- Email: [email protected]
Sensitive and specific polymerase chain reaction assays for detection of Bordetella pertussis in nasopharyngeal specimens
Sensitive and specific polymerase chain reaction assays for detection of Bordetella pertussis in nasopharyngeal specimens Qiushui He, MD, Jussi Mertsola, MD, Hanna Soini, MSc, a n d Matti K. Viljanen, MD From the National Public Health Institute, Department in Turku, and the Department of Pediatrics, Turku University Hospital, University of Turku, Turku, Finland Polymerase chain reaction (PCR) assays t h a t a m p i i f y segments of a r e p e a t e d g e n e e l e m e n t and a toxin promoter g e n e of Bordetellapertussis were c o m p a r e d with a culture established for the diagnosis of pertussis. Of 44 n a s o p h a r y n g e a l (NP) aspirates c o l l e c t e d during a pertussis outbreak, r e p e a t e d g e n e element PCR showed a positive result in 21 (48%), including all three patients with positive culture results. Results of toxin promoter g e n e PCR were positive in e i g h t (18%) cases, and the p a t h o g e n was not d e t e c t e d in one patient with a positive culture result. A more sensitive nested PCR assay, based on r e p e a t e d g e n e element PCR, was then d e v e l o p e d . During a second outbreak two different transportation systems were tested in 146 d u p l i c a t e NP swabs. Transportation of swabs in empty tubes proved to be better than in transport media for PCR. A total of 190 NP specimens from the two outbreaks were tested, and in 56 the results were shown to be positive by PCR, including all 16 cases confirmed as positive by culture. We c o n c l u d e that the PCR assay is more sensitive than culture in the diagnosis o f pertussis; NP swabbing is a simple, practical, and reliable method of collecting clinical specimens for PCR assays and cultures. (J PEDIATR1994;124: 421-6)
Definitive diagnosis of pertussis depends on the recovery of Bordetella pertussis from nasopharyngeal specimens. However, culture of this bacterium has low sensitivity even during the early phase of the disease, a4 The polymerase chain reaction technique has been widely applied to the diagnosis of various infectious diseases because of its substantial advantages in specificity and sensitivity;5, 6 P C R has potential for the diagnosis of pertussis.3, 7-11 We tested the specificities and sensitivities of three
Supported by the Academy of Finland. Submitted for publication May 7, 1993; accepted Oct. 4, 1993. Reprint requests: Qiushui He, MD, National Public Health Institute, Department in Turku, Kiinamyllynkatu 13, SF-20520 Turku, Finland. Copyright © 1994 by Mosby-Year Book, Inc. 0022-3476/94 $3.00 + 0 9/20/51946
P C R assays for the diagnosis of pertussis with nasopharyngeal aspirates and swabs. A P C R assay with the primer pair chosen from a repeated gene element and P C R assay with ET-swab Nes-PCR NP PCR Rep-PCR
TM-swab Tox-PCR
NP swab placed in an empty tube after collection Nested PCR assay based on Rep-PCR Nasopharygeal Polymerase chain reaction PCR assay using a primer pair selected from a repeated gene element of Bordetella pertussis NP swab placed in a transport medium tube after collection PCR assay using a primer pair selected from a toxin promoter gene of B. pertussis
a toxin promoter gene of B. pertussis were compared. Effects of various transportation and storage conditions of
421
422
He et al.
swabs were also studied to determine the most reliable and practical handling procedure.
METHODS Patients. Clinical specimens were collected during two outbreaks of pertussis. The first outbreak occurred in western Finland in December 1990. The NP aspirates were obtained from 44 subjects, of whom 41 had prolonged cough with vomiting, whooping, or both (mean duration of symptoms at time of sampling was 26 days; median, 14 days; range, 1 to 150 days). The remaining three patients were symptom-free contacts of a patient with a positive culture result. The second outbreak, in southwestern Finland, occurred between September and December 1992. Duplicate swabs were used to take NP specimens from each of the 109 patients with the symptoms described above (mean duration at sampling, 22 days; median, 14 days; range, 1 to 180 days) and from 37 symptom-free contacts. To assess the speeificities of the PCR assays for clinical specimens, we collected NP specimens by swab from 75 healthy subjects (60 children and 15 adults) under nonepidemic conditions and from 52 children with other respiratory infections. Clinieal specimens. During the first outbreak the NP aspirates were immediately inoculated onto charcoal agar plates (Oxoid CM 19; Uripath Co., Ogdensburg, N.Y.), supplemented with horse blood and cephalexin. 12 During the second outbreak, duplicate NP swabs (Calgiswab type 1; Spectrum Laboratories, Inc., Los Angeles, Calif.) were used to take specimens from the posterior nasopharynx through each nostril One of the two specimens was immediately inoculated onto the plate and the swab was then stored in an empty tube for PCR. The other specimen was streaked by swab onto the slant of the transport medium (same composition as the plate) and the swab was then immersed in the medium. Both tubes were transported at 4 ° C to the laboratory within 1 to 3 days of sampling. At the laboratory the TM-swab was again inoculated on the plate and stored in an empty tube for PCR. All NP aspirates and swabs were stored at - 2 0 ° C before testing. We identified B. pertussis on the plates as described earlier) Polymerase chain reaction. Details of DNA extraction from bacterial strains and NP specimens and of the PCR assay have been described previously.3 Briefly, two oligonucleotide primer pairs were chosen from a repeated gene element (for Rep-PCR 7) and one pair from a toxin promoter gene of B. pertussis (for Tox-PCRS). In the RepPCR, one primer pair was used as outer primers, as described, 3 and the other pair was used as inner primers in the nested PCR assay based on the Rep-PCR. The sequence of the inner primers was (41 bp) 5"-CGAACCGGATTT G A G A A A C T G G A A A T - 3 ' (66 pb) and (164 bp) 5'-
The Journal of Pediatrics March 1994
AATTGCTGGACCATTTCGAGTCGACG-3 ' (139 bp). The first-run PCR for both Rep-PCR and Tox-PCR was done as described previously.3, 8 For the Nes-PCR, a 50 #1 reaction volume included all reagents and 5 tA of the product amplified in the first run. A total of 15 cycles were used for further amplification. The specificities of the three PCR assays were tested with 130 bacterial strains, including 22 B. pertussis, 16 Bordetella parapertussis, and one Bordetella bronchiseptica, as described earlier for Rep-PCR. 3 Sensitivities were assessed by testing serial 10-fold dilutions (in sterile physiologic saline solution) of a fresh clinical isolate of B. pertussis. To analyze the effects of various storage conditions and reagents on PCR results, we immersed sterile calcium alginate swabs in five serial 10-fold dilutions with bacterial concentrations of 2 x 106 to 2 x 102 cells/ml for 1 minute. The swabs were then stored in empty tubes, in tubes with the transport medium (same as that used for culture), in tubes with 1 ml physiologic saline solution, and in tubes with 1 ml phosphate-buffered saline solution containing 20% fetal bovine serum and 2% polyoxyethylene sorbitan monolaurate (Tween 20) (the same as that used for antigen detection of respiratory viruses in the University of Turku Department of Virology). To mimic transport conditions, we stored the tubes at 4 ° C or 22 ° C for 2 or 4 days. A total of 160 swabs were tested. Analysis of PCR products. A 20-mer oligonucleotide probe selected from an internal region o f the amplified DNA was used for analysis of PCR products from RepPCR and Nes-PCR. After gel eletrophoresis the products were transferred to a nylon membrane (Boehringer Mannheim GmbH, Mannheim, Germany). Southern hybridization was performed with a digoxigenin-labeled probe, and immunologic detection was performed as recommended by the supplier (Boehringer, Mannheim). Statistical methods. Statistical differences between culture and PCR, as well as between PCR assays, were tested with the two-tailed Fisher Exact Test. A p value <0.05 was considered to be significant. RESULTS All three PCR assays proved to be specific for detection of B. pertussis in laboratory conditions. Further evidence for specificity was obtained by testing NP specimens collected by swab from 75 healthy persons and from 52 children with other respiratory infections. None of the results for those samples were positive for the presence of B. pertussis either in culture or in the PCR assays. No inhibitory phenomenon was detected in the specimens examined. The sensitivity of Rep-PCR was 25 bacteria per reaction tube; the sensitivity of Tox-PCR was 250 bacteria per tube (Fig. 1). By further amplification with Nes-PCR, the
The Journal o f Pediatrics Volume 124, Number 3
He et al.
423
Fi9. I. Lanes 1 to 9 show amplification of DNA extracted from 102 to 101° dilutions of starting culture of B. pertussis (2.5 X 10l° bacteria/ml). A, Amplification of 18 l-base pair segment with primer pair from a repeated element (Rep-PCR assay). Lane 6 indicates that 25 bacteria per reaction tube were detected. B, Amplification of a 191-base pair segment with primer pair from a toxin promoter gene (Tox-PCR assay). Lane 5 indicates that 250 bacteria per reaction tube were detected. Lane L contains molecular weight markers; the minus and plus signs indicate negative and positive controls, respectively.
Fig. 2. Top: Gel electrophoresis of PCR-amplified products. Bottom: Southern hybridization of PCR-amplified products detected with a digoxigenin-labeled probe. Left: First-run Rep-PCR assay. Right: Nested PCR assay. Lanes i to 6 depict the amplification of DNA extracted from serial 10-fold dilutions of B. pertussis. Lanes 7 to 12 show amplification of DNA extracted from swabs immersed in bacterial dilutions for 1 minute. For both panels, lane L contains molecular weight markers; the minus and plus signs indicate negative and positive controls, respectively.
increase in sensitivity was at least 10-fold (Fig. 2). The PCR-amplified products from the serial bacterial dilutions, containing expected fragments, reacted with the probe specific for B. pertussis (Fig. 2). To test the effects of storage conditions on P C R results, we first inoculated the swabs with different concentrations
of B. pertussis bacteria. The swabs were then stored in various media (see Methods section, above). No differences in sensitivity were found when these swabs were tested by Nes-PCR. The storage of the swabs for 2 or 4 days or at 4 ° C or 22 ° C resulted in similar sensitivities. A total of 190 N P specimens from two pertussis out-
424
He et al.
The Journal of Pediatrics March 1994
Table I. Distribution of laboratory-defined pertussis cases by age group in two outbreaks Laboratory-defined cases*
Age groups (yr)
Culture positive/total No.
%
PCR positive No.
%
Laboratory positive/ cough positivet No.
%
Laboratory positive/ symptom-free contacts:~ No.
%
Total No.
%
<1 0/9 0 4 44 4/6 67 0/3 0 4 44 1-3 0/14 0 4 29 3/12 25 1/2 50 4 29 4-6 2/37 5 9 24 9/32 28 0/5 0 9 24 7-15 8/56 14 17 30 14/43 33 3/13 23 17 30 >16 6/74 8 22 30 17/57 30 5/17 29 22 30 TOTAL 16/190 8 56 29 47/150 31 9/40 23 56 29 *Laboratory-definedpertussiscases (laboratorypositive)were confirmedby culture, PCR, or both; all culture-positivecases had positivePCR results. tOf coughingpatients, 78% bad paroxysmalcough. tAll 127 controlNP swabsfromhealthypersonsand childrenwith otherrespiratoryinfectionscollectedin nonoutbreakconditionshad negativePCR results (see Methods section).
breaks Were teste d by culture and by PCR assay; results in 56 (29%) cases were identified as positive by PCR, including all 16 cases with positive culture results (Table I). During the first outbreak a total of 21 laboratory-defined pertussis cases were identified. Of these, three (14%) were confirmed by culture, eight (38%) by Tox-PCR, and 21 (100%) by Rep-PCR (Table II). Two of the three cases with positive culture results also had a positive Tox-PCR result. The positivity rate obtained by Rep-PCR was significantly higher than that obtained by Tox-PCR (p < 0.01) and was also significantly higher than that obtained by Culture (p < 0.001). Three symptom-free contacts of a patient with a positive culture result had positive results only With RepPCR. Of the 33 patients with symptoms who had cough --<6 weeks in duration, 18 (55%) were identified as having pertussis. However, none of the eight patients with symptoms for more than 7 Weeks had diagnoses made on the basis of any of the three tests (Fig. 3). A t the onset of the second outbreak, 91 NP swabs were tested by both Rep-PCR and Nes-PCR for comparison of thefr sensitivities. The same swabs used for culture were transported to the laboratory in empty tubes Until testing by PCR. The culture results were positive in 10 cases, RepPCR detected 10 cases of pertussis, and Nes-PCR detected 24 caseS. Of the 10 culture-positive samples, 8 had positive results With Rep-PCR, and 9 had positive results with NesPCR. The culture-positive sample that was missed was found to be positive with PCR when tested by Nes-PCR with the TM-swab. A total of 146 duplicate NP swabs were collected during the second outbreak (Table III). Thirty-five laboratory-defined pertussis cases were identified by Nes-PCR and only 13 by culture (p < 0.01). All 13 culture-positive cases were also detected by the Nes-PCR assay. Of the 35 patients, diagnoses were made in 10 case (29%) with TM-swabs and in
32 (91%) with ET-swabs (p < 0.001). Of the 13 culturepositive cases, 12 (92%) were detected by PCR assay of ET-swabs and 7 (54%) by PCR assay of TM-swabs. The PCR-amplified products of 40 NP swabs, randomly selected from the 146 clinical samples, were tested by Southern hybridization. Hybridization confirmed both positive and negative results obtained with these samples. DISCUSSION Our study of two pertussis outbreaks shows that the PCR assay is more sensitive than culture in the detection of B. pertussis in NP specimens, a finding that is in agreement with previous findings. 3, 7, 9, 10 Several primer pairs derived from a repeated gene element, the toxin promoter gene, and the adenylate cyclase gene of B. pertussis have been used in PCR for diagnosis of pertussis. 3, 7-11 The sensitivities of these assays vary from 1 to 100 bacteria detected. In our study Rep-PCR was found to be more sensitive than ToxPCR in in vitro conditions and in clinical specimens. This is not surprising because multiple copies of the repeated element exist in the genome orB. Pertussis.7, 13, 14 In addition, nested PCR increases the sensitivity of the first-run assay.i5, 16 Our results confirm that the diagnostic efficacy of culture is low in populations with a high vaccination rate. 1719 Adults accounted for a large proportion of diagnosed pertussis cases in this study; this finding agree s with earlier reports from countries with high vaccination rates. 19-21Adult patients with pertussis either have atypical symptoms more often than children 19 or remain free of symptoms. 21 Of adults with positive diagnostic results in our study, about 25% had no symptoms. The role of these symptom-free adults with positive PCR results in transmission of bacteria is not clear and requires further study. However, our preliminary experience with symptom-free persons with
The Journal of Pediatrics Volume 124, Number 3
He et al.
425
Cases 20_
I---I
Total no. of NPAs
loo
BB
RepPCR-positive
15 _
~
0-1
2-3
uure°; ve ToxPCR-positive
4-6
~7
Weeks after onset of cough
Fig. 3. Results of culture, Tox-PCR assay, and Rep-PCR assay, in relation to time from onset of cough. Results from NP aspirates in 41 patients with symptoms during the first outbreak are shown.
T a b l e II. Detection of B. pertussis in NP aspirates by Rep-PCR and Tox-PCR assays compared with culture results during first outbreak Positive Rep-PCR results
N e g a t i v e Rep-PCR results
Culture results
Positive Tox-PCR results
N e g a t i v e Tox-PCR results
Positive Tox-PCR results
N e g a t i v e Tox-PCR results
Total
Positive Negative TOTAL
2 6 8
1 12 13
0 0 0
0 23 23
3 41 44
T a b l e III. Detection of B. pertussis by Nes-PCR assay of ET-swab and TM-swab compared with culture results during second outbreak Positive PCR results: ET-swab
N e g a t i v e PCR results: ET-swab
Culture results
Positive TM-swab results
N e g a t i v e TM-swab results
Positive TM-swab results
N e g a t i v e TM-swab results
Total
Positwe Negative TOTAL
7 0 7
5 20 25
1 2 3
0 111 111
13 133 146
positive PCR results is that they do not transmit pertussis. Several laboratories have had success in using swabs as primary source material for PCR. 3, 7, 9-11 The swabs (e.g., calcium alginate, dacron, and cotton) used for PCR seem to perform well, although no direct comparisons or evaluations of the transport systems have been done. Although both the ET-swab and the TM-swab seemed to be equally useful in laboratory conditions, the ET-swab was significantly better than the TM-swab in clinical conditions. For the clinical diagnostic laboratory, transport of NP swabs in an empty tube at ambient temperature is a simple and reliable tool for rapid detection of B. pertussis. The results of
PCR assay are available within 24 to 48 hours, which is usually before the appearance of the first colonies on culture plates. The PCR results can also be easily interpreted by examination of the exact size of amplified products on agarose gel. Any problems of contamination in the handling of samples and amplified products can be overcome. 22 Use of the PCR assay for the diagnosis of pertussis is gaining wide acceptance in Finland because of its sensitivity and rapidity. The cost of the PCR assay for a given clinical sample is nearly the same as that of a culture. Although this new assay is sensitive in itself, we prefer to use it in combination with culture for early diagnosis of pertussis.
426
He et al.
The Journal of Pediatrics March 1994
We thank Birgitta Aittanen for expert technical assistance, and the personnel of the National Public Health Institute, Department in Turku, and the staff of the Health Center of Paimio for their cooperation. We also thank Mikael Skurnik, PhD, for synthesizing the oligonucleotide primers. Simo Merne, MA, revised the manuscript. REFERENCES
11.
12. 13.
1. Combined Scottish Study. Diagnosis of whooping cough: comparison of serological tests with isolation of Bordetella pertussis. BMJ 1970;4:637-9. 2. Public Health Laboratory Service Epidemiological Research Laboratory and 21 Area Health Authorities. Efficacy of pertussis vaccination in England. BMJ 1982;285:357-9. 3. He Q, Mertsola J, Soini H, Skurnik M, Ruuskanen O, Viljanen MK. Comparison of polymerase chain reaction with culture and enzyme immunoassay for diagnosis of pertussis. J Clin Microbiol 1993;31:642-5. 4. Mertsola J, Kur6nen T, Turunen A, Viljanen MK, Ruuskanen O. Diagnosis of pertussis. J Infect 1984;8:149-56. 5. Eisenstein BI. The polymerase chain reaction: a new method of using molecular genetics for medical diagnosis. N Engl J Med 1990;322:178-83. 6. Peter JB. The polymerase chain reaction: amplifying our options. Rev Infect Dis 1991;13:166-71. 7. Glare EM, Paton JC, Premier RR, Lawrence A J, Nisbet IT. Analysis of a repetitive DNA sequence from Bordetella pertussis and its application to the diagnosis of pertussis using the polymerase chain reaction. J Clin Microbiol 1990;28:1982-7. 8. liouard S, Hackel C, lierzog A, Bollen A. Specific identification of Bordetella pertussis by the polymerase chain reaction. Res Microbiol 1989;140:477-87. 9. Van der Zee A, Agterberg C, Peeters M, Schellekens J, Mooi FR. Polymerase chain reaction assay for pertussis: simultaneous detection and discrimination of Bordetella pertussis and Bordetella parapertussis. J Clin Microbiol 1993;31:2134-40. 10. Schl~ipfer G, Senn liP, Berger R, Just M. Use of the polymerase chain reaction to detect Borcletella pertussis in
14.
15.
16.
17.
18.
19.
20.
21. 22.
patients with mild or atypical symptoms of infection. Eur J Clin Microbiol Infect Dis 1993;12:459-63. Douglas E, Coote JG, Parton R, McPheat W. Identification of Bordetella pertussis in nasopharyneal swabs by PCR amplification of a region of the adenylate cyclase gene. J Med Microbiol 1993;38:140-4. Regan J, Lowe F. Enrichment medium for the isolation of Bordetella. J Clin Microbiol 1977;6:303-9. McLafferty MA, Harcus DR, Hewlett EL. Nucleotide sequence and characterization of a repetitive DNA element from the genome of Bordetella pertussis with characteristics of an insertion sequence. J Gen Microbiol 1988;134:2297-306. McPheat W, McNally LT. Isolation of repeated DNA sequence from Bordetella pertussis. J Gen Microbiol 1987; 133:323-30. Aurelius E, Johansson B, Sk61denberg B, Staland A., Forsgren M. Rapid diagnosis of herpes simplex encephalitis by nested polymerase chain reaction assay of cerebrospinal fluid. Lancet 1991;337:189-92. Lipschik GY, Gill VJ, Lundgren JD, et al. Improved diagnosis of Pneumocystis cariniiinfection by polymerase chain reaction on induced sputum and blood. Lancet 1992;340:203-6. Fine PEM, Clarkson JA. The recurrence of whooping cough: possible implications for assessment of vaccine efficacy. Lancet 1982;1:666-9. Kwantes W, Joynson DHM, Williams WO. Bordetella pertussis isolation in general practice: 1977-79 whooping cough epidemic in West Glamorgan. J Hyg 1983;99:149-58. Morgan Mink C, Cherry JD, Christenson P, et al. A search for Bordetella pertussis infection in university students. Clin Infect Dis 1992;14:464-71. Robertson PW, Goldberg H, Jarvue BH, Smith DD, Whybin LR. Bordetella pertussis infection: a cause of persistent cough in adults. Med J Aust 1987;146:522-5. Mertsola J, Ruuskanen O, Eerola E, Viljanen MK. Intrafamilial spread of pertussis. J PEDIATR 1983;103:359-63. Kowk S, Higuchi R. Avoiding false positives with PCR. Nature 1989;339:237-8.
FELLOWSHIPS Fellowships available in pediatric subspecialties and those for general academic pediatric training are listed once a year, in January, in THE JOURNAL OF PEDIATRICS. Each June, forms for listing fellowships available for the academic year beginning 18 months after publication are sent to the Chairman of the Department of Pediatrics at major hospitals in the United States and Canada. In addition, a copy of the application form appears in the July, August, and September issues of THE JOURNAL (please use the current form). Should you desire to list fellowships, a separate application must be made each year for each position. All applications must be returned to M o s b y - Y e a r Book, Inc., by October 15 preceding the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing Department, M o s b y - Y e a r Book, Inc., 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/800-325-4177, ext. 4317, or 314453-4317.
Definitive diagnosis of pertussis depends on the recovery of Bordetella pertussis from nasopharyngeal specimens. However, culture of this bacterium has low sensitivity even during the early phase of the disease, a4 The polymerase chain reaction technique has been widely applied to the diagnosis of various infectious diseases because of its substantial advantages in specificity and sensitivity;5, 6 P C R has potential for the diagnosis of pertussis.3, 7-11 We tested the specificities and sensitivities of three
Supported by the Academy of Finland. Submitted for publication May 7, 1993; accepted Oct. 4, 1993. Reprint requests: Qiushui He, MD, National Public Health Institute, Department in Turku, Kiinamyllynkatu 13, SF-20520 Turku, Finland. Copyright © 1994 by Mosby-Year Book, Inc. 0022-3476/94 $3.00 + 0 9/20/51946
P C R assays for the diagnosis of pertussis with nasopharyngeal aspirates and swabs. A P C R assay with the primer pair chosen from a repeated gene element and P C R assay with ET-swab Nes-PCR NP PCR Rep-PCR
TM-swab Tox-PCR
NP swab placed in an empty tube after collection Nested PCR assay based on Rep-PCR Nasopharygeal Polymerase chain reaction PCR assay using a primer pair selected from a repeated gene element of Bordetella pertussis NP swab placed in a transport medium tube after collection PCR assay using a primer pair selected from a toxin promoter gene of B. pertussis
a toxin promoter gene of B. pertussis were compared. Effects of various transportation and storage conditions of
421
422
He et al.
swabs were also studied to determine the most reliable and practical handling procedure.
METHODS Patients. Clinical specimens were collected during two outbreaks of pertussis. The first outbreak occurred in western Finland in December 1990. The NP aspirates were obtained from 44 subjects, of whom 41 had prolonged cough with vomiting, whooping, or both (mean duration of symptoms at time of sampling was 26 days; median, 14 days; range, 1 to 150 days). The remaining three patients were symptom-free contacts of a patient with a positive culture result. The second outbreak, in southwestern Finland, occurred between September and December 1992. Duplicate swabs were used to take NP specimens from each of the 109 patients with the symptoms described above (mean duration at sampling, 22 days; median, 14 days; range, 1 to 180 days) and from 37 symptom-free contacts. To assess the speeificities of the PCR assays for clinical specimens, we collected NP specimens by swab from 75 healthy subjects (60 children and 15 adults) under nonepidemic conditions and from 52 children with other respiratory infections. Clinieal specimens. During the first outbreak the NP aspirates were immediately inoculated onto charcoal agar plates (Oxoid CM 19; Uripath Co., Ogdensburg, N.Y.), supplemented with horse blood and cephalexin. 12 During the second outbreak, duplicate NP swabs (Calgiswab type 1; Spectrum Laboratories, Inc., Los Angeles, Calif.) were used to take specimens from the posterior nasopharynx through each nostril One of the two specimens was immediately inoculated onto the plate and the swab was then stored in an empty tube for PCR. The other specimen was streaked by swab onto the slant of the transport medium (same composition as the plate) and the swab was then immersed in the medium. Both tubes were transported at 4 ° C to the laboratory within 1 to 3 days of sampling. At the laboratory the TM-swab was again inoculated on the plate and stored in an empty tube for PCR. All NP aspirates and swabs were stored at - 2 0 ° C before testing. We identified B. pertussis on the plates as described earlier) Polymerase chain reaction. Details of DNA extraction from bacterial strains and NP specimens and of the PCR assay have been described previously.3 Briefly, two oligonucleotide primer pairs were chosen from a repeated gene element (for Rep-PCR 7) and one pair from a toxin promoter gene of B. pertussis (for Tox-PCRS). In the RepPCR, one primer pair was used as outer primers, as described, 3 and the other pair was used as inner primers in the nested PCR assay based on the Rep-PCR. The sequence of the inner primers was (41 bp) 5"-CGAACCGGATTT G A G A A A C T G G A A A T - 3 ' (66 pb) and (164 bp) 5'-
The Journal of Pediatrics March 1994
AATTGCTGGACCATTTCGAGTCGACG-3 ' (139 bp). The first-run PCR for both Rep-PCR and Tox-PCR was done as described previously.3, 8 For the Nes-PCR, a 50 #1 reaction volume included all reagents and 5 tA of the product amplified in the first run. A total of 15 cycles were used for further amplification. The specificities of the three PCR assays were tested with 130 bacterial strains, including 22 B. pertussis, 16 Bordetella parapertussis, and one Bordetella bronchiseptica, as described earlier for Rep-PCR. 3 Sensitivities were assessed by testing serial 10-fold dilutions (in sterile physiologic saline solution) of a fresh clinical isolate of B. pertussis. To analyze the effects of various storage conditions and reagents on PCR results, we immersed sterile calcium alginate swabs in five serial 10-fold dilutions with bacterial concentrations of 2 x 106 to 2 x 102 cells/ml for 1 minute. The swabs were then stored in empty tubes, in tubes with the transport medium (same as that used for culture), in tubes with 1 ml physiologic saline solution, and in tubes with 1 ml phosphate-buffered saline solution containing 20% fetal bovine serum and 2% polyoxyethylene sorbitan monolaurate (Tween 20) (the same as that used for antigen detection of respiratory viruses in the University of Turku Department of Virology). To mimic transport conditions, we stored the tubes at 4 ° C or 22 ° C for 2 or 4 days. A total of 160 swabs were tested. Analysis of PCR products. A 20-mer oligonucleotide probe selected from an internal region o f the amplified DNA was used for analysis of PCR products from RepPCR and Nes-PCR. After gel eletrophoresis the products were transferred to a nylon membrane (Boehringer Mannheim GmbH, Mannheim, Germany). Southern hybridization was performed with a digoxigenin-labeled probe, and immunologic detection was performed as recommended by the supplier (Boehringer, Mannheim). Statistical methods. Statistical differences between culture and PCR, as well as between PCR assays, were tested with the two-tailed Fisher Exact Test. A p value <0.05 was considered to be significant. RESULTS All three PCR assays proved to be specific for detection of B. pertussis in laboratory conditions. Further evidence for specificity was obtained by testing NP specimens collected by swab from 75 healthy persons and from 52 children with other respiratory infections. None of the results for those samples were positive for the presence of B. pertussis either in culture or in the PCR assays. No inhibitory phenomenon was detected in the specimens examined. The sensitivity of Rep-PCR was 25 bacteria per reaction tube; the sensitivity of Tox-PCR was 250 bacteria per tube (Fig. 1). By further amplification with Nes-PCR, the
The Journal o f Pediatrics Volume 124, Number 3
He et al.
423
Fi9. I. Lanes 1 to 9 show amplification of DNA extracted from 102 to 101° dilutions of starting culture of B. pertussis (2.5 X 10l° bacteria/ml). A, Amplification of 18 l-base pair segment with primer pair from a repeated element (Rep-PCR assay). Lane 6 indicates that 25 bacteria per reaction tube were detected. B, Amplification of a 191-base pair segment with primer pair from a toxin promoter gene (Tox-PCR assay). Lane 5 indicates that 250 bacteria per reaction tube were detected. Lane L contains molecular weight markers; the minus and plus signs indicate negative and positive controls, respectively.
Fig. 2. Top: Gel electrophoresis of PCR-amplified products. Bottom: Southern hybridization of PCR-amplified products detected with a digoxigenin-labeled probe. Left: First-run Rep-PCR assay. Right: Nested PCR assay. Lanes i to 6 depict the amplification of DNA extracted from serial 10-fold dilutions of B. pertussis. Lanes 7 to 12 show amplification of DNA extracted from swabs immersed in bacterial dilutions for 1 minute. For both panels, lane L contains molecular weight markers; the minus and plus signs indicate negative and positive controls, respectively.
increase in sensitivity was at least 10-fold (Fig. 2). The PCR-amplified products from the serial bacterial dilutions, containing expected fragments, reacted with the probe specific for B. pertussis (Fig. 2). To test the effects of storage conditions on P C R results, we first inoculated the swabs with different concentrations
of B. pertussis bacteria. The swabs were then stored in various media (see Methods section, above). No differences in sensitivity were found when these swabs were tested by Nes-PCR. The storage of the swabs for 2 or 4 days or at 4 ° C or 22 ° C resulted in similar sensitivities. A total of 190 N P specimens from two pertussis out-
424
He et al.
The Journal of Pediatrics March 1994
Table I. Distribution of laboratory-defined pertussis cases by age group in two outbreaks Laboratory-defined cases*
Age groups (yr)
Culture positive/total No.
%
PCR positive No.
%
Laboratory positive/ cough positivet No.
%
Laboratory positive/ symptom-free contacts:~ No.
%
Total No.
%
<1 0/9 0 4 44 4/6 67 0/3 0 4 44 1-3 0/14 0 4 29 3/12 25 1/2 50 4 29 4-6 2/37 5 9 24 9/32 28 0/5 0 9 24 7-15 8/56 14 17 30 14/43 33 3/13 23 17 30 >16 6/74 8 22 30 17/57 30 5/17 29 22 30 TOTAL 16/190 8 56 29 47/150 31 9/40 23 56 29 *Laboratory-definedpertussiscases (laboratorypositive)were confirmedby culture, PCR, or both; all culture-positivecases had positivePCR results. tOf coughingpatients, 78% bad paroxysmalcough. tAll 127 controlNP swabsfromhealthypersonsand childrenwith otherrespiratoryinfectionscollectedin nonoutbreakconditionshad negativePCR results (see Methods section).
breaks Were teste d by culture and by PCR assay; results in 56 (29%) cases were identified as positive by PCR, including all 16 cases with positive culture results (Table I). During the first outbreak a total of 21 laboratory-defined pertussis cases were identified. Of these, three (14%) were confirmed by culture, eight (38%) by Tox-PCR, and 21 (100%) by Rep-PCR (Table II). Two of the three cases with positive culture results also had a positive Tox-PCR result. The positivity rate obtained by Rep-PCR was significantly higher than that obtained by Tox-PCR (p < 0.01) and was also significantly higher than that obtained by Culture (p < 0.001). Three symptom-free contacts of a patient with a positive culture result had positive results only With RepPCR. Of the 33 patients with symptoms who had cough --<6 weeks in duration, 18 (55%) were identified as having pertussis. However, none of the eight patients with symptoms for more than 7 Weeks had diagnoses made on the basis of any of the three tests (Fig. 3). A t the onset of the second outbreak, 91 NP swabs were tested by both Rep-PCR and Nes-PCR for comparison of thefr sensitivities. The same swabs used for culture were transported to the laboratory in empty tubes Until testing by PCR. The culture results were positive in 10 cases, RepPCR detected 10 cases of pertussis, and Nes-PCR detected 24 caseS. Of the 10 culture-positive samples, 8 had positive results With Rep-PCR, and 9 had positive results with NesPCR. The culture-positive sample that was missed was found to be positive with PCR when tested by Nes-PCR with the TM-swab. A total of 146 duplicate NP swabs were collected during the second outbreak (Table III). Thirty-five laboratory-defined pertussis cases were identified by Nes-PCR and only 13 by culture (p < 0.01). All 13 culture-positive cases were also detected by the Nes-PCR assay. Of the 35 patients, diagnoses were made in 10 case (29%) with TM-swabs and in
32 (91%) with ET-swabs (p < 0.001). Of the 13 culturepositive cases, 12 (92%) were detected by PCR assay of ET-swabs and 7 (54%) by PCR assay of TM-swabs. The PCR-amplified products of 40 NP swabs, randomly selected from the 146 clinical samples, were tested by Southern hybridization. Hybridization confirmed both positive and negative results obtained with these samples. DISCUSSION Our study of two pertussis outbreaks shows that the PCR assay is more sensitive than culture in the detection of B. pertussis in NP specimens, a finding that is in agreement with previous findings. 3, 7, 9, 10 Several primer pairs derived from a repeated gene element, the toxin promoter gene, and the adenylate cyclase gene of B. pertussis have been used in PCR for diagnosis of pertussis. 3, 7-11 The sensitivities of these assays vary from 1 to 100 bacteria detected. In our study Rep-PCR was found to be more sensitive than ToxPCR in in vitro conditions and in clinical specimens. This is not surprising because multiple copies of the repeated element exist in the genome orB. Pertussis.7, 13, 14 In addition, nested PCR increases the sensitivity of the first-run assay.i5, 16 Our results confirm that the diagnostic efficacy of culture is low in populations with a high vaccination rate. 1719 Adults accounted for a large proportion of diagnosed pertussis cases in this study; this finding agree s with earlier reports from countries with high vaccination rates. 19-21Adult patients with pertussis either have atypical symptoms more often than children 19 or remain free of symptoms. 21 Of adults with positive diagnostic results in our study, about 25% had no symptoms. The role of these symptom-free adults with positive PCR results in transmission of bacteria is not clear and requires further study. However, our preliminary experience with symptom-free persons with
The Journal of Pediatrics Volume 124, Number 3
He et al.
425
Cases 20_
I---I
Total no. of NPAs
loo
BB
RepPCR-positive
15 _
~
0-1
2-3
uure°; ve ToxPCR-positive
4-6
~7
Weeks after onset of cough
Fig. 3. Results of culture, Tox-PCR assay, and Rep-PCR assay, in relation to time from onset of cough. Results from NP aspirates in 41 patients with symptoms during the first outbreak are shown.
T a b l e II. Detection of B. pertussis in NP aspirates by Rep-PCR and Tox-PCR assays compared with culture results during first outbreak Positive Rep-PCR results
N e g a t i v e Rep-PCR results
Culture results
Positive Tox-PCR results
N e g a t i v e Tox-PCR results
Positive Tox-PCR results
N e g a t i v e Tox-PCR results
Total
Positive Negative TOTAL
2 6 8
1 12 13
0 0 0
0 23 23
3 41 44
T a b l e III. Detection of B. pertussis by Nes-PCR assay of ET-swab and TM-swab compared with culture results during second outbreak Positive PCR results: ET-swab
N e g a t i v e PCR results: ET-swab
Culture results
Positive TM-swab results
N e g a t i v e TM-swab results
Positive TM-swab results
N e g a t i v e TM-swab results
Total
Positwe Negative TOTAL
7 0 7
5 20 25
1 2 3
0 111 111
13 133 146
positive PCR results is that they do not transmit pertussis. Several laboratories have had success in using swabs as primary source material for PCR. 3, 7, 9-11 The swabs (e.g., calcium alginate, dacron, and cotton) used for PCR seem to perform well, although no direct comparisons or evaluations of the transport systems have been done. Although both the ET-swab and the TM-swab seemed to be equally useful in laboratory conditions, the ET-swab was significantly better than the TM-swab in clinical conditions. For the clinical diagnostic laboratory, transport of NP swabs in an empty tube at ambient temperature is a simple and reliable tool for rapid detection of B. pertussis. The results of
PCR assay are available within 24 to 48 hours, which is usually before the appearance of the first colonies on culture plates. The PCR results can also be easily interpreted by examination of the exact size of amplified products on agarose gel. Any problems of contamination in the handling of samples and amplified products can be overcome. 22 Use of the PCR assay for the diagnosis of pertussis is gaining wide acceptance in Finland because of its sensitivity and rapidity. The cost of the PCR assay for a given clinical sample is nearly the same as that of a culture. Although this new assay is sensitive in itself, we prefer to use it in combination with culture for early diagnosis of pertussis.
426
He et al.
The Journal of Pediatrics March 1994
We thank Birgitta Aittanen for expert technical assistance, and the personnel of the National Public Health Institute, Department in Turku, and the staff of the Health Center of Paimio for their cooperation. We also thank Mikael Skurnik, PhD, for synthesizing the oligonucleotide primers. Simo Merne, MA, revised the manuscript. REFERENCES
11.
12. 13.
1. Combined Scottish Study. Diagnosis of whooping cough: comparison of serological tests with isolation of Bordetella pertussis. BMJ 1970;4:637-9. 2. Public Health Laboratory Service Epidemiological Research Laboratory and 21 Area Health Authorities. Efficacy of pertussis vaccination in England. BMJ 1982;285:357-9. 3. He Q, Mertsola J, Soini H, Skurnik M, Ruuskanen O, Viljanen MK. Comparison of polymerase chain reaction with culture and enzyme immunoassay for diagnosis of pertussis. J Clin Microbiol 1993;31:642-5. 4. Mertsola J, Kur6nen T, Turunen A, Viljanen MK, Ruuskanen O. Diagnosis of pertussis. J Infect 1984;8:149-56. 5. Eisenstein BI. The polymerase chain reaction: a new method of using molecular genetics for medical diagnosis. N Engl J Med 1990;322:178-83. 6. Peter JB. The polymerase chain reaction: amplifying our options. Rev Infect Dis 1991;13:166-71. 7. Glare EM, Paton JC, Premier RR, Lawrence A J, Nisbet IT. Analysis of a repetitive DNA sequence from Bordetella pertussis and its application to the diagnosis of pertussis using the polymerase chain reaction. J Clin Microbiol 1990;28:1982-7. 8. liouard S, Hackel C, lierzog A, Bollen A. Specific identification of Bordetella pertussis by the polymerase chain reaction. Res Microbiol 1989;140:477-87. 9. Van der Zee A, Agterberg C, Peeters M, Schellekens J, Mooi FR. Polymerase chain reaction assay for pertussis: simultaneous detection and discrimination of Bordetella pertussis and Bordetella parapertussis. J Clin Microbiol 1993;31:2134-40. 10. Schl~ipfer G, Senn liP, Berger R, Just M. Use of the polymerase chain reaction to detect Borcletella pertussis in
14.
15.
16.
17.
18.
19.
20.
21. 22.
patients with mild or atypical symptoms of infection. Eur J Clin Microbiol Infect Dis 1993;12:459-63. Douglas E, Coote JG, Parton R, McPheat W. Identification of Bordetella pertussis in nasopharyneal swabs by PCR amplification of a region of the adenylate cyclase gene. J Med Microbiol 1993;38:140-4. Regan J, Lowe F. Enrichment medium for the isolation of Bordetella. J Clin Microbiol 1977;6:303-9. McLafferty MA, Harcus DR, Hewlett EL. Nucleotide sequence and characterization of a repetitive DNA element from the genome of Bordetella pertussis with characteristics of an insertion sequence. J Gen Microbiol 1988;134:2297-306. McPheat W, McNally LT. Isolation of repeated DNA sequence from Bordetella pertussis. J Gen Microbiol 1987; 133:323-30. Aurelius E, Johansson B, Sk61denberg B, Staland A., Forsgren M. Rapid diagnosis of herpes simplex encephalitis by nested polymerase chain reaction assay of cerebrospinal fluid. Lancet 1991;337:189-92. Lipschik GY, Gill VJ, Lundgren JD, et al. Improved diagnosis of Pneumocystis cariniiinfection by polymerase chain reaction on induced sputum and blood. Lancet 1992;340:203-6. Fine PEM, Clarkson JA. The recurrence of whooping cough: possible implications for assessment of vaccine efficacy. Lancet 1982;1:666-9. Kwantes W, Joynson DHM, Williams WO. Bordetella pertussis isolation in general practice: 1977-79 whooping cough epidemic in West Glamorgan. J Hyg 1983;99:149-58. Morgan Mink C, Cherry JD, Christenson P, et al. A search for Bordetella pertussis infection in university students. Clin Infect Dis 1992;14:464-71. Robertson PW, Goldberg H, Jarvue BH, Smith DD, Whybin LR. Bordetella pertussis infection: a cause of persistent cough in adults. Med J Aust 1987;146:522-5. Mertsola J, Ruuskanen O, Eerola E, Viljanen MK. Intrafamilial spread of pertussis. J PEDIATR 1983;103:359-63. Kowk S, Higuchi R. Avoiding false positives with PCR. Nature 1989;339:237-8.
FELLOWSHIPS Fellowships available in pediatric subspecialties and those for general academic pediatric training are listed once a year, in January, in THE JOURNAL OF PEDIATRICS. Each June, forms for listing fellowships available for the academic year beginning 18 months after publication are sent to the Chairman of the Department of Pediatrics at major hospitals in the United States and Canada. In addition, a copy of the application form appears in the July, August, and September issues of THE JOURNAL (please use the current form). Should you desire to list fellowships, a separate application must be made each year for each position. All applications must be returned to M o s b y - Y e a r Book, Inc., by October 15 preceding the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing Department, M o s b y - Y e a r Book, Inc., 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/800-325-4177, ext. 4317, or 314453-4317.