JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 2008, p. 1510–1513 0095-1137/08/$08.00⫹0 doi:10.1128/JCM.01694-07 Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 46, No. 4
GeneXpert Enterovirus Assay: One-Year Experience in a Routine Laboratory Setting and Evaluation on Three Proficiency Panels䌤 Katja Seme,1 Tina Mocˇilnik,1 Kristina Fujs Komlosˇ,1 Ana Doplihar,1 David H. Persing,2 and Mario Poljak1* Institute of Microbiology and Immunology, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia1, and Cepheid, Sunnyvale, California2 Received 25 August 2007/Returned for modification 25 October 2007/Accepted 25 January 2008
A prospective unblinded comparative evaluation of three assays for the detection of enteroviral RNA performed on 83 positive and 79 negative cerebrospinal fluid samples showed initial and resolved sensitivities of 90.4% and 98.8%, respectively, for the Cepheid GeneXpert enterovirus assay; 94.0% and 97.6%, respectively, for the Argene enterovirus consensus kit; and 100% and 100%, respectively, for an in-house real-time PCR. The initial and resolved specificities were 100% for all three assays. detection of enterovirus RNA in CSF. It combines automated nucleic acid sample preparation, amplification, and real-time detection of enteroviral RNA in a disposable, macro/microfluidic cartridge using the GeneXpert Dx system instrument. To date, only one evaluation of GXEA has been published in peer-reviewed journals. A multicenter beta trial with 102 CSF samples obtained from patients with suspected meningitis (34 of whom were enterovirus positive) showed that GXEA had a sensitivity and a specificity of 97.1% and 100%, respectively, and that it is suitable for rapid, on-demand testing (10). In analytical studies, GXEA detected 63 of the enterovirus serotypes for which tests were conducted, with limits of detection ranging from 0.0002 to 200 50% tissue culture infective doses (TCID50s)/ml or 50% lethal doses/ml. It showed no crossreactivity with 24 different microorganisms known to cause meningitis-like symptoms and tolerated the influence of interfering substances on test performance (10). We prospectively evaluated GXEA over a 1-year period and compared it with two other PCR-based assays for the detection of enteroviral RNA in CSF. When we assess a new molecular diagnostic assay in the laboratory, the candidate assay is usually first evaluated on proficiency panels, like the Quality Control for Molecular Diagnostics (QCMD) panel. If the candidate assay shows successful performance on the QCMD panels, it is then evaluated prospectively in routine settings by comparison of its results with those of the existing diagnostic methods for a certain period of time (usually 6 to 12 months). GXEA was first evaluated on three Enterovirus QCMD panels from the years 1999/2000, 2002, and 2004, consisting of 35 panel samples in total. A plasma sample containing coxsackievirus A9 (QCMD 2002) was excluded from the analyses since GXEA is recommended for the testing of CSF only. All 8 enterovirus-negative panel samples tested GXEA negative, while enterovirus RNA was detected by GXEA in 25 of 26 enterovirus-positive panel samples. Enterovirus-positive samples consisted of different 10-fold dilution series of poliovirus 2; coxsackieviruses A9, A16, and B5; ehoviruses 6, 9, and 11; and enterovirus 71 at concentrations ranging from 0.03 to 25,000 TCID50s/ml. The detailed compositions of the three enterovirus QCMD panels included in the study are available
Enteroviruses comprise a large group of immunologically distinct serotypes of viruses belonging to the family Picornaviridae (18). Infection with enteroviruses is associated with protean clinical manifestations, ranging from asymptomatic or mild febrile illness to severe and potentially fatal syndromes, including paralysis, aseptic meningitis, encephalitis, myocarditis, and neonatal systemic infection (1). Enteroviruses are the most common cause of aseptic meningitis in both children and adults and may cause up to 90% of cases of aseptic meningitis for which an etiology is identified (1, 14, 18, 19). The rapid detection and the rapid characterization of enteroviral meningitis are essential for making decisions for patient management and treatment (5, 15, 17, 20). The provision of enterovirus test results on a daily basis can have a substantial impact on health care and has been shown to be highly cost-effective (9, 12, 17). Since the conventional means of diagnosis of enteroviral meningitis by cell culture from cerebrospinal fluid (CSF) is timeconsuming and expensive and lacks sensitivity, some commercial and several in-house PCR protocols as well as nucleic acid sequence-based amplification assays have been developed during the last 15 years in order to improve the ability to detect enterovirus from CSF (2–4, 6, 7, 11, 13, 16, 20–22, 24–26). The most widely used PCR-based assays for the routine diagnosis of enteroviral meningitis are those that detect PCR products in microtiter wells (6, 21, 22) and those that use the real-time PCR technology (2, 3, 13, 24–26). However, all of these approaches require specially trained laboratory staff, thus limiting the ability of the “real-time” technology to deliver patient results that would be most useful for making patient management decisions STAT. The latest development in the field of the molecular diagnosis of enteroviral meningitis is a fully automated real-time multiplex, reverse transcription-PCR assay, the GeneXpert enterovirus assay (GXEA; Cepheid, Sunnyvale, CA) (10). GXEA is, at present, the only FDA-approved assay for the qualitative * Corresponding author. Mailing address: Institute of Microbiology and Immunology, Faculty of Medicine, Zalosˇka 4, Ljubljana 1000, Slovenia. Phone: 386 1 543 7453. Fax: 386 1 543 7418. E-mail: mario
[email protected]. 䌤 Published ahead of print on 6 February 2008. 1510
VOL. 46, 2008
NOTES
TABLE 1. Initial and resolved test results for enterovirus RNA by GXEA, the in-house real-time PCR, and the Argene assay with 162 CSF samples No. of samples Test results
Initial result
Resolved result
GXEA positive/real-time PCR positive/ Argene assay positive GXEA negative/real-time PCR negative/ Argene assay negative GXEA positive/real-time PCR positive/ Argene assay negative GXEA negative/real-time PCR positive/ Argene assay positive
70
80
79
79
5
2
8
1
on the QCMD home page (http://www.qcmd.org/). A single false-negative GXEA result was obtained when panel sample EV-B07 from the 1999/2000 QCMD panel containing 0.036 TCID50s/ml of coxsackievirus A9 was tested. This result was not unexpected, since this particular sample tested enterovirus negative in all three QCMD reference laboratories and was reported to be enterovirus negative by 40 (81.6%) of 49 participating laboratories (23). Following the successful performance of GXEA on QCMD panels, the assay was implemented in the daily routine testing of CSF samples. GXEA was evaluated in parallel with an in-house real-time PCR assay, as described previously (8). Briefly, after RNA isolation from 140 l of CSF by use of the QIAamp viral RNA mini kit (Qiagen, Hilden, Germany), 5 l of RNA was amplified with previously described primers targeting 120 bp of the 5⬘ noncoding region and with a LightCycler RNA master hybridization probes kit (Roche Applied Science, Mannheim, Germany) on a LightCycler (version 2.0) instrument (Roche Applied Science). Thus, each CSF sample was tested for the presence of enterovirus RNA by using GXEA and an in-house real-time PCR on the same day. Since the in-house real-time PCR assay had been implemented in the daily routine only 7 months before the start of the GXEA evaluation and because of a lack of an internal control for the in-house real-time PCR, all CSF samples were additionally tested once weekly by using the enterovirus consensus kit (Argene SA, Varilhes, France), following the manufacturer’s instructions, as described previously (6). Briefly, 10 l of RNA previously extracted for the in-house real-time PCR and stored for a maximum of 1 week at ⫺70°C was amplified by using a one-step reverse-transcription PCR and primers targeting a 425-bp region of the 5⬘ noncoding region, followed by detection of PCR products in a microtiter plate by use of a biotinylated probe. The in-house real-time PCR and Argene assay were performed by two technicians with substantial experience
1511
in molecular diagnostics, while GXEA testing was performed by eight operators who mainly did not have any experience in molecular diagnostics. A total of 162 CSF samples were tested prospectively for the presence of enterovirus RNA by using three different PCR assays from 15 July 2006 until 14 July 2007. Fifteen of 162 samples had to be retested due to different performance problems in the initial run. Seven samples were retested by using GXEA due to an initial invalid run because of internal control failure, while 11 samples (3 of them retested by GXEA) were retested by the Argene assay due to initial negative internal control results (8 samples) or indeterminate enterovirus results (3 samples). The enteroviral RNA status of all 15 samples with initial performance problems by GXEA and the Argene assay was resolved successfully after repeat testing. As shown in Table 1, 149 (91.9%) of 162 samples initially tested concordantly by all three assays. The initial sensitivity and specificity of GXEA, the Argene assay, and the in-house real-time PCR are shown in Table 2. All samples with discordant results were retested by all three assays. Of five initially Argene assaynegative samples, three samples became positive on repeat testing, while two samples remained enterovirus negative by repeat testing by the Argene assay (Table 1). The results for the last two samples which repeatedly tested negative by the Argene assay but which repeatedly tested positive by GXEA and the in-house reverse transcription-PCR were finally considered to be false negative by the Argene assay. For both patients, these results were supported by clinical diagnoses of enteroviral meningitis; and their CSF samples tested negative for herpes simplex virus DNA, cytomegalovirus DNA, varicella-zoster virus DNA, Epstein-Barr virus DNA, and human herpesvirus 6 DNA. These samples were also negative for immunoglobulin M and RNA tick-borne encephalitis virus. Of eight samples initially negative by GXEA, seven tested enterovirus positive, while one sample remained GXEA negative upon repeat testing (Table 1). The result for the latter sample, which repeatedly tested GXEA negative but which repeatedly tested Argene assay and real-time PCR positive, was finally considered to be a false-negative result by GXEA. This result was supported by a clinical diagnosis of enteroviral meningitis, and CSF and blood samples tested negative for all previously listed viruses. The most surprising finding in this evaluation was that seven initially GXEA-negative samples all became highly positive upon repeat testing. Although these seven samples were initially reported to be enterovirus negative by the GeneXpert Dx instrument, we noticed after the first such discrepant result (due to the nonblinded nature of our comparative evaluation) that the end-point fluorescence in these samples showed some evidence of an enteroviral amplification signal, with the signals ranging from 3 to 15. In contrast, the
TABLE 2. Initial and resolved sensitivities and specificities of GXEA, the in-house real-time PCR, and the Argene assay for detection of enterovirus RNA in 162 CSF samples GXEA
Real-time PCR
Argene assay
Testing
Initial Resolved
Sensitivity (%)
Specificity (%)
Sensitivity (%)
Specificity (%)
Sensitivity (%)
Specificity (%)
90.4 98.8
100 100
100 100
100 100
94.0 97.6
100 100
1512
NOTES
results for the concordant enterovirus-negative specimens and negative controls invariably showed enterovirus-specific endpoint fluorescence values of 0. GXEA incorporates a coamplified control for the detection of inhibition; this comprises an “armored RNA” in-process control which is added at low levels to the sample before it is extracted, in order to control for the efficiency of both extraction and amplification. Lower-thanexpected end-point values could be the result of the presence of PCR inhibitors or operator error (pipetting errors, incomplete or incorrect reagent dispensing into the GeneXpert cartridge). The presence of a PCR inhibitor(s) extremely labile to freezing-thawing might be more likely, since repeat testing after a freeze-thaw cycle when freezing lasted at least 1 h resolved the results for all the discrepant samples with endpoint fluorescence values above 0 but below the threshold for a positive result. In contrast, repeat testing of these samples on the same day or without a freeze-thaw cycle or sample dilution 1:10 again resulted in end-point fluorescence values above 0 but below the threshold for a positive result. According to our results, all enterovirus-negative samples with end-point fluorescence values above 0 should be checked, and such values should be used as an indication for the need for a repeat test. In our experience, this additional check took less than a minute and could be performed reliably by individuals without a strong background in molecular diagnostics. Further evaluation of GXEA results with lower-than-expected end-point values from other parts of the world (if they are recognized) could be important, since this phenomenon might be due to different lumbar puncture practices in different regions or other factors. After repeat testing, the results of the three assays agreed for 159 (98.1%) of 162 samples (Table 1). The resolved sensitivities and specificities of GXEA, the Argene assay, and the in-house real-time PCR are shown in Table 2. As described in detail earlier, simple manual checks of all “enterovirus-negative” GXEA results significantly improved the sensitivity of the GXEA from 90.4% to 98.8% (Table 2). Therefore, we plan to monitor all GXEA-negative samples prospectively to determine if this parameter remains useful for the determination of the need for repeat testing. In conclusion, GXEA, the only FDA-approved assay for the detection of enteroviral RNA in CSF, is an important new tool for the diagnosis of enteroviral meningitis. It delivers enterovirus PCR results on a STAT basis and meets a critical patient need for definitive diagnostic results in the evaluation of meningitis. Our study, the first to be published for a series of European patients, confirms the previously described levels of sensitivity and specificity of GXEA (10). In our experience, further improvements in the performance of GXEA were obtained by manually checking enterovirus end-point fluorescence values. For infrequent results for which enterovirus endpoint values are above 0 but below the GXEA threshold for a positive result, we suggest repeating the test after a short freeze-thaw cycle. After 1 year of experience with GXEA in a routine laboratory setting, we consider GXEA to be an excellent system for routine or “on-demand” testing for enteroviral meningitis due to its complete automation and its rapid-result capability.
J. CLIN. MICROBIOL. REFERENCES 1. Abzug, M. J. 2004. Presentation, diagnosis, and management of enterovirus infections in neonates. Paediatr. Drugs 6:1–10. 2. Beld, M., R. Minnaar, J. Weel, C. Sol, M. Damen, H. van der Avoort, P. Wertheim-van Dillen, A. van Breda, and R. Boom. 2004. Highly sensitive assay for detection of enterovirus in clinical specimens by reverse transcription-PCR with an armored RNA internal control. J. Clin. Microbiol. 42: 3059–3064. 3. Corless, C. E., M. Guiver, R. Borrow, V. Edwards-Jones, A. J. Fox, E. B. Kaczmarski, and K. J. Mutton. 2002. Development and evaluation of a ‘real-time’ RT-PCR for the detection of enterovirus and parechovirus RNA in CSF and throat swab samples. J. Med. Virol. 67:555–562. 4. Ginocchio, C. C., F. Zhang, A. Malhotra, R. Manji, P. Sillekens, H. Foolen, M. Overdyk, and M. Peeters. 2005. Development, technical performance, and clinical evaluation of a NucliSens basic kit application for detection of enterovirus RNA in cerebrospinal fluid. J. Clin. Microbiol. 43:2616–2623. 5. Hamilton, M. S., M. A. Jackson, and D. Abel. 1999. Clinical utility of polymerase chain reaction testing for enteroviral meningitis. Pediatr. Infect. Dis. 18:533–537. 6. Jacques, J., J. Carquin, V. Brodard, H. Moret, D. Lebrun, M. Bouscambert, J. Motte, G. Re´my, and L. Andre´oletti. 2003. New reverse transcription-PCR assay for rapid and sensitive detection of enterovirus genomes in cerebrospinal fluid specimens of patients with aseptic meningitis. J. Clin. Microbiol. 41:5726–5728. 7. Kammerer, U., B. Kunkel, and K. Korn. 1994. Nested PCR for specific detection and rapid identification of human picornaviruses. J. Clin. Microbiol. 32:285–291. 8. Kares, S., M. Lonnrot, P. Vuorinen, S. Oikarinen, S. Taurianen, and H. Hyoty. 2004. Real-time PCR for rapid diagnosis of entero- and rhinovirus infections using LightCycler. J. Clin. Virol. 29:99–104. 9. King, R. L., S. A. Lorch, D. M. Cohen, R. L. Hodinka, K. A. Cohn, and S. S. Shah. 2007. Routine cerebrospinal fluid enterovirus polymerase chain reaction testing reduces hospitalization and antibiotic use for infants 90 days of age or younger. Pediatrics 120:489–496. 10. Kost, C. B., B. Rogers, M. S. Oberste, C. Robinson, B. L. Eaves, K. Leos, S. Danielson, M. Satya, F. Weir, and F. S. Nolte. 2007. Multicenter beta trial of the GeneXpert enterovirus assay. J. Clin. Microbiol. 45:1081–1086. 11. Landry, M. L., R. Garner, and D. Ferguson. 2005. Real-time nucleic acid sequence-based amplification using molecular beacons for detection of enterovirus RNA in clinical specimens. J. Clin. Microbiol. 43:3136–3139. 12. Marshall, G. S., M. A. Hauck, G. Buck, and G. P. Rabalais. 1997. Potential cost savings through rapid diagnosis of enteroviral meningitis. Pediatr. Infect. Dis. J. 16:1086–1087. 13. Nijhuis, M., N. van Maarseveen, R. Schuurman, S. Verkuijlen, K. de Vos Hendriksen, and A. M. van Loon. 2002. Rapid and sensitive routine detection of all members of the genus enterovirus in different clinical specimens by real-time PCR. J. Clin. Microbiol. 40:3666–3670. 14. Nolte, F. S. 2006. Case studies in cost effectiveness of molecular diagnostics for infectious diseases: pulmonary tuberculosis, enteroviral meningitis, and BK virus nephropathy. Clin. Infect. Dis. 43:1463–1467. 15. Pe´rez-Ve´lez, C. M., M. S. Anderson, C. C. Robinson, E. J. McFarland, W. A. Nix, M. A. Pallansch, M. S. Oberste, and M. P. Glode´. 2007. Outbreak of neurologic enterovirus type 71 disease: a diagnostic challenge. Clin. Infect. Dis. 45:950–957. 16. Petitjean, J., F. Freymuth, H. Kopecka, J. Brouard, P. Boutard, and P. Lebon. 1994. Detection of enteroviruses in cerebrospinal fluids: enzymatic amplification and hybridization with a biotinylated riboprobe. Mol. Cell. Probes 8:15–22. 17. Robinson, C. C., M. Willis, A. Meagher, K. E. Gieseker, H. Rotbart, and M. P. Glode. 2002. Impact of rapid polymerase chain reaction results on management of pediatric patients with enteroviral meningitis. Pediatr. Infect. Dis. 21:283–286. 18. Romero, J. R. 2007. Enteroviruses and parechoviruses, p. 1392–1404. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. L. Landry, and M. A. Pfaller (ed.), Manual of clinical microbiology, 9th ed. ASM Press, Washington, DC. 19. Rotbart, H. A. 1990. Diagnosis of enteroviral meningitis with the polymerase chain reaction. J. Pediatr. 117:85–89. 20. Rotbart, H. A. 1990. Enzymatic RNA amplification of the enteroviruses. J. Clin. Microbiol. 28:438–442. 21. Rotbart, H. A., M. H. Sawyer, S. Fast, C. Lewinski, N. Murphy, E. F. Keyser, J. Spadoro, S. Y. Kao, and M. Loeffelholz. 1994. Diagnosis of enteroviral meningitis by using PCR with a colorimetric microwell detection assay. J. Clin. Microbiol. 32:2590–2592. 22. Stellrecht, K. A., I. Harding, F. M. Hussain, N. G. Mishrik, R. T. Czap, M. L. Lepow, and R. A. Venezia. 2007. A one-step RT-PCR assay using an enzymelinked detection system for the diagnosis of enterovirus meningitis. J. Clin. Virol. 17:143–149. 23. van Vliet, K., A. M. van Loon, P. Muir, J. M. Echevarria, P. E. Klapper, and
VOL. 46, 2008 G. M. Cleator. Second QCCA enterovirus proficiency panel: summary of results. http://www.qcmd.org/Index2.htm. 24. Verstrepen, W. A., S. Kuhn, M. M. Kockx, M. E. Van De Vyvere, and A. H. Mertens. 2001. Rapid detection of enterovirus RNA in cerebrospinal fluid specimens with a novel single-tube real-time reverse transcription-PCR assay. J. Clin. Microbiol. 39:4093–4096.
NOTES
1513
25. Verstrepen, W. A., P. Bruynseels, and A. H. Mertens. 2002. Evaluation of a rapid real-time RT-PCR assay for detection of enterovirus RNA in cerebrospinal fluid specimens. J. Clin. Virol. 25(Suppl.):S39–S43. 26. Watkins-Riedel, T., M. Woegerbauer, D. Hollemann, and P. Hufnagl. 2002. Rapid diagnosis of enterovirus infections by real-time PCR on the LightCycler using the TaqMan format. Diagn. Microbiol. Infect. Dis. 42:99–105.
