Received 24 March 2011/Returned for modification 6 May 2011/Accepted 6 July 2011

Transcription

1 JOURNAL OF CLINICAL MICROBIOLOGY, Sept. 2011, p Vol. 49, No /11/$12.00 doi: /jcm Copyright 2011, American Society for Microbiology. All Rights Reserved. Evaluation and Verification of the Seeplex Diarrhea-V ACE Assay for Simultaneous Detection of Adenovirus, Rotavirus, and Norovirus Genogroups I and II in Clinical Stool Specimens Rachel R. Higgins, 1 * Melissa Beniprashad, 1 Mark Cardona, 1 Steven Masney, 1 Donald E. Low, 1,2,3 and Jonathan B. Gubbay 1,2,3,4 Ontario Agency for Health Protection and Promotion, Toronto, Ontario, Canada 1 ; Mount Sinai Hospital, Toronto, Ontario, Canada 2 ; University of Toronto, Ontario, Canada 3 ; and The Hospital for Sick Children, Toronto, Ontario, Canada 4 Received 24 March 2011/Returned for modification 6 May 2011/Accepted 6 July 2011 Acute viral gastroenteritis is an intestinal infection that can be caused by several different viruses. Here we describe the evaluation and verification of Seeplex Diarrhea-V ACE (Seeplex DV), a novel commercial multiplex reverse transcription-pcr (RT-PCR) assay that detects 5 diarrheal pathogens, including adenovirus, rotavirus, norovirus genogroup I (GI) and GII, and astrovirus. We describe a retrospective study of 200 clinical specimens of which 177 were stool specimens previously tested for the presence of gastrointestinal viruses by electron microscopy (EM) and/or real-time RT-PCR (rrt-pcr). The remaining 23 specimens comprised other human pathogens of viral or bacterial origin. Discordant norovirus GI and GII results were resolved using a commercial kit; discordant adenovirus and rotavirus results were resolved using a home brew multiplex rrt-pcr assay. Diagnostic sensitivities and specificities were calculated before and after discordant analysis. After discordant analysis, estimated diagnostic sensitivities were 100% for adenovirus, rotavirus, and norovirus GI and 97% for norovirus GII. Diagnostic specificities after discordant analysis were 100% for adenovirus, rotavirus, and norovirus GI and 99.4% for norovirus GII. The 95% limits of detection were 31, 10, 2, and 1 genome equivalent per reaction for adenovirus, rotavirus, and norovirus GI and GII, respectively. The results demonstrate that the Seeplex DV assay is sensitive, specific, convenient, and reliable for the simultaneous detection of several viral pathogens directly in specimens from patients with gastroenteritis. Importantly, this novel multiplex PCR assay enabled the identification of viral coinfections in 12 (6.8%) stool specimens. Acute viral gastroenteritis is the second most common infectious disease worldwide (20, 5), affecting humans of all age groups (23, 17) but mostly the young, the elderly, and people in enclosed communities, such as hospitals, nursing homes, military bases, and cruise ships. Known enteric viral pathogens include adenovirus group F serotypes 40 and 41, rotavirus, norovirus, and astrovirus, with rotavirus and norovirus being the predominant causative agents of gastroenteritis (3, 7, 30). In recent years the number of reported gastroenteritis outbreaks of suspected viral etiology has increased, hence the need for fast, sensitive, and reliable diagnostic assays. The established method for detection of enteric viruses at the Ontario Agency for Health Protection and Promotion (OAHPP) Public Health Laboratories (PHL) relies on the conventional and labor-intensive electron microscopy (EM). In recent years, home brew real-time reverse transcription-pcr (rrt-pcr) assays were implemented for the detection of norovirus GI and GII (11). Similar home brew rrt-pcr assays for the detection of adenovirus (6) and rotavirus (35) have been described. These rrt-pcr assays enable the detection of a specific virus by the amplification of a unique genomic sequence within its RNA or DNA. However, the simultaneous detection of several viruses can be accelerated and simplified * Corresponding author. Mailing address: Ontario Agency for Health Protection and Promotion, 81 Resources Road, Toronto, ON M9P 3T1, Canada. Phone: (416) Fax: (416) rachel.higgins@oahpp.ca. Published ahead of print on 20 July by the use of a multiplex RT-PCR assay. Multiplex PCR encompasses multiple primer pairs in a single tube, making it possible to amplify multiple target sequences. Here we describe the verification and evaluation of the Seeplex DV assay (Seegene, South Korea), a newly developed commercial multiplex RT-PCR assay for the simultaneous detection of adenovirus, rotavirus, norovirus GI and GII, and astrovirus. The use of this novel multiplex RT-PCR assay offers the advantage of reduced reagents and labor cost, as well as a faster turnaround time. The Seeplex DV assay integrates an internal control, which facilitates the detection of PCR inhibitors and reduces false-negative results. Importantly, a unique advantage of a multiplex assay is the detection and identification of coinfecting pathogens within the same patient specimen. (Part of this work was presented at the 26th annual Clinical Microbiology Symposium, Daytona Beach, FL, April 2010.) MATERIALS AND METHODS Definitions. The following definitions were used for analysis: PHL tests, previous tests done at OAHPP, namely, EM (for adenovirus and rotavirus) and/or rrt-pcr (for norovirus GI and GII); concordant specimen, agreement between PHL test result and Seeplex DV assay result; discordant specimen, disagreement between PHL test result and Seeplex DV assay result; resolved specimen, agreement in result between any two of the three assays (PHL test, Seeplex DV assay, and reference assay [RealStar for norovirus GI and GII and home brew multiplex rrt-pcr for adenovirus and norovirus]). Clinical specimens. Samples included in the study were from patients ages 2 months to 96 years old, of which 54% were 5 years old or younger and 43% were female. Seventy-four (42%) of the clinical specimens were from 47 local outbreaks (as defined by the Ontario Ministry of Health) (25), while the remaining 3154

2 VOL. 49, 2011 MULTIPLEX MOLECULAR GASTROINTESTINAL VIRUS ASSAY 3155 TABLE 1. Seasonal distribution of 74 stool specimens from 47 outbreak investigations and 103 sporadic stool specimens submitted for testing between July 2007 and May 2010 a Pathogen Total O/B No. of spec Sporadic Season Season Total W Sp Su F W Sp Su F Adenovirus Rotavirus Norovirus GI Norovirus GII Astrovirus Negative Total a spec, specimens; O/B, outbreak; W, winter; Sp, spring; Su, summer; F, fall. TABLE 2. Distribution of viral and bacterial pathogens tested Pathogen No. of specimens tested Source Adenovirus 29 Stool Rotavirus 52 Stool Norovirus GIl 19 Stool Norovirus GII 37 Stool Astrovirus a 3 Stool Negative enteric specimens 29 Stool Salmonella enterica serovar Paratyphi 2 Stool Salmonella enterica serovar Typhi 1 Stool Shigella dysenteriae type 2 5 Stool Escherichia coli 0:157:H71a 1 Culture Salmonella enterica serovar Culture Enteritidis b Shigella flexneri b 1 Culture Shigella sonnei b 1 Culture Vibrio cholerae c 1 Culture Campylobacter jejuni d 1 Culture Influenza virus B 2 Nasopharyngeal swabs Influenza virus A/H1N1 5 Nasopharyngeal swabs Influenza virus A/H3N2 1 Nasopharyngeal swabs Picornavirus 2 Nasopharyngeal swabs Aeromonas hydrophila b 1 Culture Neisseria meningitidis e 1 Culture Streptococcus pyogenes b 2 Culture Streptococcus pneumoniae b 2 Culture Staphylococcus epidermidis b 1 Culture Total 200 a Three astrovirus-positive specimens were tested in the Seeplex DV assay but not included in the evaluation study due to the lack of sufficient positive specimens. b Cultured in blood agar at 37 C with CO 2. c Cultured in plain Columbia agar at 37 C with CO 2. d Cultured in charcoal-based selective medium agar at 42 C. e Cultured in New York City agar at 37 C with CO (58%) specimens were from sporadic cases of gastroenteritis (Table 1). Table 2 shows the distribution of specimens used in this study. A total of 200 clinical specimens and bacterial cultures, including 177 randomly selected stool specimens submitted for testing from February 2009 until May 2010 for adenovirus, rotavirus, norovirus GI, norovirus GII, astrovirus, and enteric bacteria, were used. Also, a specificity panel consisting of 10 nasopharyngeal (NP) swabs containing respiratory viruses, 14 enteric stool/bacterial cultures, and 7 nonenteric bacterial cultures was included to measure specificities and cross-reactivities of primer sets. The 3 astrovirus-positive samples were not included in the study analysis due to the lack of sufficient samples containing this target. Laboratory test methods. Routine testing of suspected adenovirus, rotavirus, and norovirus infections at OAHPP relies on the use of EM (Philips CM10 or FEI Morgagni 268). Outbreak specimens (as defined by the Ontario Ministry of Health) (25) are initially tested for norovirus by a home brew rrt-pcr assay, and if negative for norovirus, the specimens are then tested by EM (11). In this study, we evaluated the Seeplex DV assay, a commercial multiplex viral diarrhea assay (Seegene, South Korea). This assay consists of reverse transcription and PCR, followed by resolution and detection of amplified DNA products by agarose gel (1.5%) or capillary electrophoresis using equipment such as the MCE- 202 MultiNA microchip electrophoresis system (Schimadzu BioTech, Japan). Samples with discordant norovirus results were also tested with the RealStar 2.0 norovirus kit (Astra Diagnostics, Hamburg, Germany), which detects and differentiate norovirus GI and GII. To resolve discrepant adenovirus and rotavirus results, we developed a home brew multiplex rrt-pcr assay for these viruses and all specimens were retested in this rrt-pcr assay (R. R. Higgins, presented at the 2011 Clinical Virology Symposium; unpublished data). We did not evaluate the Seeplex DV assay for the detection of astrovirus due to the lack of sufficient astrovirus-positive specimens. RNA and DNA extraction. Total nucleic acid was extracted from all specimens using an automated NucliSens easymag system (Biomeriuex, France). Stool specimens were homogenized (20% [wt/vol]) in sterile water and centrifuged for 20 min at 3,000 rpm, and 250 l of the clarified supernatants were subjected to automated nucleic acid extraction. Total nucleic acid was eluted in a final volume of 60 l, of which 5 l was used for PCR amplification. Nucleic acid was extracted directly from respiratory specimens collected with nasopharyngeal swabs in universal transport medium (Copan Diagnostics Inc.). Bacterial cultures grown in suggested culture medium (Table 2) (14) were diluted to 0.5 McFarland standard (10) in phosphate buffered saline prior to automated extraction on the easymag system. The efficiency of nucleic acid extraction was measured by real-time PCR amplification of Bacteroides fragilis DNA (19) and/or the human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene (1). PCR amplification. Details of primers and probes used are shown in Table 3. Amplification of nucleic acid using the Seeplex DV assay was performed with the Icycler system (Bio-Rad Laboratories Inc.) according to the manufacturer s insert. PCR-inhibitory effects were assessed by coamplification of an internal control included in the Seeplex primer mix. PCR products were resolved by capillary electrophoresis on the MultiNA. Amplification of norovirus GI and GII targets was performed on the ABI7900HT Fast real-time PCR system (Applied Biosystems, California) using home brew rrt-pcr assays (11). Norovirus specimens, including discordant ones, were retested on the RealStar 2.0 kit from Astra Diagnostics (Hamburg, Germany) with amplification on the ABI7500 HT Fast real-time PCR instrument according to the manufacturer s insert. Adenovirus and rotavirus specimens, including discordant ones, were retested on the ABI7500 HT Fast real-time PCR instrument using internally controlled triplex home brew rrt-pcr assays developed for this study (Higgins, 2011 Clin. Virol. Symp.). Analytical sensitivity, specificity, and limit of detection. The sensitivity of the Seeplex DV assay was determined in comparison to the PHL test (see definitions

3 3156 HIGGINS ET AL. J. CLIN. MICROBIOL. TABLE 3. Primers and probes used for target amplification Target Primer and probe sequences b Amp size (bp) c Position PCR method Detection system B. fragilis f: GAG AGG AAG GTC CC rrt-pcr ABI r: CGC TAC TTG GCT GG p: FAM-CCA TTG ACC AAT ATT CCT CAC TGC TGC CT-BHQ Norovirus GI f: GCC ATG TTC CGI TGG ATG rrt-pcr ABI r: TCC TTA GAC GCC ATC ATC AT p: FAM- AGA TCG CGG TCT CCT GTC CA-BHQ Norovirus GII f: CAA GAG TCA ATG TTT AGG TGG ATG AG rrt-pcr ABI r: TCG ACG CCA TCT TCA TTC ACA p: CY5- TGG GAG GGC GAT CGC AAT CT-BHQ Adenovirus a f: CAG GAC GCC TCG GRG TAY CTS AG 103 d rrt-pcr ABI r: GGA GCC ACV GTG GGR TT p: FAM-CCG GGT CTG GTG CAG TTT GCC CGC- BHQ Rotavirus a f:cca TCT WCA CRT RAC CCT CTA TGA G rrt-pcr ABI r:ggt CAC ATA ACG CCC CTA TAG C p: CY5-AGT TAA AAG CTA ACA CTG TCA AA-BHQ Norovirus GI Astra diagnostics (RealStar assay) NA NA rrt-pcr ABI7500 NA Norovirus GII Astra diagnostics (RealStar assay) NA NA rrt-pcr ABI7500 NA Adenovirus Seegene (Seeplex DV assay) 411 NA 2-Step RT-PCR multina NA Astrovirus Seegene (Seeplex DV assay) 650 NA 2-Step RT-PCR multina NA Rotavirus Seegene (Seeplex DV assay) 541 NA 2-Step RT PCR multina NA Norovirus GI Seegene (Seeplex DV assay) 304 NA 2-Step RT PCR multina NA Norovirus GII Seegene (Seeplex DV assay) 214 NA 2-Step RT PCR multina NA Reference a The primers used for both rotavirus and adenovirus are degenerate and allow for detection of multiple serotypes: R, A/G; Y, C/T; S, C/G; V, A or C or G; W, A/T. b f, forward; r, reverse; p, probe; FAM, 6-carboxyfluorescein. c Amp, amplicon. d, base pair position for the adenovirus amplicon varies in different serotypes. above). To determine the sensitivity of the Seeplex DV assay, results were scored against the PHL test method (EM for adenovirus and rotavirus and/or rrt-pcr for norovirus GI and GII). To resolve discordant results, these specimens were retested using an rrt-pcr assay. For norovirus GI and GII, we employed the RealStar 2.0 rrt-pcr assay from Astra Diagnostics (Hamburg, Germany), and for adenovirus and rotavirus, we used home brew rrt-pcr assays (Higgins, 2011 Clin. Virol. Symp.) (Table 4). P values after discordant analysis were determined by chi-square analysis using the software program Stata 11 (Stata- Corp LP, Texas). Analytical specificity of the Seeplex DV assay was determined by testing a set TABLE 4. Results obtained for verification of Seeplex DV assay Pathogen tested or result No. of specimens and cultures with result by: PHL test a Seeplex DV rrt-pcr No. of coinfecting pathogens Adenovirus (norovirus GII) b Rotavirus (norovirus GII), b 1 (adenovirus) c Norovirus GI Norovirus GII (rotavirus), c 3 (adenovirus) c Negative Total a PHL test: EM for adenovirus and rotavirus; EM and/or rrt-pcr for norovirus GI and GII. b Coinfections with norovirus GII were confirmed by RealStar assay. c Coinfections with adenovirus or rotavirus were confirmed with home brew rrt-pcr. of commonly encountered viral and bacterial pathogens (Tables 1 and 2). Viral and bacterial samples were initially characterized by culture, electron microscopy, or home brew norovirus rrt-pcr during routine clinical testing. To confirm PHL culture results for enteric bacteria, these specimens were also tested for enteric bacteria in the Seeplex Diarrhea ACE test kit. The limit of detection (LOD) for each virus in the Seeplex DV assay was determined using tissue culture fluid (TCF) for all viral targets except norovirus GI and plasmid DNA for all targets. Quantified plasmid DNA (Seegene, South Korea) and/or TCF (Zeptometrix Corporation) for each virus was serially diluted 10-fold to of the starting material in PCR-grade water and tested in the Seeplex DV assay. The respective LODs were assessed by testing a minimum of 4 replicates of each of the 12 dilution steps. The 95% limit of detection (and 95% confidence interval) for each target was determined using probit regression (SPSS software regression package, version 18.0; SPSS, Chicago, IL). Assay reproducibility and TAT. Inter- and intra-assay reproducibilities were assessed using data generated from the LOD calculation and from testing 110 samples on 3 different days. Concordance for inter- and intra-assay reproducibility was calculated. Turnaround time (TAT) was calculated for the Seeplex DV assay and home brew rrt-pcr based on 5 runs, each of 96 samples, completed by one laboratory staff member. For EM, TAT was extrapolated from timing 3 days work. The final TAT reflected the average time to process 96 samples, including reception, hands-on, assay, and resulting/reporting times. To directly compare assays, TAT for each assay was divided by the number of targets tested to give a TAT per target for 96 samples. RESULTS Table 4 displays the results obtained for each of the 197 clinical specimens and bacterial cultures included in this study (excluding astrovirus). A pictorial example of results obtained for a subset of clinical specimens, representing each of the four

4 VOL. 49, 2011 MULTIPLEX MOLECULAR GASTROINTESTINAL VIRUS ASSAY 3157 FIG. 1. Detection of Seeplex DV PCR products on the MultiNA. Representative raw (A) or analyzed (B) data for Seeplex DV test results are depicted. Fifteen specimens representing each of the 4 viral targets evaluated in this study are shown. Results with dual infections are shown in lanes 1 (adenovirus and norovirus GII), 2 (rotavirus and norovirus GII), and 3 (rotavirus and adenovirus). The identity of the pathogen is shown at the top. AD, adenovirus; RV, rotavirus; NVI, norovirus GI; NVII, norovirus GII; NPC and PPC refer to negative and positive PCR controls, respectively. Note that the Seeplex DV assay also amplifies an internal control for a total of 6. Numbers in the table reflect band intensities. targets evaluated, before and after analyses with the Seegene viewer is depicted in Fig. 1A and B. As shown in Table 4, when comparing PHL to Seeplex DV results, there is one discordant result for each of the 4 viral pathogens tested and two discordant results in the negative enteric pool. This indicates that the performances of all three test methods are comparable. The Seeplex DV assay detected coinfecting pathogens in 12 specimens of 177 primary stool specimens tested (Table 4). Coinfections with norovirus GI were not detected in this study, likely due to the small sample

5 3158 HIGGINS ET AL. J. CLIN. MICROBIOL. TABLE 5. Calculated specificity and sensitivity of the Seeplex DV assay before discordant analysis Result by Seeplex DV No. of specimens with PHL test result EM rrt-pcr True False True False % sensitivity a % specificity b Adenovirus Positive 28 1 Negative Total (28/29) 99.4 (168/169) Rotavirus Positive 51 1 Negative Total (51/52) 99.3 (145/146) Norovirus GI Positive 18 1 Negative Total (18/19) 99.4 (178/179) Norovirus GII Positive 37 1 Negative Total (37/38) 99.4 (159/160) a Numbers in parentheses are no. positive by Seeplex DV/total no. true by PHL test. b Numbers in parentheses are no. negative by Seeplex DV/total no. false by PHL test. size of norovirus GI-positive specimens tested and the low prevalence of norovirus GI infection (12, 18, 21). Among the coinfecting pathogens detected, there were 4 cases of adenovirus/norovirus GII, 7 cases of rotavirus/norovirus GII, and 1 case of rotavirus/adenovirus mixed infections (Table 4). Examples of mixed infections are depicted in Fig. 1A and B, showing adenovirus/norovirus GII, rotavirus/adenovirus, and rotavirus/adenovirus in lanes 1 to 3, respectively. Diagnostic sensitivity and specificity of the Seeplex DV assay. To determine the diagnostic sensitivity and specificity of the Seeplex DV assay, results were scored against the PHL test method (EM for adenovirus and rotavirus and home brew rrt-pcr for norovirus GI and GII (see Tables 1, 2, and 4). To resolve discordant results, all specimens were retested using rrt-pcr assays: home brew multiplex for adenovirus and rotavirus and RealStar 2.0 for norovirus GI and GII. Diagnostic specificity was determined by testing a set of precharacterized viruses and bacterial specimens or cultures (Tables 1 and 2). No cross-reaction was observed in the specificity panel including other viruses and bacteria (Tables 4, 5, and 6). In particular, no cross-reaction was observed with enteric bacterial pathogens. PHL culture results for enteric bacteria were concordant with those of the Seeplex Bacterial Diarrhea (B1and B2) ACE multiplex assay (data not shown). Overall there were 5 discordant results between the Seeplex DV and PHL tests (Table 7); 4 were resolved in favor of the Seeplex DV assay and 1 in favor of the PHL test. One norovirus GII specimen tested negative in the Seeplex DV assay but positive in both the PHL and RealStar assays. One adenovirus specimen tested positive by EM but negative by both the Seeplex DV assay and home brew adenovirus-rotavirus rrt-pcr. Two specimens, norovirus GI and rotavirus, tested positive in TABLE 6. Calculated specificity and sensitivity of the Seeplex DV assay after discordant analysis Assay, virus, and result No. of specimens with Seeplex DV result True False % sensitivity a % specificity b Home brew multiplex rrt-pcr Adenovirus Positive 28 1 Negative Total (28/29) 99.4 (168/169) Rotavirus Positive 51 0 Negative Total (51/51) 100 (146/146) RealStar Norovirus GI Positive 18 0 Negative Total (18/18) 100 (179/179) Norovirus GII Positive 38 0 Negative Total (38/38) 100 (159/159) a Numbers in parentheses are no. positive by reference assay/total no. true by Seeplex DV. b Numbers in parentheses are no. negative by reference assay/total no. false by Seeplex DV. the PHL test but negative in the Seeplex DV assay, possibly due to repeated freeze-thaw cycles and/or a long storage period at 80 C. One norovirus GII specimen tested negative in the PHL test but positive in both the Seeplex DV and RealStar assays. In total, the Seeplex DV assay detected 135 positive specimens, compared to 137 for the PHL tests. Prior to discordant analysis, the sensitivity and specificity of the Seeplex DV multiplex assay for each target were as follows: 97% and 99.4% for adenovirus, 98% and 99.3% for rotavirus, 95% and 99.4% for norovirus GI, and 97% and 99.4% for norovirus GII, respectively (Table 5). Following discordant analysis and using classification of specimens as per definitions described above, the sensitivity and specificity of the Seeplex DV multiplex assay are 100% and 100%, respectively, for adenovirus, rotavirus, and norovirus GI. The sensitivity and specificity for norovirus Specimen no. TABLE 7. Summary of discordant results Result PHL test Seeplex DV rrt-pcr (C T a ) S45 Negative Adenovirus Adenovirus (9) S25 Rotavirus Negative Negative S81 Norovirus GI Negative Negative S52 Norovirus GII Negative Norovirus GII (33.59) S54 Negative Norovirus GII Norovirus GII (29.32) a Threshold cycle.

6 VOL. 49, 2011 MULTIPLEX MOLECULAR GASTROINTESTINAL VIRUS ASSAY 3159 FIG. 2. Determination of the limit of detection for the Seeplex DV assay using quantified plasmid DNA. Raw (A) or analyzed (B) data for serial 10-fold dilutions of plasmid DNA for rotavirus, adenovirus, norovirus GI, and norovirus GII are depicted. Detection of amplified PCR product is evident at 45, 4.5, and 0.45 copies per reaction. Note that only a subset of the experiment representing each target is shown. In this experiment, triplicates of each dilution were tested in the same run. Note also that plasmid dilutions were retested on different days. Numbers in the table reflect band intensities. GII however, remained at 97% and 99.4%, respectively (Table 6). P values were not calculated for rotavirus, norovirus GI, and norovirus GII where there was exact agreement between the Seeplex DV and PHL tests. P values for both adenovirus sensitivity and specificity following discordant analysis were 0.31, indicating no significant difference between the two tests. Consistent with its multiplex format, the Seeplex DV assay detected 12 additional positive coinfecting pathogens, including 6 of norovirus GII, 4 adenoviruses, and 2 rotaviruses (Table 4). Ninety-five percent LOD. Ninety-five percent limit of detection (LOD) indicates the lowest concentration of genome target at which a pathogen will be detected with a probability of 95%. Ninety-five percent LODs for each of the 4 targets were derived by testing a minimum of 4 replicates of each of the 12 dilution steps. Figure 2 depicts partial results (replicates not shown) of a typical LOD experiment using 10-fold serial dilutions of quantified plasmid DNA. As seen in Fig. 2, adenovirus and norovirus GI and GII are detected at 0.45 copies per reaction, and rotavirus is detected at 4.5 copies per reaction. Similarly, Fig. 3 depicts partial results of a typical LOD experiment using a 10-fold serial dilution of tissue culture fluid. As seen in Fig. 3, detection varies among the 3 targets tested, most likely due to variability in the titer of virus present in the different culture fluids. Using tissue culture fluid, norovirus was detected at , rotavirus at and adenovirus at % tissue culture infectious doses (TCID 50 )/ml. Using plasmid DNAs, the estimated 95% LODs are as follows: 30.8 genome equivalents (GEq)/reaction for adenovirus, 10.2 GEq/ reaction for rotavirus, 1.7 GEq/reaction for norovirus GII, and 2.2 GEq/reaction for norovirus GI (Fig. 2 and 3). Assay reproducibility and TAT. Inter- and intra-assay reproducibilities were both 100%. Average TATs for processing 96 samples in one run were as follows: 9 to 10 h for the Seeplex DV assay, 6 h for rrt-pcr, and 3 days for electron microscopy. TATs per target for 96 samples were 1.8 h for the Seeplex DV assay, 3 h for rrt-pcr, and 9 h for EM. DISCUSSION We describe the evaluation of the Seeplex DV assay (Seegene, South Korea), a commercial kit for the detection of pathogens associated with viral gastroenteritis in clinical stool specimens, which is approved by CE and Health Canada but not the FDA. The Seeplex DV assay is the only commercial multiplex RT-PCR product currently available for the simul-

7 3160 HIGGINS ET AL. J. CLIN. MICROBIOL. FIG. 3. Determination of the limit of detection for the Seeplex DV assay using quantified tissue culture fluid. Raw (A) or analyzed (B) data for serial 10-fold dilutions of tissue culture fluid for adenovirus, rotavirus, and norovirus GII are depicted. Detection of adenovirus amplicons is evident at a 10 6 dilution of the original concentration, whereas amplicons corresponding to rotavirus are detected at a 10 5 dilution and norovirus GII at a 10 4 dilution. Note that only a subset of the experiment representing each target is shown. In this experiment, triplicates of each dilution were tested in the same run. Note also that dilutions were retested on different days. Numbers in the table reflect band intensities. taneous detection of up to 5 viral targets, including adenovirus (group F serotypes 40 and 41), rotavirus, norovirus GI, norovirus GII, and astrovirus. We show that the Seeplex DV assay has sensitivity equivalent to those of other methods currently in use at PHL and reduces the TAT, especially when high volumes of samples are being tested. Viral gastroenteritis outbreaks occur frequently in nursing homes, health care institutions, and cruise ships, where they cause considerable morbidity and mortality; they have also been reported in schools and prisons (9, 33, 34). Although norovirus is recognized as a common cause of gastroenteritis, other enteric viruses also contribute to the syndrome. Hence, there may be a role for multiplex assay formats that will detect norovirus and in the same assay also test for other viral pathogens. Here we reconfirmed that norovirus GII causes the majority of enteric outbreaks in Ontario (11), and we have verified this assay only for outbreaks caused by norovirus GI or GII (Table 1). Future prospective studies will allow further characterization of the assay and verification for outbreaks caused by adenovirus and rotavirus. Current molecular diagnostic technologies rely on home brew tests. These tests are considerably advanced and require additional funds and staff for up-front research and development, quality assurance, and laboratory accreditation (8, 13, 15, 28). In recent years, commercial products such as those identified in this communication have allowed for the implementation of molecular testing without the added costs of research and development and additional quality assurance costs. The Seeplex DV assay demonstrates several strengths as a commercial assay for use in a wider range of clinical laboratory settings outside a highly resourced center. First, this assay

8 VOL. 49, 2011 MULTIPLEX MOLECULAR GASTROINTESTINAL VIRUS ASSAY 3161 offers a rapid, sensitive, and specific method for the diagnosis of adenovirus, rotavirus, norovirus GI, norovirus GII, and astrovirus. Second, being a commercial product, the Seeplex DV assay can be readily implemented in the clinical diagnostic laboratory (if approved by local licensing authorities). It also offers established quality controls and more-streamlined ongoing validation compared to a home brew PCR assay. Third, the Seeplex DV assay is a multiplex test, which allows the detection of multiple viral targets in one reaction, instead of multiple reactions as when using monoplex assays. This makes the assay easier to perform and utilizes less patient specimen. Fourth, compared to existing testing methodologies (rrt-pcr and EM), the reagent cost per target of the Seeplex DV multiplex assay is cheaper and the assay less laborious. The reagent cost is approximately $25.00 CAD per test or $5.00 CAD per virus target. There are several limitations to this study, including its retrospective nature. We expected to detect more positive specimens with the Seeplex DV assay than with EM. Hence, the sensitivity of the Seeplex DV assay was not higher than that of EM for detection of rotavirus and adenovirus. Nevertheless, the results of the Seeplex DV assay were confirmed with newly developed home brew adenovirus/rotavirus rrt-pcr with no increase in positivity. Another limitation of the Seeplex DV is the TAT, which requires 9 to 10 h and is longer than that of rrt-pcr (6 h). Yet the Seeplex DV assay offers detection of 5 viral targets plus an internal control for each patient. In the past 5 years, several rrt-pcr assays were developed for the detection of enteric viruses. These assays were demonstrated to have better sensitivities and specificities than electron microscopy or antigen-detection assays (4, 21, 22, 26, 27, 29). Of these studies, two have described the use of multiplex assays for the detection of viral gastroenteritis (2, 31). However, both assays lack internal controls in the PCR and therefore are of limited use in a clinical routine diagnostic setting. The limit of detection reported (21, 22, 26, 27) varied between 1 and 100 copies per reaction and is similar to the results shown in this study. A recent study described the application of two internally controlled multiplex real-time PCR assays for the simultaneous detection of adenovirus, rotavirus, norovirus GI and GII, astrovirus, and sapovirus (32). In that study, the sensitivity and specificity of the assay were clinically compared to those of the Vikia rotavirus/adenovirus chromatographic immunoassay, with a significant increase in pathogen detection from 14% to 45%. This multiplex PCR assay consists of two separate reactions; one targets astrovirus, rotavirus, and norovirus GI, and the other targets adenovirus and norovirus GII. While the manuscript was being reviewed, two studies appeared which describe home brew multiplex assays for the detection of viral gastroenteritis (16, 24). The study by Liu et al. (24) combined a one-step rrt-pcr with microspherebased fluorescence technology to detect 6 viral targets, including adenovirus, rotavirus, norovirus GI and GII, astrovirus, and sapovirus. The study by Khamrin et al. (16) combines two-step rrt-pcr and agarose gel electrophoresis to detect 10 virus targets, including adenovirus, rotavirus, norovirus GI, norovirus GII, astrovirus, sapovirus, enterovirus, parechovirus, and aichi virus. In conclusion, the Seeplex DV assay is sensitive, specific, convenient, and a reliable tool for the direct detection of enteric viruses in stool specimens from patients with gastroenteritis. This assay can be effectively implemented in a clinical microbiology setting within and outside reference laboratories and has the attractive feature of improving testing methods while also being more economical than previously used methods, in particular those using EM and home brew rrt-pcrs. ACKNOWLEDGMENTS We thank Inverness Medical (Canada) for providing Seeplex kits and loaning the MultiNA instrument. Astra Diagnostics provided RealStar kits for norovirus GI and GII testing. REFERENCES 1. Barber, R. D., D. W. Harmer, R. A. Coleman, and B. J. Clark GAPDH as a housekeeping gene: analysis of GAPDH mrna expression in a panel of 72 human tissues. Physiol. Genomics 21: Beuret, C Simultaneous detection of enteric viruses by multiplex real-time RT-PCR. J. Virol. Methods 115: Blanton, L. H., et al Molecular and epidemiologic trends of caliciviruses associated with outbreaks of acute gastroenteritis in the United States, J. Infect. Dis. 193: Chan, M. C., et al Sapovirus detection by quantitative real-time RT- PCR in clinical stool specimens. J. Virol. Methods 134: Christensen, M. L Human viral gastroenteritis. Clin. Microbiol. Rev. 2: Damen, M., et al Real-time PCR with an internal control for detection of all known human adenovirus serotypes. J. Clin. Microbiol. 46: de Wit, M. A., M. P. Koopmans, and Y. T. van Duynhoven Risk factors for norovirus, Sapporo-like virus, and group A rotavirus gastroenteritis. Emerg. Infect. Dis. 9: Espy, M. J., et al Real-time PCR in clinical microbiology: applications for routine laboratory testing. Clin. Microbiol. Rev. 19: Fankhauser, R. L., et al Epidemiologic and molecular trends of Norwalk-like viruses associated with outbreaks of gastroenteritis in the United States. J. Infect. Dis. 186: Finegold, S. M., and E. J. Baron Bailey and Scott s diagnostic microbiology, 7th ed. C. V. Mosby, St. Louis. MO. 11. Fisman, D. N., A. L. Greer, G. Brouhanski, and S. J. Drews Of gastro and the gold standard: evaluation and policy implications of norovirus test performance for outbreak detection. J. Transl. Med. 7: Gallimore, C. I., M. Iturriza-Gomara, J. Xerry, J. Adigwe, and J. J. Gray Inter-seasonal diversity of norovirus genotypes: emergence and selection of virus variants. Arch. Virol. 152: Gammie, A. J., For the proposition: for the diagnosis of viral infections, commercial assays provide more reliable results than do in-house assays. Rev. Med. Virol. 18: Garcia, L., and H. Isenberg Clinical microbiology procedures handbook, 2nd ed. ASM Press, Washington, DC. 15. James, V Against the proposition: for the diagnosis of viral infections, commercial assays provide more reliable results than do in-house assays. Rev. Med. Virol. 18: Khamrin, P., et al A single-tube multiplex PCR for rapid detection in feces of 10 viruses causing diarrhea. J. Virol. Methods 173: Koopmans, M Outbreaks of viral gastroenteritis: what s new in 2004? Curr. Opin. Infect. Dis. 18: Kroneman, A., et al Increase in norovirus activity reported in Europe. Euro Surveill. 11:E Layton, A., et al Development of Bacteroides 16S rrna gene Taq- Man-based real-time PCR assays for estimation of total, human, and bovine fecal pollution in water. Appl. Environ. Microbiol. 72: Logan, C., and N. O Sullivan Detection of viral agents of gastroenteritis: norovirus, sapovirus and astrovirus. Future Virol. 3: Logan, C., J. J. O Leary, and N. O Sullivan Real-time reverse transcription PCR detection of norovirus, sapovirus and astrovirus as causative agents of acute viral gastroenteritis. J. Virol. Methods 146: Logan, C., J. J. O Leary, and N. O Sullivan Real-time reverse transcription-pcr for detection of rotavirus and adenovirus as causative agents of acute viral gastroenteritis in children. J. Clin. Microbiol. 44: Lopman, B. A., et al Viral gastroenteritis outbreaks in Europe, Emerg. Infect. Dis. 9: Liu, J., et al Multiplex reverse transcription PCR Luminex assay for detection and quantification of viral agents of gastroenteritis. J. Clin. Virol. 50: Ontario Ministry of Health and Long-Term Care Appendix B: provincial case definitions for reportable disease. In Infections diseases protocol. Queen s Printer for Ontario, Toronto, Canada. /idprotocol/appendixb/appendix_b.pdf.

9 3162 HIGGINS ET AL. J. CLIN. MICROBIOL. 26. Pang, X. L., et al Increased detection of rotavirus using a real time reverse transcription-pcr (RT-PCR) assay in stool specimens from children with diarrhea. J. Med. Virol. 72: Pang, X. L., J. K. Preiksaitis, and B. Lee Multiplex real time RT-PCR for the detection and quantitation of norovirus genogroups I and II in patients with acute gastroenteritis. J. Clin. Virol. 33: Rabenau, H. F., et al Verification and validation of diagnostic laboratory tests in clinical virology. J. Clin. Virol. 40: Royuela, E., A. Negredo, and A. Sanchez-Fauquier Development of a one step real-time RT-PCR method for sensitive detection of human astrovirus 1. J. Virol. Methods 133: Svraka, S., et al Etiological role of viruses in outbreaks of acute gastroenteritis in The Netherlands from 1994 through J. Clin. Microbiol. 45: Svraka, S., et al Novel approach for detection of enteric viruses to enable syndrome surveillance of acute viral gastroenteritis. J. Clin. Microbiol. 47: van Maarseveen, N. M., E. Wessels, C. S. de Brouwer, A. C. Vossen, and E. C. Claas Diagnosis of viral gastroenteritis by simultaneous detection of adenovirus group F, astrovirus, rotavirus group A, norovirus genogroups I and II, and sapovirus in two internally controlled multiplex real-time PCR assays. J. Clin. Virol. 49: White, P. A., et al Norwalk-like virus 95/96-US strain is a major cause of gastroenteritis outbreaks in Australia. J. Med. Virol. 68: Widdowson, M. A., et al Outbreaks of acute gastroenteritis on cruise ships and on land: identification of a predominant circulating strain of norovirus United States, J. Infect. Dis. 190: Zeng, S. Q., et al One-step quantitative RT-PCR for the detection of rotavirus in acute gastroenteritis. J. Virol. Methods 153: