Multicenter Clinical Evaluation of the Automated Aries Group A Strep PCR Assay from Throat Swabs

Group A Streptococcus (GAS) is one of the leading causes of bacterial pharyngitis. Early GAS diagnosis is critical for appropriate antibiotic administration that reduces the risk of GAS sequelae and limits spread of the infection.

ABSTRACT

Group A Streptococcus (GAS) is one of the leading causes of bacterial pharyngitis. Early GAS diagnosis is critical for appropriate antibiotic administration that reduces the risk of GAS sequelae and limits spread of the infection. The Aries Group A Strep (GAS) assay (Luminex, Austin, TX) is a fully automated PCR assay for direct detection of GAS in throat swab specimens in less than 2 h with minimum hands-on time. This multicenter prospective study evaluated the clinical performance of the Aries GAS assay compared to that of Streptococcus pyogenes culture. Subjects with symptoms consistent with pharyngitis were enrolled across four sites in the United States, and a throat swab in liquid Amies medium was obtained. Aries and reference testing was performed within 72 and 48 h after sample collection, respectively. Of 623 throat swab specimens from patients with pharyngitis (93.6% <18 years old, 54.3% female), the reference method yielded valid results for 618 specimens. Reference and Aries assay testing showed GAS-positive results for 160 (25.9%) and 166 (26.9%) specimens, respectively. Compared to the reference method, Aries assay sensitivity was 97.5% (95% confidence interval [CI], 93.7% to 99.0%), specificity was 97.8% (95% CI, 96.0 to 98.8%), positive predictive value was 94.0% (95% CI, 89.3% to 96.7%), and negative predictive value was 99.1% (95% CI, 97.7% to 99.7%). There were 10 false-positive and four false-negative detections with the Aries assay. Discrepant analysis with bidirectional sequencing yielded concordant results with the Aries assay for nine of 14 discordant samples. The Aries assay had high sensitivity and specificity for qualitative detection of group A Streptococcus from patients with pharyngitis.

INTRODUCTION

Streptococcus pyogenes, also known as group A β-hemolytic Streptococcus (GAS), is a common bacterial pathogen that causes a broad spectrum of diseases in humans (1). Invasive GAS colonization can cause serious conditions, including rheumatic heart disease, toxic shock syndrome, necrotizing fasciitis, and septicemia, whereas more commonly, it is the leading cause of pharyngitis and tonsillitis worldwide (1). While GAS-related pharyngitis only occasionally progresses to serious disease, it by itself poses a considerable burden on health care resources (2, 3). The prevalence of GAS infection is highest in children 3 to 18 years old; an estimated 15% of school-age children in developed countries experience a symptomatic case of S. pyogenes pharyngitis each year, and the incidence in emerging nations may be severalfold higher (4, 5). In the United States, acute pharyngitis and tonsillitis account for more than 13 million ambulatory care visits per annum (6), and GAS is the causative agent in at least 20% to 30% of pediatric cases and 5% to 20% in adults. This is a mere fraction of the more than 600 million cases of GAS pharyngitis estimated to occur globally each year (5). Prompt diagnosis is essential to initiate appropriate antibiotic therapy, prevent GAS sequelae, and limit spread of disease.

Contemporary options to confirm S. pyogenes involvement in pharyngitis cases include bacterial culture, immunoassays, and molecular evaluations of throat swab specimens. Culture of a throat swab on a sheep blood agar plate is the established gold standard technique for documenting GAS pharyngitis, and if performed correctly, it provides 90% to 95% sensitivity; however, it requires 24 to 48 h to obtain results, is labor-intensive, and must be done in a clinical laboratory (7). Rapid antigen detection tests (RADTs) for group A Streptococcus are commonly used and have good specificity, but their sensitivity at an average 86% remains suboptimal and is highly variable across different assay kits (8, 9). In the United States, the Infectious Diseases Society of America and American Heart Association guidelines indicate that negative tests in children and adolescents should be submitted to bacterial culture for confirmation (7, 10). New molecular assays for GAS detection provide superior sensitivity and specificity compared to currently available RADTs and are continuously improving in specimen turnaround time and simplification of the assay protocol (11).

The Aries Group A Strep (GAS) assay (Luminex Corp., Austin, TX) is a sample-to-answer automated nucleic acid amplification assay indicated for use with Aries systems (Luminex Corp.). The Aries cassette is a disposable single-use device that contains all the reagents for GAS detection by targeting a conserved region of the S. pyogenes sdaB gene. The assay is designed to rapidly detect S. pyogenes in throat swab specimens with high specificity and sensitivity in less than 2 h. The current multicenter prospective study evaluated the clinical performance of the GAS assay in detecting S. pyogenes in clinical specimens obtained from subjects with diagnosed pharyngitis.

MATERIALS AND METHODS

Study design.

This multicenter clinical trial was conducted on prospectively collected deidentified throat swab specimens from subjects presenting with pharyngitis at four geographically diverse U.S. sites (California, Missouri, Virginia, and Wisconsin) for which their doctor ordered a GAS test (Table 1). Fresh throat swab specimens from both male and female patients were collected from January 2017 to May 2017. One enrollment site enrolled leftover throat specimens that were submitted for GAS testing for routine clinical care. The other three sites prospectively enrolled subjects with symptoms of pharyngitis during their visit to the health care facilities. Written informed consent was obtained from all participating subjects prior to study enrollment and collection of throat swabs (Table 1). Inclusion criteria were that (i) the subject or a legal guardian provided informed consent, (ii) the subject had pharyngitis consistent with GAS infection, and (iii) the subject provided a throat swab specimen collected in Amies medium with a minimum total volume of 850 µl. The sole exclusion criterion was concurrent use of antibiotics for any reason within 24 h prior to specimen collection. The study was designed to evaluate the performance of the Aries assay results compared to bacterial culture followed by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) identification. The study protocols were approved by each hospital’s institutional review board.

TABLE 1.

Demographics of the clinical study population

Characteristic	No. (%) by site (n)
	Site 1 (105)	Site 2 (240)	Site 3 (130)	Site 4 (148)	Total (623)
Sex
Male	45 (42.9)	115 (47.9)	54 (41.5)	71 (48.0)	285 (45.7)
Female	60 (57.1)	125 (52.1)	76 (58.5)	77 (52.0)	338 (54.3)
Age (yr)
≤18	105 (100.0)	240 (100.0)	90 (69.2)	148 (100.0)	583 (93.6)
>18	0 (0.0)	0 (0.0)	40 (30.8)	0 (0.0)	40 (6.4)
Subject status
Hospitalized	0 (0.0)	0 (0.0)	0 (0.0)	4 (2.7)	4 (0.6)
Emergency department	105 (100.0)	0 (0.0)	0 (0.0)	144 (97.3)	249 (40.0)
Outpatient clinic	0 (0.0)	240 (100.0)	130 (100.0)	0 (0.0)	370 (59.4)
Immune status
Compromised	1 (1.0)	0 (0.0)	1 (0.8)	2 (1.4)	4 (0.6)
Competent	104 (99.0)	240 (100.0)	129 (99.2)	146 (98.6)	619 (99.4)

Specimens.

A total of 623 subjects at the four different sites were included in the study. A summary of the subjects enrolled and GAS prevalence across the four sites is provided in Table 1. Patients from ages 1 to 81 years with symptoms consistent with pharyngitis were included. Throat swab specimens were collected in liquid Amies medium and refrigerated at 2 to 8°C within 48 h of collection. Three sites used Copan ESwabs (product no. 480C), and one site used BD ESwabs (product no. 220245; BD Biosciences, San Jose, CA). At each study site, specimens were deidentified and were assigned a unique site-specific study identifier. Four specimen aliquots were prepared and stored at 2 to 8°C before testing or were cryopreserved at −70°C. Testing was performed with the Aries assay at each study site with one sample aliquot. The second aliquot was shipped at 2 to 8°C to one of the study sites that acted as a reference testing center for performing the reference culture testing. The 3rd aliquot was utilized to rerun the Aries assay if invalid results were obtained with the first run. The 4th aliquot was shipped to the company for discordant analysis. The results from the Aries and reference culture testing were compared to determine the performance of the Aries assay.

Aries Group A Strep assay.

One specimen aliquot stored at room temperature or at 2 to 8°C (if >48 h after collection) was tested with the Aries assay within 72 h of collection at each clinical study site, in accordance with the manufacturer’s instructions. All study sites used the Aries system with GAS investigational use only (IUO) assay cassettes (product no. 50-10039; Luminex). Sample testing involved adding 200 μl of the specimen to the sample chamber of the Aries cassette. A cassette has all reagents required for testing of a sample. The Aries cassette was inserted into the magazine (part of the Aries system that holds the Aries cassette), and the magazine was loaded into the Aries system. Up to 12 samples could be batch tested on the Aries system at a single time. Results were displayed on the Aries system as positive, negative, or invalid at the conclusion of the run in less than 2 h. External controls included a GAS positive-control culture (ATCC #700294) and a medium-only negative control. An extractable sample processing control (SPC) for PCR extraction and amplification fidelity is included within the Aries assay cassette. Specimens testing negative for GAS were required to have a positive result for the SPC (threshold cycle [C_T] and melt detection in appropriate channel) for the negative result to be considered valid. Operators at each study site tested external positive and negative controls every day prior to running the Aries assay on the study samples.

Reference testing.

The second aliquots from all samples were shipped at 2 to 8°C to the reference testing center. Reference method testing of all clinical specimens was performed by trained personnel within 48 h of collection. Samples were cultured on both group A-selective strep agar containing 5% sheep blood and Trypticase soy agar (TSA) containing 5% sheep blood for 18 to 48 h under anaerobic conditions at 35 to 37°C. At 18 to 48 h, plates were examined for β-hemolytic colonies indicative of GAS. If no β-hemolytic growth was observed by 48 h, the specimen was deemed GAS negative. Positive β-hemolytic colonies were subcultured onto a new TSA II plate for an additional 18 to 48 h. If no β-hemolytic growth was observed on the subculture at 48 h, a second subculture from the original colony was performed and read at 24 and 48 h. If no β-hemolytic growth was observed on the two sequentially repeated subculture plates, the specimen was disqualified. Culture-positive β-hemolytic colonies were confirmed to be GAS by MALDI-TOF MS using a Bruker MALDI Biotyper CA system and FDA-cleared claim 2 library (Bruker Corp., Billerica, MA), as per the manufacturer’s instructions. MALDI confidence values of <2.00 were considered inconclusive.

Discordant analysis.

Specimens with discordant findings between the reference method and the GAS assay were tested by bidirectional sequencing using a reserved sample aliquot. Bidirectional sequencing targeted a genomic region distinct from that of the Aries GAS assay.

Statistical analysis.

(i) Sample size calculation. To achieve ≥95% sensitivity and specificity with a lower bound of the two-sided 95% CI of >90% (the assay acceptability criteria), based on a prevalence rate of 20% for S. pyogenes in the study population, at least 500 specimens were necessary to be enrolled, with a minimum of 100 positive specimens.

(ii) Descriptive statistics. Overall characteristics of the subjects enrolled and the GAS prevalence at each individual site were determined. GAS prevalence was also determined by sex, age, subject status, and immune status across each of the four sites.

(iii) Analytical statistics. Two-by-two data tables were utilized to determine the performance characteristics (sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]) of the Aries assay in comparison with the culture reference method. Data are presented as value ± 95% confidence intervals (CI). The analysis was performed using the VassarStats website (http://vassarstats.net/clin1.html).

RESULTS

A total of 735 unique throat swab specimens were collected from subjects presenting with indication of pharyngitis at four study sites during January to May 2017. Of these 735 specimens, 112 specimens were excluded from performance calculations, primarily due to (i) deviation from the reference method culture protocol (n = 62) and (ii) patients using antibiotics (n = 24), leaving 623 evaluable specimens. All except 5 samples were held at room temperature and refrigerated within 4 h of sample collection. These 5 samples were refrigerated at >4 h but <21 h post-sample collection. However, sample stability testing performed independently by the company verified that specimens can be held at room temperature for up to 48 h with no effect on assay performance (data not shown). Individual sites provided between 17% and 40% of the total study samples. The majority of the subjects were <18 years old, 54% were female, and nearly 60% visited outpatient clinics. Detailed patient demographic information is listed in Table 1.

Of the 623 unique specimens, 26.8% (n = 167) were reported as positive by the Aries GAS assay. Six specimens initially yielded invalid results due to run failure or instrument error and were retested (6/623; invalid rate, 1.0%); all six specimens generated valid results (positive or negative) after retest. Reference culture/MALDI-TOF MS yielded positive results for 160 samples (25.7%); five specimens (0.80%) generated inconclusive results by the reference method, with a MALDI score less than 2, and were excluded from performance calculations (Table 2). The performance characteristics of the Aries GAS assay were calculated from valid results obtained from both tests on 618 samples. The Aries GAS assay showed a sensitivity of 97.5% (95% CI, 93.7 to 99.0%) and a specificity of 97.8% (95% CI, 96.0 to 98.8%) (versus culture) (Table 3). The PPV and NPV were 94.0% (95% CI, 89.3 to 96.7%) and 99.1% (95% CI, 97.7 to 99.7%), respectively.

TABLE 2.

Site-wide GAS prevalence by reference culture

Site	No. of specimens	Bacterial culture followed by MALDI-TOF MS result (no. [%])a
		Positive	Negative	Inconclusive
1	105	37 (35.2)	65 (61.9)	3 (2.9)
2	240	60 (25.0)	178 (74.2)	2 (0.8)
3	130	35 (26.9)	95 (73.1)	0 (0.0)
4	148	28 (18.9)	120 (81.1)	0 (0.0)
Total	623	160 (25.7)	458 (73.5)	5 (0.8)

^aFive specimens generated inconclusive results by culture plus identification with matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) (MALDI score, <2).

TABLE 3.

Clinical performance of the Aries Group A Strep assay^a

Aries GAS assay result (no. [%])	Bacterial culture plus MALDI-TOF MS result (no.)
	Positive	Negative
Positive (166 [26.9])	156	10b
Negative (452 [73.1])	4c	448
Total (618)d	160	458

^aThe Aries GAS assay showed a sensitivity of 97.5% (95% CI, 93.7 to 99.0%), a specificity of 97.8% (95% CI, 96.0 to 98.8%) (versus culture), a PPV of 94.0% (95% CI, 89.3 to 96.7%), and an NPV of 99.1% (95% CI, 97.7 to 99.7%).

^bSeven Aries GAS-positive specimens that were negative by bacterial culture (i.e., false positive) were positive by bidirectional sequencing using primers targeting genomic regions distinct from those by the Aries GAS assay.

^cTwo Aries GAS-negative specimens that were positive by bacterial culture (i.e., false negative) were GAS negative by bidirectional sequencing.

^dFive specimens (0.80%) generated inconclusive results by comparator culture and were excluded from performance calculations.

Discrepant analysis was performed by repeat testing of a frozen aliquot of the discrepant sample with bidirectional sequencing between 22 and 113 days post-sample collection. Seven out of 10 Aries GAS-positive specimens that were negative by bacterial culture (i.e., false positive) were positive by bidirectional sequencing using primers targeting genomic regions (exotoxin B gene [speB]) distinct from the Aries GAS assay. Two out of four false-negative Aries GAS specimens were GAS negative by bidirectional sequencing. Detailed performance parameters for the Aries GAS assay are presented in Table 3.

DISCUSSION

The aim of this multisite clinical trial was to evaluate the performance characteristics of the Aries GAS assay compared to the reference method composed of culture followed by MALDI-TOF MS identification. This is the first study evaluating the performance of the Aries GAS assay, and the results demonstrated excellent sensitivity and specificity of the Aries GAS sample-to-answer real-time PCR-based diagnostic assay for evaluating S. pyogenes in throat swabs from patients with acute pharyngitis.

Immunoassay-based GAS RADTs are useful for positively diagnosing GAS, but their low sensitivity compared to those of culture and nucleic acid amplification assays means that negative GAS RADT findings often require confirmatory testing, usually by bacterial culture (7, 10, 11). With an average sensitivity of 86%, GAS RADTs will miss detecting S. pyogenes in many patients with actual infection, potentially leading to undertreatment and spread of GAS pharyngitis within the community. These missed diagnoses might be particularly problematic in populations that have a higher rate of sequelae due to GAS pharyngitis, such as rheumatic fever (8, 12). Molecular diagnostic tests for detecting GAS provide more reliable results than currently available GAS RADTs, and nucleic acid assay complexity and turnaround times are continuously being reduced. The economic burden of pediatric pharyngitis in the United States alone is likely in excess of $500 million annually (13), without considering the additional societal costs of overprescribing antibiotics (14). While the costs of GAS RADT and reflex culture are currently lower than those of molecular assays (15), this gap is closing (16). Thus, nucleic acid analysis is poised to potentially supplant both GAS RADTs and bacterial culture for routine evaluation of GAS in pharyngitis cases (17).

There are several FDA-cleared molecular assays, including the Simplexa Group A Strep Direct kit (Diasorin Molecular), cobas Strep A assay (Roche Molecular Diagnostics), and Xpert Xpress Strep A (Cepheid) that are real-time PCR assays and target GAS only. Published literature with prospective multicenter study design on Simplexa Group A Strep Direct kit and cobas Strep A assay demonstrated sensitivities of 97.4% and 97.7% and specificities of 95.2% and 93.3%, respectively (18, 19). The sensitivity (97.5%) and specificity (97.8%) of the Aries GAS assay obtained in our study are comparable to those of these other real-time PCR assays. Unlike the GAS RADTs, all of these real-time assays do not require test results to be confirmed by a culture. There are some other FDA-cleared molecular assays available for GAS detection. The illumigene group A Streptococcus (Meridian Bioscience, Inc.) utilizes loop-mediated isothermal amplification (LAMP) technology and has reported sensitivity and specificity of 82% to 99% and 93% to 99.6%, respectively (20, 21). The Quidel Corporation has the Solana Strep Complete assay for group A and group C/G detection and the Solana GAS assay (sensitivity, 91 to 98%; specificity, 97 to 98% [22, 23]) that utilize helicase-dependent amplification for GAS detection, and the Lyra Direct Strep assay, which is a multiplex real-time PCR assay for detection of group A and C or G Streptococcus (sensitivity and specificity, 100%) (24). The Alere i Strep A and Alere i Strep A2 assays (Abbott Laboratories) use the isothermal nicking enzyme amplification reaction (NEAR) and have reported sensitivity and specificity of 96% to 100% and 91.3% to 100%, respectively (25–27). It has been suggested that molecular assays be considered a replacement for GAS RADTs followed by culture or as a replacement of the confirmatory culture assay for quicker definitive diagnosis (18).

In the analytical assessments in this study, 99% of samples (617/623) gave valid results. The six samples that initially gave invalid readings provided a definitive outcome when retested. Seven of 10 Aries GAS-positive specimens that were negative by bacterial culture (i.e., apparent false positives) were positive by bidirectional sequencing using primers targeting genomic regions distinct from those by the Aries assay. This may have been due to samples containing S. pyogenes DNA in the absence of living bacteria, possibly from individuals with resolving infection or who had been taking antibiotics prior to the specimen collection date (28). This result may also reflect the ability of molecular assays to detect low bacterial loads that culture might not detect (29). Two out of the four false-negative samples that were classified positive by bidirectional sequencing and the culture method may have been due to the presence of viable organisms at concentrations below the assay detection limit.

Strengths of this study included the multicenter prospective design, a sufficiently sized cohort that allows generalization to the younger population typically affected by S. pyogenes-associated pharyngitis, and employment of a culture reference method with MALDI-TOF MS identification. Study limitations include the inability of the qualitative Aries assay to discern active GAS infection from an incidental low-level carrier state, though this limitation applies to all current means of detection, including culture. The likelihood of misidentifying an S. pyogenes carrier as an actively infected individual decreases when GAS testing is preceded by judicious evaluation of clinical symptoms (30). Another limitation is the inability to identify viral etiologies or to detect other less-common bacterial causes of pharyngitis, such as group C and G Streptococcus (31).

The Aries platform can simultaneously evaluate up to 12 samples, which allows for either batch testing or sequential determinations. The self-contained assay cartridges contain all reaction components and incorporate an internal control to monitor assay fidelity, so the test is effectively a one-step procedure that might be adapted for near-patient use in moderately complex urgent care settings or hospital outpatient departments, or it could potentially be further developed to reduce turnaround time (TAT) suitable for use in point-of-care (POC) diagnostics (32). The Aries Group A Strep assay provides outstanding sensitivity and specificity that may be sufficient to allow its use as a standalone method for evaluating GAS status in patients with pharyngitis.