NEW YORK (GenomeWeb News) – A pair of cancer genome sequencing studies appearing online in the Journal of the American Medical Association last night are illustrating the potential of using sequencing and analyses in a clinical setting.
Both stemmed from a broader effort to understand the genetics of acute myeloid leukemia — from de novo AML to chemotherapy-related forms of the disease, co-senior author Timothy Ley, associate director of Washington University's Genome Institute, told GenomeWeb Daily News.
For one of the studies, researchers used whole-genome sequencing to assess leukemia samples from a woman suspected of having an inherited susceptibility to cancer, despite a dearth of cancer cases in her family history.
The woman had been diagnosed with stage II triple positive (estrogen receptor, progesterone receptor, and Her2 positive) breast cancer when she was 37-years old, developed recurrent ovarian cancer in the following years, and ultimately succumbed to treatment-related acute myeloid leukemia at the age of 42.
"The patient was pretty typical of someone with a cancer susceptibility syndrome of some type," Washington University Genome Institute Director Richard Wilson, who was corresponding author on both new studies, told GWDN.
Nevertheless, Wilson explained, she did not carry mutations in BRCA1 or BRCA2 that would explain her propensity to develop breast and ovarian cancer at a young age, nor did she have mutations in any other cancer genes tested.
Researchers decided to look for other genetic clues to her condition, using Illumina paired-end sequencing to sequence the genome of her leukemia and a matched normal skin sample to nearly 29 and 30 times haploid coverage, respectively.
They also used Roche 454 sequencing to verify some potential mutations, along with SNP genotyping to find copy number changes, RNA expression profiling with the Affymetrix Exon 1.0 array to gauge gene expression patterns, and spectral karyotyping to look for larger chromosomal glitches.
During their genome sequencing and analyses of the woman's tumor and normal skin samples, though, researchers found a heterozygous germline deletion in DNA from her skin sample that nipped three exons out of one copy of her TP53 tumor suppressor gene.
The genome from her leukemia sample, meanwhile, was not only missing the other, wild type copy of the gene, but also harbored 26 more coding SNPs, two small insertions or deletions, a dozen copy number changes, and chromosomal rearrangements and translocations not found in the skin genome.
"To find a mutation in TP53 wasn't all that surprising, until we realized that it was also present in her germline genome," Wilson said.
Genotyping TP53 in her mother's blood sample did not uncover the mutation, suggesting it may have arisen spontaneously in the patient. Although samples were not available for her father, the available family history information seemed to confirm this theory.
"The best guess is that the patient's TP53 mutation likely occurred sometime in her early development," Wilson said. "Really, the biggest concern was whether or not this was something that could have been passed along to her children."
Because their functional experiments and past research both support the notion that the woman's original TP53 mutation contributed to her elevated cancer risk, the patient's family was informed of her germline mutation and of possible risks for her three children.
"We advised the family that the children should be screened for early detection of cancer," Ley said.
In a second study, the team provided proof-of-principle results showing that whole-genome sequencing can provide information for guiding a patient's treatment while preserving anonymity.
Using massively parallel whole-genome sequencing, the team tracked down a cryptic gene fusion between the chromosome 15 gene PML and the chromosome 17 gene RARA in a 39-year-old woman with an unknown sub-type of AML.
There were some clinical clues that the woman had an AML sub-type called acute promyelocytic leukemia. But metaphase cytogenetic and interphase fluorescence in situ hybridization analyses did not uncover the characteristic PML-RARA gene fusion found in most APL cases.
Accurately classifying the disease was important, researchers explained, because the PML-RARA fusion can be effectively targeted with all-trans retinoic acid, or ATRA, chemotherapy, sparing patients the risks of bone marrow transplantation, which is also more complicated and expensive.
"When you have this fusion, it basically supercedes all the other genetic information — it's dominant," Ley said. "We knew that if that's what she had it would allow us to make an accurate decision about her care."
In their effort to return clinically relevant information in a time to guide treatment, researchers completed their genome sequencing, validation, and analysis for the patient within about seven weeks.
Indeed, when the team sequenced the woman's bone marrow tumor and matched normal skin samples, they found that a bit of chromosome 15 had become inserted in chromosome 17, producing three gene fusions: one involving RARA and LOXL1 genes, another involving LOXL1 and PML, and a third between PML and RARA that resembled the classical APL fusion.
Consequently, the woman was treated with ATRA instead of with stem cell transplantation. She is currently in remission.
Moreover, when the team screened samples from another 10 diagnostically challenging AML cases from their database, they found two more with similar insertional fusions, Ley noted, suggesting that looking for the insertion in the future may help in finding and treating more APL cases.
"Figuring out that it was an insertional fusion as opposed to a translocation was enormous insight for us," he said, "and has allowed for new FISH probes to be used that will help other pathologists make the [APL] diagnosis more accurately."
Together, the genome sequencing, validation, analyses, and interpretation costs came to around $40,000 per tumor-normal genome pair in each new study, researchers explained.
In an accompanying editorial in JAMA, University of Alabama at Birmingham Comprehensive Cancer Center Deputy Director Boris Pasche and HudsonAlpha Institute for Biotechnology researcher Devin Absher called the studies "remarkable examples of the power that … genomic data hold for patients with a diagnosis of cancer."
"Today, sequencing a tumor genome is still expensive and requires an infrastructure that is incompatible with a clinical setting, but the trend suggests that we are a lot closer to cost-effective, clinical genomics than most physicians realize," they added.
For their part, the Washington University team is continuing to sequence and characterize AML genomes in a research capacity, while simultaneously looking at strategies for selectively incorporating whole-genome sequencing into a clinical setting by then end of this year and early next year.
That will likely include sequencing studies to guide clinical decision making for some AML patients with intermediate cytogenetic phenotypes, Ley noted.
Even so, while the current studies are a "nice proof-of-principle," Wilson explained, there is more research yet to be done before genome sequencing becomes more routine.
"We want to refine the pipeline a bit so that we actually have a good, up-front filter to decide which leukemia patients are good candidates for whole-genome sequencing and which are going to be best informed by cytogenetic analysis or perhaps even by sequencing a relatively small number of genes," he explained.