After two years of targeted next-generation sequencing-based diagnostic testing and almost 2,000 patient reports, the Leeds Clinical Molecular Genetics Laboratory is in a good position to analyze how the new technology has affected its costs, turnaround time, and test capacity.
The UK-based laboratory, which is part of the Yorkshire Regional Clinical Genetics Service at St. James's University Hospital in Leeds, offers DNA testing for the UK's National Health Service for families and individuals at risk of genetic diseases. Testing is offered for a variety of conditions, ranging from familial breast cancer to cystic fibrosis, and comprises about 50 genes in total. The lab has been accredited by CPA, the UK's accreditation service.
In March 2010, the lab introduced its first next-generation sequencing test for breast and ovarian cancer, which analyzes the coding regions of the BRCA1 and BRCA2 genes, on the Illumina GAIIx platform.
Since then, it has added seven more NGS panels for the following conditions: hereditary nonpolyposis colorectal cancer or Lynch syndrome, including three genes; pheochromocytoma, including nine genes; Marfan syndrome, including the FBN1 gene; Loeys-Dietz syndrome, including the TGFBR1 and TGFBR2 genes; hypertrophic cardiomyopathy, including four genes; Li-Fraumeni syndrome, including the TP53 gene; and Aicardi-Goutières syndrome, including five genes. Besides sequencing, four of these NGS tests also include dosage analysis by multiplex ligation-dependent probe amplification.
Three additional panels — for familial adenomatous polyposis, including the APC and MUTYH genes; optic atrophy, including the OPA1 gene; and familial exudative vitreoretinopathy, including five genes — are about to be transferred from Sanger to next-gen sequencing this summer.
Overall, next-gen sequencing makes up about 70 percent of the current workload, according to David Cockburn, deputy head of the laboratory, who gave a presentation about his team's experience with diagnostic next-gen sequencing at the European Society of Human Genetics annual meeting in Nuremberg, Germany, last month.
The remainder of the workload is split between familial Sanger sequencing, for example if a familial variant has been detected by NGS; targeted mutation testing by an oligonucleotide ligation assay, for example for cystic fibrosis; Y-chromosome microdeletion analysis by multiplex PCR; and CAG repeat sizing for Huntington's disease by direct PCR. In addition, some services have not yet been transferred from Sanger to next-gen sequencing because the genes are small; the genetics are complex, involving, for example, pseudogenes; or because the disorders are very heterogeneous, so other NGS enrichment methods than the one the lab currently uses would be better suited.
Overall, the lab has issued more than 1,900 NGS reports so far, more than 1,200 of them for BRCA1/2 testing. Over the last couple of years, the number of test reports has steadily increased, not only because of the introduction of new NGS tests, but also because the lab has been receiving samples from other UK centers outside the region it serves and from overseas.
Two main benefits of NGS-based testing have been reductions in price and turnaround time. The price of Aicardi-Goutières syndrome testing, for example, has fallen by more than 40 percent, from £920 ($1,430) to £530, Cockburn said. Other price reductions were "more modest," he added, and price cuts are generally determined by the size, number, and complexity of the genes involved.
The majority of NGS-based tests are currently priced at £530, and those prices are valid within the National Health Service. So far, the lab has charged users outside the NHS the same, Cockburn said, though this policy may be subject to revision.
For BRCA1/2 testing, the turnaround time has fallen from about 45 working days with existing methods to about 28 working days since the introduction of next-gen sequencing, which is well below the UK government's target of 40 days, Cockburn noted. In the last few months, turnaround time has crept up to more than 30 days, however, which is due to an increase in sample load and highlights "the importance of managing demand with resources for testing," he said.
The trend of reduced turnaround time is similar for the other NGS-based tests, he said. For pheochromatoma, for example, patients used to be tested sequentially for the nine genes that are now covered in a single NGS panel.
At present, the lab amplifies all genes by long-range PCR, followed by library preparation, sequencing on the Ilumina GAIIx, data analysis using SoftGenetics' NextGene program, data processing in customized spreadsheets, and semi-automated variant evaluation. For about a year, the lab has also been using Beckman Coulter's SPRIworks system to semi-automate a number of library preparation steps.
The data analysis requires a minimum coverage of 50 reads, and all variants except for common polymorphisms are validated by Sanger sequencing, a practice that might be dropped in the future because so far, every NGS variant was confirmed by Sanger. Standard PCR followed by Sanger sequencing also serves as a backup in case the long-range PCR fails or next-gen sequencing does not provide enough coverage.
Prior to introducing its first NGS test, the Leeds team validated the analysis of TP53, BRCA1, and BRCA2 by next-gen sequencing in 65 patients that had previously been tested by Sanger sequencing and found 100 percent concordance between the results, a study that was published in 2010.
It also tested its first 53 patients with both the NGS and Sanger workflows and found that the results were the same. For new NGS panels, it conducts smaller-scale validation studies.
"What we're doing isn't perhaps state of the art at the moment, but it's what we've quite happily been using for two years, and it works," Cockburn said.
Going forward, the lab intends to move some of its tests to the Illumina MiSeq and HiSeq platforms and plans to introduce workflows for both systems over the next few months, although details have not been decided yet.
While the HiSeq would allow for further cost reductions and larger-scale tests, for example whole exomes, the MiSeq would enable even faster turnaround times, which is important in some cases.
For a newly diagnosed breast cancer patient, for example, rapid BRCA screening in under four weeks allows doctors to consider all alternatives, including prophylactic surgery, and for some patients, this reduces the overall number of surgeries, Cockburn said.
The lab is also looking into alternative enrichment methods. While long-range PCR has proven to be flexible, allowing the lab to add new genes to its testing services, other methods might be better suited to "maximize the potential" of the HiSeq, according to Cockburn.
Finally, the lab seeks to further improve the quality of its processes, for example to minimize human error and concentrate human evaluation in areas that matter most, an issue it will be exploring in more detail during a best practices meeting it is hosting this month.