Most research on cancer therapies focuses on single genetic mutations. But chromosomal instability, which leads to losses, gains and rearrangements of DNA, represents another common feature of cancer that’s been less studied.
In two Nature papers, two international teams of scientists have profiled the patterns of chromosomal instability in thousands of human tumors and categorized them into what’s known as “copy number signatures.”
Because changes in the copy number of sequences or regions of DNA are prevalent in multiple cancer types, the teams argue their DNA structure categories could help scientists identify new cancer drug targets and allow doctors to adopt precise treatments for patients.
When making copies, chromosomes can become unstable as segments of DNA may be duplicated, deleted or reshuffled. Accumulation of chromosomal instability is a hallmark of cancer, occurring in about 80% of tumors. But characterizing all the complex genomic changes seen in chromosomal instability can be challenging.
In one Nature study, a team led by Cancer Research UK Cambridge Institute and the Spanish National Cancer Research Center examined chromosomal instability in 7,880 tumors across 33 tumor types from The Cancer Genome Atlas (TCGA).
Among the 6,335 samples that had detectable chromosomal instability, the scientists used computational tools to identify 17 different copy number signatures based on their patterns of chromosomal structural changes. These signatures were able to not only predict how tumors might respond to drugs but also to help identify potential drug targets.
The team linked the signatures to 49 novel druggable targets as potential avenues for future drug design. For example, one signature’s activity was found to be correlated with changes in ACTL6A and TERF1, both of which are required for normal chromosome separation during cell replication. Therefore, inhibiting the genes represents a promising strategy to kill off cancer, the scientists said.
The research has led the Cancer Research UK Cambridge Institute to form a spinout called Tailor Bio. The company aims to develop “pan-cancer precision therapeutics,” which are drugs for patient groups based on their genomic features.
In a separate Nature study that also utilized the TCGA database, a team of researchers led by the University College London and the University of California, San Diego built its own computational algorithm to group copy number signatures.
Based on 9,873 tumor samples representing 33 cancer types, the team identified 21 signatures. These signatures could explain patterns in copy number in 97% of the samples studied.
Signatures identified by both teams are not restricted to specific cancer types defined by location. In addition to their potential as inspirations for development of precision medicine, they could also help design more precise diagnostics so physicians can better choose patients who may respond to treatment, the two teams suggest.
The first team found that copy number signatures with impaired homologous recombination were good predictors of response to platinum-based chemotherapy. A companion diagnostic incorporating the signatures could therefore choose patients to get platinum-based therapies in a more targeted manner, the researchers said.