This article has been updated to reflect that the clinical sequencing services will be integrated into existing CLIA laboratories.
The Dr. John T. Macdonald Foundation Department of Human Genetics of the University of Miami Miller School of Medicine is developing next-generation sequencing-based clinical testing, initially focusing on cancer, pediatrics, and disease carrier screening for genetic disorders that are common in the Jewish population.
Testing services, which are scheduled to start by the end of this year or early next year, will be incorporated into existing clinical laboratories within the department.
The department's Division of Clinical and Translational Genetics, headed by Olaf Bodamer, already provides clinical and diagnostic in- and out-patient services for patients with genetic disorders through the Clinical Molecular Genetics Diagnostic Laboratory and the Biochemical Genetics Diagnostic Laboratory.
The university is currently in negotiations with additional institutions in the Miami area that might join its clinical sequencing initiative. The new services will initially be offered to local hospitals but might become available nationwide in the future.
According to Toumy Guettouche, director of research and development for the CLIA labs and a research assistant professor in the human genetics department, the university is currently considering the Ion Torrent PGM and the Illumina MiSeq as testing platforms. The department already has a PGM in house and is waiting to receive the MiSeq.
It is not considering the 454 GS Junior because of its high running costs, or the Illumina HiSeq, which requires too much infrastructure for data storage and analysis, he said.
In order to decide which system is more suitable for clinical sequencing, the researchers plan to conduct a side-by-side comparison of the MiSeq and PGM, analyzing well-characterized cancer cell lines with known mutations multiple times on both platforms.
They will judge the instruments by a number of criteria, including data accuracy, reproducibility of the results, and ease of workflow, Guettouche told Clinical Sequencing News.
"I think neither the MiSeq nor the PGM are optimal clinical sequencing instruments — they are research instruments that have been pressed into the clinical sequencing space," he said. While these are currently the best instruments available for clinical sequencing, he said, new platforms with better data accuracy and easier workflows will likely replace them in the future.
"The workflows required for these instruments are not really that suitable for a clinical lab. They are still too complicated, there are too many unknowns, there is too much art involved," he said. Compared to technologies like qPCR, "the barrier of entry into clinical sequencing is still pretty high."
Validation of mutations will also still require an independent method, for example Sanger sequencing or qPCR.
For cancer diagnostics, the laboratory will offer cancer panel sequencing and is currently considering different options for the panel size and the enrichment method. "It's probably going to be a mix of a pretty comprehensive panel that contains all the favorite genes of the local doctors, and then some additional ones," Guettouche said. Narrow panels that only cover gene sections with hotspot mutations are probably not sufficient, he added, because they do not allow newly discovered hotspots to be added.
The enrichment strategy will probably be amplification-based rather than hybridization-based, he said, because that produces less off-target sequence, which suits the limited capacity of clinical sequencing better.
The researchers are currently considering amplification products from Life Technologies and Illumina, such as the Ion AmpliSeq Cancer Panel v1, which includes hotspot regions of 46 cancer genes; the Ion AmpliSeq Comprehensive Cancer Panel, which targets the exons of 409 cancer genes; and the Illumina TruSeq Amplicon Cancer Panel, which includes mutational hotspots in 48 cancer genes.
In addition, they are collaborating with Kailos Genetics, a company based in Huntsville, Ala., that specializes in target enrichment for next-gen sequencing, to design their own cancer panel. This would allow them to remain agnostic to the sequencing platform, Guettouche said.
For inherited disorder testing in the Jewish population, the researchers have not yet decided whether they will offer a comprehensive gene panel, similar to what Stephen Kingsmore's group at Children's Mercy Hospitals and Clinics has developed for newborn testing (CSN 8/9/2011); a narrower panel; or whether they will use exome sequencing.
In terms of sample preparation, the lab will offer DNA extraction from FFPE cancer samples as a service, and the researchers are currently building their expertise around FFPE sequencing.
The bioinformatics analysis of the results is going to be "the most difficult part to get your head around," Guettouche said, and the scientists are considering several commercial offerings, including analysis platforms from Life Tech, Illumina, and Kailos.
"I'm most worried about the software options for a clinical lab," he said. "The software solutions are still in development, they're not clinical grade yet, and you probably have to do your own modifications to them."
Data storage is also going to be a challenge, for example how long to keep clinical sequencing data, and whether to set up hardware in house or use a cloud-based solution. While there are privacy concerns over cloud-based storage of clinical data, "you have to consider it if you're looking for solutions," he said.
While prices for the planned tests have not been determined yet, they are likely going to be in the hundreds and up to more than $1,000. Reagent costs for a comprehensive cancer panel alone, for example, are around $1,000, Guettouche said.