How Metagenomics Is Advancing Infectious Disease Diagnostics

Rapid and advanced results solve medical mysteries, save lives

Carrie Arnold, MA, MPH


August 04, 2017

The patient of Lauge Farnaes, MD, PhD, had already endured more than most other 8-year-olds have. A bone marrow transplant had cured his leukemia, but now the boy was back at Rady Children's Hospital in San Diego and Dr Farnaes had no idea what was wrong this time. Abscesses in the child's muscles and nodules in his lungs pointed to a fungal infection, but even after running every diagnostic test he could think of, Dr Farnaes was at a loss. A chance conversation with a colleague prompted him to try a new diagnostic test that just might solve the mystery. Within 48 hours of sending off the blood test, Dr Farnaes learned that his patient was infected with a fungus called Coccidioides posadasii.

These new diagnostic tests, designed to identify infectious disease, rely on metagenomics, a scientific process that sequences all of the DNA found in a sample to identify the various organisms it contains. Adapted from environmental microbiology research that explored soil, water, and even the human body for resident microbes, scientists and physicians are now using the same strategy to diagnose infectious diseases. What makes metagenomic diagnostics different from other tests, explained Charles Chiu, MD, PhD, an infectious disease physician at the University of California, San Francisco, is that clinicians can get answers even if they don't know what microbe may be sickening their patients. "I've received intense interest and enthusiasm from both physicians and patients," Dr Chiu told Medscape.

Historically, microbiologists diagnosed infectious disease by trying to culture the pathogen in flasks or on Petri dishes, a time-consuming and laborious process. Over the past decade, however, doctors have begun turning to DNA-based tests, such as multiplex PCR, to speed up and simplify the process. These tests give a straight yes or no answer: Yes, the test detected the pathogen; or no, it didn't. Interpreting these results was easy, but the physician ordering the test still needed an idea of what to look for. This didn't pose much of a problem for patients with commonplace or easy-to-detect infections such as Staphylococcus aureus, Klebsiella pneumoniae, or Clostridium difficile. Immunocompromised patients prone to rare and unusual infections and those with meningitis and encephalitis continued to resist prompt diagnosis and rapid treatment.

Dr Chiu and other researchers saw the impact of metagenomics as part of the Human Microbiome Project and in environmental microbiology, but they had to jump several hurdles before they could use it for patients. Metagenomic analysis first chops all of the DNA in a sample into fragments just a few hundred base pairs long. After sequencing this genomic soup, a computer must reassemble the fragments into complete genomes. The technique is the same regardless of whether researchers want to identify all of the microbes in a few crumbs of soil or whether they want to identify what pathogen might be making someone sick.

The plummeting cost of whole genome sequencing has quite possibly had the biggest impact on the development of metagenomic diagnostics. Researchers also had to build better reference databases of microbial genome sequences before they could identify the organism to which the various pieces of DNA belonged. The most challenging step, however, has been isolating the small number of pathogen sequences from the overwhelming number of human sequences and the normal array of microbes in and on the body. Contamination remains a prominent issue as well.

"This technique sequences everything in a sample—and I mean everything," said Anne Piantadosi, MD, PhD, a postdoc and infectious disease fellow at Massachusetts General Hospital who is working to develop metagenomics diagnostics for emerging diseases in low-resource settings. Minute amounts of DNA from technicians or reagent contaminants can confound results.

It's why Dr Chiu began testing his SURPI (Sequence-based Ultra-Rapid Pathogen Identification) platform on cerebrospinal fluid (CSF), which is largely sterile. He had medical reasons, too: For more than half of people with meningitis and encephalitis, a cause is never identified. A 2014 study in the New England Journal of Medicine profiled the use of SURPI in the diagnosis of leptospirosis in a teenage boy.[1] The team also just completed a year-long study at eight hospitals around the United States to see how well SURPI could diagnose meningitis and encephalitis patients. Other metagenomics diagnostics platforms have shown success, too. The University of Utah's IDbyDNA and its database search engine Taxonomer recently obtained $9 million in venture capital funding to bring its test to the clinic. Bay Area startup Karius Diagnostics, the only CLIA-approved metagenomics test, is already available to hospitalized inpatients. Unlike SURPI and IDbyDNA, Karius sequences pieces of DNA freely floating in the bloodstream, which it says helps to enrich the amount of genetic material from pathogens in the sample.

"You don't need much blood, which is important for pediatric cases. When a child is dying, you don't have time to generate a hypothesis. You need an answer now," Mickey Kertesz, CEO of Karius, stressed.

Metagenomic diagnostics are beginning to prove their worth in public health settings. Health officials had difficulty controlling an outbreak of food poisoning in Germany because traditional tests couldn't identify the culprit, an unusual strain of Escherichia coli. When researchers went back a year later, they unearthed that culprit in just a few days, rather than the several months it took using the old method.[2] Dr Piantadosi's work on samples from the West African Ebola outbreak has shown that metagenomic tests can be useful in detecting illnesses from new and emerging infections, all while providing physicians and scientists with basic genomic information about the pathogen.

The tests aren't yet perfect, however. Without FDA approval, obtaining insurance reimbursement could prove difficult, and the tests aren't cheap; Dr Chiu estimates the cost of his to be several thousand dollars if patients had to pay for it out of pocket. The road to FDA approval may be long and winding. "There's no precedent for this kind of test. We need to get feedback on what's required for FDA approval," Dr Chiu reflected.

Karius works directly with hospitals to provide its metagenomics test, with insurance indirectly picking up the tab through the bundled payment hospitals receive when providing acute hospital inpatient care. Although Dr Farnaes believes that the tests more than pay for themselves by providing a quicker diagnosis, he doesn't yet have the data to support his claim.

The thorniest issue Dr Piantadosi anticipates will be interpreting the test results. Clinical labs rarely have an in-house bioinformatics specialist to help parse out confusing answers, and clinicians aren't typically trained in how to understand sequencing results. What's more, it's still not clear how to pass results to the clinical team. Unlike tests that provide a simple positive or negative, it can be hard to know what the result is, Dr Piantadosi explained.

"Is it the number of reads? A part of the genome that could be assembled? It's not totally clear," she added.

Even though metagenomic tests aren't yet standard, they have already begun to have an impact on patient care. In January, Eliza Lape and her husband came home from their Maui honeymoon with a most unwanted souvenir. Searing, stabbing pains in her body kept Eliza from sleeping, and she felt like crawling out of her own skin. She and her husband were sent home from the ER with a diagnosis of the flu. "I've had the flu several times, and this wasn't it," said Eliza.

Ten days later, the newlyweds arrived at Dr Chiu's clinic at UCSF for an evaluation. By then, both were so sick that they were hospitalized immediately, and their physician ordered metagenomic testing on their lumbar punctures to confirm his hunch that Lape and her husband had been infected with rat lungworm (Angiostrongylus cantonensis) while in Hawaii. In less than 48 hours, they had their answer and were started on a course of high-dose prednisone to lessen the symptoms as their bodies fought off the parasite.

"Being sick and not having a diagnosis is really scary," Eliza said. "This type of test seems like it's addressing a lot of these problems."


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.