NEW YORK – Routine clinical infectious disease diagnostics based on next-generation sequencing has moved closer to a true tipping point as two of the biggest players in those markets — BioMérieux and Oxford Nanopore Technologies (ONT), respectively — announced last week that they are working together to explore new applications.
The nonexclusive agreement, timed with the opening of the European Congress of Clinical Microbiology and Infectious Diseases (ECCMID) in Copenhagen, will be the first major foray into NGS-based infectious disease diagnostics for both companies. It comes a year after Oxford Nanopore announced the formation of a diagnostics subsidiary, with infectious disease diagnostics as one of its goals.
Mark Miller, executive VP and chief medical officer at BioMérieux, said in an interview at ECCMID that an international in vitro diagnostics company like BioMérieux "can't be in this business and ignore sequencing."
The partnership is commencing with three projects that vary in timeline, Miller said. The first project is for infectious disease surveillance and outbreak monitoring, with the companies validating BioMérieux's Episeq epidemiological interpretation software with data generated by ONT sequencers. BioMérieux already has some familiarity with this topic: In 2014, it collaborated with Illumina to validate and comarket Episeq with Illumina sequencing platforms for surveillance and clinical research.
This project may wrap up within the next 12 months, Miller said, and will be an early milestone showing that the companies can successfully work together and that nanopore sequencing is suitable for identifying and typing infectious organisms.
In a medium-term project, the partners plan to clinically validate and potentially prepare for regulatory submission of a sequencing-based platform for tuberculosis drug resistance testing.
ONT already has a workflow for rapid drug-resistant tuberculosis profiling, developed in collaboration with the Quadram Institute Bioscience, which BioMérieux will help develop further for research use first, and then for IVD by adding sample prep components and data interpretation and reporting, then helping to generate the necessary data and documentation "to take it down that IVD path."
Emma Stanton, VP clinical at ONT, said in an email that regarding TB, "existing techniques provide either very slow multi-drug readouts (culture) or fast single-drug readouts (PCR). Our goal is to deliver the best of both worlds with this TB-MDR assay."
PCR, she noted, is limited in the number of pathogens and antimicrobial resistance markers it can assay. "For example, the [Cepheid] Xpert MTB/RIF PCR test for TB can detect [fewer than] 100 mutations in one gene associated with resistance to one anti-TB drug," she said. "Nanopore sequencing can detect [more than] 1,400 mutations in 24 genes associated with resistance to 16 anti-TB drugs."
Longer term, BioMérieux and ONT will explore the development of a test to identify the presence of bacteria and determine antibiotic resistance in normally sterile body fluid samples. This could encompass patients presenting with a bloodstream infection, an abscess, or fluid from an infected site, like peritoneal fluid, pleural fluid, synovial fluid, or cerebrospinal fluid — "anywhere we need sequencing-based ID and resistance determination," Miller said. "That requires quite a bit of work on the specimen prep side … [and] sensitivity and interpretation. Huge data banks will be required for interpretation of the sequencing information."
The partnership reflects a market trend of exploring the potential of sequencing to provide much more diagnostic information than the gold standard of molecular testing, PCR, making it particularly attractive for applications like antimicrobial resistance testing that require more than a yes or no answer.
"Many of the clinicians that I have spoken to at ECCMID this year can see potential applications for nanopore sequencing across a range of clinical areas," said ONT's Stanton. "They appreciate the rapid and accurate identification of microbial pathogens and associated antimicrobial resistance directly from clinical samples. … The ability to cost-effectively test small numbers of patient samples in hours while providing more comprehensive information than other rapid tests (e.g., PCR) is common to all the applications mentioned."
For all three of their projects, the companies will "develop solutions using the range of Oxford Nanopore and BioMérieux products based on what's appropriate to meet the needs of the particular focus area," Stanton added, noting, however, that it is likely ONT's GridIon sequencing platform will be a "leading device for these analyses."
The three needs of clinical labs
When BioMérieux began exploring ways to dive more deeply into infectious disease sequencing, "we knew that clinical labs require three important things, whether it's sequencing or not — they want accuracy, speed, and affordability," Miller said. "And when we began looking at sequencing internally for infectious disease diagnostics … those three factors kept coming up."
In the last few years, ONT has made great strides in fulfilling all of those criteria, he said, and BioMérieux "began to realize that in terms of a clinical IVD solution for infectious disease today, the ONT system is much more suitable for labs. We recognize the excellence of Illumina, but for the key criteria for an infectious disease IVD, nanopore is much more suitable."
In terms of competing with other sequencing platforms, it's no secret that ONT has lagged behind in raw read accuracy. But, Miller noted, the company continues to improve in this area, even though it may not be as crucial a characteristic for many infectious disease diagnostic applications.
"In infectious disease … everything comes down to 'is it good enough for this application?'" Miller said. "For many applications, you don't need incredible depth [and] coverage … you need targeted 'good-enough' coverage, which nanopore does stupendously in a short period of time."
At an ONT-sponsored symposium at ECCMID, Alban Ramette, principal investigator at the Institute for Infectious Diseases at the University of Bern in Switzerland, echoed some of these points while presenting some early benchmarking data his research group has generated as an early-access user of ONT's Rapid Barcoding Kit 24 v 14 on an unspecified ONT sequencer.
Ramette, who noted that his group is an ONT service provider that often combines nanopore sequencing with other sequencing platforms, said that for genomic epidemiology in a clinical setting, the gold standard is still Illumina's MiSeq. However, long-read technology like ONT's wins most head-to-head comparisons in terms of time, cost, and flexibility, though it lags in accuracy.
Testing the newest ONT barcoding kit to analyze Corynebacterium diphtheriae and vancomycin-resistant Enterococcus, Ramette's group on average saw about one error every 100 bases, or a quality score (Q-score) of around 20 — much improved over earlier ONT sequencing, he noted, but still significantly lower than Illumina, which states on its website that in general, the vast majority of its Q scores are 30 or above.
Ramette also said that the average library prep time with the ONT kits has been about one hour for 24 samples, with sequencing times ranging anywhere from six hours to 72 hours, depending on the application and sequencing depth needed. Meanwhile, the group calculated an average cost per sample of $30 to $35.
When considering these metrics for clinical diagnostic use, six hours is a very attractive number for a technology that provides the kind of information that long-read sequencing does, but 72 hours is not all that impressive when compared to short-read sequencers like Illumina's.
BioMérieux's Miller noted, however, that ONT's real-time reporting capabilities may be a game changer, drastically reducing the total time needed when used in an IVD setting.
"Their real-time capacity is very key," he said. "You can actually stop runs when you get enough information, which is very unique to them. Even though a full run for a lot of depth may take a day or two — you may not need all that for certain applications in infectious diseases. You can stop it after three hours, five hours … which starts to be very interesting for clinical labs."
Meanwhile, overall cost may be mitigated by the fact that the flow cells used on ONT sequencers have shown the potential for reuse after being washed. "Certainly there are ways to … reuse flow cells that many investigators and current users have shown can dramatically bring down the cost," Miller said. "Right now, I can't endorse how many times to reuse a flow cell, and neither can they, because it depends on the utilization or the application, but there is enough proof out there that you can reuse flow cells sufficiently to really start reducing the cost to get into that very affordable range for IVD."
Ramette echoed this notion in his ECCMID talk, noting that his group hasn't tried this yet for whole-genome sequencing, but for amplicon sequencing, it has been able to use flow cells "five or six times."
ONT's Stanton didn't comment specifically on washing flow cells but noted in her email that "our diagnostic products will be configured with a cartridge that can run samples with a predefined number of tests. These will be affordable and accessible for clinical use globally including [low- and middle-income countries]."
It might be challenging, though, for an IVD product that can be reused for multiple tests to pass muster with regulatory agencies, and Miller said that sequencing in general is going to be a regulatory challenge because there are not yet any US Food and Drug Administration-cleared sequencing-based infectious disease IVDs, although NGS tests from Clear Labs (using ONT sequencing) and Illumina gained Emergency Use Authorization during the COVID-19 pandemic.
"Therefore this will be new territory, and we believe that's the strength of BioMérieux, and that's why we're so excited about the deal — the complementarity is so obvious," Miller said. "There are going to be long discussions." Sequencing today is mostly used to resolve discrepancies between other types of infectious disease diagnostics, he noted.
"So what happens when you come with a sequencing-based diagnostic? What's going to be used for discrepancy resolution and predicate comparison?" Miller said. "The regulatory agencies will need to tell us what they want us to compare sequencing to when it has always been used as the reference standard for comparison. It's exciting, but it is a new frontier."