an MMR gene. This is of potential use in guiding treatment options, since MMR-deficient tumours demonstrate increased sensitivity to immune checkpoint inhibitors.

This may be particularly useful in those tumours for which microsatellite instability (MSI) testing is not fully sensitive (such as with endometrial cancers), if assessment of MMR proficiency by immunohistochemistry (IHC) is not possible or inconclusive.

• Mutational signatures reflect the underlying mutational process in a tumour.
• Mutational signatures may be associated with external environmental exposures or internal biological processes such as defective DNA repair.
• Tumours of different types may share mutational signatures if they have the same underlying aetiology, while tumours of the same type may have different mutational signatures because of different aetiology.
• Mutational signature analysis is not yet readily available in clinical practice, but it is an active area of research.

For clinicians

• COSMIC (Catalogue of Somatic Mutations in Cancer): Signatures of Mutational Processes in Human Cancer
• US National Institutes of Health: Cancer Genome Atlas Program
• Wellcome Sanger Institute: Cancer’s origins revealed

References:

• Alexandrov L, Nik-Zainal S, Wedge D and others. ‘Signatures of mutational processes in human cancer’. Nature 2013: issue 500, pages 415–421. DOI: 10.1038/nature12477
• Alexandrov LB, Kim J, Haradhvala NJ and others. ‘The repertoire of mutational signatures in human cancer’. Nature 2020: issue 578, pages 94–101. DOI: 10.1038/s41586-020-1943-3
• Campbell PJ, Getz G, Korbel JO and others. ‘Pan-cancer analysis of whole genomes’. Nature 2020: issue 578, pages 82–93. DOI: 10.1038/s41586-020-1969-6
• Chopra N, Tovey H, Pearson A and others. ‘Homologous recombination DNA repair deficiency and PARP inhibition activity in primary triple negative breast cancer’. Nature Communications 2020: issue 11, page 2,662. DOI: 10.1038/s41467-020-16142-7
• Davies H, Glodzik D, Morganella S and others. ‘HRDetect is a predictor of BRCA1 and BRCA2 deficiency based on mutational signatures’. Nature Medicine 2017: issue 23, pages 517–525. DOI: 10.1038/nm.4292
• Sahin IH, Akce M, Alese O and others. ‘Immune checkpoint inhibitors for the treatment of MSI-H/MMR-D colorectal cancer and a perspective on resistance mechanisms’. British Journal of Cancer 2019: issue 121, pages 809–818. DOI: 10.1038/s41416-019-0599-y