SAN FRANCISCO (GenomeWeb) – Acute infectious diseases often go undiagnosed, including nearly half of patients with a diarrheal disease and as many as 60 percent to 80 percent of patients with encephalitis.
But, researchers at the University of California, San Francisco hope to improve the rate of infectious disease diagnosis —which would theoretically lead to better treatments and outcomes — using a metagenomic shotgun sequencing approach. This week at Cambridge Healthtech Institute's Molecular Medicine Tri-Conference, Charles Chiu, director of UCSF's Viral Diagnostics and Discovery Center presented details of an assay that he said would be launched initially as a laboratory-developed test in the next three to six months with an eye toward US Food and Drug Administration clearance.
In addition, he said that the laboratory has been exploring the use of nanopore sequencing using Oxford Nanopore Technologies' MinIon.
The metagenomic test is currently being developed on Illumina's MiSeq system and the lab is testing a variety of library prep approaches, including Illumina's Nextera XT and Rubicon Genomics' ThruPlex. It has developed an in-house bioinformatics pipeline, dubbed SURPI for sequence-based ultra-rapid pathogen identification, which helped decrease analysis time to between 10 minutes and 15 hours from several days. Analysis time is highly dependent on the complexity of the sample type and the analysis mode itself—there is a faster mode for viral and bacterial identification only or a more comprehensive mode that will align to all sequences in GenBank, Chiu explained.
The lab has diagnosed unsuspected parasites and viruses that other techniques failed to uncover, including neuroleptospirosis in a critically ill 14-year-old boy, a case that was published last year in The New England Journal of Medicine.
The test is "an all encompassing method to capture the full spectrum of agents," Chiu said during his presentation. The team uses a metagenomic shotgun sequencing approach of all the DNA or RNA present in a clinical sample. That unbiased approach enables the team to detect and reconstruct the sequences of "nearly any and all pathogens, including viruses, bacteria, fungi, and parasites," Chiu said.
Chiu's team has so far tested it on 50 patients with an infectious agent of unknown origin who had already received extensive testing to no avail. Of those, the team's metagenomic sequencing test was able to find a causative infectious agent in about 20 percent, Chiu said. Although "our yield was not as high as we'd like it to be," Chiu said, at a 30-day follow-up of the patients that did not receive a diagnosis, none had been found to have an infectious agent, suggesting that the assay did not miss anything.
Some of those patients were later diagnosed with an autoimmune disease, a form of vasculitis, or some other non-infectious cause, while some remained undiagnosed, Chiu said.
That initial patient population was "extremely biased," Chiu said, as they had already received a battery of tests, and in some cases had been put on antibiotics or other therapies.
In one case, a 75-year-old female complained of a fever that had lasted two months, abdominal pain, nausea, and vomiting. She previously had a kidney transplant and was found to have necrotic lymph nodes as well as the fungi Candida albicans. Next-gen sequencing diagnosed histoplasmosis.
Another case was a 20-year-old woman with a fever who had traveled to Australia, where there had been a recent outbreak of Ross River Virus. The patient had fever, headache, and muscle and joint pain. Lab tests were "unremarkable," including negative results for Ebola, HIV, and cytomegalovirus. However, there is no test in the US for Ross River Virus. Sequencing uncovered a "clear signature" of human herpesvirus 7, which Chiu said was unusual because most adults are immune to it. However, the antibody test was negative, suggesting it was a primary infection. They were able to rule out an unusual tropical virus and she recovered on her own in around 2 weeks.
A third case was also someone who had recently traveled. A 70-year-old male had fever, enlarged liver and spleen, and liver cysts and had tested negative in an infectious disease workup. He had recently gone on a six-week hiking trip in Spain and "remarked that he was bitten hundreds of times by insects." The NGS test uncovered a visceral leishmaniasis, which is a parasite acquired from sand fly bites. He was treated intravenously with an anti-fungal for two weeks and improved.
Chiu said that the lab is continuing to work on validating the test and locking down the protocol, after which it will begin offering it as an LDT. He said that the cost has not been set, but will likely be in the $300 range.
In addition, Chiu said the lab is working on a second version of SURPI, SURPIviz, which will speed turnaround time even more and include visualization features such as a heat map of the results.
Enhancements will also include rapid filtering of misannotated sequences, tagging of metadata to track background contamination, front-end visualization tools and a web-based interface, and taxonomic classification against all the sequences in GenBank. Under the current version of SURPI, analyzing 10 million reads takes around 40 minutes, but with the enhancements that will be reduced to five minutes, Chiu said.
In the future Chiu hopes to bring the test through FDA clearance. That test would run on the MiSeqDx. He said he is in discussions with the FDA about what would be required, and said it could include bringing the bioinformatics and analysis software through clearance.
Chiu anticipates that the LDT would run out of a core reference laboratory, at least initially, as opposed to individual hospitals. Aside from applications in diagnostics, he said it may also have utility in public health surveillance, like outbreak tracking, and he said he is currently working with the US Centers for Disease Control and Prevention.
In the research setting, Chiu's lab has been testing metagenomic sequencing on Oxford Nanopore's MinIon. He declined to provide details about results, which he said would be published soon in a peer-reviewed journal, but said that he thinks "there is a lot of potential for real-time metagenomics sequencing [on the MinIon], and we're exploring potential applications of this to do point-of-care sequencing."