External Quality Assessment for Molecular Detectio
http://www.100md.com
微生物临床杂志 2005年第1期
Academisch Ziekenhuis Vrije Universiteit Brussel
ULB-Erasme, Brussels
Institut de Pathologie et de Genetique, Gerpinnes
C.H. Regional de la Citadelle, Liege, Belgium
GG&GD, Municipal Health Laboratory, Amsterdam
St. Elisabeth Hospital, Tilburg, The Netherlands
C.H. de Versailles, Le Chesnay
Institut de Bacteriologie, Strasbourg, France
Health Protection Agency, Respiratory and Systemic Laboratory, Specialist and Reference Microbiology Division, London, United Kingdom
University Hospital, rebro, Sweden
Klinikum Krefeld, Krefeld, Germany
ABSTRACT
Although the PCR for the detection of Bordetella pertussis is routinely performed in diagnostic laboratories, no quality assessment program has so far been described. We report on the results obtained with two external quality assessment proficiency panels sent to European laboratories. The first proficiency panel contained a series of dilutions of three previously characterized B. pertussis clinical isolates and two negative controls. No false-positive results were reported by six laboratories providing seven data sets. The reported limits of detection of the three B. pertussis strains varied between 4 and 4,000, 9 and 9,000, and 3 and 30,000 CFU/ml, respectively. The second proficiency panel, composed of a series of dilutions of reference strains of B. pertussis, B. holmesii, B. hinzii, and B. bronchiseptica, as well as negative controls, was sent to nine laboratories. One laboratory reported a negative result for a sample and reported a B. parapertussis-positive sample to be positive for B. pertussis. By using the B. pertussis-specific target gene pertactin, one laboratory detected B. pertussis with 100% specificity. All other laboratories, which used IS481-based assays, reported positive results for the samples containing B. holmesii and B. bronchiseptica, species that have occasionally been recovered from human respiratory samples. These data show that the choice of the target gene is particularly critical for the species specificity of B. pertussis PCR assays.
INTRODUCTION
Despite the introduction of large-scale vaccination for pertussis in many countries, whooping cough is still an endemic disease (5, 17). Several major and minor outbreaks have been described in countries with large-scale vaccination programs, such as Australia (7), The Netherlands (11, 30), the United States (3), Norway and Sweden (24), and Israel (20, 38).
Evidence suggests that pertussis cases might be grossly underreported (8, 9, 40). Moreover, since pertussis is thought to be an uncommon disease in industrialized countries and the clinical presentation resembles that of other illnesses associated with prolonged cough, health care providers often do not consider pertussis in the differential diagnosis (10).
Confirmation of the diagnosis of pertussis in the laboratory is challenging (16). Although culture of B. pertussis is still considered the "gold standard," it is not the ideal diagnostic tool. It has 100% specificity; but its sensitivity can vary greatly and is dependent on the stage of illness at the time of specimen collection, the technique used for specimen collection, specimen adequacy and transport, and culture conditions. Moreover, culture provides no rapid diagnosis: 7 to 10 days may be required to isolate and confirm or exclude the presence of B. pertussis.
During the last decade, the application of nucleic acid detection techniques has had a major impact on the diagnosis of pertussis. Numerous applications of extraction, amplification, and detection methods have been investigated (13, 14, 18, 25, 28, 37, 43, 45, 46). An improved sensitivity over that of culture was observed in all instances. Until now, no commercial assay has been available, and this has resulted in the development and introduction of a variety of methods developed in-house. With the improved and automated means of nucleic acid isolation, as well as the availability of real-time detection methods, a new generation of assays is being developed (1, 4, 6, 12, 21, 22, 39, 42). These technological developments have resulted in assays with very short turnaround times.
Although PCR assays for the detection of B. pertussis are routinely performed in diagnostic laboratories, no external quality assessment (EQA) program has so far been described. In order to define standards for tests that are clinically relevant, the Belgian Centres for Molecular Diagnostics (CMD) organized an EQA program. The Belgian public health service Rijksinstitut voor Ziekte- en Invalid:teitsverzekering, l’Institut National d’Assurance Maladie Invalidite requires that all members of CMD that perform molecular biology-based tests participate in this program. Laboratories from other European countries joined on a voluntary basis.
Here we describe the design, the panel assembly, and the analysis of the results obtained with two independent proficiency panels used for two blinded quality control studies distributed to six and nine participating European laboratories, respectively.
One panel was used for determination of the lower limit of detection of B. pertussis isolates, while the other panel was used for determination of both analytical sensitivity and analytical specificity.
MATERIALS AND METHODS
Participating laboratories. Four laboratories participated in both EQA panels, two participated in the first panel only, and five participated in the second panel only. Each laboratory followed its standard procedures for Bordetella PCR; thus, the procedures varied from laboratory to laboratory. The participating laboratories are as follows, in alphabetical order: Academisch Ziekenhuis Vrije Universitieit Brussel, Brussels, Belgium; C.H. de Versailles, Le Chesnay, France; C.H. Regional de la Citadelle, Liege, Belgium; GG&GD, Municipal Health Laboratory, Amsterdam, The Netherlands; Health Protection Agency, London, United Kingdom; Institut de Bacteriologie, Strasbourg, France; Institut de Pathologie et de Genetique, Gerpinnes, Belgium; Klinikum Krefeld, Krefeld, Germany; St. Elisabeth Hospital, Tilburg, The Netherlands; ULB-Erasme, Brussels, Belgium; and University Hospital, rebro, Sweden.
Preparation of proficiency panels. Proficiency panel 1 contained a series of dilutions of three B. pertussis clinical isolates collected in 2000 in the organizing laboratory. The isolates were typed previously (19) by analysis of the DNA sequence polymorphism of the pertussis toxin gene (ptxS1) and the pertactin gene (prn), as well as by analysis of chromosomal polymorphism by pulsed-field gel electrophoresis (PFGE) (Table 1). Five Bordetella species were selected for proficiency panel 2: B. pertussis (ATCC 9340), B. parapertussis (ATCC 15237), B. hinzii (LMG 1872), B. holmesii (LMG 15946), and B. bronchiseptica (ATCC 4617).
Physiologic saline solution was inoculated with cells from a single B. pertussis colony grown on Regan-Lowe charcoal agar (Charcoal Agar; Oxoid Ltd., Basingstoke, England) containing 10% horse blood and 40 mg of cephalexin (Bordetella Selective Supplement; Oxoid Ltd.) per liter. One to 10 serial dilutions were prepared in physiologic saline solution, the dilutions were aliquoted, and the number of CFU in each fraction was determined by plating 100 μl on the charcoal agar.
Similarly, the cells from a single colony each of B. hinzii, B. holmesii, and B. bronchiseptica grown on 3.7% tryptone soy agar (LabM, Lancashire, United Kingdom) containing 5% horse blood were used to spike 10 ml of physiologic saline solution. After the spiked samples were aliquoted, the number of CFU per milliliter was determined.
Panel 1 consisted of 20 frozen samples (Table 2): 6 samples of each of the 1/10 series of dilutions of strain Bord191 (range, 4 to 4 x 105 CFU/ml), strain Bord200 (range, 9 to 9 x 105 CFU/ml), and strain Bord201 (range, 3 to 3 x 105 CFU/ml) and two samples of physiologic saline solution without B. pertussis (as negative controls).
Panel 2 consisted of 15 frozen samples (Table 3): 6 samples of the dilution series of B. pertussis (range, 30 to 3 x 106 CFU/ml), 2 samples of the dilution series of B. parapertussis (2 x 106 and 2 x 104 CFU/ml), 1 sample of B. holmesii (>1 x 106 CFU/ml), 1 sample of B. hinzii (>1 x 106 CFU/ml), 1 sample of B. bronchiseptica (>1 x 106 CFU/ml), and 4 samples of physiologic saline solution (as negative controls).
In order to assess sample quality and homogeneity, three and four different technologists in the organizing laboratory tested the samples from panel 1 and panel 2, respectively, before distribution.
Distribution of EQA samples. The panels were distributed on dry ice by courier service. The participants were asked to report whether the samples were positive or negative for B. pertussis and whether another Bordetella species was detected. In order to obtain additional information on the procedures used, a questionnaire was sent to each participating laboratory.
RESULTS
Proficiency panel 1. The results obtained with the seven data sets from six laboratories are shown in Tables 2 and 4. All results were obtained by in-house methods. One laboratory (laboratory 4) reported two data sets, one with and one without DNA extraction, leading to different sensitivities. The results obtained by the first technologist in the organizing laboratory (laboratory 1 in Table 2) were also included in the data sets.
No false-positive results were reported (as determined with negative control samples B-A12 and B-A21). The detection limit for the three strains varied between 4 and 4,000, 9 and 9,000, and 3 and 30,000 CFU/ml, respectively. Laboratory 4, which reported two data sets (one with data obtained by use of MagnaPure DNA extraction and one obtained by use of no DNA extraction step), reported a higher sensitivity rate after extraction of the samples. However, one laboratory (laboratory 5) that used the QIAamp extraction procedure with a concentrated sample and another laboratory (laboratory 6) that used a boiling step with sample treatment with a reducing agent (N-acetylcysteine) reported results with very low analytical sensitivities.
There were considerable variations in the procedures used. Only one laboratory (laboratory 3) performed a nested PCR. The nested procedure was less sensitive than two procedures that used a single round of PCR (laboratories 1 and 4) but was more sensitive than two other single-round PCR assays (laboratories 5 and 6).
Of the six laboratories that reported the procedures that they used, five targeted the IS481 insertion element. The other laboratory used the pTx promoter as the target gene.
Proficiency panel 2. The results obtained with the nine data sets for proficiency panel 2 are shown in Table 3. All results were obtained by in-house methods. The results obtained by the technologist performing the first predistribution test in the organizing laboratory (laboratory 1) were also included in the data sets.
One laboratory (laboratory 2) reported that one of the four negative controls (sample B-B11) was positive for B. pertussis. This laboratory uses real-time PCR, three physically separate areas for the different steps of the procedure, and dUTP and uracil glycosylase as a measure for the prevention of contamination (Table 5).
All nine laboratories detected the samples spiked with B. pertussis (samples B-B2, B-B6, B-B13, B-B12, and B-B8) to a dilution of 300 CFU/ml. Four of the nine laboratories (laboratories 2, 7, 8, and 10) reported a positive result for the sample containing 30 CFU/ml (sample B-B3). These four laboratories did not perform the PCR with sample volumes higher than those used by the five laboratories reporting a negative result for this sample (for the first set of four laboratories, mean volume, 10 μl; median volume, 8.75 μl; and volume range, 5 to 20 μl; for the second set of five laboratories, mean volume, 10 μl; median volume, 12.5 μl; volume range, 1 to 20 μl), nor did the extraction protocol influence the detection limit. Nested PCR assays demonstrated no improved analytical sensitivities compared to those of the single-round PCR assays.
The four laboratories that used the IS1001 insertion element from B. parapertussis as the target detected an amplicon in both samples spiked with B. parapertussis (samples B-B9 and B-B14). Laboratory 2 reported the results for these samples as positive and weakly positive for B. pertussis, respectively.
None of the nine laboratories detected Bordetella spp. in the sample spiked with B. hinzii (sample B-B4). One laboratory (laboratory 11) that used the pertactin gene from B. pertussis as the target correctly reported the sample spiked with B. holmesii (sample B-B10) to be negative for B. pertussis. Use of a combination of the IS481 and ptxA promoter region as targets for the detection of B. pertussis allowed another laboratory (laboratory 9) to answer "B. pertussis or B. holmesii." The remaining seven laboratories reported this sample to be B. pertussis positive.
Laboratory 11 correctly reported the sample spiked with B. bronchiseptica (sample B-B15) to be negative for B. pertussis. Two of four technologists performing the predistribution test in the organizing laboratory (laboratory 1) as well as the remaining seven laboratories reported this sample to be B. pertussis positive. These unexpected false-positive reactions for the majority of the participating laboratories suggest the presence of a homologous sequence in the B. bronchiseptica strain used.
DISCUSSION
Numerous laboratories demonstrated that nucleic acid amplification techniques improved the sensitivity and turnaround time for the detection of B. pertussis and/or B. parapertussis and have demonstrated their clinical value for the diagnosis of pertussis (13, 14, 18, 25, 28, 37, 43, 45, 46). More recently, some laboratories introduced real-time PCR and have demonstrated the reliability of this technique as well as the shorter turnaround time (1, 4, 6, 12, 21, 22, 39, 42).
As molecular analysis is increasingly used for the diagnosis of whooping cough, standardization and EQA programs are recommended (23). The EQA program described here, in which representative laboratories from several European countries participated, is the first study on this scale to evaluate the molecular biology-based methods for the diagnosis of Bordetella spp.
The 11 laboratories participating in either one or two EQA schemes used a PCR method developed in-house, as no commercial assay is available. Evaluation of the various PCR protocols for which information was available showed no apparent association between their performance and the particular variables of the PCR method used. However, the results of such a comparison must be interpreted with caution due to the relatively small number of samples, the small number of participating laboratories, and the diversity of the methods used.
The variable analytical sensitivity observed suggests that false-negative PCR results with clinical specimens may be a problem.
The results with EQA panel 2 demonstrated a good analytical sensitivity for most participating laboratories, but the two laboratories with lower detection limits with samples from EQA panel 1 did not participate in this program. One of the two laboratories modified its procedures and uses real-time-based methods targeting the IS481 and IS1001 genes of B. pertussis and B. parapertussis.
Only one laboratory (laboratory 2) reported a false-positive result for a sample containing no Bordetella spp. The information provided on the questionnaire did not point to a particular risk for false-positive results (Table 5). Similar problems of specificity have been reported for the molecular biology-based assay detection of hepatitis B virus (44), Mycobacterium tuberculosis (31), Toxoplasma gondii (33), and hepatitis C virus (36) as well. The same laboratory reported the result for a sample spiked with B. parapertussis as B. pertussis positive, respectively.
By use of the B. pertussis-specific target gene pertactin, only one laboratory (laboratory 11) performed the tests with 100% specificity for B. pertussis. The eight other laboratories participating in EQA panel 2 reported positive results for B. pertussis for the samples spiked with B. holmesii and B. bronchiseptica. These unexpected false-positive reactions for B. pertussis by the majority of the participating laboratories suggest the presence of a homologous sequence in the strains of these two species used. The cross-reactivity with B. holmesii was expected, since the presence of IS481 in B. holmesii was described previously (34, 35) and these laboratories used this insertion element as the target gene. Laboratory 9 therefore performed a supplementary PCR specific for the toxin promoter region, which allowed that laboratory to report B. pertussis in the case of a positive result and B. pertussis or B. holmesii in the case of a negative result. A negative result by a PCR specific for the toxin promoter might be indicative of the presence of B. holmesii (sample B-B10) in the sample as well as a small amount of B. pertussis (30 CFU/ml in sample B-B8) due to the lower analytical sensitivity of this PCR compared to that of the IS481-based PCR.
Sequencing of the complete genome of representative strains of B. pertussis, B. parapertussis, and B. bronchiseptica demonstrated the presence of the IS481 target gene only in B. pertussis (32). However, Gladbach et al. (15) demonstrated that 2 of 12 B. bronchiseptica isolates tested were PCR positive for IS481. Furthermore, the sequence of a 200-bp amplicon from these isolates showed 97% homology with the B. pertussis IS481 sequence. Despite the absence of the IS481 sequence from B. bronchiseptica strain RB50, the complete sequence of which is available, a positive PCR result for the B. bronchiseptica-positive sample in our panel was observed by eight of the nine participating laboratories. This is very surprising, since it was previously demonstrated that the IS481-based assay performed by some of these laboratories showed no cross-reactivity with B. bronchiseptica (2, 22). It is therefore tempting to suggest that B. bronchiseptica might be rather genetically diverse, so that some isolates contain an IS481-like sequence, which makes these isolates more similar to B. pertussis than B. bronchiseptica strain RB50, whose genome was recently sequenced (32).
The presence of the IS481 sequence in B. holmesii (34, 35) and the isolation of B. holmesii from patients with pertussis-like symptoms (29, 49) or respiratory failure (41) were described previously. Likewise, B. bronchiseptica has been isolated from AIDS patients (27, 47) and cystic fibrosis patients (48). Although both B. holmesii and B. bronchiseptica have been implicated as infrequent causes of a pertussis-like syndrome and other respiratory illnesses (26), the clinical relevance of the presence of these rare microorganisms in human respiratory samples needs further investigation. The false-positive results for B. pertussis for samples containing B. holmesii and B. bronchiseptica strains in our EQA program suggest that the specificity and the positive predictive value of IS481-based PCR assays for the diagnosis of pertussis may be compromised. It could therefore be more appropriate to follow the suggestions of Fry et al. (13) and report the results of assays based only on IS481 sequences as evidence of the presence of Bordetella species. Otherwise, the development of an additional PCR that targets other B. pertussis-specific genes, such as the pertussis toxin promoter, the porin gene, the pertactin gene, or the adenylate cyclase gene, could be evaluated further. Although IS481 is present in multiple copies, whereas the other genes are not, assays targeting either may have comparable sensitivities, as shown by laboratory 11 in this study.
Finally, this study underlines the importance of proficiency panels for monitoring the quality of the PCR protocols performed in diagnostic laboratories. This study highlights the lack of homogeneity between PCR protocols and performance and underlines the need for an external quality assurance scheme which could provide reference samples that could be used by any laboratory wanting to establish and maintain an accurate diagnostic test based on PCR.
ACKNOWLEDGMENTS
We thank John Duncan, Anita Wyns, and Linda Godau for excellent technical assistance.
The organization of the EQA program for the Belgian CMD was supported by the public health service RIZIV-INAMI.
REFERENCES
Anderson, T. P., K. A. Beynon, and D. R. Murdoch. 2003. Comparison of real-time PCR and conventional hemi-nested PCR for the detection of Bordetella pertussis in nasopharyngeal samples. Clin. Microbiol. Infect. 9:746-749.
Backman, A., B. Johansson, and P. Olcen. 1994. Nested PCR optimized for detection of Bordetella pertussis in clinical nasopharyngeal samples. J. Clin. Microbiol. 32:2544-2548.
Bass, J. W., and R. R. Wittler. 1994. Return of epidemic pertussis in the United States. Pediatr. Infect. Dis. J. 13:343-345.
Chan, E. L., N. Antonishyn, R. McDonald, T. Maksymiw, P. Pieroni, E. Nagle, and G. B. Horsman. 2002. The use of TaqMan PCR assay for detection of Bordetella pertussis infection from clinical specimens. Arch. Pathol. Lab. Med. 126:173-176.
Cherry, J. D. 1999. Epidemiological, clinical, and laboratory aspects of pertussis in adults. Clin. Infect. Dis. 28(Suppl. 2):S112-S117.
Cloud, J. L., W. C. Hymas, A. Turlak, A. Croft, U. Reischl, J. A. Daly, and K. C. Carroll. 2003. Description of a multiplex Bordetella pertussis and Bordetella parapertussis LightCycler PCR assay with inhibition control. Diagn. Microbiol. Infect. Dis. 46:189-195.
Cordova, S. P., M. T. Gilles, and M. Y. Beers. 2000. The outbreak that had to happen: Bordetella pertussis in north-west Western Australia in 1999. Commun. Dis. Intell. 24:375-379.
Crowcroft, N. S., N. Andrews, C. Rooney, M. Brisson, and E. Miller. 2002. Deaths from pertussis are underestimated in England. Arch. Dis. Child. 86:336-338.
Crowcroft, N. S., R. Booy, T. Harrison, L. Spicer, J. Britto, Q. Mok, P. Heath, I. Murdoch, M. Zambon, R. George, and E. Miller. 2003. Severe and unrecognised: pertussis in UK infants. Arch. Dis. Child. 88:802-806.
Deeks, S., G. De Serres, N. Boulianne, B. Duval, L. Rochette, P. Dery, and S. Halperin. 1999. Failure of physicians to consider the diagnosis of pertussis in children. Clin. Infect. Dis. 28:840-846.
de Melker, H. E., J. F. Schellekens, S. E. Neppelenbroek, F. R. Mooi, H. C. Rumke, and M. A. Conyn-van Spaendonck. 2000. Reemergence of pertussis in the highly vaccinated population of The Netherlands: observations on surveillance data. Emerg. Infect. Dis. 6:348-357.
Farrell, D. J., M. McKeon, G. Daggard, M. J. Loeffelholz, C. J. Thompson, and T. K. Mukkur. 2000. Rapid-cycle PCR method to detect Bordetella pertussis that fulfills all consensus recommendations for use of PCR in diagnosis of pertussis. J. Clin. Microbiol. 38:4499-4502.
Fry, N. K., O. Tzivra, Y. T. Li, A. McNiff, N. Doshi, P. A. C. Maple, N. S. Crowcroft, E. Miller, R. C. George, and T. G. Harrison. 2004. Laboratory diagnosis of pertussis infections: the role of PCR and serology. J. Med. Microbiol. 53:1-7.
Furuya, D., A. Yagihashi, T. Endoh, N. Uehara, N. Fujii, S. Chiba, and N. Watanabe. 1999. Simultaneous amplification of Bordetella repeated insertion sequences and toxin promoter region gene by polymerase chain reaction. Immunopharmacol. Immunotoxicol. 21:55-63.
Gladbach, S., S. Hanauer, U. Reischl, K. Wilson, and G. N. Sanden. 2002. Identification of IS481 in Bordetella bronchseptica: implications for B. spp. phylogeny and diagnosis of B. pertussis infections by PCR assays. 7th Int. Symp. Pertussis: genome, pathogenesis, and immunity.
Hallander, H. O. 1999. Microbiological and serological diagnosis of pertussis. Clin. Infect. Dis. 28(Suppl. 2):S99-S106.
He, Q., and J. Mertsola. 1997. Epidemiology and prevention of pertussis. Curr. Opin. Pediatr. 9:14-18.
Heininger, U., G. Schmidt-Schlapfer, J. D. Cherry, and K. Stehr. 2000. Clinical validation of a polymerase chain reaction assay for the diagnosis of pertussis by comparison with serology, culture, and symptoms during a large pertussis vaccine efficacy trial. Pediatrics 105:E31.
Hoebrekx, N., G. Muyldermans, O. Soetens, I. De Schutter, A. Malfroot, D. Pierard, and S. Lauwers. 2001. Molecular typing of Bordetella pertussis isolates in Belgium from 1987 to 2000, p. 279. Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother. American Society for Microbiology, Washington, D.C.
Klement, E., L. Uliel, I. Engel, T. Hasin, M. Yavzori, N. Orr, N. Davidovitz, N. Lahat, I. Srugo, E. Zangvil, and D. Cohen. 2003. An outbreak of pertussis among young Israeli soldiers. Epidemiol. Infect. 131:1049-1054.
Kosters, K., M. Riffelmann, and C. H. Wirsing von Konig. 2001. Evaluation of a real-time PCR assay for detection of Bordetella pertussis and B. parapertussis in clinical samples. J. Med. Microbiol. 50:436-440.
Kosters, K., U. Reischl, J. Schmetz, M. Riffelmann, and C. H. Wirsing von Konig. 2002. Real-time LightCycler PCR for detection and discrimination of Bordetella pertussis and Bordetella parapertussis. J. Clin. Microbiol. 40:1719-1722.
Lievano, F. A., M. A. Reynolds, A. L. Waring, J. Ackelsberg, K. M. Bisgard, G. N. Sanden, D. Guris, A. Golaz, D. J. Bopp, R. J. Limberger, and P. F. Smith. 2002. Issues associated with and recommendations for using PCR to detect outbreaks of pertussis. J. Clin. Microbiol. 40:2801-2805.
Lindgren, C., J. Milerad, and H. Lagercrantz. 1997. Sudden infant death and prevalence of whooping cough in the Swedish and Norwegian communities. Eur. J. Pediatr. 156:405-409.
Loeffelholz, M. J., C. J. Thompson, K. S. Long, and M. J. Gilchrist. 1999. Comparison of PCR, culture, and direct fluorescent-antibody testing for detection of Bordetella pertussis. J. Clin. Microbiol. 37:2872-2876.
Loeffelholz, M. J. 2003. Bordetella, p. 780-789. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. American Society for Microbiology, Washington, D.C.
Lorenzo-Pajuelo, B., J. L. Villanueva, J. Rodriguez-Cuesta, N. Vergara-Irigaray, M. Bernabeu-Wittel, A. Garcia-Curiel, and G. Martinez de Tejada. 2002. Cavitary pneumonia in an AIDS patient caused by an unusual Bordetella bronchiseptica variant producing reduced amounts of pertactin and other major antigens. J. Clin. Microbiol. 40:3146-3154.
Matthews, R. C., N. Golbang, W. M. Bruck, D. Owen, A. Bailey, V. Weston, and J. R. Kerr. 1999. Semiquantitative polymerase chain reaction enzyme immunoassay for the diagnosis of pertussis. Eur. J. Clin. Microbiol. Infect. Dis. 18:748-750.
Mazengia, E., E. A. Silva, J. A. Peppe, R. Timperi, and H. George. 2000. Recovery of Bordetella holmesii from patients with pertussis-like symptoms: use of pulsed-field gel electrophoresis to characterize circulating strains. J. Clin. Microbiol. 38:2330-2333.
Mertens, P. L., F. S. Stals, J. F. Schellekens, A. W. Houben, and J. Huisman. 1999. An epidemic of pertussis among elderly people in a religious institution in The Netherlands. Eur. J. Clin. Microbiol. Infect. Dis. 18:242-247.
Noordhoek, G. T., A. H. J. Kolk, G. Bjune, D. Catty, J. W. Dale, P. E. M. Fine, P. Godfrey-Faussett, S.-N. Cho, T. Shinnick, S. B. Svenson, S. Wilson, and J. D. A. van Embden. 1994. Sensitivity and specificity of PCR for detection of Mycobacterium tuberculosis: a blind comparison study among seven laboratories. J. Clin. Microbiol. 32:277-284.
Parkhill, J., M. Sebaihia, A. Preston, L. D. Murphy, N. Thomson, D. E. Harris, M. T. Holden, C. M. Churcher, S. D. Bentley, K. L. Mungall, A. M. Cerdeno-Tarraga, L. Temple, K. James, B. Harris, M. A. Quail, M. Achtman, R. Atkin, S. Baker, D. Basham, N. Bason, I. Cherevach, T. Chillingworth, M. Collins, A. Cronin, P. Davis, J. Doggett, T. Feltwell, A. Goble, N. Hamlin, H. Hauser, S. Holroyd, K. Jagels, S. Leather, S. Moule, H. Norberczak, S. O'Neil, D. Ormond, C. Price, E. Rabbinowitsch, S. Rutter, M. Sanders, D. Saunders, K. Seeger, S. Sharp, M. Simmonds, J. Skelton, R. Squares, S. Squares, K. Stevens, L. Unwin, S. Whitehead, B. G. Barrell, and D. J. Maskell. 2003. Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat. Genet. 35:32-40.
Pelloux, H., E. Guy, M. C. Angelici, H. Aspock, M. H. Bessieres, R. Blatz, M. Del Pezzo, V. Girault, R. Gratzl, M. Holberg-Petersen, J. Johnson, D. Kruger, M. Lappalainen, A. Naessens, and M. Olsson. 1998. A second European collaborative study on polymerase chain reaction for Toxoplasma gondii, involving 15 teams. FEMS Microbiol. Lett. 165:231-237.
Poddar, S. K. 2003. Detection and discrimination of B. pertussis and B. holmesii by real-time PCR targeting IS481 using a beacon probe and probe-target melting analysis. Mol. Cell. Probes 17:91-98.
Reischl, U., N. Lehn, G. N. Sanden, and M. J. Loeffelholz. 2001. Real-time PCR assay targeting IS481 of Bordetella pertussis and molecular basis for detecting Bordetella holmesii. J. Clin. Microbiol. 39:1963-1966.
Schirm, J., A. M. van Loon, E. Valentine-Thon, P. E. Klapper, J. Reid, and G. M. Cleator. 2002. External quality assessment program for qualitative and quantitative detection of hepatitis C virus RNA in diagnostic virology. J. Clin. Microbiol. 40:2973-2980.
Schmidt-Schlapfer, G., J. G. Liese, F. Porter, S. Stojanov, M. Just, and B. H. Belohradsky. 1997. Polymerase chain reaction (PCR) compared with conventional identification in culture for detection of Bordetella pertussis in 7153 children. Clin. Microbiol. Infect. 3:462-467.
Shvartzman, P., T. Swartz, T. Stopler, and J. Herman. 1991. Outbreak of pertussis in a closed population with a high vaccination rate. Isr. J. Med. Sci. 27:137-140.
Sloan, L. M., M. K. Hopkins, P. S. Mitchell, E. A. Vetter, J. E. Rosenblatt, W. S. Harmsen, F. R. Cockerill, and R. Patel. 2002. Multiplex LightCycler PCR assay for detection and differentiation of Bordetella pertussis and Bordetella parapertussis in nasopharyngeal specimens. J. Clin. Microbiol. 40:96-100.
Sutter, R. W., and S. L. Cochi. 1992. Pertussis hospitalizations and mortality in the United States, 1985-1988. Evaluation of the completeness of national reporting. JAMA 267:386-391.
Tang, Y. W., M. K. Hopkins, C. P. Kolbert, P. A. Hartley, P. J. Severance, and D. H. Persing. 1998. Bordetella holmesii-like organisms associated with septicemia, endocarditis, and respiratory failure. Clin. Infect. Dis. 26:389-392.
Templeton, K. E., S. A. Scheltinga, A. Van Der Zee, B.M. Diederen, A. M. Kruijssen, H. Goossens, E. Kuijper, and E. C. Claas. 2003. Evaluation of real-time PCR for detection of and discrimination between Bordetella pertussis, Bordetella parapertussis, and Bordetella holmesii for clinical diagnosis. J. Clin. Microbiol. 41:4121-4126.
Tilley, P. A., M. V. Kanchana, I. Knight, J. Blondeau, N. Antonishyn, and H. Deneer. 2000. Detection of Bordetella pertussis in a clinical laboratory by culture, polymerase chain reaction, and direct fluorescent antibody staining; accuracy, and cost. Diagn. Microbiol. Infect. Dis. 37:17-23.
Valentine-Thon, E., A. M. van Loon, J. Schirm, J. Reid, P. E. Klapper, and G. M. Cleator. 2001. A European proficiency testing program for molecular detection and quantitation of hepatitis B virus DNA. J. Clin. Microbiol. 39:4407-4412.
Van der Zee, A., C. Agterberg, M. Peeters, J. Schellekens, and F. R. Mooi. 1993. Polymerase chain reaction assay for pertussis: simultaneous detection and discrimination of Bordetella pertussis and Bordetella parapertussis. J. Clin. Microbiol. 31:2134-2140.
Van der Zee, A., C. Agterberg, M. Peeters, F. Mooi, and J. Schellekens. 1996. A clinical validation of Bordetella pertussis and Bordetella parapertussis polymerase chain reaction: comparison with culture and serology using samples from patients with suspected whooping cough from a highly immunized population. J. Infect. Dis. 174:89-96.
Viejo, G., P. de la Iglesia, L. Otero, M. I. Blanco, B. Gomez, D. De Miguel, A. Del Valle, and B. De la Fuente. 2002. Bordetella bronchiseptica pleural infection in a patient with AIDS. Scand. J. Infect. Dis. 34:628-629.
Wallet, F., T. Perez, S. Armand, B. Wallaert, and R. J. Courcol. 2002. Pneumonia due to Bordetella bronchiseptica in a cystic fibrosis patient: 16S rRNA sequencing for diagnosis confirmation. J. Clin. Microbiol. 40:2300-2301.
Yih, W. K., E. A. Silva, J. Ida, N. Harrington, S. M. Lett, and H. George. 1999. Bordetella holmesii-like organisms isolated from Massachusetts patients with pertussis-like symptoms. Emerg. Infect. Dis. 5:441-443.(G. Muyldermans, O. Soeten)
ULB-Erasme, Brussels
Institut de Pathologie et de Genetique, Gerpinnes
C.H. Regional de la Citadelle, Liege, Belgium
GG&GD, Municipal Health Laboratory, Amsterdam
St. Elisabeth Hospital, Tilburg, The Netherlands
C.H. de Versailles, Le Chesnay
Institut de Bacteriologie, Strasbourg, France
Health Protection Agency, Respiratory and Systemic Laboratory, Specialist and Reference Microbiology Division, London, United Kingdom
University Hospital, rebro, Sweden
Klinikum Krefeld, Krefeld, Germany
ABSTRACT
Although the PCR for the detection of Bordetella pertussis is routinely performed in diagnostic laboratories, no quality assessment program has so far been described. We report on the results obtained with two external quality assessment proficiency panels sent to European laboratories. The first proficiency panel contained a series of dilutions of three previously characterized B. pertussis clinical isolates and two negative controls. No false-positive results were reported by six laboratories providing seven data sets. The reported limits of detection of the three B. pertussis strains varied between 4 and 4,000, 9 and 9,000, and 3 and 30,000 CFU/ml, respectively. The second proficiency panel, composed of a series of dilutions of reference strains of B. pertussis, B. holmesii, B. hinzii, and B. bronchiseptica, as well as negative controls, was sent to nine laboratories. One laboratory reported a negative result for a sample and reported a B. parapertussis-positive sample to be positive for B. pertussis. By using the B. pertussis-specific target gene pertactin, one laboratory detected B. pertussis with 100% specificity. All other laboratories, which used IS481-based assays, reported positive results for the samples containing B. holmesii and B. bronchiseptica, species that have occasionally been recovered from human respiratory samples. These data show that the choice of the target gene is particularly critical for the species specificity of B. pertussis PCR assays.
INTRODUCTION
Despite the introduction of large-scale vaccination for pertussis in many countries, whooping cough is still an endemic disease (5, 17). Several major and minor outbreaks have been described in countries with large-scale vaccination programs, such as Australia (7), The Netherlands (11, 30), the United States (3), Norway and Sweden (24), and Israel (20, 38).
Evidence suggests that pertussis cases might be grossly underreported (8, 9, 40). Moreover, since pertussis is thought to be an uncommon disease in industrialized countries and the clinical presentation resembles that of other illnesses associated with prolonged cough, health care providers often do not consider pertussis in the differential diagnosis (10).
Confirmation of the diagnosis of pertussis in the laboratory is challenging (16). Although culture of B. pertussis is still considered the "gold standard," it is not the ideal diagnostic tool. It has 100% specificity; but its sensitivity can vary greatly and is dependent on the stage of illness at the time of specimen collection, the technique used for specimen collection, specimen adequacy and transport, and culture conditions. Moreover, culture provides no rapid diagnosis: 7 to 10 days may be required to isolate and confirm or exclude the presence of B. pertussis.
During the last decade, the application of nucleic acid detection techniques has had a major impact on the diagnosis of pertussis. Numerous applications of extraction, amplification, and detection methods have been investigated (13, 14, 18, 25, 28, 37, 43, 45, 46). An improved sensitivity over that of culture was observed in all instances. Until now, no commercial assay has been available, and this has resulted in the development and introduction of a variety of methods developed in-house. With the improved and automated means of nucleic acid isolation, as well as the availability of real-time detection methods, a new generation of assays is being developed (1, 4, 6, 12, 21, 22, 39, 42). These technological developments have resulted in assays with very short turnaround times.
Although PCR assays for the detection of B. pertussis are routinely performed in diagnostic laboratories, no external quality assessment (EQA) program has so far been described. In order to define standards for tests that are clinically relevant, the Belgian Centres for Molecular Diagnostics (CMD) organized an EQA program. The Belgian public health service Rijksinstitut voor Ziekte- en Invalid:teitsverzekering, l’Institut National d’Assurance Maladie Invalidite requires that all members of CMD that perform molecular biology-based tests participate in this program. Laboratories from other European countries joined on a voluntary basis.
Here we describe the design, the panel assembly, and the analysis of the results obtained with two independent proficiency panels used for two blinded quality control studies distributed to six and nine participating European laboratories, respectively.
One panel was used for determination of the lower limit of detection of B. pertussis isolates, while the other panel was used for determination of both analytical sensitivity and analytical specificity.
MATERIALS AND METHODS
Participating laboratories. Four laboratories participated in both EQA panels, two participated in the first panel only, and five participated in the second panel only. Each laboratory followed its standard procedures for Bordetella PCR; thus, the procedures varied from laboratory to laboratory. The participating laboratories are as follows, in alphabetical order: Academisch Ziekenhuis Vrije Universitieit Brussel, Brussels, Belgium; C.H. de Versailles, Le Chesnay, France; C.H. Regional de la Citadelle, Liege, Belgium; GG&GD, Municipal Health Laboratory, Amsterdam, The Netherlands; Health Protection Agency, London, United Kingdom; Institut de Bacteriologie, Strasbourg, France; Institut de Pathologie et de Genetique, Gerpinnes, Belgium; Klinikum Krefeld, Krefeld, Germany; St. Elisabeth Hospital, Tilburg, The Netherlands; ULB-Erasme, Brussels, Belgium; and University Hospital, rebro, Sweden.
Preparation of proficiency panels. Proficiency panel 1 contained a series of dilutions of three B. pertussis clinical isolates collected in 2000 in the organizing laboratory. The isolates were typed previously (19) by analysis of the DNA sequence polymorphism of the pertussis toxin gene (ptxS1) and the pertactin gene (prn), as well as by analysis of chromosomal polymorphism by pulsed-field gel electrophoresis (PFGE) (Table 1). Five Bordetella species were selected for proficiency panel 2: B. pertussis (ATCC 9340), B. parapertussis (ATCC 15237), B. hinzii (LMG 1872), B. holmesii (LMG 15946), and B. bronchiseptica (ATCC 4617).
Physiologic saline solution was inoculated with cells from a single B. pertussis colony grown on Regan-Lowe charcoal agar (Charcoal Agar; Oxoid Ltd., Basingstoke, England) containing 10% horse blood and 40 mg of cephalexin (Bordetella Selective Supplement; Oxoid Ltd.) per liter. One to 10 serial dilutions were prepared in physiologic saline solution, the dilutions were aliquoted, and the number of CFU in each fraction was determined by plating 100 μl on the charcoal agar.
Similarly, the cells from a single colony each of B. hinzii, B. holmesii, and B. bronchiseptica grown on 3.7% tryptone soy agar (LabM, Lancashire, United Kingdom) containing 5% horse blood were used to spike 10 ml of physiologic saline solution. After the spiked samples were aliquoted, the number of CFU per milliliter was determined.
Panel 1 consisted of 20 frozen samples (Table 2): 6 samples of each of the 1/10 series of dilutions of strain Bord191 (range, 4 to 4 x 105 CFU/ml), strain Bord200 (range, 9 to 9 x 105 CFU/ml), and strain Bord201 (range, 3 to 3 x 105 CFU/ml) and two samples of physiologic saline solution without B. pertussis (as negative controls).
Panel 2 consisted of 15 frozen samples (Table 3): 6 samples of the dilution series of B. pertussis (range, 30 to 3 x 106 CFU/ml), 2 samples of the dilution series of B. parapertussis (2 x 106 and 2 x 104 CFU/ml), 1 sample of B. holmesii (>1 x 106 CFU/ml), 1 sample of B. hinzii (>1 x 106 CFU/ml), 1 sample of B. bronchiseptica (>1 x 106 CFU/ml), and 4 samples of physiologic saline solution (as negative controls).
In order to assess sample quality and homogeneity, three and four different technologists in the organizing laboratory tested the samples from panel 1 and panel 2, respectively, before distribution.
Distribution of EQA samples. The panels were distributed on dry ice by courier service. The participants were asked to report whether the samples were positive or negative for B. pertussis and whether another Bordetella species was detected. In order to obtain additional information on the procedures used, a questionnaire was sent to each participating laboratory.
RESULTS
Proficiency panel 1. The results obtained with the seven data sets from six laboratories are shown in Tables 2 and 4. All results were obtained by in-house methods. One laboratory (laboratory 4) reported two data sets, one with and one without DNA extraction, leading to different sensitivities. The results obtained by the first technologist in the organizing laboratory (laboratory 1 in Table 2) were also included in the data sets.
No false-positive results were reported (as determined with negative control samples B-A12 and B-A21). The detection limit for the three strains varied between 4 and 4,000, 9 and 9,000, and 3 and 30,000 CFU/ml, respectively. Laboratory 4, which reported two data sets (one with data obtained by use of MagnaPure DNA extraction and one obtained by use of no DNA extraction step), reported a higher sensitivity rate after extraction of the samples. However, one laboratory (laboratory 5) that used the QIAamp extraction procedure with a concentrated sample and another laboratory (laboratory 6) that used a boiling step with sample treatment with a reducing agent (N-acetylcysteine) reported results with very low analytical sensitivities.
There were considerable variations in the procedures used. Only one laboratory (laboratory 3) performed a nested PCR. The nested procedure was less sensitive than two procedures that used a single round of PCR (laboratories 1 and 4) but was more sensitive than two other single-round PCR assays (laboratories 5 and 6).
Of the six laboratories that reported the procedures that they used, five targeted the IS481 insertion element. The other laboratory used the pTx promoter as the target gene.
Proficiency panel 2. The results obtained with the nine data sets for proficiency panel 2 are shown in Table 3. All results were obtained by in-house methods. The results obtained by the technologist performing the first predistribution test in the organizing laboratory (laboratory 1) were also included in the data sets.
One laboratory (laboratory 2) reported that one of the four negative controls (sample B-B11) was positive for B. pertussis. This laboratory uses real-time PCR, three physically separate areas for the different steps of the procedure, and dUTP and uracil glycosylase as a measure for the prevention of contamination (Table 5).
All nine laboratories detected the samples spiked with B. pertussis (samples B-B2, B-B6, B-B13, B-B12, and B-B8) to a dilution of 300 CFU/ml. Four of the nine laboratories (laboratories 2, 7, 8, and 10) reported a positive result for the sample containing 30 CFU/ml (sample B-B3). These four laboratories did not perform the PCR with sample volumes higher than those used by the five laboratories reporting a negative result for this sample (for the first set of four laboratories, mean volume, 10 μl; median volume, 8.75 μl; and volume range, 5 to 20 μl; for the second set of five laboratories, mean volume, 10 μl; median volume, 12.5 μl; volume range, 1 to 20 μl), nor did the extraction protocol influence the detection limit. Nested PCR assays demonstrated no improved analytical sensitivities compared to those of the single-round PCR assays.
The four laboratories that used the IS1001 insertion element from B. parapertussis as the target detected an amplicon in both samples spiked with B. parapertussis (samples B-B9 and B-B14). Laboratory 2 reported the results for these samples as positive and weakly positive for B. pertussis, respectively.
None of the nine laboratories detected Bordetella spp. in the sample spiked with B. hinzii (sample B-B4). One laboratory (laboratory 11) that used the pertactin gene from B. pertussis as the target correctly reported the sample spiked with B. holmesii (sample B-B10) to be negative for B. pertussis. Use of a combination of the IS481 and ptxA promoter region as targets for the detection of B. pertussis allowed another laboratory (laboratory 9) to answer "B. pertussis or B. holmesii." The remaining seven laboratories reported this sample to be B. pertussis positive.
Laboratory 11 correctly reported the sample spiked with B. bronchiseptica (sample B-B15) to be negative for B. pertussis. Two of four technologists performing the predistribution test in the organizing laboratory (laboratory 1) as well as the remaining seven laboratories reported this sample to be B. pertussis positive. These unexpected false-positive reactions for the majority of the participating laboratories suggest the presence of a homologous sequence in the B. bronchiseptica strain used.
DISCUSSION
Numerous laboratories demonstrated that nucleic acid amplification techniques improved the sensitivity and turnaround time for the detection of B. pertussis and/or B. parapertussis and have demonstrated their clinical value for the diagnosis of pertussis (13, 14, 18, 25, 28, 37, 43, 45, 46). More recently, some laboratories introduced real-time PCR and have demonstrated the reliability of this technique as well as the shorter turnaround time (1, 4, 6, 12, 21, 22, 39, 42).
As molecular analysis is increasingly used for the diagnosis of whooping cough, standardization and EQA programs are recommended (23). The EQA program described here, in which representative laboratories from several European countries participated, is the first study on this scale to evaluate the molecular biology-based methods for the diagnosis of Bordetella spp.
The 11 laboratories participating in either one or two EQA schemes used a PCR method developed in-house, as no commercial assay is available. Evaluation of the various PCR protocols for which information was available showed no apparent association between their performance and the particular variables of the PCR method used. However, the results of such a comparison must be interpreted with caution due to the relatively small number of samples, the small number of participating laboratories, and the diversity of the methods used.
The variable analytical sensitivity observed suggests that false-negative PCR results with clinical specimens may be a problem.
The results with EQA panel 2 demonstrated a good analytical sensitivity for most participating laboratories, but the two laboratories with lower detection limits with samples from EQA panel 1 did not participate in this program. One of the two laboratories modified its procedures and uses real-time-based methods targeting the IS481 and IS1001 genes of B. pertussis and B. parapertussis.
Only one laboratory (laboratory 2) reported a false-positive result for a sample containing no Bordetella spp. The information provided on the questionnaire did not point to a particular risk for false-positive results (Table 5). Similar problems of specificity have been reported for the molecular biology-based assay detection of hepatitis B virus (44), Mycobacterium tuberculosis (31), Toxoplasma gondii (33), and hepatitis C virus (36) as well. The same laboratory reported the result for a sample spiked with B. parapertussis as B. pertussis positive, respectively.
By use of the B. pertussis-specific target gene pertactin, only one laboratory (laboratory 11) performed the tests with 100% specificity for B. pertussis. The eight other laboratories participating in EQA panel 2 reported positive results for B. pertussis for the samples spiked with B. holmesii and B. bronchiseptica. These unexpected false-positive reactions for B. pertussis by the majority of the participating laboratories suggest the presence of a homologous sequence in the strains of these two species used. The cross-reactivity with B. holmesii was expected, since the presence of IS481 in B. holmesii was described previously (34, 35) and these laboratories used this insertion element as the target gene. Laboratory 9 therefore performed a supplementary PCR specific for the toxin promoter region, which allowed that laboratory to report B. pertussis in the case of a positive result and B. pertussis or B. holmesii in the case of a negative result. A negative result by a PCR specific for the toxin promoter might be indicative of the presence of B. holmesii (sample B-B10) in the sample as well as a small amount of B. pertussis (30 CFU/ml in sample B-B8) due to the lower analytical sensitivity of this PCR compared to that of the IS481-based PCR.
Sequencing of the complete genome of representative strains of B. pertussis, B. parapertussis, and B. bronchiseptica demonstrated the presence of the IS481 target gene only in B. pertussis (32). However, Gladbach et al. (15) demonstrated that 2 of 12 B. bronchiseptica isolates tested were PCR positive for IS481. Furthermore, the sequence of a 200-bp amplicon from these isolates showed 97% homology with the B. pertussis IS481 sequence. Despite the absence of the IS481 sequence from B. bronchiseptica strain RB50, the complete sequence of which is available, a positive PCR result for the B. bronchiseptica-positive sample in our panel was observed by eight of the nine participating laboratories. This is very surprising, since it was previously demonstrated that the IS481-based assay performed by some of these laboratories showed no cross-reactivity with B. bronchiseptica (2, 22). It is therefore tempting to suggest that B. bronchiseptica might be rather genetically diverse, so that some isolates contain an IS481-like sequence, which makes these isolates more similar to B. pertussis than B. bronchiseptica strain RB50, whose genome was recently sequenced (32).
The presence of the IS481 sequence in B. holmesii (34, 35) and the isolation of B. holmesii from patients with pertussis-like symptoms (29, 49) or respiratory failure (41) were described previously. Likewise, B. bronchiseptica has been isolated from AIDS patients (27, 47) and cystic fibrosis patients (48). Although both B. holmesii and B. bronchiseptica have been implicated as infrequent causes of a pertussis-like syndrome and other respiratory illnesses (26), the clinical relevance of the presence of these rare microorganisms in human respiratory samples needs further investigation. The false-positive results for B. pertussis for samples containing B. holmesii and B. bronchiseptica strains in our EQA program suggest that the specificity and the positive predictive value of IS481-based PCR assays for the diagnosis of pertussis may be compromised. It could therefore be more appropriate to follow the suggestions of Fry et al. (13) and report the results of assays based only on IS481 sequences as evidence of the presence of Bordetella species. Otherwise, the development of an additional PCR that targets other B. pertussis-specific genes, such as the pertussis toxin promoter, the porin gene, the pertactin gene, or the adenylate cyclase gene, could be evaluated further. Although IS481 is present in multiple copies, whereas the other genes are not, assays targeting either may have comparable sensitivities, as shown by laboratory 11 in this study.
Finally, this study underlines the importance of proficiency panels for monitoring the quality of the PCR protocols performed in diagnostic laboratories. This study highlights the lack of homogeneity between PCR protocols and performance and underlines the need for an external quality assurance scheme which could provide reference samples that could be used by any laboratory wanting to establish and maintain an accurate diagnostic test based on PCR.
ACKNOWLEDGMENTS
We thank John Duncan, Anita Wyns, and Linda Godau for excellent technical assistance.
The organization of the EQA program for the Belgian CMD was supported by the public health service RIZIV-INAMI.
REFERENCES
Anderson, T. P., K. A. Beynon, and D. R. Murdoch. 2003. Comparison of real-time PCR and conventional hemi-nested PCR for the detection of Bordetella pertussis in nasopharyngeal samples. Clin. Microbiol. Infect. 9:746-749.
Backman, A., B. Johansson, and P. Olcen. 1994. Nested PCR optimized for detection of Bordetella pertussis in clinical nasopharyngeal samples. J. Clin. Microbiol. 32:2544-2548.
Bass, J. W., and R. R. Wittler. 1994. Return of epidemic pertussis in the United States. Pediatr. Infect. Dis. J. 13:343-345.
Chan, E. L., N. Antonishyn, R. McDonald, T. Maksymiw, P. Pieroni, E. Nagle, and G. B. Horsman. 2002. The use of TaqMan PCR assay for detection of Bordetella pertussis infection from clinical specimens. Arch. Pathol. Lab. Med. 126:173-176.
Cherry, J. D. 1999. Epidemiological, clinical, and laboratory aspects of pertussis in adults. Clin. Infect. Dis. 28(Suppl. 2):S112-S117.
Cloud, J. L., W. C. Hymas, A. Turlak, A. Croft, U. Reischl, J. A. Daly, and K. C. Carroll. 2003. Description of a multiplex Bordetella pertussis and Bordetella parapertussis LightCycler PCR assay with inhibition control. Diagn. Microbiol. Infect. Dis. 46:189-195.
Cordova, S. P., M. T. Gilles, and M. Y. Beers. 2000. The outbreak that had to happen: Bordetella pertussis in north-west Western Australia in 1999. Commun. Dis. Intell. 24:375-379.
Crowcroft, N. S., N. Andrews, C. Rooney, M. Brisson, and E. Miller. 2002. Deaths from pertussis are underestimated in England. Arch. Dis. Child. 86:336-338.
Crowcroft, N. S., R. Booy, T. Harrison, L. Spicer, J. Britto, Q. Mok, P. Heath, I. Murdoch, M. Zambon, R. George, and E. Miller. 2003. Severe and unrecognised: pertussis in UK infants. Arch. Dis. Child. 88:802-806.
Deeks, S., G. De Serres, N. Boulianne, B. Duval, L. Rochette, P. Dery, and S. Halperin. 1999. Failure of physicians to consider the diagnosis of pertussis in children. Clin. Infect. Dis. 28:840-846.
de Melker, H. E., J. F. Schellekens, S. E. Neppelenbroek, F. R. Mooi, H. C. Rumke, and M. A. Conyn-van Spaendonck. 2000. Reemergence of pertussis in the highly vaccinated population of The Netherlands: observations on surveillance data. Emerg. Infect. Dis. 6:348-357.
Farrell, D. J., M. McKeon, G. Daggard, M. J. Loeffelholz, C. J. Thompson, and T. K. Mukkur. 2000. Rapid-cycle PCR method to detect Bordetella pertussis that fulfills all consensus recommendations for use of PCR in diagnosis of pertussis. J. Clin. Microbiol. 38:4499-4502.
Fry, N. K., O. Tzivra, Y. T. Li, A. McNiff, N. Doshi, P. A. C. Maple, N. S. Crowcroft, E. Miller, R. C. George, and T. G. Harrison. 2004. Laboratory diagnosis of pertussis infections: the role of PCR and serology. J. Med. Microbiol. 53:1-7.
Furuya, D., A. Yagihashi, T. Endoh, N. Uehara, N. Fujii, S. Chiba, and N. Watanabe. 1999. Simultaneous amplification of Bordetella repeated insertion sequences and toxin promoter region gene by polymerase chain reaction. Immunopharmacol. Immunotoxicol. 21:55-63.
Gladbach, S., S. Hanauer, U. Reischl, K. Wilson, and G. N. Sanden. 2002. Identification of IS481 in Bordetella bronchseptica: implications for B. spp. phylogeny and diagnosis of B. pertussis infections by PCR assays. 7th Int. Symp. Pertussis: genome, pathogenesis, and immunity.
Hallander, H. O. 1999. Microbiological and serological diagnosis of pertussis. Clin. Infect. Dis. 28(Suppl. 2):S99-S106.
He, Q., and J. Mertsola. 1997. Epidemiology and prevention of pertussis. Curr. Opin. Pediatr. 9:14-18.
Heininger, U., G. Schmidt-Schlapfer, J. D. Cherry, and K. Stehr. 2000. Clinical validation of a polymerase chain reaction assay for the diagnosis of pertussis by comparison with serology, culture, and symptoms during a large pertussis vaccine efficacy trial. Pediatrics 105:E31.
Hoebrekx, N., G. Muyldermans, O. Soetens, I. De Schutter, A. Malfroot, D. Pierard, and S. Lauwers. 2001. Molecular typing of Bordetella pertussis isolates in Belgium from 1987 to 2000, p. 279. Abstr. 41st Intersci. Conf. Antimicrob. Agents Chemother. American Society for Microbiology, Washington, D.C.
Klement, E., L. Uliel, I. Engel, T. Hasin, M. Yavzori, N. Orr, N. Davidovitz, N. Lahat, I. Srugo, E. Zangvil, and D. Cohen. 2003. An outbreak of pertussis among young Israeli soldiers. Epidemiol. Infect. 131:1049-1054.
Kosters, K., M. Riffelmann, and C. H. Wirsing von Konig. 2001. Evaluation of a real-time PCR assay for detection of Bordetella pertussis and B. parapertussis in clinical samples. J. Med. Microbiol. 50:436-440.
Kosters, K., U. Reischl, J. Schmetz, M. Riffelmann, and C. H. Wirsing von Konig. 2002. Real-time LightCycler PCR for detection and discrimination of Bordetella pertussis and Bordetella parapertussis. J. Clin. Microbiol. 40:1719-1722.
Lievano, F. A., M. A. Reynolds, A. L. Waring, J. Ackelsberg, K. M. Bisgard, G. N. Sanden, D. Guris, A. Golaz, D. J. Bopp, R. J. Limberger, and P. F. Smith. 2002. Issues associated with and recommendations for using PCR to detect outbreaks of pertussis. J. Clin. Microbiol. 40:2801-2805.
Lindgren, C., J. Milerad, and H. Lagercrantz. 1997. Sudden infant death and prevalence of whooping cough in the Swedish and Norwegian communities. Eur. J. Pediatr. 156:405-409.
Loeffelholz, M. J., C. J. Thompson, K. S. Long, and M. J. Gilchrist. 1999. Comparison of PCR, culture, and direct fluorescent-antibody testing for detection of Bordetella pertussis. J. Clin. Microbiol. 37:2872-2876.
Loeffelholz, M. J. 2003. Bordetella, p. 780-789. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. American Society for Microbiology, Washington, D.C.
Lorenzo-Pajuelo, B., J. L. Villanueva, J. Rodriguez-Cuesta, N. Vergara-Irigaray, M. Bernabeu-Wittel, A. Garcia-Curiel, and G. Martinez de Tejada. 2002. Cavitary pneumonia in an AIDS patient caused by an unusual Bordetella bronchiseptica variant producing reduced amounts of pertactin and other major antigens. J. Clin. Microbiol. 40:3146-3154.
Matthews, R. C., N. Golbang, W. M. Bruck, D. Owen, A. Bailey, V. Weston, and J. R. Kerr. 1999. Semiquantitative polymerase chain reaction enzyme immunoassay for the diagnosis of pertussis. Eur. J. Clin. Microbiol. Infect. Dis. 18:748-750.
Mazengia, E., E. A. Silva, J. A. Peppe, R. Timperi, and H. George. 2000. Recovery of Bordetella holmesii from patients with pertussis-like symptoms: use of pulsed-field gel electrophoresis to characterize circulating strains. J. Clin. Microbiol. 38:2330-2333.
Mertens, P. L., F. S. Stals, J. F. Schellekens, A. W. Houben, and J. Huisman. 1999. An epidemic of pertussis among elderly people in a religious institution in The Netherlands. Eur. J. Clin. Microbiol. Infect. Dis. 18:242-247.
Noordhoek, G. T., A. H. J. Kolk, G. Bjune, D. Catty, J. W. Dale, P. E. M. Fine, P. Godfrey-Faussett, S.-N. Cho, T. Shinnick, S. B. Svenson, S. Wilson, and J. D. A. van Embden. 1994. Sensitivity and specificity of PCR for detection of Mycobacterium tuberculosis: a blind comparison study among seven laboratories. J. Clin. Microbiol. 32:277-284.
Parkhill, J., M. Sebaihia, A. Preston, L. D. Murphy, N. Thomson, D. E. Harris, M. T. Holden, C. M. Churcher, S. D. Bentley, K. L. Mungall, A. M. Cerdeno-Tarraga, L. Temple, K. James, B. Harris, M. A. Quail, M. Achtman, R. Atkin, S. Baker, D. Basham, N. Bason, I. Cherevach, T. Chillingworth, M. Collins, A. Cronin, P. Davis, J. Doggett, T. Feltwell, A. Goble, N. Hamlin, H. Hauser, S. Holroyd, K. Jagels, S. Leather, S. Moule, H. Norberczak, S. O'Neil, D. Ormond, C. Price, E. Rabbinowitsch, S. Rutter, M. Sanders, D. Saunders, K. Seeger, S. Sharp, M. Simmonds, J. Skelton, R. Squares, S. Squares, K. Stevens, L. Unwin, S. Whitehead, B. G. Barrell, and D. J. Maskell. 2003. Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat. Genet. 35:32-40.
Pelloux, H., E. Guy, M. C. Angelici, H. Aspock, M. H. Bessieres, R. Blatz, M. Del Pezzo, V. Girault, R. Gratzl, M. Holberg-Petersen, J. Johnson, D. Kruger, M. Lappalainen, A. Naessens, and M. Olsson. 1998. A second European collaborative study on polymerase chain reaction for Toxoplasma gondii, involving 15 teams. FEMS Microbiol. Lett. 165:231-237.
Poddar, S. K. 2003. Detection and discrimination of B. pertussis and B. holmesii by real-time PCR targeting IS481 using a beacon probe and probe-target melting analysis. Mol. Cell. Probes 17:91-98.
Reischl, U., N. Lehn, G. N. Sanden, and M. J. Loeffelholz. 2001. Real-time PCR assay targeting IS481 of Bordetella pertussis and molecular basis for detecting Bordetella holmesii. J. Clin. Microbiol. 39:1963-1966.
Schirm, J., A. M. van Loon, E. Valentine-Thon, P. E. Klapper, J. Reid, and G. M. Cleator. 2002. External quality assessment program for qualitative and quantitative detection of hepatitis C virus RNA in diagnostic virology. J. Clin. Microbiol. 40:2973-2980.
Schmidt-Schlapfer, G., J. G. Liese, F. Porter, S. Stojanov, M. Just, and B. H. Belohradsky. 1997. Polymerase chain reaction (PCR) compared with conventional identification in culture for detection of Bordetella pertussis in 7153 children. Clin. Microbiol. Infect. 3:462-467.
Shvartzman, P., T. Swartz, T. Stopler, and J. Herman. 1991. Outbreak of pertussis in a closed population with a high vaccination rate. Isr. J. Med. Sci. 27:137-140.
Sloan, L. M., M. K. Hopkins, P. S. Mitchell, E. A. Vetter, J. E. Rosenblatt, W. S. Harmsen, F. R. Cockerill, and R. Patel. 2002. Multiplex LightCycler PCR assay for detection and differentiation of Bordetella pertussis and Bordetella parapertussis in nasopharyngeal specimens. J. Clin. Microbiol. 40:96-100.
Sutter, R. W., and S. L. Cochi. 1992. Pertussis hospitalizations and mortality in the United States, 1985-1988. Evaluation of the completeness of national reporting. JAMA 267:386-391.
Tang, Y. W., M. K. Hopkins, C. P. Kolbert, P. A. Hartley, P. J. Severance, and D. H. Persing. 1998. Bordetella holmesii-like organisms associated with septicemia, endocarditis, and respiratory failure. Clin. Infect. Dis. 26:389-392.
Templeton, K. E., S. A. Scheltinga, A. Van Der Zee, B.M. Diederen, A. M. Kruijssen, H. Goossens, E. Kuijper, and E. C. Claas. 2003. Evaluation of real-time PCR for detection of and discrimination between Bordetella pertussis, Bordetella parapertussis, and Bordetella holmesii for clinical diagnosis. J. Clin. Microbiol. 41:4121-4126.
Tilley, P. A., M. V. Kanchana, I. Knight, J. Blondeau, N. Antonishyn, and H. Deneer. 2000. Detection of Bordetella pertussis in a clinical laboratory by culture, polymerase chain reaction, and direct fluorescent antibody staining; accuracy, and cost. Diagn. Microbiol. Infect. Dis. 37:17-23.
Valentine-Thon, E., A. M. van Loon, J. Schirm, J. Reid, P. E. Klapper, and G. M. Cleator. 2001. A European proficiency testing program for molecular detection and quantitation of hepatitis B virus DNA. J. Clin. Microbiol. 39:4407-4412.
Van der Zee, A., C. Agterberg, M. Peeters, J. Schellekens, and F. R. Mooi. 1993. Polymerase chain reaction assay for pertussis: simultaneous detection and discrimination of Bordetella pertussis and Bordetella parapertussis. J. Clin. Microbiol. 31:2134-2140.
Van der Zee, A., C. Agterberg, M. Peeters, F. Mooi, and J. Schellekens. 1996. A clinical validation of Bordetella pertussis and Bordetella parapertussis polymerase chain reaction: comparison with culture and serology using samples from patients with suspected whooping cough from a highly immunized population. J. Infect. Dis. 174:89-96.
Viejo, G., P. de la Iglesia, L. Otero, M. I. Blanco, B. Gomez, D. De Miguel, A. Del Valle, and B. De la Fuente. 2002. Bordetella bronchiseptica pleural infection in a patient with AIDS. Scand. J. Infect. Dis. 34:628-629.
Wallet, F., T. Perez, S. Armand, B. Wallaert, and R. J. Courcol. 2002. Pneumonia due to Bordetella bronchiseptica in a cystic fibrosis patient: 16S rRNA sequencing for diagnosis confirmation. J. Clin. Microbiol. 40:2300-2301.
Yih, W. K., E. A. Silva, J. Ida, N. Harrington, S. M. Lett, and H. George. 1999. Bordetella holmesii-like organisms isolated from Massachusetts patients with pertussis-like symptoms. Emerg. Infect. Dis. 5:441-443.(G. Muyldermans, O. Soeten)