By Julia Karow
As the cost of human whole-genome sequencing is plummeting, researchers are beginning to explore its usefulness in the clinic. Earlier this month at the Personal Genomes conference at Cold Spring Harbor Laboratory, several groups presented some of the first examples of how whole-genome sequencing is helping doctors select a therapy or make a diagnosis.
While these initial cases appear promising, researchers pointed out that no therapy was selected based on the presence of genome mutations alone; rather; this information provided one more result for doctors to take into consideration for treatment. Also, the examples involved cases where all other options to diagnose or treat the disease had been exhausted.
Most of the groups who presented results analyzed their patient samples using Illumina's sequencing technology, including one that was sequenced in Illumina's CLIA-certified lab.
A Novel Fusion
Rick Wilson, co-director of the Genome Center at Washington University School of Medicine, presented the case of a 39-year-old woman with acute promyelocytic leukemia. Most APL patients have a translocation between two chromosomes that result in the fusion of the PML and RARA genes, and patients with a PML-RARA fusion protein benefit from the drug all-trans retinoic acid, or ATRA. The translocation is normally diagnosed by karyotyping and fluorescence in situ hybridization, but cytogenetics of this patient showed no translocation, and FISH indicated a RARA-PML fusion but not a PML-RARA fusion, so she received conventional therapy but no ATRA.
In order to find out whether the translocation was present in her cancer despite the cytogenetics results — and whether she might therefore benefit from ATRA — the WashU researchers proceeded to sequence her tumor and a normal control from her skin. Each sample was analyzed in a single run on a HiSeq 2000 that generated on the order of 200 gigabases of data, or more than 60-fold coverage. It took about a week to sequence the samples, and the entire effort — including library construction, analysis, and validation — was completed in less than six weeks from the date the researchers received the sample, Wilson said.
The analysis revealed a novel transposition event, where a chunk of DNA from chromosome 17 had replaced a small segment in chromosome 15, which resulted in a PML-RARA fusion. Based on this finding, the patient received ATRA as part of her consolidation therapy and is currently doing well, according to Wilson. “The correct clinical decision was made because we were able to sequence and determine her unique oncogenotype within the required clinical time frame,” he said.
The researchers are now studying additional atypical APL cases with normal cytogenetics to see if they harbor similar gene fusions resulting from an insertional event, rather than a translocation or deletion.
“It is time to utilize whole-genome sequencing as a diagnostic approach for potentially understanding atypical cancer cases,” Wilson said. “That is not to say that everybody who has cancer should be sequenced,” he added, although that might be the case sometime in the future. Others at the meeting suggested, however, that in the case of the APL patient, a more targeted approach might have sufficed to discover the gene fusion.
Wilson and his colleagues at WashU have implemented a process to bring in atypical cancer patients for genome sequencing, he said, demonstrating that “it is possible to navigate the [institutional review board] waters.”
Following his talk, Wilson told In Sequence via e-mail that this patient was the first at WashU "to be sequenced specifically for the purpose of hoping to better direct treatment." He added that there will "certainly be others" in the future.
In his talk, Wilson said that the cost of sequencing patient genomes is becoming “pretty minimal” — $5,000 for sequencing one flow cell on the HiSeq. However, cancer patient genome sequencing projects still require a “substantial” team of experts to analyze the genome and arrive at a diagnosis. “That's the hard part of it,” he said.
He told In Sequence his team hasn't performed a cost analysis for the entire process, which would include analysis, though he noted that "the cost and effort required to effectively understand a patient's tumor genome to an actionable end is significantly more than the cost and effort that is now required to produce the data."
Steven Jones, associate director and head of bioinformatics at the Genome Sciences Centre at the BC Cancer Agency in Vancouver, showed how genome sequencing provided insights into a rare tumor and helped select a treatment for the patient. He presented the case of an 80-year-old man with an adenocarcinoma of the tongue, a type of tumor for which he said no established treatment protocol exists.
Using the Illumina GAIIx, Jones and his team sequenced the genomes of the tumor and normal DNA from the patient's blood, as well as the transcriptome of initial lung metastases and normal cells. They then tried to correlate genomic features of the cancer, such as somatic mutations, copy number alterations, and expression data, with drug target information from the DrugBank database, which contains detailed information about almost 4,800 drugs.
The researchers found several thousand genes in amplified regions of the tumor, as well as hundreds of genes with increased expression and four with somatic protein-coding mutations. Their analysis suggested that the tumor cells were driven by the RET oncogene, which was highly expressed in the tumor. They also found that several genes whose products are targeted by known RET inhibitors were either amplified or highly expressed in the tumor, indicating that it might respond to these drugs. As a result, the patient received a RET inhibitor — sunitinib — which stabilized his disease for four months. After that, a second round of treatment with another RET inhibitor , sorafenib , combined with another drug sulindac , stabilized the disease for another three months, after which new lesions developed. An analysis of a new skin metastasis showed that the cancer had developed novel somatic protein-coding mutations and amplifications that explained the drug resistance.
Jones, who recently published the project in Genome Biology, told In Sequence that his group is working on several similar cases, adding that genome sequencing is appropriate in cases where there are no other therapeutic options. For more common tumor types with well-established protocols and standards of care, on the other hand, it might take several more years before genomic approaches will be of use.
Not Only Cancer
Whole-genome sequencing is also proving to be useful in diseases other than cancer. For example, Liz Worthey from the Medical College of Wisconsin and the Children's Hospital of Wisconsin spoke about using whole-exome and whole-genome sequencing to identify causative mutations in children with life-threatening diseases, one of the aims being to guide their therapy.
All cases selected for genome sequencing by the Wisconsin researchers have been reviewed and approved by an institutional review board, she said. The selection process starts with a nomination by two doctors who have determined that genome sequencing is warranted and appropriate, and that all other clinical assessments have been made. The case is then reviewed to determine the medical necessity of sequencing, taking into account whether it will benefit the patient and family, the likelihood of altering the treatment, whether all pertinent testing has been completed, and how medical knowledge can be advanced. So far, all cases — just a few at this point — have been funded through private sources.
For exome sequencing, the researchers have been using NimbleGen arrays in conjunction with 454 sequencing but are now moving to NimbleGen's in-solution capture method. In addition, they have had one genome sequenced by Illumina's CLIA lab through the firm's Individual Genome Sequencing service.
The Wisconsin team has also developed its own clinical variant annotation tool, called Carpe Novo, to help geneticists identify mutations associated with disease.
The researchers received their first genome sequencing request last year, Worthey said, for a child suffering from a severe version of inflammatory bowel syndrome that was labeled “non-standard” in terms of the age of onset, symptoms, severity, and progression. At the time, Worthey said, whole-genome sequencing was still too expensive, so the team proceeded with exome sequencing on the 454 platform instead. Their analysis identified a single candidate gene, an inhibitor of apoptosis, and the result was confirmed by a diagnostic lab. Functional analyses of the child's cells confirmed that they were more sensitive to apoptosis. Partly in response to this finding, the child received a bone marrow transplant and is recovering well, Worthey reported.
In another case, the Wisconsin researchers analyzed the genome of a child with severe combined immune deficiency, which was sequenced in Illumina's CLIA lab. So far, the analysis has revealed “a handful” of candidate mutations likely associated with the disease, she said.
While genome sequencing is beginning to prove useful in the clinic, Worthey agreed with Wilson that each case currently requires a large team of players, including researchers involved with the sequencing, bioinformatics, IT and systems support, and follow-up assays, as well as several physicians, genetic counselors, and ethicists.
Finally, a group at the Institute of Clinical Molecular Biology at the University of Kiel in Germany has used whole-genome sequencing to better understand the disease of a child suffering from a severe form of Crohn's disease.
The researchers, led by Stefan Schreiber, sequenced the genomes of both the unaffected parents and the child using Life Technologies' ABI SOLiD 4 platform, and analyzed gene expression in blood cells and intestinal tissue of the patient using RNA-seq. They reported several hundred variants with potential functional effects for which the child was homozygous, but the parents were heterozygous. Several non-synonymous SNPs were in genes involved in immune system pathways, suggesting a certain therapy might help. By the end of this year, Schreiber and his team want to sequence the genomes of five patients and 70 exomes, he said.