Skip to main content
Premium Trial:

Request an Annual Quote

CHOP's Joint Genome Center Turns to Sequencing for MDx and Drug Repositioning


By Monica Heger

Since its opening last November, the Children's Hospital of Philadelphia's Joint Genome Center has embarked on a number of projects that aim to bring sequencing into the clinic by molecularly diagnosing pediatric diseases, integrating sequence data with electronic medical records, and discovering mutations that might be targeted with existing drugs.

The center's director, Hakon Hakonarson, discussed some of these projects at the Future of Genomic Medicine conference held in San Diego, Calif., earlier this month.

The center, created under a partnership with China's BGI, is equipped with around 15 high-throughput sequencers with the ability to scale up to 40 or 50. The sequencing lineup includes five Illumina HiSeq 2000s with five more on the way, eight SOLiD 5500s, and several SOLiD 4 instruments that are being phased out. Additionally, it has ordered several Ion Torrent PGMs and will beta-test an Ion Proton, and will also likely purchase a MiSeq, said Hakonarson.

CHOP currently has around 150,000 genotyped patient samples, and Hakonarson said the eventual goal is to sequence all of those samples. In the nearer term, though, he said the center is focused on five disease areas: autism, attention deficit hyperactivity disorder, type 1 diabetes, and inflammatory bowel disease.

Autism MDx

In a collaboration with RainDance Technologies and the Center for Autism Research at CHOP, researchers at the Joint Genome Center are doing a targeted sequencing study of around 2,000 autism patients. The project has added additional genes to RainDance's autism panel, which targets around 50 genes, and plans to add more.

"We have all kinds of new research findings and we'll add those genes to the panel," Hakonarson said. "There's no limitation" to the number of genes that can be added, he said.

The goal of the project is to identify the genetic causes of autism. Hakonarson said that the CHOP team will "report back to the patient positive findings of validated [copy number variations] or mutations." A genetic counselor will go over the report with the patient, and then incorporate it into the patient's medical records.

Even though molecularly diagnosing patients will not lead to a cure for autism, it could still impact patient management and the family's perception of the disease, said Hakonarson.

"A lot of times the family blames themselves for having done something wrong," he said. A diagnosis helps alleviate that blame, particularly if the mutation is de novo.

Knowing whether the mutation is inherited or spontaneous can also help the family make more informed decisions about whether to have more children, and what the risk would be of additional children being born with autism.

In some cases, there may even be therapeutic implications, Hakonarson said. For instance, last year, CHOP researchers identified CNVs in a group of genes known as GRM, which fall within the glutamate signaling pathway, that were associated with ADHD.

Looking at the same pathway in autism cases, Hakonarson found that network of genes, while "not quite as impactful as in ADHD, were still highly significant."

Hakonarson said that the team next looked for compounds that would target this pathway, leading them to fasoracetam, a glutamate agonist that had been developed for a different disease area but failed in phase III trials. He has now started enrolling patients with either deletions or duplications to the GRM genes into a clinical trial for the drug. Any autism patients with that specific mutation would be eligible for the trial, he said.

Even for patients without GRM mutations, Hakonarson said that an early diagnosis of autism "can make a huge impact" on prognosis and severity. "In my view, [diagnostic] information is always going to be helpful. And every parent is asking for it," he said.

IBD Exome Sequencing

In a separate project, CHOP researchers are sequencing the exomes of 900 inflammatory bowel disease patients. Hakonarson said they are looking at families with more than one affected individual, and sequencing the affected individuals, both parents, and an unaffected sibling, when available.

This study is purely research-oriented, he said, so the team will not return results to patients, though the goal is to eventually validate the significant findings and consent the patients so that the results can be returned.

So far, exome sequencing has uncovered 12 novel SNPs that are not in any public databases. Several of those SNPs functionally impaired the DcR3 receptor, which has also been found to be involved in rheumatoid arthritis and some tumors.

The mutations appear to be driving the disease, he said, and there are also compounds that can act on the gene.

For example, the gene regulates cytokine responses and is actually a decoy receptor, so when it binds to its target, it does not elicit a response. One of the genes it binds to TLA1, has also been implicated in Crohn's disease. TLA1 antibodies have been shown in vitro and in mouse models to effectively inhibit TLA1 in cases where the DcR3 receptor is impaired, offering a potential treatment option for this subset of patients, said Hakonarson.

Other mutations were found in known IBD genes, including IL27, IL12, and IL23. Examining that pathway identified 18 hits, suggesting the "possibility of intervening at different levels of the pathway," Hakonarson said.

He said his team continues to do exome sequencing of IBD families and is also following up on the therapeutic opportunities it has identified by doing further testing in animal models. The group is also looking to partner with pharmaceutical companies on this work.

Repositioning Drugs

One of the goals of all the sequencing projects at the center is to look for new indications for compounds that are either on the market, or that have been developed but never brought all the way through clinical development, Hakonarson said.

Using next-gen sequencing to molecularly subtype patients and identify mutations or pathways that can be targeted with existing compounds is one way to bring "clinical development out of academia," he said.

For instance, in the example of the ADHD and autism patients with GRM mutations, it took the CHOP team around 30 months to have a proof of concept for a clinical trial for fasorecetam. "Compared to traditional drug development, this is a lot faster," he said.

Additionally, because the drug's target is known and only patients with the molecular subtype will be enrolled in the trial, the trial can be much smaller and "a lot more cost-effective than a large-scale conventional phase I trial," he added.

Hakonarson added that the patients in the CHOP biobank have all consented to be re-contacted about future studies, so "we can go straight to the individual and ask them if they want to participate," he said.

This is a strategy that he plans to use on all sequencing studies at the center, he said. "We'll look for therapeutic targets," he said, first looking to see if there are available drugs or whether it would require a drug to be developed.

One thing he thinks will start to play out as more sequencing studies are done is that "patients with different diseases but similar molecular phenotypes will respond to the same drugs."

This has already been reported in a number of cancer sequencing studies, where small subsets of patients have been found to have targetable mutations for drugs that were originally developed for a cancer originating from a different site. Hakonarson thinks the same thing will happen with other diseases, as well.

Incorporating Sequence Data with EMRs

Another goal of the center, Hakonarson said, is to start incorporating sequence data with patients' electronic medical records.

He said that he has applied for a National Institutes of Health grant to do this for several different disease areas, including ADHD, obesity, and asthma.

For this project, he said that a report with only the relevant variants would be included with the medical record, such as any disease-causing variant and variants that affect drug response or are associated with drug side effects.

"We can't have it too complicated, especially at the beginning," he said. "So we are just going to report on the most clinically relevant information."

While he said all variants would first be validated with Sanger sequencing or another clinical test, once the accuracy of the sequence data has been demonstrated, "we'll just start floating the data straight into the records," he said.

Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.