NEW YORK (GenomeWeb) – A multidisciplinary group of researchers from the Netherlands is developing a multiplex assay to quickly detect bloodstream infections (BSIs) caused by bacterial and fungal pathogens in patients grappling with sepsis. If successful, the researchers believe it could be used as a point-of-care test in intensive care units.
In collaboration with Microbiome, a Dutch molecular diagnostics startup, the researchers are developing the test based on a PCR assay that targets pathogen biomarkers. The team sought to develop a multiplex assay that covered at least 95 percent of the most common sepsis-associated pathogens, includingl bacterial and fungal species could cause BSIs in vulnerable patients in hospitals.
"There was discussion with clinicians to tackle ... the morbidity of sepsis, [as] it was quite high," explained VU University Amsterdam medical microbiology professor Paul Savelkoul, an author on a recently published study describing the test. "One of the major issues was that the time to result from the diagnostics was too slow, and they contacted us to... improve patient care."
In the proof-of-concept study, published last month in the European Journal of Clinical Microbiology and Infectious Diseases, the researchers designed three novel multiplex assays in order to to detect specific pathogens at the species level, as well as an additional broad PCR assay, called molecular Gram stain, to discriminate clinically relevant Gram-negative specimens from Gram-positive specimens. Bacterial pathogens included Escherichia coli, Enterococcus faecium, E. faecalis, Acinetobacter baumannii, and Staphylococcus aureus. In addition, the researchers included probes for Candida species, Aspergillus, and the resistance markers mecA and CTX-M1,9.
The researchers collected 5 milliliters of blood from critically ill patients across multiple years for both blood culture and PCR testing. They used 347 blood-culture positive samples — representing up to 50 instances for each pathogen covered by the assay — as well as 200 blood-culture negative samples in order to compare PCR results.
"We collected blood and processed it once a week, which might not be optimal for clinical usage," explained Martine Bos, an author on the study and a scientist at Microbiome. "[But we] hope that if you apply it in the clinic, it would be processed much faster, and that bacterial samples may be preserved better before PCR testing."
After sample collection, the team added a buffer solution and performed centrifugation on the samples, isolating 7 microliters to 10 microliters of pathogen DNA per sample for PCR testing.
"With the low amount of DNA, we [could] still test for 20 or so targets including a group of bacterial, yeast, and antibiotic resistance markers," Bos explained.
Running the PCR assays on a Roche Diagnostics LightCycler system, the authors tested PCRs in a monoplex setting and found them "to be highly specific across a broad panel of clinically relevant microorganisms." The team integrated the 19 monoplex assays into the 4 multiplex PCRs, each containing an internal amplification control.
According to Bos, the entire process — from sample preparation and blood draw to PCR analysis — required about three hours.
The researchers discovered that sensitivity for bacterial species-specific PCR ranged from 65 to 100 percent, with gram-positive PCR sensitivity of 26 percent and gram-negative PCR sensitivity of 32 percent.
The team noted that the sensitivity for different yeast PCRs ranged from 0 to 7 percent. However, when placed in a smaller setup, they saw that yeast detection improved to 40 percent.
The group highlighted that there was no overall link between BSI-PCR sensitivity and time to positivity of blood culture. Overall, they believe that sensitivity of the BSI-PCR is promising for individual bacterial pathogens, but still inadequate for yeasts and generic PCRs.
The investigators noted that the specificity of individual PCRs ranged from 89 to 100 percent in blood culture-negative samples. However, they also noted that there were 174 discordant positive species-specific PCR results that occured among 4,598 individual PCRs performed, yielding an overall clinical specificity of 96 percent.
Besides the issues with clinical sensitivity and sensitivity, Bos noted that her team was surprised to find so few results in the blood samples, "since we extensively tested the whole procedure with spiked blood samples before." In a previous study, published in PLOS One in 2013, Bos and her colleagues had developed a technique to enrich samples from bloodstream infections in blood culture samples.
In the current study however, the team realized that blood samples from critically ill patients produced dramatically reduced results compared to blood culture. In addition, Bos believes that the team will need to develop a separate assay for Candida species in the future in order to improve the yeast's minimal detection rate.
"There might be an issue with pathogen loads, if they're really low, we might still miss it," Bos explained. As such they are experimenting with collecting higher volumes of blood, from 5 to 10 milliliters, to reduce the chances of missing target pathogens.
The team also found that most PCRs did not produce false-positive signals in over 100 isolation controls and negative control blood samples, besides the molecular Gram stain and the occasional A. baumannii PCR.
Umer Hassan, a research scientist at the department of bioengineering at the University of Illinois Urbana-Champaign and a point-of-care diagnostics expert who was not affiliated with the study, noted in an email that while the technology could provide crucial information about critically ill patients, the research does not address the translation of the technology into a point-of-care assay.
However, Hassan believes the technology is critically important because traditional blood culture requires several days to produce results, and providing a pathogen ID in hours will transform current clinical sepsis management.
A clinical diagnostic test using the BSI-PCR technology could be available in two to three years, said VU University Amsterdam's Savelkoul, who is also cofounder and CEO of Microbiome. Along with researchers at the VU University Amsterdam University Medical Center, Savelkoul began the firm in 2006 to commercialize molecular diagnostic technology that they believed was not transitioning fast enough to patient care.
While the firm is currently using the labs at VU University Medical Center for its research, Savelkoul said that Microbiome plans to separate from the university, as he wants to build a independent company and increase the potential for commercial growth. Salvokou and his team have also partnered with Dutch hospitals to run preliminary tests using the BSI-PCR assay.
"We are currently in the middle of starting research trials," Savelkoul explained. "The idea is, with all the regulation needed, we'll start [in] the first half of 2019 for validation and real [clinical] tests."
Since starting Microbiome in 2006, Savelkoul and his firm have filed for and been awarded multiple patents linked to the assay, mostly centering on microorganism collection, isolation, and sample amplification.
Microbiome previously partnered with Biocartis in 2015 to develop the integrated, multiplex BSI-PCR assay on Biocartis' Idylla Enrich platform, a pre-enrichment platform that isolated pathogens from a large volume blood. While Biocartis has shifted its focus toward oncology in the past few years, Savelkoul said that Microbiome is currently negotiating with the firm on the next steps to further develop the BSI-PCR assay on the platform.
Savelkoul declined to comment on potential commercial partners besides Biocartis. However, he noted that Microbiome will initially apply its technology in both the European and US research and commercial markets.
While Microbiome's BSI-PCR technology may produce results in a rapid timeframe, the firm is entering a crowded sepsis diagnostic space. Multiple companies, including T2 Biosystems, BioMérieux, Roche Diagnostics, OpGen, and Accelerate Diagnostics are either developing or marketing advanced sepsis detection assays.
Declining to comment on the potential price of the BSI-PCR test, Savelkoul instead emphasized that the assay's ease of handling, improved sensitivity, and quick turnaround time will help distinguish it from competitors in the field.
In addition to BSI detection assays, Microbiome has also developed PCR tests for periodontal and STD infections. According to Savelkoul, the assays can use a variety of clinically relevant samples, from blood to saliva samples.
Depending on the type of disease, Savelkoul foresees his firm's PCR technology being used in multiple areas within the clinical space, such as public health institutions and peripheral labs, university hospitals, and at the point of care or even an individual patient's home.
In future studies, Bos' team aims to evaluate the test's sensitivity and automation, as well as determine the clinical relevance of its discordant positive results. They hope to ultimately implement the test as part of a point-of-care device, since it currently requires a researcher familiar with the technology to produce results.
Bos envisions the technology being used for any diseases that deal with bacterial infections in the bloodstream, such as endocarditis. While her team has yet to validate the test for other sites of bacterial infection in the body, Bos believes the assay could eventually be used with different liquid samples containing bacterial or fungal pathogens.
"If it was automated, anyone could operate it, and it could be done more quickly and efficiently at places like the ICU," Bos explained.