NEW YORK (GenomeWeb News) – A team of researchers from Johns Hopkins University and Life Technologies has developed a sequencing-based method for finding rearrangements in individual tumors — identifying biomarkers that could subsequently be used to track cancer using patient blood samples.
The approach, dubbed personalized analysis of rearranged ends, or PARE, relies on the use of massively parallel sequencing to find translocations and rearrangements in solid tumors. After finding such rearrangements in a handful of breast and colorectal cancer samples, the researchers used PARE to track cancer treatment response, recurrence, and metastasis in one of the colorectal cancer patients. The research appears online today in the journal Science Translational Medicine.
"[W]e feel that this is an important step in bringing new genome sequencing technologies to personalized patient care," senior author Victor Velculescu, an oncology researcher at Johns Hopkins and co-director of the institute's cancer biology program, said in a statement.
The new approach stemmed from past studies in which researchers tried to find recurrent rearrangements in cancer, Velculescu told reporters during a press briefing at the American Association for the Advancement of Science annual meeting today. Such recurrent mutations were often difficult or impossible to find.
Instead, researchers noticed that most tumors did contain rearrangements, but the location of these rearrangements varied from one individual to the next — making them good biomarker candidates.
For the current study, Velculescu and his colleagues used the Applied Biosystems SOLiD platform to sequence four matched tumor-normal samples — two from individuals with colon cancer and two from individuals with breast cancer — specifically looking for rearrangements.
"In sequencing individuals' genomes in the past, we focused on single-letter changes, but in this study, we looked for the swapping of entire sections of the tumor genome," co-author Bert Vogelstein, an oncology researcher at Johns Hopkins and co-director of the Ludwig Institute for Cancer Genetics and Therapeutics, said in a statement.
After mapping about 40 million mate-pair sequence reads per sample to version 18 of the human reference genome, the team sifted through the sequence looking for signs of rearrangements, such as copy number changes or situations in which mate pair tags mapped to different parts of the genome.
In the process, they found an average of 14 rearrangements per tumor. Of these, roughly half were intra-chromosomal rearrangements and half were inter-chromosomal.
When the researchers expanded their analysis to include two more colorectal tumor samples, they found an average of nine rearrangements per sample or between four and 15 rearrangements in each tumor.
And, the team noted, they were able to find rearrangements in two of the colorectal cancer genomes even without sequencing matched normal samples.
They then designed primers corresponding to rearrangement breakpoints in the tumor sequence and investigated whether they could detect these sequences when tumor DNA was diluted by normal DNA — as it would be in blood or other bodily fluids. Indeed, the researchers reported, they could detect one cancer genome equivalent even when dwarfed by some 390,000 normal copies of the genome.
They also showed that they could use the same PCR-based approach to detect rearrangements in actual blood samples from two of the colorectal cancer patients.
For instance, by testing blood samples from one colorectal cancer patient before and after surgery, the team was able to follow the individual's treatment.
Before surgery, DNA from the patient's blood sample contained a telltale chromosome 4-chromosome 8 fusion corresponding to the individual's cancer. Immediately after surgery, levels of the mutant DNA fusion dropped off but increased a few days later as the cancer bounced back. When the individual underwent chemotherapy, this treatment again curbed the levels of the fusion biomarker, though a small amount remained due to metastatic lesions in the individual's liver.
Because most clinically important tumors are thought to contain DNA rearrangements, researchers noted, the PARE approach holds promise for finding patient-specific biomarkers that can be used to improve the treatment of a variety of cancers.
For the current study, the cost of PARE was reportedly around $5,000 per assay, though the cost is expected to go down as sequencing prices drop and read quality and length improve, researchers noted.
"As PARE becomes affordable, it will be a helpful addition for physicians to tailor patient care and may become a useful supplement to traditional monitoring by imaging or other approaches," lead author Rebecca Leary, a graduate student at Johns Hopkins, said in a statement.
Those involved say the method holds potential for monitoring cancer and guiding treatment. For instance, it may help differentiate between individuals whose cancers are cured by surgery alone and those who require follow-up with aggressive chemotherapy or radiation following surgery, co-author Luis Diaz, an oncology researcher at Johns Hopkins, told reporters.
Velculescu predicted that PARE will be available for many cancer patients within roughly two to three years, though he noted that that time line will depend largely on sequencing costs.
The team plans to test and tweak PARE in more cancer patients and eventually intends to develop a commercially available product based on the method. As such, a patent related to the research has already been filed.
Velculescu, Vogelstein, and Kenneth Kinzler, who is co-director of the Ludwig Center at Johns Hopkins and an author on the paper, will receive part of the royalties on the sale of products stemming from the research through a licensing agreement between Johns Hopkins and the Cambridge, Massachusetts-based biotechnology company Genzyme.