By Deborah Borfitz
March 18, 2022 | Merging biotech and infotech, life science company Karius has laid the groundwork for dramatically advancing molecular diagnostics for infectious diseases on a global scale, according to Chief Medical Officer (CMO) Brad Perkins, M.D. Using technology similar to what’s being used for detecting circulating tumor DNA, the Karius “liquid biopsy” test looks for microbial cell-free DNA (cfDNA) to accurately identify more than 1,000 pathogens from a single blood draw no matter where the infection may originate.
Karius was founded in 2014 on the premise that the molecular approach has many attractive characteristics allowing microbial DNA signals to be enhanced and that of background human DNA to be removed, he explains. “Once you do that, the microbial fragments you’re left with are really quite well suited to direct sequencing.”
The Karius Test is like other liquid biopsies currently in use—for detecting cancer as well as fetal abnormalities and organ transplant rejection— because it detects cell-free DNA fragments in blood. But, unlike other liquid biopsies, the Karius Test detects microbial DNA (not human DNA) to help identify infections, Perkins says. Technology developed by Karius, involving metagenomics and bioinformatics, can line up the 1,000-plus suspects, based on their cfDNA signatures, and identify the culprit, which in this case is the cause of infection.
“In that process, we can identify things the standard of care might miss, so you get much better yield, much faster identification and, importantly, because we’re using plasma, we are not having to subject people to invasive biopsies to obtain a diagnostic specimen,” says Perkins. “Anywhere you have a serious infection, even if it is deep-seated, we see those microbial DNA fragments leaking into the plasma.”
The standard of care for diagnosing infections typically requires a specimen from the site of an infection to make a diagnosis, Perkins continues. Among patients who are hospitalized for suspected infections, it is common for doctors to use needles or scopes to biopsy tissue at the infection site, which might be the lung, brain, cerebrospinal fluid, or liver.
These invasive tests can be slow, and come with multiple potential complications, he says. The diagnostic yield can also be low. “The Karius Test is positioned to overcome all of those challenges.”
This is a send-out diagnostic intended for in-patient use only, says Perkins. It is a laboratory developed test (LDT) that is certified under the Clinical Laboratory Improvement Amendments and accredited by the College of American Pathologists, in addition to meeting the stringent approval requirements of New York State.
As required of LDTs, Karius operates out of a single laboratory so it can control test quality. Results are produced within 24 hours of specimen receipt and are immediately reportable via the Karius app, Perkins says. The company is exploring the possibility of eventually moving to a distributed format so that other clinical labs can prepare the specimens and load them onto their sequencer using Karius’ bioinformatics capabilities to run the pathogen identification.
Perkins joined Karius only last May, having most recently served as cofounder and CMO at The Commons Project Foundation, a nonprofit that builds digital platforms and services for the common good. In the late 1980s and early 1990s, he worked for the Centers for Disease Control and Prevention (CDC), where he led the investigation into anthrax attacks in the U.S. and later served as the agency’s chief strategy and innovation officer.
It was during his time at the CDC that microbiology made its critical transition from pure phenotypic approaches to detecting previously unrecognized bacteria to looking at their metabolism, growth, and, ultimately, DNA to identify and taxonomically classify the organisms. “I had some early success in this space and have been hooked ever since, and what Karius is doing is really a culmination of that to scale global impact on infectious diseases.”
When Human Longevity launched in 2013, Perkins was its first CMO. Perkins says he worked with Craig Venter, one of the company’s founders, early in his CDC tenure when microbial agents were the proving ground for genomics. Venter was the first to do a full genomic sequence on a free-living life form, specifically the Haemophilus influenzae strain that often causes pneumonia.
“Essentially, I have always been doing the same thing, which is trying to find new science and technology that scales and improves health, and ultimately helps to reorient the healthcare system to be much more focused on early recognition and prevention of illness… [rather than] reactive sick care,” says Perkins.
The first place cfDNA found traction was in noninvasive pregnancy testing, says Perkins, since fetal DNA can be differentiated from maternal DNA in the mother’s bloodstream and used to screen for abnormalities such as Down syndrome. It’s now an alternative to amniocentesis, an invasive test that comes with a small but concerning miscarriage risk.
This was followed by application of cfDNA testing to cancer detection as well as risk of organ transplant rejection based on increasing levels of cfDNA from the donor, he says. It was in the organ transplant context that Karius cofounders (then-Stanford postdocs in Steve Quake’s lab), Mickey Kertesz and Tim Blauwkamp, recognized the utility of microbial cell-free DNA as a biologic analyte in infectious disease diagnosis and management and began their pioneering work.
The Karius liquid biopsy test has been validated in over 50 peer-reviewed publications (notably, Nature Microbiology, DOI: 10.1038/s41564-018-0349-6) and is now being used by more than 200 hospitals around the country, says Perkins. “Physicians are cautious in their adoption of new advanced molecular tests for a variety of reasons, which I think are sound, and they like to wait for high-quality evidence of clinical and in some cases economic value [e.g., Applied Health Economics and Health Policy, DOI: 10.1007/s40258-020-00611-7] and we are in the process of gathering that evidence.”
As Karius explains on its website, the liquid biopsy test involves extracting cfDNA fragments from the plasma specimen, converting them to shotgun DNA sequencing libraries. Next, the DNA is sequenced using Illumina NextSeq technology.
During analysis, the absolute concentration of cfDNA from each pathogen is calculated and compared to controls to determine which microbes are contributing significant levels of cfDNA to the specimen. Reporting consists of confirming several quality control indicators and, upon acceptable performance, faxing the result back to the ordering institution.
With the more than 1,000 suspects identified to date, the Karius liquid biopsy test has the landscape of recognized DNA-based microbial pathogens well covered, says Perkins. Importantly, that includes parasites, many of which are extremely difficult to diagnose and, in some cases, only via serology testing where the long wait on results diminishes the clinical utility of the findings. The test also picks up two groups of fungal pathogens—molds (pneumoconiosis) and yeasts (e.g., Candida)—as well as DNA-based viral pathogens often seen in immunocompromised patients, he continues.
The current standard of care “completely fails to simultaneously identify” a large group of infection-causing parasites and fungi identified by the Karius Test, Perkins notes. “This is one of the advantages of hypothesis-free infectious disease diagnostics. Right now, doctors rely on clinical signals in the form of history [e.g., recent travel], evolution of the disease and symptoms, and [organism-specific] laboratory tests.”
That “hit-and-miss cycle” of testing can be lengthy and often unsuccessful and, if it results in patients getting broad-spectrum antibiotics, can contribute to the creation of antimicrobial resistance, he says. “Many of the pathogens we identify aren’t even covered with those empiric antibiotics [so] it is a failure of the standard of care on many levels.”
Since Perkins came on board, Karius has narrowed its evidence generation work primarily to immunocompromised patient populations at most serious risk for infectious diseases. These include cancer patients, who account for about 70% of immunocompromised admissions for suspected infectious diseases in the U.S., and people who have undergone a solid organ transplant.
Toward that end, Karius has just completed an important, 10-center study of immunocompromised patients with pneumonia where its plasma-based test was directly compared to bronchoscopy, which is an expensive and complicated but frequently used test done to obtain lung specimens for culture. Study results are expected around mid-year, Perkins says. Similar studies will be done for other clinical syndromes affecting cancer patients and solid organ transplant patients.
The liquid biopsy test currently has high usage among immunocompromised patients as well as those with bloodstream infections such as sepsis (which occurs in both immunocompromised and non-immunocompromised patients) and endocarditis (infections of the heart), reports Perkins. A newly published article in the New England Journal of Medicine (DOI: 10.1056/NEJMra2111003) should also create demand for the test in diagnosing fever of unknown origin, a classic infectious disease syndrome characterized by prolonged fever. The use of a plasma genomic test was part of the recommended workup.
Although Karius now has a few competitors, “none of them have the experience of running their test commercially since 2016,” says Perkins. That head start is particularly important given that artificial intelligence (AI)—what Perkins prefers to call augmented intelligence—plays a key role in the testing platform. Machine learning is used to match the “millions and millions” of generated data fragments to the lineup of 1,000-plus pathogens to ensure high-quality test results.
Karius is constantly increasing the number of pathogens in the lineup, he adds, as a reflection of growing commercial experience with the test. As it is, the test can identify pathogens causing focal infections “essentially everywhere in the body”—a substantial advance from the status quo of extracting tissue from the area and trying to culture it.
In the future, Perkins says, Karius anticipates being able to use AI and machine learning to help physicians make better decisions based on the data from its liquid biopsy test and potentially other data accessible in electronic medical records.
More immediately, the company hopes to add a group of RNA-based viral pathogens (including SARS-CoV-2) to its suspect list, he adds. Karius plans to soon introduce the identification of antimicrobial resistance as a feature of its testing platform. “Many of the antimicrobial resistance mechanisms are genomically based and, because the code for those resistance mechanisms is written into the DNA, we can… identify those particular fragments.”