NEW YORK – A UK research team has detected a slew of new mutational signatures by analyzing tumor and matched normal sequence data representing a wide range of cancer types profiled through the Genomics England 100,000 Genomes Project (100kGP).
As they reported in Science on Thursday, investigators with whole-genome sequence data for more than 15,800 matched 100kGP tumor and normal samples profiled from patients enrolled through the UK's National Health Service Genomic Medicine Centres, focused in on sequences from 12,222 of the prospectively collected Genomics England tumors from 11,585 participants with 19 cancer types.
They also considered genome sequence data for another 3,001 cases profiled through the International Cancer Genome Consortium and 3,417 metastatic cancer cases assessed by the Hartwig Medical Foundation.
With the help of a bioinformatics pipeline for identifying somatic mutations, the team tracked down and analyzed hundreds of millions of DNA sequence substitutions, small insertions or deletions, and rearrangements in tumors from each of the affected sites considered, uncovering 82 single-base substitution mutational signatures and 27 double-base substitution signatures with a range of frequencies in the tumors.
"With thousands of mutations per cancer, we have unprecedented power to look for commonalities and differences across NHS patients, and in doing so we uncovered 58 new mutational signatures and broadened our knowledge of cancer," first author Andrea Degasperi, a medical genetics researcher at the University of Cambridge School of Clinical Medicine and its Early Cancer Institute, said in a statement.
These included previously unreported 40 mutational signatures marked by single-base substitutions, the investigators explained, along with 18 new mutational signatures that involved double-base substitutions. Together, these mutational signatures reflected everything from past exposures to mutagens such as smoking, ultraviolet light, platinum chemotherapy, or aristolochic acid to endogenous errors in DNA repair or other cellular processes.
The mutational signatures also provided a look at the common and much rarer signatures that reflect a tumor's tissue of origin. While a core set of common mutational signatures turned up across the cohorts considered, the team noted that increasing sample size boosted the detection of relatively rare mutational signatures.
In addition, the mutational signatures offered insights that were specific to a tumor's site of origin, despite some shared mutational contributors. In 16 tissue types represented across the 100kGP, ICGC, and Hartwig Medical Foundation cohorts, the authors noted that "signatures from the same organ in different cohorts were more similar to each other than to those in other tissue types, providing reassuring evidence that mutational signatures in each organ are highly reproducible, have tissue specificities, and were detectable regardless of sequencing platform or mutation-calling algorithm."
They suggested the ability to detect mutational signatures may help to develop yet-untapped treatment targets, therapeutic strategies, and disease management options, prompting the development of a new "Signature Fit Multi-Step" (FitMS) algorithm for reliably uncovering mutational signatures in tumors — an approach tested in 100 simulated genome sequences containing several common or rare single-base substitution signatures identified in breast cancer.
The study's senior author Serena Nik-Zainal, a genomic medicine and bioinformatics researcher affiliated with the University of Cambridge and Cambridge University Hospitals, noted that mutational signatures "can highlight abnormalities that may be targeted with specific drugs or may indicate a potential 'Achilles' heel' in individual cancers."