NEW YORK (GenomeWeb) — Spurred by evidence that PCR of sputum, in addition to commonly tested nasal and pharyngeal swabs, can allow more sensitive diagnosis of respiratory viruses, researchers from the University of Rochester have published a study demonstrating their method for processing sputum to work with automated multiplex PCR instruments.
Without additional processing, sputum cannot be successfully analyzed using automated RT-PCR because of its viscosity, according to the study authors, so automated systems instead rely on swabs of the nose and throat.
However, research by the Rochester group and others has found that viruses with severe lower respiratory activity may be absent in the upper respiratory secretions sampled by nose and throat swabs, or NTS, and thus missed by NTS-based automated multiplex RT-PCR assays. Thus, it would be beneficial to have a way to test sputum, despite its stickiness, in order to increase the chance of detecting a virus.
While some other studies have looked at multiplex PCR of sputum, they have relied on RNA extraction from the sputum sample before processing.
The Rochester researchers, in their study published last week in the Journal of Clinical Microbiology, set out to evaluate a much simpler processing method to allow use of sputum in automated multiplex RT-PCR. Angela Branche, the study's first author, and her colleagues call the strategy, which involves dunking a cotton swab in sputum and swirling it in sterile water, "dunk and swirl."
"When people have tried to use [sputum] in automated systems it's generally been unsuccessful because there's no way to process it so it doesn't gunk up the machine and invalidate the assays," Branche told PCR Insider this week.
"We had earlier work showing that sputum could be a valuable tool and may be better in some cases than nasopharyngeal swabs, but we didn't know how to use it in an automated system till we developed this strategy," she said.
In the team's study, Branche and her colleagues analyzed archived samples prospectively collected from 965 adults with respiratory illnesses at the Rochester General Hospital from 2008 to 2012. The group first conducted uniplex PCR on both NTS and straight — un-dunked and un-swirled — sputum to compare the presence of virus in the two sample types.
The team reported that 295 of the 956 patients were positive for a viral infection in at least one type of sample. Out of this group, 124 of the viral infections were positive in both NTS and sputum. Another 105 were positive in sputum alone, and 72 were positive by NTS alone, although the majority of these, 65 percent, were patients for whom sputum samples were not available at all. Overall, the group calculated that testing sputum added significantly, about 11 percent, to the viral diagnostic yield over NTS alone.
The researchers then set out to test a subset of the sputum samples using BioFire Diagnostics' FilmArray automated multiplex PCR system.
The team randomly chose 108 of the sputum samples that were positive for viral infection in straight, un-dunked sputum to process using their dunk and swirl preparation method for FilmArray. About half the samples were from patients who were PCR-negative in their NTS samples and half were from patients who were positive for viruses in both NTS and sputum.
According to the researchers, FilmArray successfully identified virus in 99 of the 108 dunked sputum samples. Only four of those nine negative samples were due to interrupted processing, likely due to lingering sputum viscosity, the authors wrote.
Overall, the group calculated the sensitivity of FilmArray on dunked and swirled sputum at between 91 percent and 100 percent for the four most frequently isolated viruses: Flu A, OC42, RSV, and HMPV.
Branch said she and her colleagues believe that the results bear out the utility of the simple dunk and swirl, and hope the study communicates to other laboratories that rely on automated multiplex PCR how they could add sputum analysis into their viral diagnostic processes without having to perform complicated extraction.
"We hope this demonstrates to other microbiologists and physicians that they should be considering sputum if a nasopharyngeal swab is negative, because they could be potentially missing a diagnosis," Branche said. "It also [shares] this method that is very simple and practical and easy for any lab to do."
In the study, Branche and her colleagues also performed quantitative PCR to analyze the viral loads in both straight sputum and dunked sputum for those samples that were FilmArray positive as well as corresponding uniplex PCR-positive NTS samples.
The authors reported that for the majority of samples, across all four viruses, viral loads in straight sputum were slightly higher than in dunked sputum samples. Both types of sputum samples had higher viral load than the NTS samples overall.
According to the authors, this suggests that dunked sputum, when available, might allow viral detection when the viral load in an NTS sample is below the limit of detection.
The necessity for sputum analysis may vary between viruses, the group also noted. For example, in the study, the researchers found that RSV viral loads were much lower in nasal samples than the other viruses studied. This suggests that sputum testing may be particularly beneficial in detecting RSV.
The authors also discussed in the study that combining sputum and NTS samples, by dunking in sputum and then into an NTS sample, may be an efficient and cost-effective way to increase sensitivity of multiplex PCR viral testing. However, demonstrating that this combined process works will require further study.
According to Branche and her coauthors, the results cannot be generalized to other automated mutliplexed PCR systems. The FilmArray platform was available to the group at Rochester, and Branche said she and her team don't plan to explore dunk and swirl with other platforms, but they hope that other groups will.
Branche said the team is also interested in going back and analyzing additional samples of some of the virus types that were less represented in the study cohort. "We want to make sure this is true for all viruses and not just the ones where we had the most samples," she said.
The researchers are also interested in doing a prospective study to evaluate the effectiveness of dunk and swirl sputum testing.