BALTIMORE – Thermo Fisher Scientific this week launched a new sequencing system called the Ion Torrent Genexus, which it says offers a sample-to-report automated workflow with almost no hands-on steps that delivers results in a single day at comparatively low cost.
The research-use-only system, which will be fully available in 2020, reflects Thermo Fisher's vision of a future of disseminated implementation of NGS in which customers can validate and implement gene panels easily in house.
During a company-sponsored workshop here at the Association for Molecular Pathology annual meeting, Thermo Fisher VP of product management Andy Felton said the goal of Genexus is to enable even small local hospitals to adopt NGS testing in house, something that other company representatives and speakers at the meeting suggested may be necessary to realize the promise of precision medicine.
Disucssing how genomics has transformed the treatment of lung cancer during the Thermo Fisher workshop, Eric Vail, director of molecular pathology at Cedar's Sinai, highlighted a study published by Flatiron Health last year that revealed that significant numbers of individuals who receive comprehensive genomic testing and have an actionable variant still don't get treated with targeted therapy — at least not in a community setting.
Some have hypothesized that test turnaround time or confusion among clinicians ordering tests from outside vendors may contribute to this. Thermo said it believes its new machine can help solve that issue by bringing NGS systems to more hospital labs. It also hopes it can address other hurdles, like the large sample input requirements that it says many central labs have.
"We want to enable all the stakeholders — any physician, pathologist, oncologist, cancer care team, the clinical researcher, and eventually the patient — to take advantage of the best possible result when it comes to molecular testing," said Luca Quagliata, Thermo Fisher's senior director of medical affairs.
Thermo is not the only company to promise a sample-to-answer workflow, though. Qiagen launched its GeneReader sequencer several years ago as part of a "sample-to-insight" model that arranges a chain of integrated instruments and informatics tools with NGS at their center. Although Qiagen has now decided to discontinue updating the GeneReader itself, the rest of the components can still be connected to either end of another sequencing instrument.
Felton said in an interview that unlike previous attempts at sample-to-answer systems, Genexus is a much more integrated system: contained in two machines, and fully automated to make it more attractive for smaller community hospitals, whereas other solutions that may be marketed as sample-to-answer still require a daisy chain of three or more machines.
Genexus still comprises two separate boxes. The first, which won't be available until next year, is an automated nucleic acid extraction, purification, and quantitation system. This machine hasn't been priced yet. The second, which will be available later this year, performs everything from library preparation to sequencing to clinical reporting and has a list price of $299,000.
Felton and others stressed that though the Genexus machines are currently RUO products, the system — and the assays that will run on it — have been designed and are being manufactured with a plan to seek regulatory approval for in vitro diagnostic use from the US Food and Drug Administration.
According to Felton, the front end sample instrument will be able to handle a variety of different materials including FFPE tissue, plasma, bone marrow, and blood. "It's a magnetic-based extraction purification system with full automation, cartridge-based reagents, five minutes of setup time and a two-hour workflow," he said.
Out of the first box come two plates: an archival plate to store and a working plate containing quantitated nucleic acids that become the starting point for the integrated sequencer workflow.
Each GX5 sequencing chip has four independent lanes, each of which can generate 12 to 15 million reads. "We've also incorporated fully automated library preparation for both Ampliseq and Ampliseq HD [methods], and that can handle up to 32 libraries" that can go in a single run, Felton said.
Each chip can support up to four different assays simultaneously, and if the chemistry is the same, users can run DNA and RNA assays alongside one another.
The system can complete its full process in as little as 14 hours for a single lane and takes between 24 hours and about 30 hours for a full chip run, he said.
Genexus overall is designed to operate rapidly, bringing the turnaround time for an NGS result closer to that of single-gene test and other lab procedures, he added.
Design of the sequencer also allows for timing flexibility, with reagents having several weeks of on-instrument stability. Currently, the company has tested this for up to two weeks, but Thermo Fisher believes it can increase that to at least four weeks. "You can use lane one on week one ... lane two on week two, lane three on week three, and potentially up to lane four on week four," Felton said.
Right now, he said, it's not feasible for labs dealing with low numbers of samples to invest in NGS. "If they've only got five samples in a week, they say, 'I'll just wait and send it out.' But that's another four weeks for patients. And the ideal solution is doing this close to the patient and getting the answer to them quickly."
For analysis and reporting of results, Genexus is using new software, which Felton said represents the first major update since the launch of the original Ion Torrent PGM back in 2010.
"Essentially, we've put what used to be our Torrent suite software and Ion Reporter together in a single piece of software called the Genexus system software, and that goes from run setup ... through variant calling, annotation, and then it will go into our Oncomine Knowledgebase reporter tool, and that gives you the clinical report," Felton said.
"What we wanted to build was an end-to-end solution, which basically does all the variant calling, annotation, and then connects the variants to the therapies, the clinical trials … and you get out this one-pager that your clinician needs to go have the conversation with their patient [about]."
In addition, the software will track data from the beginning to the end of the workflow and provides options to view the reagents that were used on the system. "Each reagent is tracked ... to make sure that it is in the right position, so the user cannot make any mistakes when they are setting up the platform," Felton said.
He added that Thermo Fisher put in "a huge effort" to make sure that the system can be run cost effectively with low numbers of samples while still having flexibility to go to higher levels of samples.
"As compared to our current GeneStudio [with Ion] Chef platform, we anticipate about a 50-percent lower cost per sample at a batch size of four on this system," Felton said. "And even batches of three or two or one are going to be significantly less expensive ... compared to the current platform technologies."
According to Felton, Genexus will be able to run various assay panels, including the company's existing Oncomine comprehensive panel and others. "We've tested a variety of different assays, including our custom Ampliseq assays ... our inherited disease portfolio ... and our long-read capability with our TCR-beta panel ... that requires up to 400 base pair sequencing … and as we go through 2020, we will add more assays to the portfolio, including myeloid, BRCA, and there's more to come," he said.
In addition, Thermo Fisher is introducing a new instrument-specific panel for the sequencer, the 50-gene Oncomine Precision Assay, which is the first assay the company has created that works for both FFPE tissue and liquid biopsy samples. It uses a novel fusion detection approach, as well as the company's current targeted isoform detection method.
During his presentation, Felton showed results from test runs using the new Oncomine Precision panel, comparing FFPE and liquid biopsy samples.
"If you're running a single sample per lane in the liquid biopsy workflow, that entire process will take you about 16.5 hours. Four samples across the whole chip would be 22.5 hours," Felton said. "The corresponding FFPE times would be 16.5 hours for three samples and 28.5 hours for 12 samples per run," he added.
Jose Luis Costa, an early user of Genexus and principal investigator at the Institute of Molecular Pathology and Immunology of the University of Porto in Portugal, described some of his team's initial experiences with the system at the AMP meeting.
"We've been sort of playing and stressing the system [a little bit] … and what we did was initially to use commercial control samples, both Horizon and Seracare FFPE and cell-free DNA reference material, which we just pushed through the system," he said.
Using the Oncomine comprehensive assay, the group then moved on to test some clinical samples, in their case lung cancer tissue and liquid biopsy samples that they had previously characterized, to compare how they would behave in the new system.
Studying tissue samples, for the most part, Costa said, "driver mutations were actually all detected at similar levels with the new system," with some discordances that the group is still studying.
"But things really pushed my happiness in the lab when I started to test the liquid biopsies," Costa said. He shared an example of one sample the lab ran. Investigators began prepping the sample (they are currently still doing this manually) at 9AM.
"At 4PM, it went onto the Genexus system … in 10 minutes [it was] being sequenced ... and the next morning when I got back to the lab, [we] had a full report completely characterizing the sample," he said.
Felton said reaching users who have not previously owned a sequencer is a goal for Thermo Fisher. However, he added, labs with established sequencing infrastructure may also want the Genexus as an adjunct to their existing instruments, potentially for clinical scenarios where an ultrafast turnaround for clinical results might complement longer-term exploratory sequencing.