Article In Brief
Using metagenomics, investigators were able to diagnose causes of meningitis and encephalitis that were otherwise undetected, and the analysis turned up several unexpected results.
Metagenomic next generation sequencing continues to prove its clinical utility in diagnosing the hardest cases of meningitis and encephalitis, according to a new study in the June 13 edition of the New England Journal of Medicine (NEJM).
In the prospective study meant to mimic real-world clinical challenges, researchers using the technique identified the causative agent in about one-fifth of patients with an infectious cause not otherwise diagnosed, and in about half those cases, the result was useful for guiding therapy.
“This is a step forward for metagenomic sequencing in terms of becoming a more practical tool for clinicians,” commented David B. Clifford, MD, FAAN, professor of medicine and neurology at Washington University in Saint Louis, who was not involved in the study. “They demonstrated in this work that is was possible to collect and analyze samples in real time to give results that were clinically meaningful to very sick, challenging patients, and to provide answers in a number that were unique and not obtained by other tests.”
“Metagenomics” refers to the search for and analysis of many different genomes at once. This is made possible by combining next generation sequencing—a comprehensive read of all the DNA or RNA in the sample—with bioinformatic comparison of the sample sequences to a large and growing number of microorganism genomes in publicly available databases.
“There are more than 100 different infectious causes of meningitis or encephalitis,” said Charles Chiu, MD, PhD, senior author on the study and professor of laboratory medicine at the University of California, San Francisco (UCSF) School of Medicine. “In our testing, we are looking for everything at once.”
Dr. Chiu's colleague and first author on the new study Michael Wilson, MD, also at UCSF, previously published a report showing the potential for metagenomics in chronic meningitis and encephalitis. In the new paper, to further explore the practical clinical utility of the technique, they performed a one-year prospective study in patients with more acute illness. Testing was done in a CLIA-certified clinical laboratory, rather than the research laboratory used previously, and results could be used by physicians for managing and treating patients.
Study Design, Findings
The authors enrolled 204 patients from eight hospitals, about half of whom were in the intensive care unit. Patients received typical diagnostic testing for meningitis or encephalitis, including culture, PCR, and/or serology of cerebrospinal fluid, under the care of the local treating physician, and CSF samples were sent to the UCSF lab for clinical metagenomic sequencing as well.
Among all patients, a diagnosis was established by all forms of testing in just over half. Of these, 28 percent were infectious, 8 percent were autoimmune, and the rest a mix of neoplastic, toxin-induced, vascular, and other causes. Among the 57 cases of infectious meningitis or encephalitis, both conventional testing and metagenomic sequencing achieved a diagnosis in 19 cases. Conventional testing but not metagenomic testing diagnosed 26 cases, and metagenomics but not conventional testing diagnosed 13 cases (23 percent).
The fact that metagenomic sequencing missed some diagnoses is not surprising, Dr. Chiu said. “This is why I believe this test should be viewed as a complement to standard testing, to catch missed diagnoses for which conventional testing has failed.”
One reason that metagenomic sequencing fell short was a high level of human DNA in the CSF, which swamped the signal from the pathogen. Another reason was a very low level of pathogen DNA or RNA, a problem which may be mitigated in the future by a fall in costs for sequencing.
“We are looking for a needle in a haystack,” Dr. Chiu said, “but as prices come down, it will become possible to perform deeper and deeper sequencing to, in effect, make the needle shine a bit brighter among the hay.”
In several cases where metagenomic sequencing made a diagnosis missed by conventional testing, it turned up some surprises. “Many of the infectious organisms we found were unusual,” Dr. Chiu said. “Some were organisms that a neurologist would not even entertain as a possible cause,” including one case of encephalitis caused by hepatitis E virus. Although the patient had hepatitis, and in fact was scheduled for a liver transplant, “the clinician didn't consider it because neurological disease from this organism is extremely rare.” The diagnosis allowed the clinician to begin effective antiviral treatment, and spared the patient's liver.
Other unusual infections discovered by metagenomic sequencing included Candida tropicalis and Enterobacter aerogenes, along with a range of viruses. In cases where both standard and metagenomic testing arrived at the same diagnosis, with no likely infectious organism found, the results often gave physicians extra confidence to discontinue empirical therapy and seek another cause—most often autoimmune—and begin appropriate treatment.
There are still advances to be made, Dr. Chiu said. “One of the limitations of the test is that it takes about 48 to 72 hours to get results.” But progress in sequencing is setting the stage for a four to six-hour turnaround time, “so that in the future, we are going to have the ability to diagnose infections on the same day the sample is collected.”
“The test is also limited by the size of the reference database,” the repository of all known microbial sequences called GenBank, maintained by the National Center for Biotechnology Information at the US National Library of Medicine. While extensive, it still lacks many genomic sequences, including large numbers of fungi and parasites, some of which cause meningitis or encephalitis.
“As more sequences are deposited in GenBank, the sensitivity of the test will improve,” Dr. Chiu said. In the study, two cases were shown to be due to consuming raw fruit containing Angiostrongylus, a parasitic nematode whose complete genomic sequence was only added to GenBank in 2016. “If we had looked at their data before that, we may not have detected it.”
“We have to give this group, and Dr. Michael Wilson in particular, tremendous credit for bringing this technique forward and making it part of what will become our routine clinical acumen for identifying pathogens for brain and spinal cord infections,” commented Karen L. Roos, MD, FAAN, the John and Nancy Nelson professor of neurology at the University of Indiana School of Medicine in Indianapolis.
One of the main goals of testing for encephalitis and meningitis, beyond identifying the culprit pathogen when there is one, “is to support our clinical decision to stop empirical therapy,” Dr. Roos said, when tests indicate it is not appropriate, which was seen in this study. Additionally, she noted, the test expedited initiating therapy for non-infectious causes, “and that is a very important result. We want tests with results that are clinically actionable, and this certainly helps us to get there.”
Echoing Dr. Chiu's comments, Dr. Roos emphasized that metagenomic sequencing should not be thought of as a stand-alone test, “but if you do all these tests, and include metagenomic sequencing, you have a much higher likelihood of identifying the pathogen than if you don't do all of them.”
Dr. Clifford of Washington University said, “I think for the time being, this method is likely to be used in a minority of patients,” namely in those challenging patients in whom the obvious candidates are not found. “After you've checked off a short list of the more common causes, then the chance of finding a diagnosis [by standard methods] starts to fall, and those are the cases where this approach is most likely to yield some meaningful results.”
Dr. Chiu reports grants from the California Initiative to Advance Precision Medicine, the Charles and Helen Schwab Foundation, the National Institutes of Health the UC Center for Accelerated Innovation, the George and Judy Marcus Innovation Fund, the Sandler Foundation, and William K. Bowes, Jr. Foundation. In addition, Dr. Chiu has a patent pending for Pathogen Detection using Next-Generation Sequencing, a patent SF2016-185 mNGS control material pending, and a patent SF2017-147 mNGS contaminant database pending. Drs. Roos and Clifford disclosed no conflicts.
The Science Explained
WHAT IT IS
Metagenomics is the analysis of all the genetic material—RNA or DNA or both—in any sample, to identify all the organisms present and to characterize their abundance.
HOW IT WORKS
A sample is analyzed with “next-generation” sequencing, in which the genetic material is fragmented to allow for rapid, massively parallel sequencing. The resulting sequences are then built back up, by aligning overlapping regions, to reveal the sequences of the original genetic material. The organisms contributing these sequences are then identified by comparing them to databases of the genomes of all known organisms. The relative abundance of each organism present is determined by comparing the total number of sequence fragments from each organism in the dataset.
HOW IT IS APPLIED
By identifying the full breadth of species present, metagenomics has revolutionized the understanding of microbial communities wherever they occur. Metagenomics has been used to analyze microbial life in deep sea vents, toxic waste dumps, seawater, the human gut, and many other environments. In each case, the diversity and complexity of the microbial communities was greater than previously thought.