Recent research have revealed a number of genes that accumulate somatic, or non-inherited, mutations and alterations in estrogen receptor positive (ER-positive) breast cancer. Although a few alterations are quite common and relatively well understood, many genes are mutated in less than 5 per cent of cases. As a result, little has been known about how uncommon mutations impact disease outcome.
To address this problem, researchers from UBC, Baylor College of Medicine, and Washington University in St. Louis conducted the largest analysis to date, involving more than 2,500 ER-positive breast cancer cases, to establish associations between rare somatic mutations and prognosis.
The study, published in Nature Communications, analyzed 625 samples from postmenopausal women and 328 from premenopausal women diagnosed over two decades ago, most of them patients who were treated by BC Cancer or through the Canadian Cancer Trials Group. The samples provided the opportunity to examine the relation between cancer mutations and the long-term risk of relapse.
Using DNA hybrid capture, deep sequencing of 83 genes, and rigorous bioinformatics analysis, the researchers were able to compensate for the old age of the samples and lack of matched normal DNA, which is normally required for studies of this kind.
“Most breast cancers are estrogen receptor positive, and are treated with hormone therapies like tamoxifen,” says Torsten Nielsen, Professor in the Department of Pathology and Laboratory Medicine, who directed Canada’s contribution to this international study. “However, these cancers have different underlying mutations, and new technologies now allow us to figure out which mutations are present and how they affect the chance of successful treatment.”
The research team found consistent associations with outcomes for two relatively common mutations in two tumor suppressor genes called MAP3K1 and TP53, which had opposing effects. MAP3K1 is associated with favorable outcomes, and TP53 with poor prognosis. These results were expected from previous studies.
However, early relapse and deaths were found to be associated with uncommon mutations in three genes not previously well-characterized in breast cancer: DDR1, a collagen receptor, PIK3R1, a signaling suppressor, and NF1, a negative regulator of the RAS oncogene.
The results were validated through analysis of another, independently published cohort of breast cancer specimens and genetic sequences, most of which had been previously generated by other UBC researchers, including Sam Aparicio and Sohrab Shah.
“Low frequency mutations, such as DDR1 and NF1, need to be paid attention to, because although these mutations are considered rare, they associate with early relapse and death, which makes them much more common in patients who unfortunately die from the disease and, thus, could be critical therapeutic targets. Their identification also gives us very important molecular clues into the nature of aggressive tumor behavior,” said Matthew Ellis, McNair Scholar and director of the Lester and Sue Smith Breast Center, part of the NCI-designated Dan L Duncan Comprehensive Cancer Center at Baylor.
“When we don’t cure any disease, it is because we have not made the right diagnosis at a fundamental molecular level. ER-positive breast cancer has a wide variety of outcomes and molecular characteristics, and consequently there simply is not a ‘one size fits all’ treatment solution. Ultimately, progress depends on a fundamental understanding of what is driving the cancer in each individual case, and tumor DNA sequencing is going to be part of that diagnostic puzzle,” Dr. Ellis added.
This work was supported by the Canadian Cancer Society Research Institute, the Susan G. Komen Foundation, the Cancer Research and Prevention Institute of Texas, the Department of Defense, the McNair Foundation, and the National Cancer Institute.