NEW YORK (GenomeWeb) – New research has found that a mutational signature associated with inherited mutations in the breast and ovarian cancer risk genes BRCA1 and BRCA2 may be more widespread than previously appreciated. This suggests that individuals with an appropriate mutational signature in their tumors, as well as individuals with specific BRCA1/2 mutations, may benefit from targeted PARP inhibitor treatments.
"[O]ur study shows that there are many more people who have cancers that look like they have the same signatures and same weakness as patients with faulty BRCA1 and BRCA2 genes. We should explore if they could also benefit from PARP inhibitors," senior author Serena Nik-Zainal, a researcher at the Wellcome Trust Sanger Institute, said in a statement. "The results suggest that clinical trials now need to look at cancer patients who share the same genetic signature in their cancer. This could change how clinical trials are designed in the future."
As they reported in Nature Medicine today, Nik-Zainal and her colleagues came up with a weighted mathematical model called HRDetect, based on genome sequences for breast cancer tumors, to recognize mutational signatures that stem from defects in homologous recombination-mediated double-strand break repair — a hallmark of tumors that form in individuals with risky BRCA1 and/or BRCA2 germline mutations.
"Unlike most biomarkers, these multiple mutational signatures are the direct consequence of abrogation of [double-strand break] repair pathways," the authors wrote. "Thus, in the current analysis, we exploit this observation to quantitatively define the genomic features of BRCA1/BRCA2 deficiency and present a [whole-genome sequencing]-based predictor with remarkable performance for detection of [homologous recombination] deficient tumors."
After demonstrating that HRDetect could identify BRCA1/2-related DNA repair deficient tumors nearly 99 percent of the time when tumor whole-genome sequences were available, the researchers used it to assess samples from 560 individuals with breast cancer, uncovering 22 cases involving inherited BRCA1/2 alterations and another 22 cases marked by somatic mutations in these genes. But 47 tumors seemed to share the same mutational signature, despite appearing BRCA1/2 mutation-free.
In follow-up analyses on samples from hundreds of other individuals with breast, ovarian, or pancreatic cancers, the team saw hints that the BRCA1/2 mutational signature may occur in as many as one-fifth of breast tumors. In contrast, germline mutations to BRCA1 or BRCA2 are typically present in just 1 percent to 5 percent of breast cancer cases.
Consequently, the team argued that there may be a benefit to testing the PARP inhibitors used to treat BRCA1/2-mutated tumors in a wider swath of breast cancer patients than are currently eligible for the treatment.
"This work uses mutational signatures to identify the complete set of cancers that will respond to certain drugs that are already known to be effective in a subset," co-author Michael Stratton, director of the Wellcome Trust Sanger Institute and CEO of the Wellcome Genome Campus, said in a statement. "To translate these results into treatments, further sequencing of cancer genomes and more clinical trials are urgently needed, but this is a most promising start."
The authors cautioned that HRDetect's sensitivity does drop off dramatically when applied to tumor exome sequences rather than whole-genome sequences, though they were able to retrain the model to bump up the sensitivity somewhat when exome sequences were used as input information. The approach also appeared amenable to testing both formalin fixed, paraffin-embedded and needle biopsy samples.