By Monica Heger
Looking at the genomic changes of single cells may be more effective in the clinical setting than studying bulk tumors, according to researchers from Cold Spring Harbor Laboratory who are conducting a single-cell sequencing project of prostate cancer patients.
By sequencing single cells from biopsy samples, surgical samples, and circulating tumor cells, the researchers hope to find copy number variations that will help inform earlier diagnoses, guide treatment, piece together the evolutionary history of the tumor, and detect drug resistance.
Most genomic analysis that has been done in cancer to date has looked at "bulk tumors, which are a mixture of everything," Lloyd Trotman, one of the study's investigators, told Clinical Sequencing News. "But the question that's still open is, 'Are there individual cells or certain foci, which behave differently within the bulk … that are too rare to spot by [bulk] sequencing?'"
There are three main arms of the project, which the team at Cold Spring Harbor is conducting in collaboration with Howard Scher's oncology team at Memorial Sloan-Kettering Cancer Center: single-cell sequencing from biopsy samples; surgical samples; and circulating tumor cells from metastatic disease.
The sequencing is being done at CSHL by Mike Wigler's lab, which has developed a technique for sequencing single cells on the Illumina platform (IS 5/18/2010).
The team will be honing in on copy number variation rather than focusing on point mutations because "we've found that we can trace the tumor's evolutionary lineage by the breakpoints in the chromosome rearrangements that are revealed by copy number profiling," James Hicks, a research leader of the project at CSHL, told Clinical Sequencing News.
The Illumina HiSeq 2000 will be used for the sequencing, which will be done at low coverage — only about 5 percent of the genome — because that is all that is necessary to evaluate copy number, said Hicks.
For the first portion of the project, the team will sequence between 24 and 48 cells from tumor biopsies of 20 patients. They will focus on those patients whose Gleason scores indicate an aggressive form of prostate cancer as well as those that indicate an indolent form to try and see if there are genomic differences between the two.
For each sample, Trotman said the team will focus on isolating cells from an area of the tumor that has a "high likelihood of harboring the more aggressive cells," and will then look for copy number changes within those cells.
"The idea is to try and correlate what we find in our copy number analysis with the Gleason scores and pathology," Trotman said.
The team will also follow up on the results of a previous study in which Trotman found the co-deletion of tumor suppressors PTEN and PHLPP1 only in patients with metastatic disease. He also found that the co-deletion is highly correlated with the deletion of the tumor suppressor TP53.
Now, Trotman is using single-cell sequencing to see if those changes can be seen earlier on. "Our approach is to ask if the changes that are found in the metastasis are already present in single cells" from the primary tumor, he said.
If so, that information can be used to "help pathologists and give them information that will refine their analysis critically so they can make a better decision as to who should receive an operation and who should not."
Additionally, those genes are located within the PI 3-kinase pathway, and there are a number of drugs in clinical trials that target that pathway, such as a GDC-0941, being developed by Genentech, and BEZ235 by Novartis.
The second arm of the project will examine surgical samples from around 10 patients who have received a prostatectomy. The researchers will sample different sites from the same sample, to see if the cancer had a single lineage or multiple lineages. For each patient, they plan to sequence between 25 and 100 cells.
"Often when you do multiple biopsies [of one patient] you find cancer in multiple places in the prostate, rather than a single tumor," explained Hicks. "And very often you find a wide range of grades of tumors." Whether those different grades of tumors are all from the same lineage of cancer cells that have simply moved around the gland, or whether they are from different cancers, is what Hicks aims to discover through sequencing.
Finally, the team plans to sequence between 25 to 100 circulating tumor cells from around 25 patients who have gone on to develop metastatic disease post-prostatectomy.
The goal of this portion of the project is to help guide treatment, figure out what copy number changes are leading to drug resistance, determine whether a drug is effective, discern the differences between the metastatic and primary tumor, and to eventually understand the mechanism of how tumors develop drug resistance.
Looking at copy number changes at this stage is also critical, said Hicks, because typically tumors from prostate cancer, as well as breast and ovarian cancers, do not continuously acquire point mutations. Instead, they "rearrange their genomes in order to acquire resistance or grow in certain parts of the body."
"We're trying to figure out those patterns," he said, and in doing so, figure out the mechanism. Additionally, said Hicks, sequencing the circulating tumor cells to look for signs of drug resistance could help inform physicians as to whether a drug is effective or not.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.