A pair of studies published last week in Cell highlight the clonal nature of cancer, adding further evidence that treatment should take into account not only dominant mutational clones, but also the subclones.
The studies also identified mutational patterns in different tumors, including a pattern specific to inherited forms of breast cancer.
Researchers from the Wellcome Trust Sanger Institute and elsewhere sequenced the whole genomes of 21 breast cancers. In one study, the researchers used the sequence data to piece together the evolution of the cancer, showing the progression of both a dominant clone and subclones. In a second study, published in the same Cell issue, the team evaluated the whole spectrum of mutations in each tumor, finding that different subtypes of breast cancer have different mutational patterns.
Understanding both the mutational patterns and the evolution of cancer, including the different subclones, could potentially have implications for treatment.
Currently, "we think and treat cancer in terms of a dominant clone, but if we understand its complexity and that there are processes coming on at different times, maybe we will be able to treat the cancers more effectively … and not just treat the dominant clone," Serena Nik-Zainal, the lead author of both papers and a clinical fellow at the Sanger Institute, told Clinical Sequencing News.
Additionally, Nik-Zainal said that the team identified eight or nine different mutational patterns. "If we can start to identify those patterns and understand why those patterns are there, then we might be able to develop more specific treatments to individuals' cancers."
The idea of clonal evolution in cancer — that a founding mutation or clone triggers the rapid multiplication of further cancer cells until a separate mutation enables divergence from that clone, creating a subclone — has been increasingly studied using next-generation sequencing.
Researchers from Washington University's Genome Institute and the BC Cancer Agency recently published separate studies about clonal evolution in acute myeloid leukemia (CSN 3/28/2012). Additionally, researchers from Cold Spring Harbor have tackled clonal evolution using single-cell sequencing, showing that different cells within the same tumor have distinct mutational profiles (IS 5/18/2010 and IS 5/15/2012).
The recent Cell studies add further evidence to the importance of clonal evolution in cancers, particularly when it comes to designing personalized cancer therapies.
To look at clonal evolution, the Sanger team sequenced the whole genomes of 21 primary breast cancer tumors on the Illumina HiSeq. They included four cases each of ER-positive, HER2-positive, and BRCA2-positive breast cancer; three cases of triple negative breast cancer, five cases of BRCA1-positive breast cancer; and one ER-positive tumor with a distinctive mutator phenotype.
Next, the team used newly developed bioinformatic algorithms to piece together the genomic history of the tumors by building phylogenic trees for each tumor, pinpointing where throughout the tumor's evolutionary process each mutation popped up.
The researchers identified a dominant clone in about half of all cases, said Nik-Zainal, and also found that each tumor had around three to four subclonal populations.
Importantly, in several cases, the researchers identified subclonal populations with known drug sensitivities, including one subclone that would be sensitive to Genentech's Herceptin (trastuzumab) and another that would be sensitive to tamoxifen. In such cases, they noted, it might make sense to use a combination of therapies — either a generalized chemotherapy or a drug targeted at the dominant clone, as well as Herceptin or tamoxifen to treat the subclone.
"If you treat the dominant clone and kill those cells, you've now changed the environment of the cancer … and a minor subclone could then become the dominant clone," Nik-Zainal said. But, if you knew there were multiple populations, and had a way to treat both at once, that could be a better strategy, she said.
In the second Cell study, the researchers examined the whole-genome sequence data of the same 21 tumors, but this time scoured the genomes for mutational patterns.
The team identified eight or nine distinct mutational patterns, said Nik-Zainal, including a pattern that was specific to cancers caused by germline mutations in the BRCA1 and 2 genes and distinguishable from the other cancers. "That's quite interesting to us because we know that the BRCA 1 and 2 cancers are sensitive to PARP inhibitors," she said.
If other cancers that are sensitive to PARP inhibitors display the same mutational pattern, the pattern itself could become a marker of drug sensitivity, she said. "This is the earliest steps to using mutational patterns to predict sensitivity to therapy," she added.
The team also identified a mutational pattern that was present in 13 out of the 21 cancers, which they named kataegis, the Greek word for a rain shower, because the mutations were all clustered together. The clusters were also predominantly C-to-T mutations, and they occurred in a TpC context. This pattern also coincided with rearrangements, "so it's as if something that is generating the substitutions is also generating the rearrangements, or vice versa," Nik-Zainal said.
She said the team's next step is to try and better understand the different patterns, what is causing them, and what they mean. "If we can understand that, we're in a better position to try and treat them."
Many more genomes will also have to be sequenced to validate the findings and to search for additional patterns.
Nik-Zainal said the team would also want to do whole-genome sequencing on tumor samples both before and after treatment to see how the drugs impact both the mutational pattern and the clonal populations.