NEW YORK (GenomeWeb) – In a paper appearing online this week in the Proceedings of the National Academy of Sciences, Yale University researchers discussed the phylogenetic approach they took to unraveling the origins and evolution of metastatic tumor lineages in more than a dozen cancer types.
The team used exome sequencing to assess primary tumor samples, metastatic tumor samples, and matched normal tissue samples from 40 individuals with cancer. Based on these samples, which represented 13 cancer types, the group concluded that metastases tend to involve lineages that follow branched rather than linear evolution — including metastatic lineages that branch off early on in primary tumor development.
"We show that metastatic lineages separate very early on in the development of the cancer," senior author Jeffrey Townsend, a biostatistics researcher at the Yale School of Public Health, said in a statement.
"The major known cancer driver genes are often mutated in the primary tumor and all metastases," Townsend added, "and successfully targeting them in therapy could provide widespread therapeutic benefit."
The researchers focused on sequence divergence patterns in primary and metastatic tumor samples in an effort to tease apart heterogeneous tumor clones and their phylogenetic relationships with one another.
More than half of the participants included in the study were affected by lung cancer, the researchers noted, and another seven individuals had pancreatic cancer.
All told, the team analyzed samples from individuals with 13 cancer types, using NimbleGen SeqCap kits to capture protein-coding sequences from each sample that were subsequently sequenced to an average depth of almost 200-fold on an Illumina HiSeq instrument.
Based on exome sequences from 32 primary tumors and 139 metastatic samples — representing between two and seven metastases for each of the 40 cancer patients included in the analysis — the team identified some 20 to 5,370 somatic mutations per person.
After putting together a multiple sequence alignment for each individual, the researchers took a look at tumor phylogenies, which were calibrated with help from clinical clues such as individuals' diagnosis, treatment, and metastasis times as well as estimated cell division times for each tumor type.
From the patterns detected in the two dozen best characterized cases, the team concluded that metastases do not generally arise through rare mutations in a single lineage that advance in a stepwise manner. Instead, the study's author explained, "[o]ur phylogenies clearly demonstrate a branched evolution model of tumorigenesis and metastasis."
The researchers found further evidence for branched evolution, with metastatic and primary tumor lineages arising and evolving side by side, in eight more phylogenies from cases with trickier-to-assign primary tumor lineages.
Still, their findings suggest that some lineages are more prone to metastases than others, including some lineages cropping up early in the tumor formation process, prompting more detailed analyses of so-called internodes between lineages and the order in which driver gene mutations arise.
"[T]his evolutionary approach provides a method that accurately characterizes which genes are being mutated early and late," Townsend said. "That is extremely useful in terms of prioritizing which mutations should be targeted in order to produce therapies that are going to work."