In this study, we compared the performance characteristics of two commercial multiplex assays for the detection of respiratory viruses. The most common viruses that are considered significant respiratory pathogens in immunocompromised hosts include influenza A virus, influenza B virus, RSV, PIV1 to PIV3, hMPV, and picornaviruses (rhinoviruses and enteroviruses) (11
). Newly discovered viruses, including bocavirus, PIV4, novel coronaviruses, and rhinoviruses, are increasingly being recognized as causes of respiratory illness in this population. However, the ability to detect these novel viruses as well as other significant pathogens by conventional virologic methods is limited.
Molecular assays have been shown to be superior to traditional methods for virus detection and further allow the identification of novel viruses, most of which are not cultivable by techniques routinely used in diagnostic laboratories (15
). In our study, xTAG RVP Fast and FA RVP had significantly higher sensitivities for the detection of respiratory viruses than traditional methods, such as DFA and viral culture. The sensitivity of FA RVP was generally higher than that of xTAG RVP Fast, which translated as a lower specificity, since the additional viruses not detected by all other methods could be considered false positives. The difference in sensitivity between the two assays remained statistically significant when only confirmed positive isolates were considered. The increased sensitivity of the FA RVP assay might be due to the incorporation of nested PCR as the amplification method in the FA RVP assay, increasing the likelihood of the detection of a low viral burden. Additionally, there is more than one reaction site in each pouch for the second-stage PCR for each virus, further increasing the sensitivity of the assay. The sensitivity of the FA RVP was closer to that of the Resplex II assay, in which a novel template-enhanced multiplexing process was used for amplification (13
). In a previous study by Rand and colleagues, FA RVP was also shown to detect more total viruses than the xTAG RVP assay (20
). In persons with cancer, where a positive nasopharyngeal swab for respiratory viruses has substantial impact on oncologic care and outcome, the implementation of tests with increased sensitivity is imperative.
The lower sensitivity of xTAG RVP Fast was especially noted for Flu B and RSV. A previous report by Pabbaraju et al. also noted a decreased sensitivity for Flu B in a study comparing xTAG RVP Fast to xTAG RVP (18
). As the authors of that study suggested, sequence variation in the hemagglutinin gene of Flu B could result in primer mismatch and lower sensitivity. Recent mutations in RSV have resulted in the lower detection of the virus with the current FDA-cleared version of the xTAG assay during the 2010 to 2011 respiratory season (R. T. Horvat, personal communication). Although the primers and targets are different in the xTAG RVP Fast version (per the manufacturer), the lower sensitivity of the assay for RSV detection compared to that of FA RVP and DFA could similarly be attributed to the genetic evolution of the virus. The detection rate for Flu A was similar in both assays, with the added advantage of the specific detection of the 2009 H1N1 Flu A by FA RVP as opposed to being labeled an untypeable result, as was the case for xTAG RVP Fast. This can have important clinical implications in outbreak settings and in the selection of antiviral therapy for treatment as various susceptibilities in circulating influenza virus strains occur.
The detection of PIV was higher with FA RVP than xTAG RVP Fast. For PIV3, FA RVP was sensitive enough to detect the prolonged shedding of the virus in one patient for an additional month after DFA, culture, and xTAG RVP became negative. This discrepancy could have been due to low viral load, as suggested in a recent study comparing FA RVP to xTAG RVP. Rand et al. showed that most of the FA RVP-positive, xTAG RVP-negative specimens for RSV had high cycle threshold values (20
). Although no cycle values were available for our results, a similar explanation could apply to our results for PIV3 and RSV. This is particularly important specifically for the detection of PIV and RSV, where the asymptomatic shedding of virus is known to be the source of several outbreaks and the detection of low viral burden is likely to have a clinical impact (10
Both assays detected a higher number of rhinovirus/enterovirus infections and several coronaviruses not routinely detected by DFA or culture. This result is similar to that reported by Rand et al. for rhinovirus detection by FA RVP and xTAG RVP (20
). Their comparison did not include coronaviruses, which are not detected by the xTAG RVP assay. Approximately 30% of pediatric patients were positive for rhinovirus (20%) and/or coronavirus (10%). The detection rate of these two viruses by molecular assays was twice that of culture. Similar infection rates have been reported previously among pediatric patients with leukemia and allogeneic HSCT recipients, where infection often occurs in the first 100 days after transplant (12
). As persons with cancer and HSCT recipients can develop devastating complications from respiratory virus infections, the timely and sensitive diagnosis of these viruses is particularly important to develop a better understanding of their clinical impact and to develop effective treatment and control strategies.
Human bocavirus was included in both RVP assays. Bocavirus is a recently identified parvovirus (1
), and its clinical significance as a respiratory pathogen has not been fully elucidated (22
). In our study, 3 patients were infected with confirmed bocavirus, and 2 of these patients had dual infections with rhinovirus, coronavirus, and Flu B. Although more than 90% of our pediatric patients were symptomatic (), including the cases with bocavirus, the presence of additional coinfecting viruses makes it challenging to ascertain the role of bocavirus as a sole respiratory pathogen.
With the implementation of rapid and highly sensitive molecular assays, one of the biggest impacts will be on infection control practices. Currently at MSKCC, patients are placed on droplet precautions until they are asymptomatic and the DFA and culture results are negative (minimum isolation duration is 3 days irrespective of results). With the implementation of PCR, diagnostic yield in symptomatic patients being tested for viral infection will increase. A large number of patients will be found positive for viruses other than the usual suspects (i.e., influenza virus and RSV), infections will likely be detected early, and prolonged shedding will be commonly encountered. Shedding for up to 3 months has been described in 13% of HSCT recipients with rhinovirus and coronavirus infection (16
). The replication competence of viruses among prolonged shedders is often debated, and the best infection control practice in this situation has not been determined. While the overall impact of the transition to PCR-based testing is favorable, some practical challenges will arise in the process.
In conclusion, each assay evaluated in this study had advantages and disadvantages, but overall the performance of FA RVP was superior to that of xTAG RVP Fast for the majority of viruses in the panels. In addition to detecting almost two to three times as many viruses in our pediatric population than DFA/culture, these two assays significantly cut down on the turnaround time for results from several days (e.g., 2 to 3 days for Flu A and up to 7 days for rhinoviruses) to a few hours for FA RVP and up to 24 h for xTAG RVP Fast. Several reports have established the superiority of molecular assays for respiratory virus detection; however, to our knowledge, this study is the first to compare the RUO version of these two assays. FA RVP was FDA cleared in May 2011 for the detection of 15 viruses and subtypes, including influenza A virus, influenza A virus subtype H1, influenza A virus subtype H1 (2009), influenza A virus subtype H3, influenza B virus, RSV, hMPV, rhinovirus/enterovirus, adenovirus, parainfluenza viruses 1 to 4, coronavirus OC43, and coronavirus HKU1. xTAG RVP Fast was FDA cleared in July 2011 but only for the detection of 8 viruses and subtypes, including influenza A virus, influenza A virus subtype H1, influenza A virus subtype H3, influenza B virus, RSV, hMPV, rhinovirus, and adenovirus. Additionally, both assays are FDA cleared only for nasopharyngeal swabs, leaving other specimen types (sputum, bronchoscopic washes, lavages, etc.) to be validated by the individual laboratories.
FA RVP was more sensitive than xTAG RVP Fast, with a turnaround time of approximately 1 h for each specimen. The simplicity and the random-access characteristic make it an excellent choice in a hospital with low to medium volume and on-demand urgent service.