Scandinavian Journal of Infectious Diseases, 2014; 46: 897–901
SHORT COMMUNICATION
Comparison of the FilmArray assay and in-house real-time PCR for detection of respiratory infection
MARIA E. ANDERSSON, SIGVARD OLOFSSON & MAGNUS LINDH Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
Department of Infectious Diseases, University of Gothenburg, Gothenburg, Sweden
Abstract Recently, molecular methods capable of detecting almost all microbial agents that may cause acute respiratory infection have been introduced. The FilmArray Respiratory Panel assay, which integrates nucleic acid extraction, nested amplification and detection in a reaction pouch preloaded with all reagents required for detection of 17 viruses and 3 bacteria, was compared with an in-house real-time PCR that detects these agents in 8 parallel amplifications. When 128 clinical samples representing 18 of these agents were analysed by both assays the agreement was excellent, with kappa values ranging between 0.54 and 1.0. Discordances were mainly observed for adenovirus, but not when version 1.7 of FilmArray was used. The results show that these assays detect a wide range of pathogens with similar performance. FilmArray provides results after approximately 1 h, including ≈ 5 min hands-on time, and does not require advanced equipment or expertise in molecular diagnostics, making it a useful point-of-care-test for acute respiratory infections.
Keywords: Acute respiratory infection, real-time PCR, point-of care testing, FilmArray, molecular diagnostics
Introduction A wide range of viral and bacterial pathogens can cause acute respiratory infections. It is difficult to conclude or even guess the causative agent from symptoms because the clinical presentation is very variable, and microbial agents associated with common cold can also cause serious lower respiratory tract infections [1]. Therefore, simultaneous testing for many respiratory agents is important and can be performed by multiplex PCR assays that target many pathogens by using a mixture of many primers and probes [2]. Amplified products can then be distinguished, for example, as light emitted from beads that bind amplicons [3,4], but a disadvantage with this technique is that it requires post-PCR processing of amplified material, and therefore is time-consuming and poses a risk that amplified products may spread in the laboratory. These problems are avoided in real-time PCR where amplicons are detected in closed reaction wells during amplification. However, multiplexing is usually limited to three to five targets per reaction by this method, due to limitations of technical performance of the instrument and the risk
of interference when several primers and probes are mixed. To detect many pathogens by such real-time PCR each sample has to be run in several parallel reactions. A novel assay, FilmArray (BioFire Diagnostics, Salt Lake City, UT), has refined this technique to allow highly sensitive detection of a large number of targets. This assay integrates nucleic acid extraction, amplification and detection of amplicons in a reaction pouch that contains all necessary reagents, and thus, provides a point-of-care testing approach suitable for smaller laboratories and healthcare settings. In the present study we compared the performance of the FilmArray assay with a multiple real-time PCR that is run in eight parallel amplifications.
Materials and methods Specimens Clinical samples (63 nasopharyngeal swabs, 55 mixed materials containing nasopharyngeal swabs and throat swabs, 5 throat swabs and 5 broncho-alveolar
Correspondence: Maria E. Andersson, Department of Infectious Diseases, University of Gothenburg, Guldhedsgatan 10B, 413 36 Gothenburg, Sweden. Tel: ⫹ 46 709 771595. Fax: ⫹ 46 31 827032. E-mail: [email protected] (Received 19 March 2014 ; accepted 16 July 2014 ) ISSN 0036-5548 print/ISSN 1651-1980 online © 2014 Informa Healthcare DOI: 10.3109/00365548.2014.951681
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
898
M. E. Andersson et al.
lavages) sent for clinical testing to the Virological Diagnostic Laboratory at Sahlgrenska University Hospital were included in the study. Except for age (which ranged between 1 month and 96 years, median 40 years), no clinical information about the patients was registered. Because patient and sample identities were anonymised ethical approval was not required. In a prospective part of the study, 75 samples were analysed in parallel by a laboratory-developed multiplex real-time PCR assay and the FilmArray Respiratory Panel assay. All targeted viruses and bacteria did not appear or were rarely detected in these prospectively collected samples. Therefore, an additional 53 samples that had previously been analysed by the real-time PCR assay in clinical diagnostics (and were stored at –70°C) were included for retrospective analysis by FilmArray. These samples were selected so as to include all agents targeted by the FilmArray panel, aiming at a minimum of four samples per pathogen. Samples containing parainfluenza 4 and Bordetella pertussis were not included because parainfluenza 4 is not included in the in-house assay and because no stored samples with Bordetella pertussis were available. All samples were analysed by FilmArray Respiratory Panel kit version 1.6. Fifteen samples were retested with version 1.7 to evaluate whether adenovirus detection was improved.
Nucleic acid extraction and real-time PCR by the in-house assay Nucleic acid from 200 μl specimen was extracted by a MagNA Pure LC instrument (Roche Diagnostics, Mannheim, Germany) using the Total Nucleic Acid isolation kit. The nucleic acids were eluted in 100 μl volume, and 5 μl of this were used for real-time PCR. Real-time PCR was performed in an ABI 7900 384well system (Applied Biosystems, Foster City, CA, USA) in 8 parallel 20 μl reactions containing oligonucleotides described in Table I [5–9], and Taqman Fast Virus 1-step Mastermix (ABI, for RNA targets) or Universal Mastermix (ABI, for DNA targets). After a reverse transcription step at 46°C for 30 min followed by 10 min of denaturation at 95°C, 45 cycles of two-step PCR was performed (15 s at 95°C, 60 s at 58°C). For optimal performance the ramp rate was set to 80% during all 45 cycles. The performance for each multiplex reagent mixture had been evaluated using pUC57 plasmids with inserts of the targeted viral or bacterial sequences, synthesized by GenScript Corp. (Piscataway, NJ, USA). During this development, combinations were only accepted if the
Ct (threshold cycle) value in multiplex was not more than two cycles higher than when amplification was carried out in separate reactions. The same plasmids with target sequences were subsequently used in each run parallel with patient specimens to verify the performance of each target PCR (master mix control). In addition, one positive control was processed with each set of samples, from extraction of nucleic acids through the detection of amplified products. This in-house real-time PCR targets parainfluenzavirus (PIV) 1–3, respiratory syncytial virus (RSV), influenza A virus (IfA), influenza B virus (IfB), human coronaviruses (CoV NL63, HKU1, OC43, 229E), human enterovirus (EV), human rhinovirus (RV), adenovirus (AdV), bocavirus (BV), human metapneumovirus (hMpn), Mycoplasma pneumoniae, Chlamydophila pneumoniae and Bordetella pertussis. FilmArray respiratory panel The FilmArray analyses were performed according to the manufacturer’s instructions. First, 1 ml of the hydration solution was injected with a syringe to rehydrate biochemical reagents that are needed for the reactions. In the second injection port, 300 μl of a mixture including one part of the sample to be tested and two parts of denaturing sample buffer were added with a second syringe. The loaded pouch was then inserted into the FilmArray instrument, and the pouch and sample were registered using a bar code reader. All further processing was carried out in the instrument without further intervention. After mechanical lysis using zirconium beads and nucleic acid purification using magnetic beads, nested PCR was performed, including an initial multiplex step and a second step in which a small volume from the first step was further amplified in separate reaction mixtures containing a cyanine dye, yielding results within 65 min. The instrument interprets socalled melting curves and for each pathogen presents the result as detected or not detected. Melting curve data can be inspected using separate software. Two internal controls are included; an RNA control that follows the sample through the whole process and a DNA control that is present in the second-stage PCR [10].
Results Of the 75 prospectively analysed samples, any viral or bacterial pathogen was detected in 49 samples by real-time PCR, and in 51 by the FilmArray assay. Thus, 24 samples were negative for all agents by
FilmArray and real-time PCR respiratory panels
899
Table I. Primers and probes for in-house real-time PCR.
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
Pathogen
Mix
Oligo type
Oligo sequence
Parainfluenzavirus 1
1
Parainfluenzavirus 2
1
Parainfluenzavirus 3
1
Respiratory syncytial virus
2
Influenza B virus
2
Coronavirus NL63
2
Metapneumovirus
3
Enterovirus
3
Chlamydophila pneumoniae
3
Influenza A virus
4
Coronavirus HKU1
4
Coronavirus OC43
4
Rhinovirus
5
Coronavirus 229E
5
Adenovirus
6
Mycoplasma pneumoniae
7
Bocavirus
8
F R P F R P F R P F R P F R P F R P F R P F R P F R P F R P F R P F R P F F F R P F R P F R P F R P F R P
CAACAGGAAATCATGTTCTGTAATAGC TCACAGTGGGCAAGGAGCA NED-CTGGTAACCCCTTGTTCCTG-MGB GCATTTCCAATCTTCAGGACTATGA ACCTCCTGGTATAGCAGTGACTGAAC JOE-CCATTTACCTAAGTGATGGAATCAATCGCAAA-BHQ1 AAATGATCTGATTTATGCTTATACCTC TCAGGTACCAAGTCTGAGTTTACA FAM-CGAGGTTGYCAGGATATAGGAAAATCA-BHQ1 GCAAATATGGAAACATACGTGAACA GCACCCATATTGTWAGTGATGCA FAM-CTTCACGAAGGCTCCACATACACAGCWG-BHQ1 AAATACGGTGGATTAAATAAAAGCAA CCAGCAATAGCTCCGAAGAAA VIC-CCATAGGAAATTGCC-MGB ACGTACTTCTATTATGAAGCATGATATTAA AGCAGATCTAATGTTATACTTAAAACTACG NED-CCAAGGCTCCTAAACG-MGB ATGTCTCTTCAAGGGATTCACCT AMAGYGTTATTTCTTGTTGCAATGATGA FAM-CATGCTATATTAAAAGAGTCTCARTAC-MGB AGGTGYGAAGAGYCTATTGAGCTA GGACACCCAAAGTAGTCGGTTC VIC-TCCGGCCCCTGAATG-MGB CAAGGGCTATAAAGGCGTTGCT ATGGTCGCAGACTTTGTTCCA NED-CCCCTTGCCAACAGA-MGB AAGACCAATCCTGTCACCTCTGA CAAAGCGTCTACGCTGCAGTCC FAM-TTTGTGTTCACGCTCACCGT-MGB AAATGTGATCGTGCTATGCCAA CTTAACATAATAGCAACCGCCACA VIC-CCTTGCGAATGAATG-MGB CGATGAGGCTATTCCGACTAGGT CCTTCCTGAGCCTTCAATATAGTAACC NED-CCTGGCACGGTACTC-MGB GGTGTGAAGAGCCSCRTGTGCT GGTGTGAAGACTCGCATGTGCT GGGTGYGAAGAGYCTANTGTGCT GCAGGGTTTRGGTTAGCCRCATT VIC-TCCGGCCCCTGAATG-MGB CAGTCAAATGGGCTGATGCA AAAGGGCTATAAAGAGAATAAGGTATTCT FAM-CCTGACGACCACGTTGT-MGB GCCACGGTGGGGTTTCTAAACTT GCCCCAGTGGTCTTACATGCACATC HEX-TGCACCAGACCCGGGCTCAGGTACTCCGA-BHQ1 GGAATCCCAATGCACAAGAACA GCTTTGGTCAACACATCAACCTT NED-CAAACCCAGCCTTCA-MGB CGGGCTCATATCATCAGGAAC ATCACTTGGTCTGAGGTCTTCGA FAM-CAATCAGCCACCTATC-MGB
Reference no. This study
9
Unpublisheda
7
This study
6 (modified)
7
7 (modified)
7 (modified)
8
This study
6 (modified)
7 (modified)
6 (modified)
5
7 (modified)
This study
F, forward primer; R, reverse primer, P, probe; NED, JOE, FAM, HEX and VIC are names of fluorophores; MGB, minor groove-binding modification. aDesigned by Lars P. Nielsen, Center for Geogenetik, Copenhagen, Denmark.
both methods. FilmArray detected bocavirus in one, rhinovirus in two and RSV in two more samples than did real-time PCR. On the other hand, real-time PCR detected CoV 229E in one and adenovirus in
two more samples than FilmArray. When the two discordant samples positive for RSV by FilmArray were re-analysed by real-time PCR, one was positive with a Ct value ⬎ 40.
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
900
M. E. Andersson et al.
At least one viral or bacterial pathogen was detected in 49 of the 53 samples that were analysed retrospectively by FilmArray because they had been reactive by real-time PCR. FilmArray detected bocavirus in one and metapneumovirus in one more sample than did real-time PCR, but did not detect IfA H1/09 in two, CoV HKU1 in one, bocavirus in one, IfB in one, Chlamydophila pneumoniae in one or adenovirus in seven samples in which these agents were detected by real-time PCR. Most cases with negative FilmArray results had Ct values above 37 by the in-house real-time PCR, indicating low pathogen concentration, except for adenovirus and IfA H1/09, where Ct values were 31.8–37.0 and 34.0–35.5, respectively. When the results for all the 128 samples were compared the agreement between the two assays was excellent, with kappa values ranging from 0.54 to 1.0, as shown in Table II. There was significant disagreement only for adenovirus detection. Of the 15 adenovirus-positive samples, only 6 (40%) were identified by FilmArray v 1.6. However, when these samples were retested by FilmArray v 1.7, adenovirus was detected in all of them. Comparison including adenovirus results by FilmArray v 1.7 showed a complete agreement in pathogen detection in 88.4% of all samples. Among the 79 samples with one detected pathogen (by either assay), this agent was detected by both assays in 92.4%. In 25 samples (20%) either of the assays identified two (n ⫽ 21) or three (n ⫽ 4) agents. In 17 of these samples both assays identified the same 2–3 agents; in the remaining 8, FilmArray detected 1 more agent in 6, and 1 less agent in 2 than did real-time PCR.
Discussion This study compared the performance of two assays for highly sensitive detection of agents causing acute respiratory infection. When 128 clinical samples were analysed, at least 1 agent was detected in 99 samples by any of the assays, and a total of 133 respiratory pathogens were identified. The agreement between the assays was very good, with identical results in 81.4% (by FilmArray v 1.6). If detection by either assay was considered true, then the sensitivity ranged between 71% and 100% (mean 96%) for both FilmArray v 1.6 and real-time PCR. The initial testing showed a poor performance for adenovirus by FilmArray (version 1.6), in agreement with previous evaluations [4,11–13]. When the false negative samples were retested by version 1.7 all adenovirus infections were identified, indicating that this version is accurate also for adenovirus detection. Version 1.7 also differs by not including bocavirus. Considering that the role of bocavirus for respiratory infections is debated [14,15], this limitation of the new version is of uncertain clinical importance. FilmArray did not detect two samples with IfA H1N1 pdm09. The real-time PCR Ct values for these samples were 34–35, indicating that the viral load was relatively low, but still at a level that should be detected. A suboptimal sensitivity for IfA H1N1 pdm09 was observed in a previous evaluation [11], indicating that future versions of FilmArray might need to be modified. Some agents were detected by FilmArray and not by the real-time PCR assay, including two rhino/ enterovirus, three RSV and one bocavirus. This could
Table II. Respiratory agents detected by FilmArray or real-time PCR.
Agent Parainfluenzavirus 1–3 Influenza A virus, H3N2 Influenza A virus, H1N1 Influenza B virus Coronavirus NL63 Coronavirus HKU1 Coronavirus 229E Coronavirus OC43 Bocavirus Respiratory syncytial virus Rhinovirus Enterovirus Metapneumovirus Mycoplasma pneumoniae Chlamydophila pneumoniae Adenovirus
Detected by either assay Detected by FilmArray 11 8 7 14 5 4 5 6 7 13 16 6 6 5 5 15
Psp, prospective; Rsp, retrospective. 1.6. b Version 1.78. a Version
11 (100%) 8 (100%) 5 (71%) 13 (93%) 5 (100%) 3 (75%) 4 (80%) 6 (100%) 6 (86%) 13 (100%) 16 (100%) 6 (100%) 6 (100%) 5 (100%) 4 (80%) 6 (40%)a; 15 (100%)b
Psp
Rsp
3 7 3 13 3 0 1 5 2 8 9 0 2 1 1 2
8 1 2 0 2 3 3 1 4 5 7 6 4 4 3 4
Detected by real-time PCR 11 8 7 14 5 4 5 6 5 11 14 6 5 5 5 15
(100%) (100%) (100%) (100%) (100%) (100%) (100%) (100%) (71%) (85%) (88%) (100%) (83%) (100%) (100%) (100%)
Ct value (range) 20.4–40.2 19.0–43.0 29.2–40.2 18.0–39.6 29.2–36.1 24.8–36.1 20.6–37.4 19.6–37.6 24.8–40.5 19.2–38.1 25.7–41.2 26.2–36.5 24.4–34.8 27.2–37.9 30.1–40.6 20.5–39.8
Psp Rsp Kappa value 3 7 3 13 3 0 2 5 1 6 7 0 2 1 1 4
8 1 4 1 2 4 3 1 4 5 7 6 3 4 4 11
1.00 1.00 0.83 0.96 1.00 0.85 0.89 1.00 0.74 0.91 0.93 1.00 0.91 1.00 0.89 0.54a; 1.00b
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
FilmArray and real-time PCR respiratory panels reflect that FilmArray is more sensitive for these agents. However, FilmArray detected them all as coinfections together with another agent that was detected also by real-time PCR with a low Ct value (high viral load). This observation suggests that some of these cases might reflect cross-reactivity, particularly as the package insert declares the specificities of RSV and rhino/enterovirus to be 89.1% and 94.6%, respectively. However, it is difficult to draw any conclusions from the disconcordant results as the number of samples was relatively low. Overall, this study shows that the performance of FilmArray is comparable to real-time PCR run in eight parallel reactions. This is in line with previous evaluations [12,16] of the FilmArray assay and confirms that it has a high sensitivity for all relevant agents causing acute respiratory infection. This, in combination with its user friendliness (hands-on time less than 5 min, no demand for molecular diagnostics expertise), the fully enclosed platform without requirement of special work area, and rapid delivery of result (≈ 1 h) makes FilmArray suitable for point-of-care testing [17]. The main disadvantage is the low throughput, as only a single sample can be processed on the instrument at one time. Thus, during the time 46 samples are processed with the in-house real-time PCR, 1 FilmArray system will process only 5 samples. On the other hand, results are obtained without delay and there is almost no cost for personnel. In many cases these advantages justify a higher cost for testing, because a rapid result may limit costs for the healthcare-related procedures and help to prevent the spread of respiratory infections.
Acknowledgment We thank Kirsten Kriz at Biotech-IgG A/S for supplying FilmArray assays. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References [1] Olofsson S, Brittain-Long R, Andersson L-M, Westin J, Lindh M. PCR for detection of respiratory viruses: seasonal variations of virus infections. Exp Rev Anti Infect Ther 2011;9:615–26. [2] Mahony JB. Detection of respiratory viruses by molecular methods. Clin Microbiol Rev 2008;21:716–47.
901
[3] Gadsby NJ, Hardie A, Claas ECJ, Templeton KE. Comparison of the Luminex Respiratory Virus Panel Fast Assay with in-house real-time PCR for respiratory viral infection diagnosis. J Clin Microbiol 2010;48:2213–16. [4] Popowitch EB, O’Neill SS, Miller MB. Comparison of the Biofire FilmArray RP, Genmark eSensor RVP, Luminex xTAG RVPv1, and Luminex xTAG RVP Fast Multiplex Assays for detection of respiratory viruses. J Clin Microbiol 2013;51:1528–33. [5] Heim A, Ebnet C, Harste G, Pring-Åkerblom P. Rapid and quantitative detection of human adenovirus DNA by realtime PCR. J Med Virol 2003;70:228–39. [6] Gunson RN, Collins TC, Carman WF. Real-time RT-PCR detection of 12 respiratory viral infections in four triplex reactions. J Clin Virol 2005;33:341–4. [7] Brittain-Long R, Nord S, Olofsson S, Westin J, Anderson L-M, Lindh M. Multiplex real-time PCR for detection of respiratory tract infections. J Clin Virol 2008; 41:53–6. [8] Ward CL, Dempsey MH, Ring CJA, Kempson RE, Zhang L, Gor D, et al. Design and performance testing of quantitative real time PCR assays for influenza A and B viral load measurement. J Clin Virol 2004;29:179–88. [9] Watzinger F, Suda M, Preuner S, Baumgartinger R, Ebner K, Baskova L, et al. Real-time quantitative PCR assays for detection and monitoring of pathogenic human viruses in immunosuppressed pediatric patients. J Clin Microbiol 2004; 42:5189–98. [10] Poritz MA, Blaschke AJ, Byington CL, Meyers L, Nilsson K, Jones DE, et al. FilmArray, an automated nested multiplex PCR system for multi-pathogen detection: development and application to respiratory tract infection. PLoS One 2011; 6:e26047. [11] Renaud C, Crowley J, Jerome KR, Kuypers J. Comparison of FilmArray Respiratory Panel and laboratory-developed real-time reverse transcription-polymerase chain reaction assays for respiratory virus detection. Diagn Microbiol Infect Dis 2012;74:379–83. [12] Pierce VM, Elkan M, Leet M, McGowan KL, Hodinka RL. Comparison of the Idaho Technology FilmArray system to real-time PCR for detection of respiratory pathogens in children. J Clin Microbiol 2012;50:364–71. [13] Couturier MR, Barney T, Alger G, Hymas WC, Stevenson JB, Hillyard D, et al. Evaluation of the FilmArray ®Respiratory Panel for clinical use in a large children’s hospital. J Clin Lab Anal 2013;27:148–54. [14] Don M, Soderlund-Venermo M, Hedman K, Ruuskanen O, Allander T, Korppi M. Don’t forget serum in the diagnosis of human bocavirus infection. J Infect Dis 2011;203: 1031–2. [15] Zhao B, Yu X, Wang C, Teng Z, Wang C, Shen J, et al. High human bocavirus viral load is associated with disease severity in children under five years of age. PLoS One 2013;8: e62318. [16] Loeffelholz MJ, Pong DL, Pyles RB, Xiong Y, Miller AL, Bufton KK, et al. Comparison of the FilmArray Respiratory Panel and Prodesse Real-Time PCR assays for detection of respiratory pathogens. J Clin Microbiol 2011;49: 4083–8. [17] Moore C. Point-of-care tests for infection control: should rapid testing be in the laboratory or at the front line? J Hosp Infect 2013;85:1–7.
SHORT COMMUNICATION
Comparison of the FilmArray assay and in-house real-time PCR for detection of respiratory infection
MARIA E. ANDERSSON, SIGVARD OLOFSSON & MAGNUS LINDH Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
Department of Infectious Diseases, University of Gothenburg, Gothenburg, Sweden
Abstract Recently, molecular methods capable of detecting almost all microbial agents that may cause acute respiratory infection have been introduced. The FilmArray Respiratory Panel assay, which integrates nucleic acid extraction, nested amplification and detection in a reaction pouch preloaded with all reagents required for detection of 17 viruses and 3 bacteria, was compared with an in-house real-time PCR that detects these agents in 8 parallel amplifications. When 128 clinical samples representing 18 of these agents were analysed by both assays the agreement was excellent, with kappa values ranging between 0.54 and 1.0. Discordances were mainly observed for adenovirus, but not when version 1.7 of FilmArray was used. The results show that these assays detect a wide range of pathogens with similar performance. FilmArray provides results after approximately 1 h, including ≈ 5 min hands-on time, and does not require advanced equipment or expertise in molecular diagnostics, making it a useful point-of-care-test for acute respiratory infections.
Keywords: Acute respiratory infection, real-time PCR, point-of care testing, FilmArray, molecular diagnostics
Introduction A wide range of viral and bacterial pathogens can cause acute respiratory infections. It is difficult to conclude or even guess the causative agent from symptoms because the clinical presentation is very variable, and microbial agents associated with common cold can also cause serious lower respiratory tract infections [1]. Therefore, simultaneous testing for many respiratory agents is important and can be performed by multiplex PCR assays that target many pathogens by using a mixture of many primers and probes [2]. Amplified products can then be distinguished, for example, as light emitted from beads that bind amplicons [3,4], but a disadvantage with this technique is that it requires post-PCR processing of amplified material, and therefore is time-consuming and poses a risk that amplified products may spread in the laboratory. These problems are avoided in real-time PCR where amplicons are detected in closed reaction wells during amplification. However, multiplexing is usually limited to three to five targets per reaction by this method, due to limitations of technical performance of the instrument and the risk
of interference when several primers and probes are mixed. To detect many pathogens by such real-time PCR each sample has to be run in several parallel reactions. A novel assay, FilmArray (BioFire Diagnostics, Salt Lake City, UT), has refined this technique to allow highly sensitive detection of a large number of targets. This assay integrates nucleic acid extraction, amplification and detection of amplicons in a reaction pouch that contains all necessary reagents, and thus, provides a point-of-care testing approach suitable for smaller laboratories and healthcare settings. In the present study we compared the performance of the FilmArray assay with a multiple real-time PCR that is run in eight parallel amplifications.
Materials and methods Specimens Clinical samples (63 nasopharyngeal swabs, 55 mixed materials containing nasopharyngeal swabs and throat swabs, 5 throat swabs and 5 broncho-alveolar
Correspondence: Maria E. Andersson, Department of Infectious Diseases, University of Gothenburg, Guldhedsgatan 10B, 413 36 Gothenburg, Sweden. Tel: ⫹ 46 709 771595. Fax: ⫹ 46 31 827032. E-mail: [email protected] (Received 19 March 2014 ; accepted 16 July 2014 ) ISSN 0036-5548 print/ISSN 1651-1980 online © 2014 Informa Healthcare DOI: 10.3109/00365548.2014.951681
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
898
M. E. Andersson et al.
lavages) sent for clinical testing to the Virological Diagnostic Laboratory at Sahlgrenska University Hospital were included in the study. Except for age (which ranged between 1 month and 96 years, median 40 years), no clinical information about the patients was registered. Because patient and sample identities were anonymised ethical approval was not required. In a prospective part of the study, 75 samples were analysed in parallel by a laboratory-developed multiplex real-time PCR assay and the FilmArray Respiratory Panel assay. All targeted viruses and bacteria did not appear or were rarely detected in these prospectively collected samples. Therefore, an additional 53 samples that had previously been analysed by the real-time PCR assay in clinical diagnostics (and were stored at –70°C) were included for retrospective analysis by FilmArray. These samples were selected so as to include all agents targeted by the FilmArray panel, aiming at a minimum of four samples per pathogen. Samples containing parainfluenza 4 and Bordetella pertussis were not included because parainfluenza 4 is not included in the in-house assay and because no stored samples with Bordetella pertussis were available. All samples were analysed by FilmArray Respiratory Panel kit version 1.6. Fifteen samples were retested with version 1.7 to evaluate whether adenovirus detection was improved.
Nucleic acid extraction and real-time PCR by the in-house assay Nucleic acid from 200 μl specimen was extracted by a MagNA Pure LC instrument (Roche Diagnostics, Mannheim, Germany) using the Total Nucleic Acid isolation kit. The nucleic acids were eluted in 100 μl volume, and 5 μl of this were used for real-time PCR. Real-time PCR was performed in an ABI 7900 384well system (Applied Biosystems, Foster City, CA, USA) in 8 parallel 20 μl reactions containing oligonucleotides described in Table I [5–9], and Taqman Fast Virus 1-step Mastermix (ABI, for RNA targets) or Universal Mastermix (ABI, for DNA targets). After a reverse transcription step at 46°C for 30 min followed by 10 min of denaturation at 95°C, 45 cycles of two-step PCR was performed (15 s at 95°C, 60 s at 58°C). For optimal performance the ramp rate was set to 80% during all 45 cycles. The performance for each multiplex reagent mixture had been evaluated using pUC57 plasmids with inserts of the targeted viral or bacterial sequences, synthesized by GenScript Corp. (Piscataway, NJ, USA). During this development, combinations were only accepted if the
Ct (threshold cycle) value in multiplex was not more than two cycles higher than when amplification was carried out in separate reactions. The same plasmids with target sequences were subsequently used in each run parallel with patient specimens to verify the performance of each target PCR (master mix control). In addition, one positive control was processed with each set of samples, from extraction of nucleic acids through the detection of amplified products. This in-house real-time PCR targets parainfluenzavirus (PIV) 1–3, respiratory syncytial virus (RSV), influenza A virus (IfA), influenza B virus (IfB), human coronaviruses (CoV NL63, HKU1, OC43, 229E), human enterovirus (EV), human rhinovirus (RV), adenovirus (AdV), bocavirus (BV), human metapneumovirus (hMpn), Mycoplasma pneumoniae, Chlamydophila pneumoniae and Bordetella pertussis. FilmArray respiratory panel The FilmArray analyses were performed according to the manufacturer’s instructions. First, 1 ml of the hydration solution was injected with a syringe to rehydrate biochemical reagents that are needed for the reactions. In the second injection port, 300 μl of a mixture including one part of the sample to be tested and two parts of denaturing sample buffer were added with a second syringe. The loaded pouch was then inserted into the FilmArray instrument, and the pouch and sample were registered using a bar code reader. All further processing was carried out in the instrument without further intervention. After mechanical lysis using zirconium beads and nucleic acid purification using magnetic beads, nested PCR was performed, including an initial multiplex step and a second step in which a small volume from the first step was further amplified in separate reaction mixtures containing a cyanine dye, yielding results within 65 min. The instrument interprets socalled melting curves and for each pathogen presents the result as detected or not detected. Melting curve data can be inspected using separate software. Two internal controls are included; an RNA control that follows the sample through the whole process and a DNA control that is present in the second-stage PCR [10].
Results Of the 75 prospectively analysed samples, any viral or bacterial pathogen was detected in 49 samples by real-time PCR, and in 51 by the FilmArray assay. Thus, 24 samples were negative for all agents by
FilmArray and real-time PCR respiratory panels
899
Table I. Primers and probes for in-house real-time PCR.
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
Pathogen
Mix
Oligo type
Oligo sequence
Parainfluenzavirus 1
1
Parainfluenzavirus 2
1
Parainfluenzavirus 3
1
Respiratory syncytial virus
2
Influenza B virus
2
Coronavirus NL63
2
Metapneumovirus
3
Enterovirus
3
Chlamydophila pneumoniae
3
Influenza A virus
4
Coronavirus HKU1
4
Coronavirus OC43
4
Rhinovirus
5
Coronavirus 229E
5
Adenovirus
6
Mycoplasma pneumoniae
7
Bocavirus
8
F R P F R P F R P F R P F R P F R P F R P F R P F R P F R P F R P F R P F F F R P F R P F R P F R P F R P
CAACAGGAAATCATGTTCTGTAATAGC TCACAGTGGGCAAGGAGCA NED-CTGGTAACCCCTTGTTCCTG-MGB GCATTTCCAATCTTCAGGACTATGA ACCTCCTGGTATAGCAGTGACTGAAC JOE-CCATTTACCTAAGTGATGGAATCAATCGCAAA-BHQ1 AAATGATCTGATTTATGCTTATACCTC TCAGGTACCAAGTCTGAGTTTACA FAM-CGAGGTTGYCAGGATATAGGAAAATCA-BHQ1 GCAAATATGGAAACATACGTGAACA GCACCCATATTGTWAGTGATGCA FAM-CTTCACGAAGGCTCCACATACACAGCWG-BHQ1 AAATACGGTGGATTAAATAAAAGCAA CCAGCAATAGCTCCGAAGAAA VIC-CCATAGGAAATTGCC-MGB ACGTACTTCTATTATGAAGCATGATATTAA AGCAGATCTAATGTTATACTTAAAACTACG NED-CCAAGGCTCCTAAACG-MGB ATGTCTCTTCAAGGGATTCACCT AMAGYGTTATTTCTTGTTGCAATGATGA FAM-CATGCTATATTAAAAGAGTCTCARTAC-MGB AGGTGYGAAGAGYCTATTGAGCTA GGACACCCAAAGTAGTCGGTTC VIC-TCCGGCCCCTGAATG-MGB CAAGGGCTATAAAGGCGTTGCT ATGGTCGCAGACTTTGTTCCA NED-CCCCTTGCCAACAGA-MGB AAGACCAATCCTGTCACCTCTGA CAAAGCGTCTACGCTGCAGTCC FAM-TTTGTGTTCACGCTCACCGT-MGB AAATGTGATCGTGCTATGCCAA CTTAACATAATAGCAACCGCCACA VIC-CCTTGCGAATGAATG-MGB CGATGAGGCTATTCCGACTAGGT CCTTCCTGAGCCTTCAATATAGTAACC NED-CCTGGCACGGTACTC-MGB GGTGTGAAGAGCCSCRTGTGCT GGTGTGAAGACTCGCATGTGCT GGGTGYGAAGAGYCTANTGTGCT GCAGGGTTTRGGTTAGCCRCATT VIC-TCCGGCCCCTGAATG-MGB CAGTCAAATGGGCTGATGCA AAAGGGCTATAAAGAGAATAAGGTATTCT FAM-CCTGACGACCACGTTGT-MGB GCCACGGTGGGGTTTCTAAACTT GCCCCAGTGGTCTTACATGCACATC HEX-TGCACCAGACCCGGGCTCAGGTACTCCGA-BHQ1 GGAATCCCAATGCACAAGAACA GCTTTGGTCAACACATCAACCTT NED-CAAACCCAGCCTTCA-MGB CGGGCTCATATCATCAGGAAC ATCACTTGGTCTGAGGTCTTCGA FAM-CAATCAGCCACCTATC-MGB
Reference no. This study
9
Unpublisheda
7
This study
6 (modified)
7
7 (modified)
7 (modified)
8
This study
6 (modified)
7 (modified)
6 (modified)
5
7 (modified)
This study
F, forward primer; R, reverse primer, P, probe; NED, JOE, FAM, HEX and VIC are names of fluorophores; MGB, minor groove-binding modification. aDesigned by Lars P. Nielsen, Center for Geogenetik, Copenhagen, Denmark.
both methods. FilmArray detected bocavirus in one, rhinovirus in two and RSV in two more samples than did real-time PCR. On the other hand, real-time PCR detected CoV 229E in one and adenovirus in
two more samples than FilmArray. When the two discordant samples positive for RSV by FilmArray were re-analysed by real-time PCR, one was positive with a Ct value ⬎ 40.
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
900
M. E. Andersson et al.
At least one viral or bacterial pathogen was detected in 49 of the 53 samples that were analysed retrospectively by FilmArray because they had been reactive by real-time PCR. FilmArray detected bocavirus in one and metapneumovirus in one more sample than did real-time PCR, but did not detect IfA H1/09 in two, CoV HKU1 in one, bocavirus in one, IfB in one, Chlamydophila pneumoniae in one or adenovirus in seven samples in which these agents were detected by real-time PCR. Most cases with negative FilmArray results had Ct values above 37 by the in-house real-time PCR, indicating low pathogen concentration, except for adenovirus and IfA H1/09, where Ct values were 31.8–37.0 and 34.0–35.5, respectively. When the results for all the 128 samples were compared the agreement between the two assays was excellent, with kappa values ranging from 0.54 to 1.0, as shown in Table II. There was significant disagreement only for adenovirus detection. Of the 15 adenovirus-positive samples, only 6 (40%) were identified by FilmArray v 1.6. However, when these samples were retested by FilmArray v 1.7, adenovirus was detected in all of them. Comparison including adenovirus results by FilmArray v 1.7 showed a complete agreement in pathogen detection in 88.4% of all samples. Among the 79 samples with one detected pathogen (by either assay), this agent was detected by both assays in 92.4%. In 25 samples (20%) either of the assays identified two (n ⫽ 21) or three (n ⫽ 4) agents. In 17 of these samples both assays identified the same 2–3 agents; in the remaining 8, FilmArray detected 1 more agent in 6, and 1 less agent in 2 than did real-time PCR.
Discussion This study compared the performance of two assays for highly sensitive detection of agents causing acute respiratory infection. When 128 clinical samples were analysed, at least 1 agent was detected in 99 samples by any of the assays, and a total of 133 respiratory pathogens were identified. The agreement between the assays was very good, with identical results in 81.4% (by FilmArray v 1.6). If detection by either assay was considered true, then the sensitivity ranged between 71% and 100% (mean 96%) for both FilmArray v 1.6 and real-time PCR. The initial testing showed a poor performance for adenovirus by FilmArray (version 1.6), in agreement with previous evaluations [4,11–13]. When the false negative samples were retested by version 1.7 all adenovirus infections were identified, indicating that this version is accurate also for adenovirus detection. Version 1.7 also differs by not including bocavirus. Considering that the role of bocavirus for respiratory infections is debated [14,15], this limitation of the new version is of uncertain clinical importance. FilmArray did not detect two samples with IfA H1N1 pdm09. The real-time PCR Ct values for these samples were 34–35, indicating that the viral load was relatively low, but still at a level that should be detected. A suboptimal sensitivity for IfA H1N1 pdm09 was observed in a previous evaluation [11], indicating that future versions of FilmArray might need to be modified. Some agents were detected by FilmArray and not by the real-time PCR assay, including two rhino/ enterovirus, three RSV and one bocavirus. This could
Table II. Respiratory agents detected by FilmArray or real-time PCR.
Agent Parainfluenzavirus 1–3 Influenza A virus, H3N2 Influenza A virus, H1N1 Influenza B virus Coronavirus NL63 Coronavirus HKU1 Coronavirus 229E Coronavirus OC43 Bocavirus Respiratory syncytial virus Rhinovirus Enterovirus Metapneumovirus Mycoplasma pneumoniae Chlamydophila pneumoniae Adenovirus
Detected by either assay Detected by FilmArray 11 8 7 14 5 4 5 6 7 13 16 6 6 5 5 15
Psp, prospective; Rsp, retrospective. 1.6. b Version 1.78. a Version
11 (100%) 8 (100%) 5 (71%) 13 (93%) 5 (100%) 3 (75%) 4 (80%) 6 (100%) 6 (86%) 13 (100%) 16 (100%) 6 (100%) 6 (100%) 5 (100%) 4 (80%) 6 (40%)a; 15 (100%)b
Psp
Rsp
3 7 3 13 3 0 1 5 2 8 9 0 2 1 1 2
8 1 2 0 2 3 3 1 4 5 7 6 4 4 3 4
Detected by real-time PCR 11 8 7 14 5 4 5 6 5 11 14 6 5 5 5 15
(100%) (100%) (100%) (100%) (100%) (100%) (100%) (100%) (71%) (85%) (88%) (100%) (83%) (100%) (100%) (100%)
Ct value (range) 20.4–40.2 19.0–43.0 29.2–40.2 18.0–39.6 29.2–36.1 24.8–36.1 20.6–37.4 19.6–37.6 24.8–40.5 19.2–38.1 25.7–41.2 26.2–36.5 24.4–34.8 27.2–37.9 30.1–40.6 20.5–39.8
Psp Rsp Kappa value 3 7 3 13 3 0 2 5 1 6 7 0 2 1 1 4
8 1 4 1 2 4 3 1 4 5 7 6 3 4 4 11
1.00 1.00 0.83 0.96 1.00 0.85 0.89 1.00 0.74 0.91 0.93 1.00 0.91 1.00 0.89 0.54a; 1.00b
Scand J Infect Dis Downloaded from informahealthcare.com by Fudan University on 05/12/15 For personal use only.
FilmArray and real-time PCR respiratory panels reflect that FilmArray is more sensitive for these agents. However, FilmArray detected them all as coinfections together with another agent that was detected also by real-time PCR with a low Ct value (high viral load). This observation suggests that some of these cases might reflect cross-reactivity, particularly as the package insert declares the specificities of RSV and rhino/enterovirus to be 89.1% and 94.6%, respectively. However, it is difficult to draw any conclusions from the disconcordant results as the number of samples was relatively low. Overall, this study shows that the performance of FilmArray is comparable to real-time PCR run in eight parallel reactions. This is in line with previous evaluations [12,16] of the FilmArray assay and confirms that it has a high sensitivity for all relevant agents causing acute respiratory infection. This, in combination with its user friendliness (hands-on time less than 5 min, no demand for molecular diagnostics expertise), the fully enclosed platform without requirement of special work area, and rapid delivery of result (≈ 1 h) makes FilmArray suitable for point-of-care testing [17]. The main disadvantage is the low throughput, as only a single sample can be processed on the instrument at one time. Thus, during the time 46 samples are processed with the in-house real-time PCR, 1 FilmArray system will process only 5 samples. On the other hand, results are obtained without delay and there is almost no cost for personnel. In many cases these advantages justify a higher cost for testing, because a rapid result may limit costs for the healthcare-related procedures and help to prevent the spread of respiratory infections.
Acknowledgment We thank Kirsten Kriz at Biotech-IgG A/S for supplying FilmArray assays. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References [1] Olofsson S, Brittain-Long R, Andersson L-M, Westin J, Lindh M. PCR for detection of respiratory viruses: seasonal variations of virus infections. Exp Rev Anti Infect Ther 2011;9:615–26. [2] Mahony JB. Detection of respiratory viruses by molecular methods. Clin Microbiol Rev 2008;21:716–47.
901
[3] Gadsby NJ, Hardie A, Claas ECJ, Templeton KE. Comparison of the Luminex Respiratory Virus Panel Fast Assay with in-house real-time PCR for respiratory viral infection diagnosis. J Clin Microbiol 2010;48:2213–16. [4] Popowitch EB, O’Neill SS, Miller MB. Comparison of the Biofire FilmArray RP, Genmark eSensor RVP, Luminex xTAG RVPv1, and Luminex xTAG RVP Fast Multiplex Assays for detection of respiratory viruses. J Clin Microbiol 2013;51:1528–33. [5] Heim A, Ebnet C, Harste G, Pring-Åkerblom P. Rapid and quantitative detection of human adenovirus DNA by realtime PCR. J Med Virol 2003;70:228–39. [6] Gunson RN, Collins TC, Carman WF. Real-time RT-PCR detection of 12 respiratory viral infections in four triplex reactions. J Clin Virol 2005;33:341–4. [7] Brittain-Long R, Nord S, Olofsson S, Westin J, Anderson L-M, Lindh M. Multiplex real-time PCR for detection of respiratory tract infections. J Clin Virol 2008; 41:53–6. [8] Ward CL, Dempsey MH, Ring CJA, Kempson RE, Zhang L, Gor D, et al. Design and performance testing of quantitative real time PCR assays for influenza A and B viral load measurement. J Clin Virol 2004;29:179–88. [9] Watzinger F, Suda M, Preuner S, Baumgartinger R, Ebner K, Baskova L, et al. Real-time quantitative PCR assays for detection and monitoring of pathogenic human viruses in immunosuppressed pediatric patients. J Clin Microbiol 2004; 42:5189–98. [10] Poritz MA, Blaschke AJ, Byington CL, Meyers L, Nilsson K, Jones DE, et al. FilmArray, an automated nested multiplex PCR system for multi-pathogen detection: development and application to respiratory tract infection. PLoS One 2011; 6:e26047. [11] Renaud C, Crowley J, Jerome KR, Kuypers J. Comparison of FilmArray Respiratory Panel and laboratory-developed real-time reverse transcription-polymerase chain reaction assays for respiratory virus detection. Diagn Microbiol Infect Dis 2012;74:379–83. [12] Pierce VM, Elkan M, Leet M, McGowan KL, Hodinka RL. Comparison of the Idaho Technology FilmArray system to real-time PCR for detection of respiratory pathogens in children. J Clin Microbiol 2012;50:364–71. [13] Couturier MR, Barney T, Alger G, Hymas WC, Stevenson JB, Hillyard D, et al. Evaluation of the FilmArray ®Respiratory Panel for clinical use in a large children’s hospital. J Clin Lab Anal 2013;27:148–54. [14] Don M, Soderlund-Venermo M, Hedman K, Ruuskanen O, Allander T, Korppi M. Don’t forget serum in the diagnosis of human bocavirus infection. J Infect Dis 2011;203: 1031–2. [15] Zhao B, Yu X, Wang C, Teng Z, Wang C, Shen J, et al. High human bocavirus viral load is associated with disease severity in children under five years of age. PLoS One 2013;8: e62318. [16] Loeffelholz MJ, Pong DL, Pyles RB, Xiong Y, Miller AL, Bufton KK, et al. Comparison of the FilmArray Respiratory Panel and Prodesse Real-Time PCR assays for detection of respiratory pathogens. J Clin Microbiol 2011;49: 4083–8. [17] Moore C. Point-of-care tests for infection control: should rapid testing be in the laboratory or at the front line? J Hosp Infect 2013;85:1–7.
