Evaluation of RIDA^®GENE norovirus GI/GII real time RT-PCR using stool specimens collected from children and adults with acute gastroenteritis

Abstract

Background:

Norovirus is the leading cause of epidemic and sporadic acute gastroenteritis (AGE) in the United States. Widespread prevalence necessitates implementation of accurate norovirus detection assays in clinical diagnostic laboratories.

Objective:

To evaluate RIDA^®GENE norovirus GI/GII real-time RT-PCR assay (RGN RT-PCR) using stool samples from patients with sporadic AGE.

Study design:

Patients between 14 days to 101 years of age with symptoms of AGE were enrolled prospectively at four sites across the United States during 2014–2015. Stool specimens were screened for the presence of norovirus RNA by the RGN RT-PCR assay. Results were compared with a reference method that included conventional RT-PCR and sequencing of a partial region of the 5′end of the norovirus ORF2 gene.

Results:

A total of 259 (36.0%) of 719 specimens tested positive for norovirus by the reference method. The RGN RT-PCR assay detected norovirus in 244 (94%) of these 259 norovirus positive specimens. The sensitivity and specificity (95% confidence interval) of the RGN RT-PCR assay for detecting norovirus genogroup (G) I was 82.8% (63.5–93.5) and 99.1% (98.0–99.6) and for GII was 94.8% (90.8–97.2) and 98.6% (96.9–99.4), respectively. Seven specimens tested positive by the RGN-RT PCR that were negative by the reference method. The fifteen false negative samples were typed as GII.4 Sydney, GII.13, GI.3, GI.5, GI.2, GII.1, and GII.3 in the reference method.

Conclusions:

The RGN RT-PCR assay had a high sensitivity and specificity for the detection of norovirus in stool specimens from patients with sporadic AGE.

1. Background

Globally, norovirus contributes to 18% of all diarrheal diseases and is associated with approximately 200,000 deaths each year [1]. Individuals from all age groups can be infected by norovirus [2]. Recent studies indicate that with the introduction of rotavirus vaccines (70–80% uptake), norovirus has become the leading etiology for acute gastroenteritis (AGE) among children in the United States [3–5]. Although norovirus gastroenteritis is for most people a self-limiting disease, it may be life-threatening for young children, the elderly, and immunocompromised patients [2].

Noroviruses are a genetically and antigenically diverse group of viruses belonging to the family Caliciviridae [6]. They can be classified into at least six established and one recently proposed genogroups (GI-GVII) with more than 40 genotypes [1,2,6,7]. GI and GII viruses are the leading cause of infections in humans over the last decade. Specifically, GII.4 viruses are responsible for the majority of norovirus infections both globally and in the USA [8,9]. Enzyme immunoassays and rapid immuno-chromatographic assays have been developed but sensitivity could range between 31.6%–92.0% and 17.0%–90.2%, respectively [2]. PCR based methods are currently considered the gold standard method for norovirus detection [2].

The RIDA^®GENE norovirus GI/GII real-time RT-PCR assay (RGN RT-PCR; R-Biopharm AG, Darmstadt, Germany) can simultaneously detect and differentiate norovirus GI and GII in stool specimens. This RGN RT-PCR targets the relatively conserved ORF1/ORF2 junction region. Although few groups internationally have evaluated the RGN RT-PCR [10,11], an assessment of the assay in the United States has not been reported.

The primary goal of this study was to evaluate the performance characteristics of RIDA^®GENE norovirus GI/GII RT-PCR assay using stool samples from pediatric and adult patients with AGE.

2. Materials and Methods

Study design and sample collection:

This study was conducted at four sites including two primary care medical facilities and two public health laboratory and a reference site (Centers for Disease Control and Prevention [CDC]) from February 2014 to December 2015 (Table 1). The study protocol was approved by the respective institutional review boards. Specimens came from male and female patients between the ages of 14 days to 101 years. Individual’s greater than 18 years were considered adults for study analysis purposes. Two enrollment site prospectively enrolled subjects with symptoms of AGE during their visit to the health care facilities. Written consent was obtained and stool specimens were collected exclusively for this study. The other two facilities (public health laboratories) exclusively enrolled left-over stool specimens that were submitted for gastroenteritis testing for routine clinical care (Table 1). Stool specimens were obtained within ten days from onset of AGE symptoms for all 4 sites. Specimens were stored at 2–8 °C until nucleic acid was extracted for performing RGN RT-PCR within a week of sample collection. Stool specimens were also tested at the reference site by a reference method/algorithm, comprised of conventional RT-PCR of region C and retesting of the region C negative samples by conventional RT-PCR of region D, followed by sequencing of the PCR products of appropriate sizes [12–14]. Sequences were compared to CDC reference databases for genotype determination.

Table 1.

Patient statistics of participating sites.

Locationa	Childrenb	Adultsb	Male/Female	GIc	GIIc
KC Metropolitan area; MO & KS	249	0	144/105	16	55
Albany, NY	11	88	51/48	11	60
Los Angeles, CA	0	215	208/7	2	3
Lansing, MI	6	150	44/112	0	112
Total	266	453	447/272	29	230

^aMO: Missouri; KS: Kansas; CA: California; MI: Michigan.

^bThe age cut-off for children and adults was 18 years.

^cGI and GII positive detections using composite reference method.

Sample preparation:

Unpreserved raw stool specimens were well mixed and homogenized (if necessary by adding few drops of deionized water). Samples were divided into 250 μl aliquots. The first aliquot was shipped on cold packs to the CDC for testing with the reference method. The second aliquot was extracted and tested in the RGN RT-PCR assay as per manufacturer’s protocol. The remainder of the raw stool specimens were kept at 4 °C until testing was completed.

RNA extraction and RIDA^®GENE norovirus GI/GII PCR assay:

The raw stool specimen was diluted 1:10 with water, the suspension was homogenized, and the homogenate was centrifuged for 1 min at 13,000 × g. Two hundred microliters of clear supernatant and 20 μl of internal control RNA (bacteriophage MS2, provided with the kit) was utilized for nucleic acid (NA) extraction using the NucliSENS^® easyMAG^® instrument (bioMérieux Inc., Durham, NC, USA) to obtain 60 μl elution volume. A negative extraction control was included in each run. NA was stored at 4 °C for up to 8 h. If RT-PCR was not performed within 8 h, NA was stored at −20 °C for up to a week. After RT-PCR testing was completed, NA was stored at −80 °C.

Applied Biosystems^®7500 FAST Dx System (Thermo Fisher Scientific Corp., MA, USA) with software version 1.4 was utilized for the one-step RGN RT-PCR assay. The reaction mixture consisted of 19.4 μl of reaction mix, 0.7 μl of enzyme mix, and 5 μl of NA or controls provided in the kit. The thermal profile included cDNA synthesis at 58 °C for 10 min; denaturation at 95 °C for 5 min; amplification with 45 cycles of 95 °C for 15 s, and 55 °C for 30 s; and cooling at 40 °C for 10 s. The assay utilized Taqman^® probes with fluorescent reporter dye (norovirus GI: Cy5, norovirus GII: FAM, and internal control RNA: VIC) for detection of the amplified targets.

Statistical analysis:

Overall prevalence of norovirus as well as age, gender and symptom distribution among the norovirus-positive and -negative groups were determined.

Two by two data tables were used to determine the sensitivity, specificity, positive and negative predictive values, of the RGN RT-PCR assay in comparison with the reference method. The true positive, true negative, false positive, and false negative categories were determined based on the reference method/algorithms. Estimates are presented together with their 95% confidence intervals (CI) as calculated from site: http://vassarstats.net/clin1.html. The median, minimum and maximum C_T values for both GI and GII were determined.

3. Results

A total of 769 samples were collected by the four study sites. Of these, 50 had to be excluded due to missing results or other exclusion criteria (extended time span or insufficient sample volume n = 24, missing age or gender information n = 6, exceeded expiry of extraction kit n = 19, no result for reference method n = 1). Tests for nine patient samples were repeated because the original test result was invalid due to internal control failure (Invalid Rate: 1.3%). The 719 study samples included 447 specimens from males and 272 from females. The median patient age was 52 years (range 14 days-101 years). The number of patients with diarrhea only, vomiting only, and both diarrhea and vomiting were 294 (40.9%), 109 (15.2%), and 273 (37.9%), respectively. Detailed patient demographic information is listed in Table 2.

Table 2.

Patient demographics and clinical symptoms.

	Reference GI/GII method	Gender	Age Median (range)years	Diarrhea only (%)a	Vomiting only (%)a	Diarrhea and vomiting (%)a	Fever (%)b
Children	Norovirus GI/GII positiv	M: 41	3	2	31	44	20
(N=266)	(n = 79; 29.7%)	F: 38	(0.17–16.1)	(2.5)	(39.2)	(55.7)	(25.3)
	Norovirus Negative	M: 110	2	26	71	87	96
	(n = 187; 70.3%)	F: 77	(0.04–18)	(13.9)	(38.0)	(46.5)	(51.3)
Adults	Norovirus GI/GII positive	M: 66	75	47	4	97	24
(N = 453)	(n = 180; 39.7%)	F: 114	(19–101)	(26.1)	(2.2)	(53.9)	(13.3)
	Norovirus Negative	M: 230	65	219	3	45	11
	(n = 273; 60.3%)	F: 43	(21–99)	(80.2)	(1.1)	(16.5)	(4.03)
Overall	Norovirus GI/GII positive	M: 107	52	49	35	141	44
(N = 719)	(n = 259; 36.0%)	F: 152	(0.17–101)	(18.9)	(13.5)	(54.4)	(16.9)
	Norovirus Negative	M: 340	51	245	74	132	107
	(n = 460; 64.0%)	F: 120	(0.04–99)	(53.3)	(16.1)	(28.7)	(23.3)

^aData missing for 34 subjects.

^bFever temperature defined as 37.2 °C or more. Data missing for 34 subjects.

Reference method and the experimental method detected norovirus in 259 (36%) and 251 (34.9%) of the 719 tested specimens. The sensitivity of the RGN RT-PCR assay compared to the composite reference method for detection of GI and GII was 82.8% (95% CI: 63.5–93.5) and 94.8% (95% CI: 90.8–97.2), respectively. The specificity of the RGN RT-PCR assay for GI and GII was 99.1% (95% CI: 98.0–99.6) and 98.6% (95% CI: 96.9–99.4), respectively. The overall sensitivity and specificity of RGN RT-PCR for norovirus detection was 94.2% (95% CI: 90.4–96.6) and 98.5% (95% CI: 96.7–99.3), respectively. The RGN RT-PCR assay detected dual infections with both GI and GII in four specimens; the reference method confirmed 3 specimens as positive for GII and one as positive for GI. There were two additional specimens (one G1 and one GII) for which the genogroups detected by the RGN RT-PCR assay did not match those detected by the reference method. Detailed RGN RT-PCR assay performance parameters for all study subjects and specific parameters for adult and pediatric population are listed in Table 3.

Table 3.

Performance characteristics of RGN RT-PCR versus reference PCR method.

	Genogroup	TP	FP	TN	FN	Sensitivity (95% CI)	Specificity (95% CI)	% PPV (95% CI)	% NPV (95% CI)
Children	GI	18	1	247	0	100 (78.1–100.0)	99.6 (97.4–99.9)	94.7 (71.9–99.7)	100 (98.1–100.0)
	GII	58	0	205	3	95.1 (85.4–98.7)	100 (97.7–100)	100 (92.3–100)	98.6 (95.5–99.6)
Adults	GI	6	5a	437	5c	54.5 (24.6–81.9)	98.9 (97.2–99.6)	54.5 (24.6–81.9)	98.9 (97.2–99.6)
	GII	160	7b	277	9d	94.7 (89.8–97.4)	97.5 (94.8–98.9)	95.8 (91.2–98.2)	96.9 (93.9–98.5)
Children and Adult	GI	24	6	684	5	82.8 (63.5–93.5)	99.1 (98.0–99.6)	80 (60.9–91.6)	99.3 (98.2–99.7)
	GII	218	7	482	12	94.8 (90.8–97.2)	98.6 (96.9–99.4)	96.9 (93.4–98.6)	97.6 (95.7–98.7)
Overall	GI and GII	244	7	453	15	94.2 (90.4–96.6)	98.5 (96.7–99.3)	97.2 (94.1–98.8)	96.8 (94.6–98.1)

TP = True positive; FP = False positive; TN = True negative; FN = False negative.

^aOf the 5 GI false positive (including 3 dual positives by the RGN RT-PCR assay), four specimens were classified as GII and one were negative by reference algorithm.

^bOf the 7 GII false positive (including one dual positive by the RGN RT-PCR assay), two specimens were classified as GI and five were negative by reference algorithm.

^cOne specimen was misclassified as GII by the RGN RT-PCR assay.

^dOne specimen was misclassified as GI by the RGN RT-PCR assay.

Fifteen specimens negative by the RGN RT-PCR assay were positive by the reference method. These included seven different genotypes: GII.4 Sydney (n = 6), GII.13 (n = 3), GI.3 (n = 2), GI.5 (n = 1), GI.2 (n = 1), GII.1 (n = 1), and GII.3 (n = 1). Seven (2 GI, 5 GII) specimens that were negative by the reference method tested positive by RGN RT-PCR; the C_T value for the two GI false positive results were 27.5 and 29.9 and the median C_T value of GII false positive results was 29.7 (range 17.0–35.9).

Overall, of the 259 specimens determined positive by the reference method, 29 (11.2%) were found to be GI and 230 (88.8%) were found to be GII. The genotypes detected by reference assay in this patient population in order of their frequency of detection for genogroup GI were GI.3, GI.2, and GI.5 and for genogroup GII were GII.4 Sydney, GII.6, GII.13, GII.7, GII.3, GII.2, GII.1, GII.8, and GII.17. The most prevalent genotype in GI was GI.3 (n = 25, 9.7%). The most prevalent genotypes in GII were GII.4 Sydney (n = 143, 55.2%), GII.6 (n = 30,11.6%) and GII.13 (n = 23, 8.9%) (Table 4).

Table 4.

Frequency distribution of genogroups and genotypes determined by the reference PCR method.

Genogroups	Prevalence	Genotypes	Frequency (N)	Percentage (%)
GI	11.2% (29/259)	GI.3	25	9.7
		GI.2	2	0.8
		GI.5	2	0.8
GII	88.8% (230/259)	GII.4 Sydney	143	55.2
		GII.6	30	11.6
		GII.13	23	8.9
		GII.7	17	6.6
		GII.3	8	3.1
		GII.2	4	1.5
		GII.1	2	0.8
		GII.8	2	0.8
		GII.17	1	0.4

4. Discussion

Reverse Transcriptase PCR (RT-PCR) is the method of choice for norovirus detection in clinical laboratories [7]. However, RT-PCR assay could yield positive results in asymptomatic individuals shedding norovirus in stool specimens [15]. RT-PCR assay results should be cautiously interpreted since a positive result may not always correlate with the clinical disease state. Both clinical symptoms along with assessment of viral load should be used to determine the casual relationship [7].

The RGN RT-PCR is a TaqMan-based real-time RT-PCR assay for the simultaneous detection of norovirus GI and GII viruses. This study reports the evaluation of the performance of this assay in the United States. In this prospective study among children and adults with AGE, we found that the RGN RT-PCR demonstrated a comparable sensitivity (GI: 82.8% and GII: 94.8%) and specificity (GI: 99.1% and GII: 98.6%) to a reference method for the detection of norovirus in clinical stool specimens. There were seven false positive results by RGN RT-PCR assay; C_T values range 17.0–35.9. Reference method included sequencing of PCR products obtained from targeting region C and D of the ORF2 region with the conventional PCR method. The target region for RGN RT-PCR assay is the ORF1/ORF2 junction region. The difference in the target region could have attributed to these discrepancies. Alternate molecular method was not implemented due to unavailability of another standalone FDA approved molecular method at the time this study was conducted. The prevalence of norovirus among patients with AGE across the four study sites was relatively high at approximately 36%, with 88.8% of the infections caused by GII and 11.2% by GI viruses. The most prevalent genotypes were GII.4 Sydney (55.2%), GII.6 (11.6%), GI.3 (9.7%), and GII.13 (8.9%).

Several PCR-based commercial assay kits have been evaluated for the detection of norovirus [2]. The reported sensitivities of these assays have been more than 90% [10,11,16]. A study in South Korea compared the RGN RT-PCR assay and the AccuPower norovirus PCR assay (Bioneer Co., South Korea) on archived residual specimens. The percent positive and negative agreement between the two assays were 99.0% and 95.1%, respectively [11].

The RGN RT-PCR was previously evaluated on specimens derived from convenience sampling of outbreak specimens in Australia [10]. The specimens came from AGE patients aged 1 year to 101 years (mean age 64 years). Similar to our findings, GII viruses were predominant (87%), with genotype GII.4 being the most prevalent genotype. That study found the assay to have both a sensitivity and a specificity of about 98%. The assay was found to be more sensitive for detecting genogroup GII (100%) than genogroup GI (85%). Our study actively enrolled adult and pediatric AGE patients who had a median age of 52 years (range 14 days-101 years) and the data generated in it showed the RGN RT-PCR assay to have a higher sensitivity for detecting genogroup GII (94.8%) than genogroup GI (82.8%). The slight difference in the performance parameters (combined for both genotypes; sensitivity: 94.2% and specificity: 98.5%) obtained in our and the Australian study may be due to differences in the diversity of genotypes circulating in the population or the different reference methods that were used. We utilized conventional PCR assays targeting the C and/or D regions followed by sequencing as a reference method, while the Australian study utilized a set of eight different protocols to determine the sensitivity and specificity of the RGN RT-PCR assay.

A recent study compared three FDA-cleared multiplex assays that include norovirus as one of the targets; Verigene^® enteric pathogen test (Nanosphere, Inc., Illinois, USA), FilmArray gastrointestinal panel (Biofire Diagnostics, Utah, USA) and Luminex xTAG^® gastrointestinal panel (Luminex Corporation, Toronto, Canada) using retrospective and prospective specimens collected from children [17]. The sensitivity of norovirus detection in the Verigene assay was assessed at 89% (95% CI: 67.0–98.7), in the Luminex 89.5% (95% CI: 67.0–98.7) and in the FilmArray assay 94.7% (95% CI: 74.0–100). Notably the study was limited to a pediatric population and a relatively low number of positive samples (n = 19), resulting in broader confidence intervals and less reliable estimates. However, it is interesting that the 94.2% (95% CI: 90.4–96.6) sensitivity of the RGN RT-PCR assay in our study, generated from 244 true positive specimens that resulted in robust sensitivity estimates, was similar to that of the Film Array. The emergence of norovirus as the leading cause of AGE calls for robust detection assays in a clinical diagnostic lab. Our data suggest that the RGN RT-PCR assay, a stand-alone norovirus detection assay, can provide a sensitive and less expensive method for norovirus detection compared to gastrointestinal pathogen multiplex assays.

A recent study evaluated the Xpert^® norovirus assay, another standalone molecular assay that can detect and differentiate GI and GII noroviruses [13]. This multicenter study tested samples collected in 2014 and used similar reference assays used in our study. The positive percent agreement for the Xpert^® norovirus assay for GI and GII viruses was 98.3% and 99.4%, respectively. The negative percent agreement for GI and GII was 98.1% and 98.2%, respectively. The Xpert^® norovirus assay had equivalent specificity but higher sensitivity when compared to the RGN RT-PCR assay. However, the majority of samples tested in the Xpert^® norovirus study were obtained from adults over 21 years of age (n = 1225, 87.3%), whereas in our study adults contributed to only 63% of the study population. Furthermore, the prevalence of GI viruses among the positive samples was lower in our study [n = 29 (11.2%)] as compared with this Xpert study [n = 138 (41.1%)]. Additional studies are required for more thorough comparisons of these commercially available norovirus detection assays.

The genotypes detected consistently (100% detection rate) by the RGN assay included GII.2, GII.6, GII.7, GII.8, and GII.17. The fifteen samples that tested false negative by the RGN RT-PCR assay were typed as GII.4 Sydney, GII.13, GI.3, GI.5, GI.2, GII.1, and GII.3. In the Australian study, the RGN RT-PCR assay similarly missed GI.2 but consistently detected G1.3 and GII.13 [10]. The inability of the assay to consistently detect some strains may be due to sequence variation between strains of the same genotype or a slightly low analytical sensitivity for detection of particular strains. Overall, the RGN RT-PCR assay detected a diverse range of genotypes present in our patient population.

Overall, our study results indicate that the RGN RT-PCR assay has a high sensitivity and specificity for the detection of norovirus in stool specimens from patients with sporadic AGE. The RGN RT-PCR assay is an easy-to-use single step, real time RT-PCR assay for detection and differentiation of GI and GII viruses.