The Children's Hospital of Philadelphia has begun offering next-gen sequencing-based HLA typing services internally.
Dimitri Monos, director of the immunogenetics laboratory within CHOP's Division of Genome Diagnostics, told Clinical Sequencing News that the NGS-based test "has basically solved the problem of HLA typing."
"We've never had a good way to characterize the genes," he said. Methods such as Sanger sequencing often lack phasing information, while PCR-based approaches and an oligonucleotide probe approach known as sequence specific oligonucleotide probe hybridization, are "laborious, time consuming, expensive, and we still never characterize the whole gene."
The HLA assay is run on Illumina's MiSeq system, and evaluates polymorphisms in the HLA-A, HLA-B, HLA-C, HLA-DQA, HLA-DQB, and HLA-DRB1 genes. With the exception of the DRB1 gene, the genes are amplified in full using long-range PCR.
He said that the laboratory evaluated a number of next-gen sequencing systems, including the MiSeq, Life Technologies' Ion Torrent PGM, Pacific Biosciences' RS, and Roche's 454 system. "All of them were able to do it," he said, "but for clinical lab and routine work, you have to account for other parameters," he said, "that don't necessarily have to do with the ability of the system to give you the typing."
Additionally, he said that the platform choice could change in the future depending on how technologies change and improve.
In creating the test, Monos said that the laboratory developed and validated its own primers that amplify the full length of the genes from the 5' to the 3' end. For now, the group decided to just hone in on an approximately 4-kb segment of the DRB1 gene, but plans to amplify and sequence the entire gene in the future, he said. While the HLA-A, -B, and -C genes are around 3.3 kilobases in length and the DQA1 and DQB1 genes are around 7 kb, the DRB1 gene is much larger, between 10 kb and 15 kb, Monos said. Additionally, amplifying the entire DRB1 gene would "require different PCR conditions" in order to amplify it as a single piece, Monos said.
Monos said that the primers were tested on samples from individuals from different populations so that they would be successful on 99.5 percent of individuals.
Sensitivity is 100 percent and specificity is around 85 percent, he added, meaning that about 85 percent of the reads are utilized. So far, the laboratory has evaluated the test on around 300 samples, and comparing it to traditional methods concordance has been 100 percent — and "we get more information," Monos said.
Turnaround time is around one week, he said. The lab does one MiSeq run per week and can batch up to 50 samples, he said. Currently, sample prep is done manually, and the researchers begin processing the samples on Wednesday and put them in the sequencer on Friday, so that they run through the weekend. Data is analyzed on Monday, and a report is issued on Tuesday, he said.
Monos said that the laboratory plans to automate its sample prep and library construction processes and will likely purchase additional machines in order to scale up as it receives more samples. Currently, he said, the lab processes around 10 to 20 samples per week, but that number will likely increase, especially if it begins offering the test outside of CHOP.
Purchasing additional machines will also enable more runs per week to be done. "The protocol is adjustable to have high volumes," he said.
The test is validated to evaluate over 10,000 known alleles, Monos said, but because the entire genes are being sequenced, novel SNPs will be discovered that may also be important for HLA typing in bone marrow transplantation, solid organ transplantation, or disease.
The HLA genes are "extremely polymorphic," Monos said, "and many diseases have been associated with HLA."
While the test will provide the same level of information as previous HLA typing tests, additional information will be collected, so that researchers will be able to go back retrospectively and perhaps identify additional SNPs associated with clinical outcome of transplantation, Monos said.
Using an NGS-based test will also help better characterize samples in donor registries, Monos said. There are around 11 million donors within the National Bone Marrow Donor Program, Monos said. However, because the current methods for HLA typing are expensive, not all of those donors have been characterized precisely.
When a physician looks for a patient within the registry, because the information is not complete, many potential matches may be found. However, those samples will need further testing before a decision can be made, which is "spent money and time," Monos said. Patients in need of a bone marrow transplant do not have time to waste. Going through the registries and typing them with next-gen sequencing will have a "very immediate impact," Monos said. The "doctor looking for a match will be able to quickly make a decision."
CHOP is not the first to begin moving into next-gen sequencing for HLA typing applications. Last year, HLA typing firm HistoGenetics purchased 20 MiSeq systems and said it would incorporate them into its HLA typing workflow. The firm at the time said it initially planned to use a hybrid approach, using the MiSeq as complementary technology to its Sanger systems in order to phase variants and provide a higher resolution.
The Red Cross Blood Transfusion Service of Upper Austria also offers HLA typing on Roche's 454 GS Junior. Last year, it received accreditation for its method.
Additionally, Illumina plans to develop an HLA typing panel that will target eight HLA genes.