Comparison of Real-Time PCR Assays with Fluorescent-Antibody Assays for Diagnosis of Respiratory Virus Infections in Children
http://www.100md.com
《微生物临床杂志》
Department of Laboratory Medicine, University of Washington, Seattle, Washington
ABSTRACT
Conventional fluorescent-antibody (FA) methods were compared to real-time PCR assays for detection of respiratory syncytial virus (RSV), influenza virus type A (FluA), parainfluenza virus types 1, 2, and 3 (PIV1, PIV2, and PIV3), human metapneumovirus (MPV), and adenovirus (AdV) in 1,138 specimens from children with respiratory illnesses collected over a 1-year period. At least one virus was detected in 436 (38.3%) specimens by FA and in 608 (53.4%) specimens by PCR (P < 0.001). Specimen quality was inadequate for FA in 52 (4.6%) specimens; 13 of these (25%) were positive by PCR. In contrast, 18 (1.6%) specimens could not be analyzed by PCR; 1 of these was positive by FA. The number of specimens positive only by PCR among specimens positive by PCR and/or FA was 18 (7.0%) of 257 for RSV, 18 (13.4%) of 134 for FluA, 25 (64.1%) of 39 for PIV1, 8 (88.9%) of 9 for PIV2, 17 (30.1%) of 55 for PIV3, and 101 (76.5%) of 132 for AdV. MPV was detected in 6.6% of all specimens and in 9.5% of the 702 specimens negative by FA. The mean number of virus copies per milliliter in specimens positive by both PCR and FA was significantly higher, at 6.7 x 107, than that in specimens positive only by PCR, at 4.1 x 104 (P < 0.001). The PCR assays were significantly more sensitive than FA assays for detecting respiratory viruses, especially parainfluenza virus and adenovirus. Use of real-time PCR to identify viral respiratory pathogens in children will lead to improved diagnosis of respiratory illness.
INTRODUCTION
Accurate detection of respiratory viruses is important to guide antiviral therapy, prevent nosocomial spread, provide surveillance, and in some cases, decrease hospital costs and lengths of stay (1, 2, 11, 21). By using standard laboratory methods, such as staining with fluorescent antibodies (FA) and isolation by culture, viruses have been detected in 13 to 45% of children with symptoms of respiratory illness (3, 8, 12, 22, 28). Disadvantages of FA include requiring multiple reagents which may vary in sensitivity, potential variability in technical reading, and the need for an adequate number of cells to examine each specimen. Several studies have shown that PCR methods appear to be more sensitive than FA and culture for the diagnosis of acute respiratory virus infections (8, 22, 23, 24, 26, 28). PCR is less affected by specimen quality and transport and provides an objective interpretation of results. Real-time PCR technology, which combines nucleic acid amplification with amplicon detection, provides results more quickly than conventional PCR, has in some cases shown improved sensitivity compared to conventional PCR, and provides a uniform platform for quantifying both single and multiple pathogens in a single sample (4, 7, 18).
In this study, separate quantitative real-time reverse transcription (RT)-PCR assays were used to detect six RNA viruses, including respiratory syncytial virus (RSV), influenza virus type A (FluA), parainfluenza virus types 1, 2, and 3 (PIV1, PIV2, and PIV3), and human metapneumovirus (MPV). A quantitative real-time PCR assay was used to detect adenovirus (AdV) DNA. All of the RNA virus assays used identical RT-PCR master mix and cycling parameters. Unique sets of PCR primers and TaqMan probes were designed to target highly conserved sequences in each viral genome. Standard curves generated by amplification of viral RNA transcripts provided absolute quantification of virus copy numbers. Specimen processing controls were included to prevent false-negative results due to reaction inhibitors or inadequate nucleic acid extraction. Results from the PCR assays (both RT-PCR and PCR real-time methods) were compared to those from a standard FA method for the ability to detect six etiologic agents of respiratory infections in specimens from children. A real-time RT-PCR for MPV was also applied to these specimens to determine the prevalence of MPV in this population; an appropriate FA was not available for MPV at the time of this study.
MATERIALS AND METHODS
Clinical specimens. From October 2003 through September 2004, 1,138 consecutive specimens (1,074 nasal wash samples, 14 nasal swabs, 44 tracheal aspirates, and 6 bronchoalveolar lavage [BAL] specimens) submitted to the University of Washington Virology Laboratory for respiratory virus FA or FA and culture were tested by PCR. The 1,138 specimens, representing approximately one-third of the total pediatric specimens submitted during this time period, were those that contained sufficient residual material for PCR testing. The median age of the patients from whom the specimens were collected was 16 months (range = 1 day to 19 years); 41.8% of patients were less than 1 year old, and 79.7% were less than 5 years old. Fifty-six percent of samples were from male patients, and 44% were from female patients. There were no significant differences between the median ages or the FA results of the patients whose samples were tested by PCR and those whose samples had insufficient volume for testing.
Respiratory virus antigen detection (FA). Specimens were tested for RSV, PIV (types 1 to 4), FluA, influenza type B (FluB), and AdV by use of an indirect fluorescent-antibody assay optimized to yield the most accurate and reliable results possible. After addition of an antibiotic solution and aspiration and expulsion with a Pasteur pipette to break up mucus, the respiratory specimens were centrifuged for 10 min at 4°C at 700 x g. If a pellet was visible, cells were dripped onto a 10-well slide and air dried. If no pellet was seen, a slide was prepared by cytocentrifugation. Additional washes were done if the cell pellet was too thick with mucus. The slides were fixed in acetone and incubated with virus-specific mouse monoclonal antibodies (Chemicon, Temecula, CA). Each lot of antiserum was titrated to its optimum dilution, and fresh reagents were prepared weekly. After a 30-min incubation at 37°C with primary antibody, goat anti-mouse fluorescein-conjugated monoclonal antibodies (ICN Biomedicals, Inc., Costa Mesa, CA) were applied to the sample wells, and the slides were incubated for an additional 30 min at 37°C. Slides were washed and read immediately using a fluorescence microscope. The presence of bright green fluorescence within intact cells was considered positive. Slides with too few intact cells were considered inadequate for analysis. Each FA slide was read twice by technologists with an average of 15 years of virology laboratory experience. This careful processing resulted in an interpretable FA slide in 95% of samples; approximately 1% of samples had a discrepant reading.
Sample preparation for PCR assays. Total nucleic acids were isolated from 200 μl of each respiratory specimen as previously described (13). To ensure that negative results were not due to poor nucleic acid extraction or inhibition of the PCR assay, 1,000 copies/PCR of the EXO specimen processing control, a 262-base RNA transcript derived from jellyfish DNA (16), were added to the lysis buffer. One low-positive control containing 200 to 1,000 copies/PCR of each respiratory virus harvested from cell culture and diluted in minimal essential medium and one negative control consisting of cultured, uninfected human epithelial cells were processed with each batch of clinical specimens.
TaqMan assays. Specimens were tested for RSV, PIV types 1, 2, and 3, FluA, MPV, and AdV using seven separate, quantitative PCR assays. Samples were analyzed without knowledge of the patient's FA result. One hundred specimens (8.8%) were not tested for AdV due to insufficient sample volume. Although PCR tests were developed for FluB and PIV type 4, they were not used to test these specimens because only one sample each was positive by FA for FluB and for PIV type 4 during this period.
The RT-PCRs were performed for the RNA viruses using a one-step RT-PCR master mix as previously described (13). AdV DNA was detected using a PCR master mix (Quantitect multiplex PCR kit; QIAGEN, Inc., Valencia, CA) and the following cycling parameters: 50°C for 2 min, 95°C for 15 min, and 45 cycles of 94°C for 1 min and 60°C for 1 min. The PCR primer and probe sequences for detection of RSV (13) and MPV (14) have been previously described. The primers and 6-carboxyfluorescein (FAM)-labeled probes for detection of FluA, PIV1, PIV2, PIV3, and AdV were designed using Primer Express software (Applied Biosystems, Foster City, CA) from multiple aligned sequences of each virus obtained from the NCBI database. The primer and probe characteristics and sequences are given in Table 1. The assays for RSV, FluA, PIV1, PIV3, and AdV were duplexed with a second primer set and VIC-labeled probe (13) that amplified and detected the exogenously added EXO specimen processing control.
RNA transcripts for each RNA virus amplicon were synthesized, purified, and quantified as previously described (13). Contaminating DNA was not detected by PCR amplification in up to 106 RNA transcript copies. Tenfold serial dilutions of 107 to 10 copies of each RNA transcript for the RNA viruses and EXO were added to the RT-PCRs in duplicate. Similar dilutions were prepared using a plasmid containing the AdV amplicon sequences. The results were used to generate standard curves for quantification of the respiratory viruses and EXO in clinical samples. The threshold cycles of samples were compared to the standard curves; results were expressed as virus copies per milliliter of original sample. All samples with negative respiratory virus results required detection of at least 200 EXO copies per reaction to be considered valid. Nucleic acid extraction (if sufficient sample volume was available) and PCR were repeated on all samples that were negative for both respiratory viruses and EXO. Specimens that did not amplify EXO after repeat analysis were considered unsatisfactory for PCR.
Assay validation. The specificity of the PCR assays was assessed by testing RNA or DNA purified from at least two culture isolates of 19 viruses that might be found in respiratory specimens, including RSV, PIV types 1, 2, 3, and 4, FluA, FluB, rhinovirus, coronavirus, enterovirus, coxsackie B virus, AdV, and herpes virus types 1 through 8. The specificity of the AdV assay was also tested on 39 AdV serotypes obtained from the ATCC and on isolates of JC virus and BK virus. Each assay detected only the target for which the primers and probes were designed. The sensitivity of each assay was determined using 10-fold serial dilutions of previously quantified positive specimens. Each assay could reliably detect 10 virus copies per reaction, providing a sensitivity of 1,000 copies/ml for the RNA viruses (10 μl of specimen added per reaction) and 500 copies/ml for AdV (20 μl of specimen added per reaction). Specimens with positive results less than 10 copies/reaction were repeated to confirm positivity. The performances of the assays for RSV (13) and MPV (14) have previously been reported. Specimens that were positive for PIV types 1, 2, or 3 and negative by FA were confirmed using alternate PCR assays for PIV types 1, 2, and 3 that target the hemagglutinin neuraminidase (HN) gene (27).
Statistical analysis. Significant differences between groups were determined by the Wilcoxon rank sum test for comparison of medians, the t test for comparison of means, and the chi-square test for comparison of proportions. Logistic regression modeling was used to determine the factors that predicted the FA result.
RESULTS
Results of respiratory virus detection by FA and PCR. The results of FA and PCR on 1,138 specimens from children with respiratory illness are shown in Table 2. One or more of the seven respiratory viruses was detected in 436 (38.3%) of the 1,138 respiratory specimens by FA and in 608 (53.4%) by PCR (P < 0.001), including 58 specimens in which only MPV was detected. Specimen quality was inadequate for FA due to insufficient cells in 52 (4.6%) of 1,138 specimens (1 swab, 3 tracheal aspirates, and 48 nasal washes). Of these 52 specimens, 13 (25%) were positive by PCR (6 for RSV, 2 for FluA, 1 for PIV3, 1 for AdV, 2 for MPV, and 1 for both PIV3 and MPV). Eighteen (1.6%) of the 1,138 specimens (all nasal washes) were considered unsatisfactory for PCR based on low EXO control amplification, 1 of which was positive for RSV by FA. Six (1.4%) of the 436 specimens positive by FA and 83 (13.7%) of the 608 specimens positive by PCR were positive for more than one virus. The nasal wash and swab specimens were more likely to be positive by either FA or PCR (426 [39.2%] and 590 [54.2%], respectively, of 1,088) than were the tracheal aspirates and BAL specimens (10 [20%] and 18 [36%], respectively, of 50) (P < 0.007).
The number of specimens positive for each of the seven respiratory viruses by FA and PCR among the 1,138 tested (1,038 for AdV) is shown in Table 3. One of the two specimens positive for RSV by FA only was unsatisfactory for amplification by PCR. Fewer positive specimens were detected by FA than by PCR for every virus. However, there were differences between the viruses in the proportion of specimens positive only by PCR. Only 18 (7.1%) of 255 and 18 (13.4%) of 134 specimens positive for RSV and FluA, respectively, by PCR were negative by FA. In contrast, 50 (48.5%) of 103 specimens positive for any type of PIV and 101 (77.7%) of 130 specimens positive for AdV by PCR were negative by FA.
The 50 specimens positive by PCR and negative by FA for any PIV (25 PIV1, 8 PIV2, and 17 PIV3) were confirmed by testing with alternate PCR assays employing primers and probes that target the HN gene mRNA of these viruses (27). All but two were positive using the HN PCR assays. The two specimens not confirmed by the HN PCR had relatively low numbers (1.22 x 104 and 2.33 x 104 copies/ml) of PIV3, whereas the other 15 PIV3 specimens had a mean of 2.63 x 106 copies/ml.
MPV was not included in the FA assay but was detected by PCR in 6.6% (75 of 1,138) of all specimens and in 9.5% (67 of 702) of the specimens negative for other respiratory viruses by FA. Two specimens reported as "probable RSV" by FA were RSV negative by PCR. However, one was positive for FluA by PCR and the other was positive for MPV by PCR. Other findings included one BAL specimen positive for PIV type 4 and one nasal swab positive for FluB by both FA and PCR (data not shown).
The number of specimens with more than one virus detected by FA and PCR is shown in Table 3. All six of the specimens positive for more than one virus by FA were positive for AdV and either RSV (three specimens), FluA (two specimens), or PIV3 (one specimen). Of the 83 specimens positive for multiple viruses by PCR, 59 (71.1%) were positive for AdV and either RSV (32 specimens), FluA (15 specimens), PIV3 (6 specimens), MPV (6 specimens), or PIV1 (4 specimens). More than one virus was detected by PCR in approximately half of all specimens containing PIV2 (5 of 9, 55.6%) or AdV (59 of 130, 45.4%). In contrast, only 16.5% (42 of 255) or 17.2% (23 of 134) of the specimens positive by PCR for RSV or FluA, respectively, contained more than one respiratory virus.
Respiratory virus detection by age. The prevalence of a documented viral respiratory illness was higher for patients less than 5 years old than for patients aged 5 years or older when detected by either method. The median ages of patients with specimens positive by FA and by PCR were 12 months and 15 months, respectively. By FA, 389 (42.9%) of 907 patients less than 5 years old and 47 (20.3%) of 231 patients aged 5 years and older were positive for any respiratory virus (P < 0.005). By PCR, 531 (58.5%) of 907 patients less than 5 years old and 77 (33.3%) of 231 patients aged 5 years and older were positive for any respiratory virus (P < 0.005). More than one virus was detected by PCR in specimens from 77 (8.5%) of 907 patients less than 5 years old compared to 6 (2.6%) of 231 patients 5 years and older (P < 0.005). All six patients with specimens positive for multiple viruses by FA were less than 5 years old. Analyzed by age, the overall respiratory virus prevalence was 10% to 24% lower by FA than by PCR among different age groups of children (Fig. 1).
FIG. 1. Prevalence of detection of any of seven respiratory viruses by FA (dark columns) and PCR (light columns) among different age groups of children.
Virus quantity was also associated with patient age. Among the 622 results that were positive by PCR, specimens from children less than 12 months old were more likely to have virus quantities greater than 106 copies/ml (78%, 208 of 267) than specimens from children 12 months and older (62%, 220 of 355) (P < 0.001).
Comparison of specimens with positive PCR results by FA result. Differences between specimens that were positive by PCR and positive or negative by FA were examined after specimens that were positive for MPV only were excluded. The mean viral load in samples that were positive for any of the six respiratory viruses by both FA and PCR was 6.1 x 107 copies/ml versus 4.1 x 104 copies/ml for 187 viral infections detected only by PCR (P < 0.001). Figure 2A shows the number of specimens that were positive or negative by FA for any respiratory virus, stratified by PCR viral load. Among the 187 cases of virus detection that were negative by FA, 157 (84%) had fewer than 1 x 106 virus copies/ml compared to only 37 (8.5%) of 435 results that were positive by FA and PCR for the same virus (P < 0.001). The number and mean number of viruses (expressed as log10 copies/ml) in specimens that were positive by PCR and positive or negative by FA are shown for the six individual viruses in Fig. 2B. For every virus, the mean number of virus copies/ml was lower in specimens that were negative by FA than in specimens that were positive by FA. The mean viral load for the 75 MPV-positive specimens was 3.8 x 107 copies/ml.
FIG. 2. (A) The number of specimens positive for any of six respiratory viruses by viral load (log10 respiratory virus copies/ml) among specimens that were negative by FA (light columns) or positive by FA (dark columns). (B) The number of log10 copies/ml for six individual respiratory viruses quantified by PCR in respiratory specimens that were positive (diamonds) and negative (triangles) by FA.
The specimens that were positive by both PCR and FA and the specimens that were positive by PCR and negative by FA also differed by patient age (Table 4). Overall, patients with specimens that were positive by FA and PCR for the same virus were significantly younger (median age = 12 months) than patients with specimens that were negative by FA and positive by PCR (median age = 19 months) (P < 0.001). This relationship was true for all of the individual viruses except AdV, although the differences were significantly different only for PIV1 (P = 0.04). Multiple logistic regression modeling was performed to assess whether the effect due to age was still present after consideration of virus quantity. Although the quantity of virus still predicted FA positivity (odds ratio = 4.7; 95% confidence interval, 3.7 to 6.1; P < 0.0001), age was no longer predictive of FA positivity (odds ratio = 0.97; 95% confidence interval, 0.89 to 1.05; P = 0.45).
DISCUSSION
The real-time PCR assays described here were significantly more sensitive than FA for the detection of respiratory viruses in clinical specimens from children, detecting 40% more positive specimens and missing only four specimens positive by FA. In fact, respiratory viruses were detected by PCR in 25% of the specimens judged inadequate for analysis by FA. The PCR assays also identified specimens with multiple viral infections with a 10-fold greater frequency than did FA. Detection of MPV, which was not included in the FA panel, accounted for about 30% of the increased sensitivity of PCR in this study. The results of the individual PCR assays showed differences among the six viruses included in the FA panel in the proportion that were also positive by FA. While about 90% of the RSV and FluA specimens positive by PCR were also positive by FA, a much smaller proportion of the specimens positive for any PIV or for AdV was detected by FA. These same respiratory viruses were also more likely to be detected in specimens containing more than one virus. PIV and AdV infections accounted for more than half of the results that were positive by PCR only and were detected in more than 90% of the specimens containing more than one virus. The specimens that were negative by FA and positive by PCR and that contained PIV1, 2, or 3 were confirmed using PCR assays that target a different PIV gene.
The AdV PCR results differed in several ways from the RNA virus results. First, the median age of patients with negative FA results was lower than that of patients with positive FA results. Second, the majority of specimens that were negative by FA and positive by PCR (101 of 187) were AdV positive, and most of these specimens contained low numbers of virus (mean log10 copies/ml of AdV-positive specimens that were negative by FA was 4.43). AdV was also the most frequently detected virus in mixed infections. AdV remains detectable in the respiratory tract after primary infection. AdV group B2 strains were detected in 49 of 50 BAL specimens from patients without AdV-related respiratory illness, indicating that AdV is present in asymptomatic patients (15). The children tested in our study reported symptoms of respiratory illness. However, it is possible that the presence of low numbers of AdV was not the cause of the illness but a coincidental presence of persistent virus. Information about the severity and duration of illness in the AdV-positive children which could help interpret these results was not available. Additional studies using quantitative PCR methods and including clinical data are being performed to determine what AdV viral load is considered significant for a pathogenic role in respiratory illness.
Detection of the respiratory viruses was less sensitive by FA than by PCR for all patient age groups. In spite of this difference in sensitivity between the methods, the median ages of patients positive for a respiratory virus were similar when the specimens were tested by either FA or PCR. Not surprisingly, respiratory viruses were more frequently detected by both FA and PCR in children less than 5 years old (17, 25) than in older children. The additional finding that infection with more than one respiratory virus was more common in children less than 5 years old has not previously been reported. Younger age was also associated with larger viral quantity.
Findings similar to those in this study have been previously reported by other investigators using conventional PCR assays for both pediatric and adult populations (3, 5, 6, 8, 9, 10, 19, 20, 22, 28). Conventional PCR assays doubled the detection rate compared to culture for six respiratory viruses in samples obtained from children less than 5 years old hospitalized with acute respiratory infections in one study (28) and increased the prevalence of seven respiratory viruses from 26.9% by FA to 33.8% in specimens from children with acute respiratory infections in another study (22). Comparisons of real-time PCR with FA and culture have previously been reported only for adult populations. Compared to conventional methods, real-time PCR increased the detection of 15 respiratory pathogens in adult patients with community-acquired pneumonia from 49.5% to 76% and the detection of mixed infections from 2.8% to 26.5% (24). Compared to culture, real-time PCR increased the detection of 10 respiratory viruses in adult stem cell transplant patients from 21% to 63% in patients with illness and from 1% to 9% in asymptomatic patients (26). In addition to a comparison of specific respiratory virus detection, the use of quantitative assays in our study provides new information about the quantity of specific viruses in specimens that are positive by PCR and negative by FA.
Most of the respiratory viruses not detected by FA had low copy numbers of viral nucleic acid present. The mean number of virus copies/ml in specimens that were positive by PCR and also positive by FA was more than 3 logs higher than the number in specimens that were negative by FA. Overall, the FA assay detected only 19% (37 of 194) of the respiratory viruses with viral loads less than 106 copies/ml. For the individual viruses, the difference in FA detection between RSV and FluA on the one hand and PIV and AdV on the other was due to differences in the quantity of virus in these specimens. The viruses most often detected in specimens that were positive by PCR and negative by FA, and those more likely to be detected in a mixed infection with another virus, such as PIV and AdV, were also more frequently present in low copy numbers (less than 106 copies/ml) and, consequently, had the lowest mean number of virus copies/ml. By univariate analysis, specimens that were positive by PCR and negative by FA were more likely to be from older patients than specimens that were positive by both PCR and FA. However, there was no association when analyzed by multiple regression.
In this study, real-time PCR was a significant improvement over FA for detection of respiratory viruses in clinical specimens from children, especially for detection of MPV, PIV, and AdV. These data are not likely due to false-positive PCR results but rather quantities of virus that are below the level of detection of FA. In addition to improved detection, in high-volume laboratories, real-time PCR can provide results faster than FA. Although reagent and instrument costs are higher for PCR than for FA (13), real-time PCR requires less hands-on time per specimen than FA, which is labor-intensive. Real-time PCR assays offer advantages over conventional PCR by providing lower risk of false-positive results due to amplicon contamination, identification of the etiologic agent in a clinically relevant time period, and quantification of viral load. Although the diagnostic value of knowing an individual's virus copy number is currently unclear for pediatric respiratory infections, quantification of respiratory viruses can provide important information about the role of these pathogens in respiratory tract infections among all patient groups. Ongoing investigations will help determine any associations between viral load and clinical signs and symptoms. These real-time PCR assays are useful tools for further investigations on the epidemiology of and disease associated with respiratory viruses and will provide information to better understand the relationship between illness and the quantity of virus being shed.
ACKNOWLEDGMENTS
We thank the staff at the University of Washington Clinical Virology Laboratory for performing the FA tests and providing specimen aliquots for PCR testing and Amalia Meier for statistical help.
This work was supported in part by National Cancer Institute grant number 5 PO1 CA018029-30.
FOOTNOTES
REFERENCES
Barenfanger, J., C. Drake, N. Leon, T. Mueller, and T. Troutt. 2000. Clinical and financial benefits of rapid detection of respiratory viruses: an outcomes study. J. Clin. Microbiol. 38:2824-2828.
Beekman, S. E., H. D. Engler, A. S. Collins, J. Canosa, D. K. Henderson, and A. Freifield. 1996. Rapid identification of respiratory viruses: impact on isolation practices and transmission among immunocompromised pediatric patients. Infect. Control Hosp. Epidemiol. 17:581-586.
Bellau-Pujol, S., A. Vabret, L. Legrand, J. Dina, S. Gouarin, J. Petitjean-Lecherbonnier, B. Pozzetto, C. Ginevra, and F. Freymuth. 2005. Development of three multiplex RT-PCR assays for the detection of 12 respiratory viruses. J. Virol. Methods 126:53-63.
Cockerill, F. R., III. 2003. Application of rapid-cycle real-time polymerase chain reaction for diagnostic testing in the clinical microbiology laboratory. Arch. Pathol. Lab. Med. 127:1112-1120.
Coiras, M. T., P. Perez-Brea, M. L. García, and I. Casas. 2003. Simultaneous detection of influenza A, B, C viruses, respiratory syncytial virus, and adenoviruses in clinical samples by multiplex reverse transcription nested-PCR assay. J. Med. Virol. 69:132-144.
Coyle, P. V., G. M. Ong, H. J. O'Neill, C. McCaughey, D. De Ornellas, F. Mitchell, S. J. Mitchell, S. A. Feeney, D. E. Wyatt, M. Forde, and J. Stockton. 2004. A touchdown nucleic acid amplification protocol as an alternative to culture backup for immunofluorescence in the routine diagnosis of acute viral respiratory tract infections. BMC Microbiol. 4:41-49.
Dagher, H., H. Donniger, P. Hutchinson, R. Ghildyal, and P. Bardin. 2004. Rhinovirus detection: comparison of real-time and conventional PCR. J. Virol. Methods 117:113-121.
Erdman, D. D., G. A. Weinberg, K. M. Edwards, F. J. Walker, B. C. Anderson, J. Winter, M. Gonzalez, and L. J. Anderson. 2003. GeneScan reverse transcription-PCR assay for detection of six common respiratory viruses in young children hospitalized with acute respiratory illness. J. Clin. Microbiol. 41:4298-4303.
Grndahl, B., W. Puppe, A. Hoppe, I. Kühne, J. A. I. Weigl, and H.-J. Schmitt. 1999. Rapid identification of nine microorganisms causing acute respiratory tract infections by single-tube multiplex reverse transcription-PCR: feasibility study. J. Clin. Microbiol. 37:1-7.
Gruteke, P., A. S. Glas, M. Dierdorp, W. B. Vreede, J.-W. Pilon, and S. M. Bruisten. 2004. Practical implementation of a multiplex PCR for acute respiratory tract infections in children. J. Clin. Microbiol. 42:5596-5603.
Henrickson, K. J. 2005. Cost-effective use of rapid diagnostic techniques in the treatment and prevention of viral respiratory infections. Pediatr. Ann. 34:24-31.
Kehl, S. C., K. J. Henrickson, W. Hua, and J. Fan. 2001. Evaluation of the Hexaplex assay for detection of respiratory viruses in children. J. Clin. Microbiol. 39:1696-1701.
Kuypers, J., N. Wright, and R. Morrow. 2004. Evaluation of quantitative and type-specific real-time PCR assays for detection of RSV in respiratory specimens from children. J. Clin. Virol. 31:123-129.
Kuypers, J., N. Wright, L. Corey, and R. Morrow. 2005. Detection and quantification of human metapneumovirus in pediatric specimens by real-time PCR. J. Clin. Virol. 33:299-305.
Leung, A. Y., M. Chan, V. C. Cheng, K. Y. Yuen, and Y. L. Kwong. 2005. Quantification of adenovirus in the lower respiratory tract of patients without clinical adenovirus-related respiratory disease. Clin. Infect. Dis. 40:1541-1544.
Limaye, A. P., M.-L. Huang, W. Leisenring, L. Stensland, L. Corey, and M. Boeckh. 2001. Cytomegalovirus (CMV) DNA load in plasma for the diagnosis of CMV disease before engraftment in hematopoietic stem-cell transplant recipients. J. Infect. Dis. 183:377-382.
Monto, A. S. 2002. Epidemiology of viral respiratory infections. Am. J. Med. 112:4S-12S.
Poon, L. L., K. H. Chan, O. K. Wong, T. K. Cheung, I. Ng, B. Zheng, W. H. Seto, K. Y. Yuen, Y. Guan, and J. S. Peiris. 2004. Detection of SARS coronavirus in patients with severe acute respiratory syndrome by conventional and real-time quantitative reverse transcription-PCR assays. Clin. Chem. 50:67-72.
Puppe, W., J. A. I. Weigl, G. Aron, B. Grndahl, H.-J. Schmitt, H. G. M. Niesters, and J. Groen. 2004. Evaluation of a multiplex reverse transcriptase PCR ELISA for the detection of nine respiratory tract pathogens. J. Clin. Virol. 30:165-174.
Rovida, F., E. Percivalle, M. Zavattoni, M. Torsellini, A. Sarasini, G. Campanini, S. Paolucci, F. Baldanti, M. G. Revello, and G. Gerna. 2005. Monoclonal antibodies versus reverse transcription-PCR for detection of respiratory viruses in a patient population with respiratory tract infections admitted to hospital. J. Med. Virol. 75:336-347.
Skerret, S. J. 1997. Diagnostic testing to establish a microbial cause is helpful in the management of community-acquired pneumonia. Semin. Respir. Infect. 12:308-321.
Syrmis, M. W., D. M. Whiley, M. Thomas, I. M. Mackay, J. Williamson, D. J. Siebert, M. D. Nissen, and T. P. Sloots. 2004. A sensitive, specific, and cost-effective multiplex reverse transcriptase-PCR assay for the detection of seven common respiratory viruses in respiratory samples. J. Mol. Diagn. 6:125-131.
Templeton, K. E., S. A. Scheltinga, M. R. C. Beersma, A. C. M. Kroes, and E. C. J. Claas. 2004. Rapid and sensitive method using multiplex real-time PCR for diagnosis of infections by influenza A and influenza B viruses, respiratory syncytial virus, and parainfluenza viruses 1, 2, 3, and 4. J. Clin. Microbiol. 42:1564-1569.
Templeton, K. E., S. A. Scheltinga, W. C. J. F. M. van den Eeden, A. W. Graffelman, P. J. van den Broek, and E. C. J. Claas. 2005. Improved diagnosis of the etiology of community-acquired pneumonia with real-time polymerase chain reaction. Clin. Infect. Dis. 41:345-351.
van Gageldonk-Lafeber, A. B., M.-L. A. Heijnen, A. I. M. Bartelds, M. F. Peters, S. M. van der Plas, and B. Wilbrink. 2005. A case-control study of acute respiratory tract infection in general practice patients in The Netherlands. Clin. Infect. Dis. 41:490-497.
van Kraaij, M. G. J., L. J. R. van Elden, A. M. van Loon, K. A. W. Hendricksen, L. Laterveer, A. W. Dekker, and M. Nijhuis. 2005. Frequent detection of respiratory viruses in adult recipients of stem cell transplants with the use of real-time polymerase chain reaction, compared with viral culture. Clin. Infect. Dis. 40:662-669.
Watzinger, F., M. Suda, S. Preuner, R. Baumgartinger, K. Ebner, L. Baskova, H. G. M. Niesters, A. Lawitschka, and T. Lion. 2004. Real-time quantitative PCR assays for detection and monitoring of pathogenic human viruses in immunosuppressed pediatric patients. J. Clin. Microbiol. 42:5189-5198.
Weinberg, G. A., D. D. Erdman, K. M. Edwards, C. B. Hall, F. J. Walker, M. R. Griffin, B. Schwartz, and the New Vaccine Surveillance Network Study Group. 2004. Superiority of reverse-transcription polymerase chain reaction to conventional viral culture in the diagnosis of acute respiratory tract infections in children. J. Infect. Dis. 189:706-710.(Jane Kuypers, Nancy Wrigh)
ABSTRACT
Conventional fluorescent-antibody (FA) methods were compared to real-time PCR assays for detection of respiratory syncytial virus (RSV), influenza virus type A (FluA), parainfluenza virus types 1, 2, and 3 (PIV1, PIV2, and PIV3), human metapneumovirus (MPV), and adenovirus (AdV) in 1,138 specimens from children with respiratory illnesses collected over a 1-year period. At least one virus was detected in 436 (38.3%) specimens by FA and in 608 (53.4%) specimens by PCR (P < 0.001). Specimen quality was inadequate for FA in 52 (4.6%) specimens; 13 of these (25%) were positive by PCR. In contrast, 18 (1.6%) specimens could not be analyzed by PCR; 1 of these was positive by FA. The number of specimens positive only by PCR among specimens positive by PCR and/or FA was 18 (7.0%) of 257 for RSV, 18 (13.4%) of 134 for FluA, 25 (64.1%) of 39 for PIV1, 8 (88.9%) of 9 for PIV2, 17 (30.1%) of 55 for PIV3, and 101 (76.5%) of 132 for AdV. MPV was detected in 6.6% of all specimens and in 9.5% of the 702 specimens negative by FA. The mean number of virus copies per milliliter in specimens positive by both PCR and FA was significantly higher, at 6.7 x 107, than that in specimens positive only by PCR, at 4.1 x 104 (P < 0.001). The PCR assays were significantly more sensitive than FA assays for detecting respiratory viruses, especially parainfluenza virus and adenovirus. Use of real-time PCR to identify viral respiratory pathogens in children will lead to improved diagnosis of respiratory illness.
INTRODUCTION
Accurate detection of respiratory viruses is important to guide antiviral therapy, prevent nosocomial spread, provide surveillance, and in some cases, decrease hospital costs and lengths of stay (1, 2, 11, 21). By using standard laboratory methods, such as staining with fluorescent antibodies (FA) and isolation by culture, viruses have been detected in 13 to 45% of children with symptoms of respiratory illness (3, 8, 12, 22, 28). Disadvantages of FA include requiring multiple reagents which may vary in sensitivity, potential variability in technical reading, and the need for an adequate number of cells to examine each specimen. Several studies have shown that PCR methods appear to be more sensitive than FA and culture for the diagnosis of acute respiratory virus infections (8, 22, 23, 24, 26, 28). PCR is less affected by specimen quality and transport and provides an objective interpretation of results. Real-time PCR technology, which combines nucleic acid amplification with amplicon detection, provides results more quickly than conventional PCR, has in some cases shown improved sensitivity compared to conventional PCR, and provides a uniform platform for quantifying both single and multiple pathogens in a single sample (4, 7, 18).
In this study, separate quantitative real-time reverse transcription (RT)-PCR assays were used to detect six RNA viruses, including respiratory syncytial virus (RSV), influenza virus type A (FluA), parainfluenza virus types 1, 2, and 3 (PIV1, PIV2, and PIV3), and human metapneumovirus (MPV). A quantitative real-time PCR assay was used to detect adenovirus (AdV) DNA. All of the RNA virus assays used identical RT-PCR master mix and cycling parameters. Unique sets of PCR primers and TaqMan probes were designed to target highly conserved sequences in each viral genome. Standard curves generated by amplification of viral RNA transcripts provided absolute quantification of virus copy numbers. Specimen processing controls were included to prevent false-negative results due to reaction inhibitors or inadequate nucleic acid extraction. Results from the PCR assays (both RT-PCR and PCR real-time methods) were compared to those from a standard FA method for the ability to detect six etiologic agents of respiratory infections in specimens from children. A real-time RT-PCR for MPV was also applied to these specimens to determine the prevalence of MPV in this population; an appropriate FA was not available for MPV at the time of this study.
MATERIALS AND METHODS
Clinical specimens. From October 2003 through September 2004, 1,138 consecutive specimens (1,074 nasal wash samples, 14 nasal swabs, 44 tracheal aspirates, and 6 bronchoalveolar lavage [BAL] specimens) submitted to the University of Washington Virology Laboratory for respiratory virus FA or FA and culture were tested by PCR. The 1,138 specimens, representing approximately one-third of the total pediatric specimens submitted during this time period, were those that contained sufficient residual material for PCR testing. The median age of the patients from whom the specimens were collected was 16 months (range = 1 day to 19 years); 41.8% of patients were less than 1 year old, and 79.7% were less than 5 years old. Fifty-six percent of samples were from male patients, and 44% were from female patients. There were no significant differences between the median ages or the FA results of the patients whose samples were tested by PCR and those whose samples had insufficient volume for testing.
Respiratory virus antigen detection (FA). Specimens were tested for RSV, PIV (types 1 to 4), FluA, influenza type B (FluB), and AdV by use of an indirect fluorescent-antibody assay optimized to yield the most accurate and reliable results possible. After addition of an antibiotic solution and aspiration and expulsion with a Pasteur pipette to break up mucus, the respiratory specimens were centrifuged for 10 min at 4°C at 700 x g. If a pellet was visible, cells were dripped onto a 10-well slide and air dried. If no pellet was seen, a slide was prepared by cytocentrifugation. Additional washes were done if the cell pellet was too thick with mucus. The slides were fixed in acetone and incubated with virus-specific mouse monoclonal antibodies (Chemicon, Temecula, CA). Each lot of antiserum was titrated to its optimum dilution, and fresh reagents were prepared weekly. After a 30-min incubation at 37°C with primary antibody, goat anti-mouse fluorescein-conjugated monoclonal antibodies (ICN Biomedicals, Inc., Costa Mesa, CA) were applied to the sample wells, and the slides were incubated for an additional 30 min at 37°C. Slides were washed and read immediately using a fluorescence microscope. The presence of bright green fluorescence within intact cells was considered positive. Slides with too few intact cells were considered inadequate for analysis. Each FA slide was read twice by technologists with an average of 15 years of virology laboratory experience. This careful processing resulted in an interpretable FA slide in 95% of samples; approximately 1% of samples had a discrepant reading.
Sample preparation for PCR assays. Total nucleic acids were isolated from 200 μl of each respiratory specimen as previously described (13). To ensure that negative results were not due to poor nucleic acid extraction or inhibition of the PCR assay, 1,000 copies/PCR of the EXO specimen processing control, a 262-base RNA transcript derived from jellyfish DNA (16), were added to the lysis buffer. One low-positive control containing 200 to 1,000 copies/PCR of each respiratory virus harvested from cell culture and diluted in minimal essential medium and one negative control consisting of cultured, uninfected human epithelial cells were processed with each batch of clinical specimens.
TaqMan assays. Specimens were tested for RSV, PIV types 1, 2, and 3, FluA, MPV, and AdV using seven separate, quantitative PCR assays. Samples were analyzed without knowledge of the patient's FA result. One hundred specimens (8.8%) were not tested for AdV due to insufficient sample volume. Although PCR tests were developed for FluB and PIV type 4, they were not used to test these specimens because only one sample each was positive by FA for FluB and for PIV type 4 during this period.
The RT-PCRs were performed for the RNA viruses using a one-step RT-PCR master mix as previously described (13). AdV DNA was detected using a PCR master mix (Quantitect multiplex PCR kit; QIAGEN, Inc., Valencia, CA) and the following cycling parameters: 50°C for 2 min, 95°C for 15 min, and 45 cycles of 94°C for 1 min and 60°C for 1 min. The PCR primer and probe sequences for detection of RSV (13) and MPV (14) have been previously described. The primers and 6-carboxyfluorescein (FAM)-labeled probes for detection of FluA, PIV1, PIV2, PIV3, and AdV were designed using Primer Express software (Applied Biosystems, Foster City, CA) from multiple aligned sequences of each virus obtained from the NCBI database. The primer and probe characteristics and sequences are given in Table 1. The assays for RSV, FluA, PIV1, PIV3, and AdV were duplexed with a second primer set and VIC-labeled probe (13) that amplified and detected the exogenously added EXO specimen processing control.
RNA transcripts for each RNA virus amplicon were synthesized, purified, and quantified as previously described (13). Contaminating DNA was not detected by PCR amplification in up to 106 RNA transcript copies. Tenfold serial dilutions of 107 to 10 copies of each RNA transcript for the RNA viruses and EXO were added to the RT-PCRs in duplicate. Similar dilutions were prepared using a plasmid containing the AdV amplicon sequences. The results were used to generate standard curves for quantification of the respiratory viruses and EXO in clinical samples. The threshold cycles of samples were compared to the standard curves; results were expressed as virus copies per milliliter of original sample. All samples with negative respiratory virus results required detection of at least 200 EXO copies per reaction to be considered valid. Nucleic acid extraction (if sufficient sample volume was available) and PCR were repeated on all samples that were negative for both respiratory viruses and EXO. Specimens that did not amplify EXO after repeat analysis were considered unsatisfactory for PCR.
Assay validation. The specificity of the PCR assays was assessed by testing RNA or DNA purified from at least two culture isolates of 19 viruses that might be found in respiratory specimens, including RSV, PIV types 1, 2, 3, and 4, FluA, FluB, rhinovirus, coronavirus, enterovirus, coxsackie B virus, AdV, and herpes virus types 1 through 8. The specificity of the AdV assay was also tested on 39 AdV serotypes obtained from the ATCC and on isolates of JC virus and BK virus. Each assay detected only the target for which the primers and probes were designed. The sensitivity of each assay was determined using 10-fold serial dilutions of previously quantified positive specimens. Each assay could reliably detect 10 virus copies per reaction, providing a sensitivity of 1,000 copies/ml for the RNA viruses (10 μl of specimen added per reaction) and 500 copies/ml for AdV (20 μl of specimen added per reaction). Specimens with positive results less than 10 copies/reaction were repeated to confirm positivity. The performances of the assays for RSV (13) and MPV (14) have previously been reported. Specimens that were positive for PIV types 1, 2, or 3 and negative by FA were confirmed using alternate PCR assays for PIV types 1, 2, and 3 that target the hemagglutinin neuraminidase (HN) gene (27).
Statistical analysis. Significant differences between groups were determined by the Wilcoxon rank sum test for comparison of medians, the t test for comparison of means, and the chi-square test for comparison of proportions. Logistic regression modeling was used to determine the factors that predicted the FA result.
RESULTS
Results of respiratory virus detection by FA and PCR. The results of FA and PCR on 1,138 specimens from children with respiratory illness are shown in Table 2. One or more of the seven respiratory viruses was detected in 436 (38.3%) of the 1,138 respiratory specimens by FA and in 608 (53.4%) by PCR (P < 0.001), including 58 specimens in which only MPV was detected. Specimen quality was inadequate for FA due to insufficient cells in 52 (4.6%) of 1,138 specimens (1 swab, 3 tracheal aspirates, and 48 nasal washes). Of these 52 specimens, 13 (25%) were positive by PCR (6 for RSV, 2 for FluA, 1 for PIV3, 1 for AdV, 2 for MPV, and 1 for both PIV3 and MPV). Eighteen (1.6%) of the 1,138 specimens (all nasal washes) were considered unsatisfactory for PCR based on low EXO control amplification, 1 of which was positive for RSV by FA. Six (1.4%) of the 436 specimens positive by FA and 83 (13.7%) of the 608 specimens positive by PCR were positive for more than one virus. The nasal wash and swab specimens were more likely to be positive by either FA or PCR (426 [39.2%] and 590 [54.2%], respectively, of 1,088) than were the tracheal aspirates and BAL specimens (10 [20%] and 18 [36%], respectively, of 50) (P < 0.007).
The number of specimens positive for each of the seven respiratory viruses by FA and PCR among the 1,138 tested (1,038 for AdV) is shown in Table 3. One of the two specimens positive for RSV by FA only was unsatisfactory for amplification by PCR. Fewer positive specimens were detected by FA than by PCR for every virus. However, there were differences between the viruses in the proportion of specimens positive only by PCR. Only 18 (7.1%) of 255 and 18 (13.4%) of 134 specimens positive for RSV and FluA, respectively, by PCR were negative by FA. In contrast, 50 (48.5%) of 103 specimens positive for any type of PIV and 101 (77.7%) of 130 specimens positive for AdV by PCR were negative by FA.
The 50 specimens positive by PCR and negative by FA for any PIV (25 PIV1, 8 PIV2, and 17 PIV3) were confirmed by testing with alternate PCR assays employing primers and probes that target the HN gene mRNA of these viruses (27). All but two were positive using the HN PCR assays. The two specimens not confirmed by the HN PCR had relatively low numbers (1.22 x 104 and 2.33 x 104 copies/ml) of PIV3, whereas the other 15 PIV3 specimens had a mean of 2.63 x 106 copies/ml.
MPV was not included in the FA assay but was detected by PCR in 6.6% (75 of 1,138) of all specimens and in 9.5% (67 of 702) of the specimens negative for other respiratory viruses by FA. Two specimens reported as "probable RSV" by FA were RSV negative by PCR. However, one was positive for FluA by PCR and the other was positive for MPV by PCR. Other findings included one BAL specimen positive for PIV type 4 and one nasal swab positive for FluB by both FA and PCR (data not shown).
The number of specimens with more than one virus detected by FA and PCR is shown in Table 3. All six of the specimens positive for more than one virus by FA were positive for AdV and either RSV (three specimens), FluA (two specimens), or PIV3 (one specimen). Of the 83 specimens positive for multiple viruses by PCR, 59 (71.1%) were positive for AdV and either RSV (32 specimens), FluA (15 specimens), PIV3 (6 specimens), MPV (6 specimens), or PIV1 (4 specimens). More than one virus was detected by PCR in approximately half of all specimens containing PIV2 (5 of 9, 55.6%) or AdV (59 of 130, 45.4%). In contrast, only 16.5% (42 of 255) or 17.2% (23 of 134) of the specimens positive by PCR for RSV or FluA, respectively, contained more than one respiratory virus.
Respiratory virus detection by age. The prevalence of a documented viral respiratory illness was higher for patients less than 5 years old than for patients aged 5 years or older when detected by either method. The median ages of patients with specimens positive by FA and by PCR were 12 months and 15 months, respectively. By FA, 389 (42.9%) of 907 patients less than 5 years old and 47 (20.3%) of 231 patients aged 5 years and older were positive for any respiratory virus (P < 0.005). By PCR, 531 (58.5%) of 907 patients less than 5 years old and 77 (33.3%) of 231 patients aged 5 years and older were positive for any respiratory virus (P < 0.005). More than one virus was detected by PCR in specimens from 77 (8.5%) of 907 patients less than 5 years old compared to 6 (2.6%) of 231 patients 5 years and older (P < 0.005). All six patients with specimens positive for multiple viruses by FA were less than 5 years old. Analyzed by age, the overall respiratory virus prevalence was 10% to 24% lower by FA than by PCR among different age groups of children (Fig. 1).
FIG. 1. Prevalence of detection of any of seven respiratory viruses by FA (dark columns) and PCR (light columns) among different age groups of children.
Virus quantity was also associated with patient age. Among the 622 results that were positive by PCR, specimens from children less than 12 months old were more likely to have virus quantities greater than 106 copies/ml (78%, 208 of 267) than specimens from children 12 months and older (62%, 220 of 355) (P < 0.001).
Comparison of specimens with positive PCR results by FA result. Differences between specimens that were positive by PCR and positive or negative by FA were examined after specimens that were positive for MPV only were excluded. The mean viral load in samples that were positive for any of the six respiratory viruses by both FA and PCR was 6.1 x 107 copies/ml versus 4.1 x 104 copies/ml for 187 viral infections detected only by PCR (P < 0.001). Figure 2A shows the number of specimens that were positive or negative by FA for any respiratory virus, stratified by PCR viral load. Among the 187 cases of virus detection that were negative by FA, 157 (84%) had fewer than 1 x 106 virus copies/ml compared to only 37 (8.5%) of 435 results that were positive by FA and PCR for the same virus (P < 0.001). The number and mean number of viruses (expressed as log10 copies/ml) in specimens that were positive by PCR and positive or negative by FA are shown for the six individual viruses in Fig. 2B. For every virus, the mean number of virus copies/ml was lower in specimens that were negative by FA than in specimens that were positive by FA. The mean viral load for the 75 MPV-positive specimens was 3.8 x 107 copies/ml.
FIG. 2. (A) The number of specimens positive for any of six respiratory viruses by viral load (log10 respiratory virus copies/ml) among specimens that were negative by FA (light columns) or positive by FA (dark columns). (B) The number of log10 copies/ml for six individual respiratory viruses quantified by PCR in respiratory specimens that were positive (diamonds) and negative (triangles) by FA.
The specimens that were positive by both PCR and FA and the specimens that were positive by PCR and negative by FA also differed by patient age (Table 4). Overall, patients with specimens that were positive by FA and PCR for the same virus were significantly younger (median age = 12 months) than patients with specimens that were negative by FA and positive by PCR (median age = 19 months) (P < 0.001). This relationship was true for all of the individual viruses except AdV, although the differences were significantly different only for PIV1 (P = 0.04). Multiple logistic regression modeling was performed to assess whether the effect due to age was still present after consideration of virus quantity. Although the quantity of virus still predicted FA positivity (odds ratio = 4.7; 95% confidence interval, 3.7 to 6.1; P < 0.0001), age was no longer predictive of FA positivity (odds ratio = 0.97; 95% confidence interval, 0.89 to 1.05; P = 0.45).
DISCUSSION
The real-time PCR assays described here were significantly more sensitive than FA for the detection of respiratory viruses in clinical specimens from children, detecting 40% more positive specimens and missing only four specimens positive by FA. In fact, respiratory viruses were detected by PCR in 25% of the specimens judged inadequate for analysis by FA. The PCR assays also identified specimens with multiple viral infections with a 10-fold greater frequency than did FA. Detection of MPV, which was not included in the FA panel, accounted for about 30% of the increased sensitivity of PCR in this study. The results of the individual PCR assays showed differences among the six viruses included in the FA panel in the proportion that were also positive by FA. While about 90% of the RSV and FluA specimens positive by PCR were also positive by FA, a much smaller proportion of the specimens positive for any PIV or for AdV was detected by FA. These same respiratory viruses were also more likely to be detected in specimens containing more than one virus. PIV and AdV infections accounted for more than half of the results that were positive by PCR only and were detected in more than 90% of the specimens containing more than one virus. The specimens that were negative by FA and positive by PCR and that contained PIV1, 2, or 3 were confirmed using PCR assays that target a different PIV gene.
The AdV PCR results differed in several ways from the RNA virus results. First, the median age of patients with negative FA results was lower than that of patients with positive FA results. Second, the majority of specimens that were negative by FA and positive by PCR (101 of 187) were AdV positive, and most of these specimens contained low numbers of virus (mean log10 copies/ml of AdV-positive specimens that were negative by FA was 4.43). AdV was also the most frequently detected virus in mixed infections. AdV remains detectable in the respiratory tract after primary infection. AdV group B2 strains were detected in 49 of 50 BAL specimens from patients without AdV-related respiratory illness, indicating that AdV is present in asymptomatic patients (15). The children tested in our study reported symptoms of respiratory illness. However, it is possible that the presence of low numbers of AdV was not the cause of the illness but a coincidental presence of persistent virus. Information about the severity and duration of illness in the AdV-positive children which could help interpret these results was not available. Additional studies using quantitative PCR methods and including clinical data are being performed to determine what AdV viral load is considered significant for a pathogenic role in respiratory illness.
Detection of the respiratory viruses was less sensitive by FA than by PCR for all patient age groups. In spite of this difference in sensitivity between the methods, the median ages of patients positive for a respiratory virus were similar when the specimens were tested by either FA or PCR. Not surprisingly, respiratory viruses were more frequently detected by both FA and PCR in children less than 5 years old (17, 25) than in older children. The additional finding that infection with more than one respiratory virus was more common in children less than 5 years old has not previously been reported. Younger age was also associated with larger viral quantity.
Findings similar to those in this study have been previously reported by other investigators using conventional PCR assays for both pediatric and adult populations (3, 5, 6, 8, 9, 10, 19, 20, 22, 28). Conventional PCR assays doubled the detection rate compared to culture for six respiratory viruses in samples obtained from children less than 5 years old hospitalized with acute respiratory infections in one study (28) and increased the prevalence of seven respiratory viruses from 26.9% by FA to 33.8% in specimens from children with acute respiratory infections in another study (22). Comparisons of real-time PCR with FA and culture have previously been reported only for adult populations. Compared to conventional methods, real-time PCR increased the detection of 15 respiratory pathogens in adult patients with community-acquired pneumonia from 49.5% to 76% and the detection of mixed infections from 2.8% to 26.5% (24). Compared to culture, real-time PCR increased the detection of 10 respiratory viruses in adult stem cell transplant patients from 21% to 63% in patients with illness and from 1% to 9% in asymptomatic patients (26). In addition to a comparison of specific respiratory virus detection, the use of quantitative assays in our study provides new information about the quantity of specific viruses in specimens that are positive by PCR and negative by FA.
Most of the respiratory viruses not detected by FA had low copy numbers of viral nucleic acid present. The mean number of virus copies/ml in specimens that were positive by PCR and also positive by FA was more than 3 logs higher than the number in specimens that were negative by FA. Overall, the FA assay detected only 19% (37 of 194) of the respiratory viruses with viral loads less than 106 copies/ml. For the individual viruses, the difference in FA detection between RSV and FluA on the one hand and PIV and AdV on the other was due to differences in the quantity of virus in these specimens. The viruses most often detected in specimens that were positive by PCR and negative by FA, and those more likely to be detected in a mixed infection with another virus, such as PIV and AdV, were also more frequently present in low copy numbers (less than 106 copies/ml) and, consequently, had the lowest mean number of virus copies/ml. By univariate analysis, specimens that were positive by PCR and negative by FA were more likely to be from older patients than specimens that were positive by both PCR and FA. However, there was no association when analyzed by multiple regression.
In this study, real-time PCR was a significant improvement over FA for detection of respiratory viruses in clinical specimens from children, especially for detection of MPV, PIV, and AdV. These data are not likely due to false-positive PCR results but rather quantities of virus that are below the level of detection of FA. In addition to improved detection, in high-volume laboratories, real-time PCR can provide results faster than FA. Although reagent and instrument costs are higher for PCR than for FA (13), real-time PCR requires less hands-on time per specimen than FA, which is labor-intensive. Real-time PCR assays offer advantages over conventional PCR by providing lower risk of false-positive results due to amplicon contamination, identification of the etiologic agent in a clinically relevant time period, and quantification of viral load. Although the diagnostic value of knowing an individual's virus copy number is currently unclear for pediatric respiratory infections, quantification of respiratory viruses can provide important information about the role of these pathogens in respiratory tract infections among all patient groups. Ongoing investigations will help determine any associations between viral load and clinical signs and symptoms. These real-time PCR assays are useful tools for further investigations on the epidemiology of and disease associated with respiratory viruses and will provide information to better understand the relationship between illness and the quantity of virus being shed.
ACKNOWLEDGMENTS
We thank the staff at the University of Washington Clinical Virology Laboratory for performing the FA tests and providing specimen aliquots for PCR testing and Amalia Meier for statistical help.
This work was supported in part by National Cancer Institute grant number 5 PO1 CA018029-30.
FOOTNOTES
REFERENCES
Barenfanger, J., C. Drake, N. Leon, T. Mueller, and T. Troutt. 2000. Clinical and financial benefits of rapid detection of respiratory viruses: an outcomes study. J. Clin. Microbiol. 38:2824-2828.
Beekman, S. E., H. D. Engler, A. S. Collins, J. Canosa, D. K. Henderson, and A. Freifield. 1996. Rapid identification of respiratory viruses: impact on isolation practices and transmission among immunocompromised pediatric patients. Infect. Control Hosp. Epidemiol. 17:581-586.
Bellau-Pujol, S., A. Vabret, L. Legrand, J. Dina, S. Gouarin, J. Petitjean-Lecherbonnier, B. Pozzetto, C. Ginevra, and F. Freymuth. 2005. Development of three multiplex RT-PCR assays for the detection of 12 respiratory viruses. J. Virol. Methods 126:53-63.
Cockerill, F. R., III. 2003. Application of rapid-cycle real-time polymerase chain reaction for diagnostic testing in the clinical microbiology laboratory. Arch. Pathol. Lab. Med. 127:1112-1120.
Coiras, M. T., P. Perez-Brea, M. L. García, and I. Casas. 2003. Simultaneous detection of influenza A, B, C viruses, respiratory syncytial virus, and adenoviruses in clinical samples by multiplex reverse transcription nested-PCR assay. J. Med. Virol. 69:132-144.
Coyle, P. V., G. M. Ong, H. J. O'Neill, C. McCaughey, D. De Ornellas, F. Mitchell, S. J. Mitchell, S. A. Feeney, D. E. Wyatt, M. Forde, and J. Stockton. 2004. A touchdown nucleic acid amplification protocol as an alternative to culture backup for immunofluorescence in the routine diagnosis of acute viral respiratory tract infections. BMC Microbiol. 4:41-49.
Dagher, H., H. Donniger, P. Hutchinson, R. Ghildyal, and P. Bardin. 2004. Rhinovirus detection: comparison of real-time and conventional PCR. J. Virol. Methods 117:113-121.
Erdman, D. D., G. A. Weinberg, K. M. Edwards, F. J. Walker, B. C. Anderson, J. Winter, M. Gonzalez, and L. J. Anderson. 2003. GeneScan reverse transcription-PCR assay for detection of six common respiratory viruses in young children hospitalized with acute respiratory illness. J. Clin. Microbiol. 41:4298-4303.
Grndahl, B., W. Puppe, A. Hoppe, I. Kühne, J. A. I. Weigl, and H.-J. Schmitt. 1999. Rapid identification of nine microorganisms causing acute respiratory tract infections by single-tube multiplex reverse transcription-PCR: feasibility study. J. Clin. Microbiol. 37:1-7.
Gruteke, P., A. S. Glas, M. Dierdorp, W. B. Vreede, J.-W. Pilon, and S. M. Bruisten. 2004. Practical implementation of a multiplex PCR for acute respiratory tract infections in children. J. Clin. Microbiol. 42:5596-5603.
Henrickson, K. J. 2005. Cost-effective use of rapid diagnostic techniques in the treatment and prevention of viral respiratory infections. Pediatr. Ann. 34:24-31.
Kehl, S. C., K. J. Henrickson, W. Hua, and J. Fan. 2001. Evaluation of the Hexaplex assay for detection of respiratory viruses in children. J. Clin. Microbiol. 39:1696-1701.
Kuypers, J., N. Wright, and R. Morrow. 2004. Evaluation of quantitative and type-specific real-time PCR assays for detection of RSV in respiratory specimens from children. J. Clin. Virol. 31:123-129.
Kuypers, J., N. Wright, L. Corey, and R. Morrow. 2005. Detection and quantification of human metapneumovirus in pediatric specimens by real-time PCR. J. Clin. Virol. 33:299-305.
Leung, A. Y., M. Chan, V. C. Cheng, K. Y. Yuen, and Y. L. Kwong. 2005. Quantification of adenovirus in the lower respiratory tract of patients without clinical adenovirus-related respiratory disease. Clin. Infect. Dis. 40:1541-1544.
Limaye, A. P., M.-L. Huang, W. Leisenring, L. Stensland, L. Corey, and M. Boeckh. 2001. Cytomegalovirus (CMV) DNA load in plasma for the diagnosis of CMV disease before engraftment in hematopoietic stem-cell transplant recipients. J. Infect. Dis. 183:377-382.
Monto, A. S. 2002. Epidemiology of viral respiratory infections. Am. J. Med. 112:4S-12S.
Poon, L. L., K. H. Chan, O. K. Wong, T. K. Cheung, I. Ng, B. Zheng, W. H. Seto, K. Y. Yuen, Y. Guan, and J. S. Peiris. 2004. Detection of SARS coronavirus in patients with severe acute respiratory syndrome by conventional and real-time quantitative reverse transcription-PCR assays. Clin. Chem. 50:67-72.
Puppe, W., J. A. I. Weigl, G. Aron, B. Grndahl, H.-J. Schmitt, H. G. M. Niesters, and J. Groen. 2004. Evaluation of a multiplex reverse transcriptase PCR ELISA for the detection of nine respiratory tract pathogens. J. Clin. Virol. 30:165-174.
Rovida, F., E. Percivalle, M. Zavattoni, M. Torsellini, A. Sarasini, G. Campanini, S. Paolucci, F. Baldanti, M. G. Revello, and G. Gerna. 2005. Monoclonal antibodies versus reverse transcription-PCR for detection of respiratory viruses in a patient population with respiratory tract infections admitted to hospital. J. Med. Virol. 75:336-347.
Skerret, S. J. 1997. Diagnostic testing to establish a microbial cause is helpful in the management of community-acquired pneumonia. Semin. Respir. Infect. 12:308-321.
Syrmis, M. W., D. M. Whiley, M. Thomas, I. M. Mackay, J. Williamson, D. J. Siebert, M. D. Nissen, and T. P. Sloots. 2004. A sensitive, specific, and cost-effective multiplex reverse transcriptase-PCR assay for the detection of seven common respiratory viruses in respiratory samples. J. Mol. Diagn. 6:125-131.
Templeton, K. E., S. A. Scheltinga, M. R. C. Beersma, A. C. M. Kroes, and E. C. J. Claas. 2004. Rapid and sensitive method using multiplex real-time PCR for diagnosis of infections by influenza A and influenza B viruses, respiratory syncytial virus, and parainfluenza viruses 1, 2, 3, and 4. J. Clin. Microbiol. 42:1564-1569.
Templeton, K. E., S. A. Scheltinga, W. C. J. F. M. van den Eeden, A. W. Graffelman, P. J. van den Broek, and E. C. J. Claas. 2005. Improved diagnosis of the etiology of community-acquired pneumonia with real-time polymerase chain reaction. Clin. Infect. Dis. 41:345-351.
van Gageldonk-Lafeber, A. B., M.-L. A. Heijnen, A. I. M. Bartelds, M. F. Peters, S. M. van der Plas, and B. Wilbrink. 2005. A case-control study of acute respiratory tract infection in general practice patients in The Netherlands. Clin. Infect. Dis. 41:490-497.
van Kraaij, M. G. J., L. J. R. van Elden, A. M. van Loon, K. A. W. Hendricksen, L. Laterveer, A. W. Dekker, and M. Nijhuis. 2005. Frequent detection of respiratory viruses in adult recipients of stem cell transplants with the use of real-time polymerase chain reaction, compared with viral culture. Clin. Infect. Dis. 40:662-669.
Watzinger, F., M. Suda, S. Preuner, R. Baumgartinger, K. Ebner, L. Baskova, H. G. M. Niesters, A. Lawitschka, and T. Lion. 2004. Real-time quantitative PCR assays for detection and monitoring of pathogenic human viruses in immunosuppressed pediatric patients. J. Clin. Microbiol. 42:5189-5198.
Weinberg, G. A., D. D. Erdman, K. M. Edwards, C. B. Hall, F. J. Walker, M. R. Griffin, B. Schwartz, and the New Vaccine Surveillance Network Study Group. 2004. Superiority of reverse-transcription polymerase chain reaction to conventional viral culture in the diagnosis of acute respiratory tract infections in children. J. Infect. Dis. 189:706-710.(Jane Kuypers, Nancy Wrigh)