MolecularMD, a diagnostics company offering reference laboratory and companion diagnostics development services, has recently expanded its menu with several targeted sequencing tests ranging from commercial cancer panels to custom targeted sequencing.
Jin Li, research director of advanced diagnostics at MolecularMD, told Clinical Sequencing News that the company's next-gen sequencing group, based in Cambridge, Mass., has validated an internal workflow using the Ion Torrent PGM AmpliSeq and Illumina MiSeq TruSeq cancer panels, and has been offering them for the last few months for research use only.
By the end of March, the company plans to move this sequencing capability to its CLIA-certified headquarters in Portland, Ore., which will offer various sequencing services for clinical use.
The company plans to add to the commercial panels a number of other targeted offerings through a custom panel service focusing on anywhere from a handful of genes down to deep sequencing of single gene mutations.
At the annual meeting of the American Society of Hematology in December, Li and colleagues from MolecularMD presented a poster highlighting one of these efforts: a test using the PGM for the detection of resistance mutations in the BCR-ABL kinase domain in patients with chronic myelogenous leukemia.
"In the study, we basically validated the limit of detection, analytic sensitivity, specificity, and the reproducibility of this next-gen sequencing test," Li said.
According to Li, while several other laboratories — mostly in Europe — offer BCR-ABL deep sequencing tests, most use a different sequencing approach.
"Based on information I learned at the [ASH] meeting, at least one academic lab is using PacBio," Li said, "but most use 454, especially in Europe. I think we are the only one running this on the PGM."
"Also, we use a different library preparation method than others in this field," he explained. "Most people use a nested PCR approach. Here we only have one PCR. I think this has certain advantages over nested PCR because it’s a simpler workflow. Also, less PCR is better; it lowers the error rate."
Li said the MolecularMD approach involves a single long-range PCR after reverse transcription, followed by randomly fragmenting the PCR products and adding adapters for sequencing on the PGM.
According to Li, the company also utilizes its own proprietary analysis pipeline coupled to the standard Ion Torrent analysis pipeline.
Li said that the overall approach isn't tied to the PGM. "You could use other platforms," he said. "But here, because our customers will be pharma or biotech — maybe some physicians — we need to have flexibility to run maybe just a few samples or more samples. So we see an advantage of PGM to have that flexibility to run with different batch sizes."
According to the group's report, BCR-ABL1 mutations in the kinase domain of ABL1 account for 40 percent to 50 percent of all cases of resistance to tyrosine kinase inhibitors. Rare mutations such as E123Q and T212R in the regulatory domain of ABL1 upstream of the KD have also been reported to lead to resistance.
Currently Sanger sequencing is the gold standard for testing, and although commercial NGS cancer panels include ABL1, "ABL1 resistance mutations should be sequenced from BCR-ABL1 fusion transcripts instead of being sequenced from genomic DNA as in the commercial panels," the researchers wrote in their poster.
In the company's ASH presentation, Li and his colleagues described MolecularMD's fusion transcript-based assay, demonstrating analytical sensitivity down to one percent using serial dilution, and showing comprehensive coverage of the kinase domain, regulatory domain, and the SH2/SH3 domains in a test of 36 clinical samples, comparing the results with Sanger sequencing to confirm that that they could accurately detect low-level mutations.
According to the group, the concordance between the Ion Torrent method and Sanger sequencing among mutations with frequency above 10 percent was 100 percent, while Sanger missed 26 low-level mutations with less than 10 percent frequency.
"Drug resistance mutations in BCR-ABL1 can be very low abundance," Li said. "So it's challenging to ensure the specificity of your result ... We showed that even if the mutation is as low as one percent, we can ensure our method is accurate. So I think based on this, we got a pretty [good] result."
According to Li, MolecularMD hopes to start offering the BCR-ABL1 sequencing assay some time in the first quarter of this year, most likely mainly for pharmaceutical and CRO customers, but also potentially for physicians for clinical use.
He also said the company is working to develop other targeted sequencing assays, and to make itself available to develop custom panels on demand for clinical research customers.
"There is a strong demand for targeted sequencing … Customers come to us not asking for a 400-gene or a 200-gene cancer panel. Some want just a few genes. This is especially clear for us for our … pharma/biotech [customers] who are running clinical trials," Li said.
One example, he said, in non-small cell lung cancer, where there is a need for panels covering EGFR, ALK, and KRAS, some groups require even deeper targeting to look deeply at a single area, like BCR-ABL1, or even a single mutation.
"We've developed a proprietary laboratory preparation method, for example, for the drug-resistant EGFR mutation T790M in non-small cell lung cancer, which is important to detect with very high sensitivity and specificity, so we can get a limit of detection as low as 0.03 percent using NGS," Li said.
According to Li, these tiered services — from the commercial AmpliSeq and TruSeq panels, to customizing a targeted sequencing assay for smaller target sets — will be expanded this year from research use only offerings to clinical services out of the company's CLIA lab in Portland.
"They already have many years of CLIA experience, so we will install the sequencers there and run the tests there," he said.