Genome Study of Aggressive Breast Cancer Yields Clues to Metastasis
To study how cancer cells change as the disease progresses, researchers have carried out a genome analysis of four DNA samples from a woman with breast cancer whose disease spread to her brain. In addition to the primary breast and metastatic brain tumors, the researchers used the patient’s normal blood cells and a tumor grown in a mouse from a sample of the original tumor (a xenograft).
A four-way comparison of the sequence and genome structure data has opened a window into the metastatic process in a single patient. The metastatic and xenograft tumors (i.e., the secondary tumors) both appeared to arise from a minority of cells in the patient’s breast tumor, researchers from the Washington University School of Medicine in St. Louis and their colleagues reported in the April 15 Nature.
The patient was a 44-year-old African American woman with inflammatory breast cancer that had the hallmarks of the basal-like subtype of breast cancer. These are aggressive tumors that disproportionately affect younger women and African Americans. Despite chemotherapy and radiation therapy, the patient died within a year of diagnosis, and her case was not unusual, the researchers noted.
To investigate the genetic changes underlying this metastatic process, a team led by Washington University’s Drs. Li Ding and Matthew Ellis profiled genetic mutations, structural changes, and differences in the number of gene copies in the four samples. Their analysis revealed 48 genetic mutations that were common to all three tumors (primary, metastatic, and xenograft), a number of structural alterations, including some large deletions, and about 7 translocations that occurred between chromosomes in the primary tumor alone, confirming previous reports that basal-like breast cancers have unstable genomes.
While additional genetic mutations and other changes did occur over the clinical course of the disease, most of the original mutations and structural variants in the breast tumor were present in the metastatic brain tumors and in the xenograft.
Tour de Force
“This study is a true tour de force in genomics,” said Dr. Patricia Steeg, who heads the Women’s Cancers Section in NCI’s Laboratory of Molecular Pharmacology and was not involved in the study. “Wouldn’t it be fascinating to see how a liver metastasis varies from a brain metastasis? Or how two metastases from the same organ do or don’t vary?”
This was only the second study to use whole-genome sequencing to compare a primary tumor and metastasis from the same patient. In the first study, published last year, the majority of mutations found in the metastasis had not been present in the primary tumor. But there were fundamental differences between the studies, including the fact that the metastatic process evolved over 9 years in the first study compared with less than a year in the current report.
“This emerging literature says that metastases are not identical to primary tumors,” noted Dr. Steeg, who is also president of the Metastasis Research Society. “They may be similar in many respects, but important differences remain. For most cancers, we are trying to develop treatments for metastatic disease, and I hope the new data will prompt more translational researchers to use metastatic models for experimental therapeutics.”
Because basal-like breast cancers have unstable genomes, researchers have wondered whether the deadly metastatic process is driven by mutations that occur after the tumor cells arrive at the distant site, or whether the primary tumor produces cells with the full complement of mutations required for metastatic growth. The findings support the latter view, but the researchers stress that more comparisons are needed to confirm the results.
Some of those studies are underway, according to Dr. Elaine Mardis, co-director of The Genome Center at Washington University and the study’s senior author. Her team has obtained DNA samples from other patients with basal breast cancers that have spread to various parts of the body.
“Sequencing more samples could help us understand the genetic landscape of the basal-subtype tumor and also to identify commonly mutated genes,” said Dr. Mardis. “These genes would be of interest for developing targeted therapies for treating these very aggressive tumors.”
Basal breast cancers are heterogeneous, noted Dr. Steeg, and the patient in this study represents one rare form.
A Remarkable Puzzle
In an accompanying editorial, Dr. Joe Gray of the Lawrence Berkeley National Laboratory noted that the researchers obtained a remarkable result when they tested the prevalence of the mutant DNA sequences in the tumors. Because the next-generation DNA sequencing machines in the study measure mutations in individual strands of DNA, the authors were able to calculate the prevalence of mutations as a fraction of the sequences at each genomic region carrying a mutation.
This analysis revealed that 16 mutations were present in almost every metastatic cancer cell in the patient and in almost every xenograft tumor cell in the mouse, although the mutations were found in a much smaller proportion of cells in the primary tumor. This finding “suggests that similar evolutionary pressures are at work on these cells in both environments,” Dr. Gray wrote.
The breast cancer cells used to seed the mouse tumor were collected before the patient received any therapy and therefore had not been affected by exposure to therapies. “The natural histories of the xenograft tumor and the metastasis were obviously quite different, yet in both tumors we see these same 16 genes rising to prevalence,” said Dr. Mardis. “This is a very interesting finding.”
The similarity of the brain metastasis specimen and the xenograft is a puzzle, said Dr. Steeg. “I am not sure, however, that these data are representative of long-term cell lines, which are the mainstay of our experimental therapeutics research,” she added.
Dr. Gray noted that further work is needed to determine which genetic changes in the tumors are driving the disease and which ones are merely present.
This study had several firsts—the first published genome of an African American and the first basal-like breast cancer genome sequence. But it will likely be one of the last genome studies by these authors to describe a single patient.
“We are rapidly transitioning from doing tumor-normal genome comparisons, and even quartet studies like this one, to multi-tumor studies and analyses,” Dr. Mardis said. The group has sequenced 50 genomes from patients with acute myeloid leukemia and a similar number of breast cancers. They are also sequencing the genomes of 600 patients with pediatric cancers, in collaboration with St. Jude Children’s Research Hospital.
An unexpected bonus from this research was considerable information about the genetics of the immune-deficient mice used in the study, which came about because mouse cells were inadvertently sequenced along with the human cells from the xenograft tumors. The researchers are planning to make this information publicly available.
—Edward R. Winstead