Laboratory diagnosis of pertussis: A survey on provincial public health laboratory methods

Abstract

Background: Pertussis, a vaccine preventable respiratory illness caused by the bacterium Bordetella pertussis (B. pertussis), has been a nationally reportable disease in Canada for over 100 years; however, cases resurged in Canada and globally in 2023–2024.

Objective: To examine the breadth and depth of pertussis strain surveillance currently being carried out across Canada.

Methods: A survey was sent to all ten provincial public health laboratories inquiring how pertussis was diagnosed or identified in the laboratory, including the polymerase chain reaction (PCR) diagnostic methods, bacteriological culture, identification and strain characterization such as molecular typing and antibiotic susceptibility testing.

Results: Nine of the ten provincial laboratories provided responses. Five provincial laboratories reported performing bacteriological culture, and some only from specimens that tested positive by PCR. Long-term storage of submitted and historical specimens took place in six laboratories. Identification of B. pertussis was commonly done through matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) analysis, though immunochemical and PCR-based methods were also used. One laboratory conducted antibiotic susceptibility testing in specific circumstances. No laboratory performed fimbriae serotyping or examined expression of other pertussis vaccine antigens. One laboratory used whole-genome sequencing for outbreak investigation. The PCR diagnostics were performed in eight of the responding laboratories and always include IS481 and pIS1001 gene targets. Some laboratories also reported using other gene targets to identify and distinguish between B. pertussis, B. parapertussis, B. holmesii and B. bronchiseptica.

Conclusion: Given the global increase in pertussis, with the emergence of macrolide-resistant and pertactin-deficient strains, strain characterization should be added to the Canadian national pertussis surveillance program.

Introduction

Pertussis, which is commonly referred to as whooping cough, is caused by the bacterium Bordetella pertussis (B. pertussis) ^{Footnote 1}. Pertussis has been a notifiable disease in Canada since 1924 ^{Footnote 2}. Surveillance of pertussis is done by provinces and territories and they in turn voluntarily report cases to the Public Health Agency of Canada (PHAC)’s Canadian Notifiable Disease Surveillance System. Data collected and validated by the Canadian Notifiable Disease Surveillance System are published annually online in the Notifiable Diseases Online website ^{Footnote 3}. The PHAC also supports the Canadian Immunization Program ACTive (IMPACT) to carry out hospital-based surveillance of paediatric pertussis cases ^{Footnote 4}.

For control of pertussis, an inactivated whole-cell vaccine was introduced in Canada in 1943, which was subsequently replaced by the adsorbed whole-cell vaccine from 1981 to 1985. Eventually, acellular pertussis vaccine was introduced in 1997 because of its less reactogenic nature ^{Footnote 5}. Despite having a vaccine, pertussis continues to cause disease in infants, young children, adolescents and adults for various reasons related to vaccine hesitancy, waning of vaccine-induced protective immunity and divergence of current circulating and vaccine strains ^{Footnote 6}.

Between 2020 and 2022, COVID-19 pandemic restrictions caused disruptions in social gatherings, which resulted in lower numbers of different respiratory infections, including pertussis ^{Footnote 7}. As with other respiratory infections ^{Footnote 8Footnote 9}, pertussis appears to have re-emerged after the pandemic to cause more cases than before the pandemic ^{Footnote 10Footnote 11}. Data from 29 European Economic Area countries reported 1,578 and 2,623 cases of pertussis in 2021 and 2022, respectively ^{Footnote 12Footnote 13}. In 2023, the number of pertussis cases reported by these countries jumped to over 25,000 cases; and in the first three months of 2024, more than 32,000 cases were reported ^{Footnote 14}. In China, pertussis increased in 2022 and 2023 with 39,781 and 38,205 cases reported, respectively ^{Footnote 15}. In the first two months of 2024, 32,380 cases, including 13 deaths, were recorded ^{Footnote 16}. Furthermore, the disease has shifted from mainly affecting infants to older children, and a new strain showing macrolide resistance and pertactin deficiency emerged ^{Footnote 17}. In South Korea, a national epidemic was described in 2024 with the highest incidence rate found in those aged 13 years, with 526.2 cases per 100,000 population ^{Footnote 18}. In addition, the macrolide-resistant strain appeared to have spread to some Asian countries ^{Footnote 19}. Besides monitoring for susceptibility to macrolides, the European Centre for Disease Prevention and Control (ECDC) also recommends performing serotyping, multi-locus antigen sequence typing and vaccine antigen expression by enzyme-linked immunosorbent assay (ELISA) or gene sequencing ^{Footnote 20}. Similar strain surveillance programs are also available at the United States (US) Centers for Disease Control and Prevention and the United Kingdom Health Security Agency.

In 2024, the Pan American Health Organization also reported increases in pertussis in multiple countries within the Americas including the US, Brazil, Mexico and Peru ^{Footnote 21}. For example, preliminary data in the US show that reported cases of pertussis in 2024 increased six-fold when compared to 2023 ^{Footnote 22}. Canada was no different from other nations, reporting increases in pertussis cases in multiple jurisdictions leading to significant case numbers not seen since prior to the introduction of the pertussis vaccines. The increases in pertussis cases occurred in nearly all provinces and territories ^{Footnote 23Footnote 24Footnote 25Footnote 26Footnote 27Footnote 28Footnote 29} (personal communication Dr. Paul Van Caeseele, March 31, 2025). For the first eleven months in 2024, Public Health Ontario reported 1,634 pertussis cases, with 1,396 as confirmed and 239 probable cases. This led to the highest incidence rates since 2017 in those younger than one year old and those between the ages of 10 and 14 years old (74.2 and 55.2 per 100,000, respectively). According to Québec public health, 16,130 cases of whooping cough were recorded January 1–October 9, 2024. In Newfoundland and Labrador, 230 confirmed cases of pertussis were recorded throughout the province by September 10, 2024.

Given the global resurgence of pertussis after the COVID-19 pandemic, with the emergence of macrolide-resistant and vaccine antigen-deficient strains ^{Footnote 30Footnote 31}, the capacity in Canada for a national surveillance program that includes strain characterizations and antibiotic susceptibility testing must be re-examined. It is with this understanding that the National Microbiology Laboratory Branch of the PHAC is working with provincial and territorial public health laboratory partners to examine how pertussis is diagnosed and characterized in Canada in order to understand the breadth and depth of pertussis surveillance in the country.

Methods

This study was intended to obtain details on strain characterization work, which can contribute to the Canadian national surveillance of pertussis. Strain characterizations are often done at the provincial public health laboratories, which also serve as reference laboratories for their provinces as well as neighboring territorial governments. The PHAC’s National Microbiology Laboratory Branch has a close working relationship with the public health agencies in all provinces and territories as partners in public health microbiology issues. This relationship is formalized as a national association of public health laboratory professionals, established in 2001 as the Canadian Public Health Laboratory Network, with a role to provide rapid, coordinated and unified laboratory response to emerging and re-emerging infectious diseases ^{Footnote 32}. As such, frontline and hospital laboratories as targets of the survey were not included.

Therefore, on November 17, 2024, an e-mail was sent to the medical microbiologists or medical directors of ten provincial public health laboratories explaining the purpose of the survey and a questionnaire with questions covering bacteriological culture of B. pertussis, polymerase chain reaction (PCR) diagnostic method and strain characterization. For strain characterization, the following questions were asked: 1) how long are cultures preserved; 2) how cultures are identified and characterized including serotyping (for expression of fimbriae antigens); 3) expression of other vaccine antigens; 4) molecular typing, and 5) antibiotic susceptibility testing. For PCR diagnostic methods, details of the method used (commercial kit/platform or laboratory developed method), the gene targets detected in the PCR assays and positivity cut-off values were included in the questionnaire. A copy of the survey questionnaire can be found in the Appendix and participants had up to December 31, 2024 to respond voluntarily.

Since the survey did not request any personal information and was done within the autonomy of the Canadian Public Health Laboratory Network for public health purposes, institutional research ethics approval was not sought nor was informed consent necessary as the response was voluntary. Data protection was only applied with laboratories identified by numbers instead of naming the laboratory linked to the data captured in the survey. The survey questions were based on questions received by the National Microbiology Laboratory Branch from public health practitioners across the country, such as antibiotic resistance patterns and strain types found.

Results

Responses to the questionnaire were received from nine of the ten provincial public health laboratories. One provincial public health laboratory did not respond to the survey.

Culture of Bordetella pertussis and subsequent testing on Bordetella pertussis isolates

Five provincial public health laboratories performed bacteriological culture of primary specimens for isolation of B. pertussis, although one of them reported doing so only rarely in recent years. Three of these five provincial laboratories disclosed that only specimens tested positive by PCR for B. pertussis were subjected to culture, including one of these laboratories only performed culture on specimens that were PCR-positive with low cycle threshold (Ct) values. Six provincial laboratories reported long term (on the scale of years) storage of B. pertussis isolates; although some laboratories had stopped performing primary cultures, historical isolates were preserved and stored. Two provincial laboratories also reported storage of primary specimens submitted for B. pertussis testing at −80°C.

Among those performing primary bacteriological culture and/or performing bacterial identification as a provincial public health reference laboratory, six laboratories (including one laboratory that did not provide primary culture service but received samples for reference diagnostic identification testing) identified B. pertussis by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF). One of these six laboratories also used biochemical tests such as oxidase, motility, growth on MacConkey agar, and genetic sequencing for identification of B. pertussis and other uncommon Bordetella species, while another laboratory also used bacterial agglutination and fluorescent antibody tests for identification of B. pertussis and B. parapertussis.

Only one laboratory reported performing antibiotic susceptibility testing for B. pertussis specimens when requested by clinicians. No laboratory was reporting any phenotypic or genetic typing, including looking at expression of vaccine antigens or sequencing of vaccine antigen genes. Only one laboratory reported performing whole-genome sequencing on B. pertussis and B. parapertussis for outbreak investigation. The scope of bacteriological culture from primary specimens, and the subsequent testing to characterize the strains in the different provincial public health laboratories are summarized in Table 1.

Polymerase chain reaction diagnosis of pertussis

Three provincial public health laboratories used commercially available test kits or platforms for the laboratory diagnosis of pertussis: the R-Biopharm AG RIDA®GENE Bordetella ^{Footnote 33}, the Diasorin Simplexa™ Bordetella Direct Test ^{Footnote 34} and the QuidelOrtho Corporation’s Solana® Bordetella Complete Assay ^{Footnote 35}. Five other provincial laboratories used laboratory developed tests (LDT), and while the gene targets detected by specific primer-probe sets in these LDTs may differ, they invariably included IS481 and pIS1001 (for the simultaneous detection and differentiation or identification of B. pertussis and B. parapertussis). Some laboratories also employed additional gene targets such as hIS1001 (for detection of B. holmesii), BHrecA (for detection of B. holmesii), bfrZ (for detection of B. bronchiseptica) and IS1002 (for detection of B. parapertussis). The gene targets used in the PCR assays by the various laboratories and their ability to identify and differentiate common Bordetella species are summarized in Table 2. Five laboratories that used LDTs also reported using Ct values to interpret the PCR results. Positive PCR results for pertussis were defined by Ct values ranging from less than or equal to 35 to less than 45 (Table 2).

Discussion

Nine of the ten provincial public health laboratories responded to this survey and the data obtained formed the basis of this report. Of the eight provincial public health laboratories that provide diagnostic PCR assays, five are able to detect B. pertussis, B. parapertussis and B. holmesii, and three were able to detect only B. pertussis and B. parapertussis. Thus, not all laboratories could definitively identify these three most important species by PCR. Bacteriological culture to recover viable B. pertussis from clinical specimens was only performed in five of the responding laboratories, and often only on PCR-positive clinical samples. The MALDI-TOF was the most common method used to identify B. pertussis cultures (used by all laboratories that provide this service of reference diagnostic/identification); however, only one laboratory carried out antibiotic susceptibility testing by E-test when requested by clinicians. Furthermore, none of the laboratories carried out routine typing of isolates to understand characteristics of the circulating B. pertussis strain. Therefore, current laboratory results do not contribute additional information towards the understanding of how strain characteristics may alter the epidemiology of pertussis in Canada.

At this time, the national surveillance of pertussis depends on provinces and territories reporting cases with minimal demographic information, including age and gender. The IMPACT surveillance of pertussis focuses on hospitalized cases with some additional clinical information ^{Footnote 4}. One deficient area is the lack of information on the B. pertussis bacteria currently circulating in Canada, which is needed to understand if the increase in Canadian pertussis cases in 2024 was due to expansion of a clonal strain or to diverse strain types. Resistance to oral macrolide antibiotics, such as erythromycin or clarithromycin, has been reported elsewhere and may also have emerged in circulating Canadian strains. It is with this understanding that the laboratory methods currently being used by our provincial partners for the surveillance of this highly contagious bacterial disease was reviewed.

Like for other common pathogens, strain characterization is an important component to an overall understanding of the evolving nature of the changing epidemiology of pertussis in Canada and globally. For example, once mainly a childhood disease, it now appears that depending on the locality, a noticeable proportion of cases occur either in older children, adolescent, adults or elderly. Several studies in the late 1990s and in early 2000s have described genetic polymorphisms in the vaccine antigen genes (e.g., pertactin and pertussis toxin) leading to the suggestion of divergence between current circulating B. pertussis strains and the strains used to manufacture vaccines ^{Footnote 36Footnote 37Footnote 38Footnote 39Footnote 40}. Genetic polymorphisms in the fim3 gene that encode fimbriae 3 or serotype 3 antigen have also been observed. In Canada, since the 1970s, the majority of B. pertussis isolates examined expressed the fimbriae 3 antigen but rarely the fimbriae 2 antigen. Additionally, non-synonymous mutations of the fim3 gene resulted in amino acid changes on a surface epitope of the fimbriae antigen led to the postulation that the polymorphisms may be subjected to immune pressure or selection from vaccine induced or naturally occurring immune response ^{Footnote 41}. Recent studies have also shown that genetic changes that favour the increase production of pertussis toxin (such as having the ptxP3 allele) have been associated with pertussis resurgence ^{Footnote 42}. Also, recent global B. pertussis isolates are frequently deficient in the expression of the pertactin antigen, including isolates in Canada ^{Footnote 43Footnote 44Footnote 45Footnote 46Footnote 47Footnote 48}. Furthermore, over a period of nine years, B. pertussis samples recovered from one province had the potential to fluctuate between different genotypes and expression of the vaccine antigen pertactin ^{Footnote 49}.

Although it is not fully understood 1) how these genetic polymorphisms within the vaccine antigen genes or 2) how strains not expressing certain vaccine antigens (such as Canadian isolates not expressing fimbriae 2 and pertactin antigens) would affect vaccine efficacy, it is expected that further antigenic drift away from the B. pertussis vaccine strains and their encoded antigens will negatively affect vaccine efficacy. Therefore, this highlights the importance of incorporating strain characterization into the overall pertussis surveillance program in Canada. Strain characterization may also mitigate potential emergence of antibiotic resistance, as in the case of a large increase in the number of pertussis cases due to the emergence of an erythromycin-resistant B pertussis strain in China ^{Footnote 50}.

In Canada, as in many other countries, most pertussis cases are diagnosed in the laboratory by PCR. This is likely because B. pertussis is a nutritionally fastidious bacteria that requires specialized and enriched culture media to support growth. The commonly used culture media for bacteriological isolation from primary specimens include both the Bordet-Gengou agar and the Regan-Lowe charcoal blood agar, which contain starch and/or charcoal to neutralize toxic fatty acids and peroxides, defibrinated horse or sheep blood, and sometimes antibiotics such as cephalexin to inhibit normal flora present in the nasopharynx. Also, bacteria can only be recovered during the first two weeks of cough, while PCR would remain positive for up to three or four weeks after disease onset. As such, bacterial culture has become less popular in frontline laboratories, which are increasingly adopting PCR assays that can simultaneously detect and identify a number of respiratory pathogens (e.g., BioFire Respiratory Panel [RP]2.1-EZ, which can detect up to 19 respiratory pathogens including B. pertussis and B. parapertussis). Also, the practice of PCR diagnostics for pertussis may varies by province (e.g., more reliance on provincial public health laboratory to provide this service in some provinces versus decentralized testing in others). Similarly, the ability to detect Bordetella species (e.g., B. holmesii and B. bronchiseptica) that may also cause cough symptoms varies across the country.

Not all the commercially available test kits and platforms or LDT for detection of pertussis by nucleic acid amplification testing (NAAT) can detect and differentiate between pertussis-causing B. pertussis and pertussis-symptoms like causing B. parapertussis and B. holmesii. To detect and differentiate these three Bordetella species, a NAAT needs to have specific primers and/or probes that target these three species separately ^{Footnote 51Footnote 52}. One such specific gene target for B. pertussis is ptxA. Bordetella bronchiseptica may also cause respiratory infections with cough in immunocompromised individuals ^{Footnote 53}. For detection and identification of B. bronchiseptica, yet another set of primers and/or probes would be required ^{Footnote 54Footnote 55}. The challenge in using NAAT with minimum numbers of primer pairs and probes is due to the fact that, for example, B. holmesii has been reported to harbor low copy numbers of IS481 ^{Footnote 56} while B. bronchiseptica has been reported to harbor low copy numbers of IS481 and IS1001 ^{Footnote 57}. The IS481 and IS1001 are often used in PCR assays to detect B. pertussis and B. parapertussis, respectively.

While NAAT may be able to detect and differentiate or identify important Bordetella species involved in causing respiratory infections in human, the sensitive nature of this laboratory method may require some additional considerations in the interpretation of the test result. First, results from a NAAT for pertussis must be interpreted in the context of the clinical setting. For example, in Canada, a laboratory-confirmed case of pertussis is defined either by bacteriological isolation of B. pertussis or by detection of B. pertussis DNA from an appropriate clinical specimen, together with at least one compatible clinical findings of either cough lasting for two weeks or longer, paroxysmal cough of any duration, cough with inspiratory “whoop” or cough ending in vomiting or gagging, which may be associated with apnea ^{Footnote 58}. Secondly, contamination from the environment with the organism or its DNA may introduce potential false positive results. For example, pseudo-outbreaks of pertussis have been described in the literature ^{Footnote 59Footnote 60Footnote 61}. Therefore, the US Centers for Disease Control and Prevention also recommends that culture confirmation of at least one case should occur during the setting of a suspected pertussis outbreak ^{Footnote 62}.

In summary, maintaining B. pertussis culture capacity as well as strain identification and characterization including antibiotic susceptibility testing should remain available at either the provincial and/or national level pending further discussions on the most cost-effective surveillance program that meets the need of Canadian public health. Previously National Microbiology Laboratory has performed serotyping using monoclonal antibodies to B. pertussis fimbriae 2 and fimbriae 3, western immunoblot to detect the expression of vaccine antigen pertactin and genotyping of vaccine antigens genes, including pertussis toxin subunit A and pertussis toxin promoter region, fimbriae 3, pertactin and filamentous hemagglutinin ^{Footnote 41Footnote 43Footnote 49}.

In 2002, the Government of Canada organized a national consensus conference on pertussis with recommendations concerning laboratory diagnosis and surveillance ^{Footnote 63}. As a follow up, the National Microbiology Laboratory gathered Canadian experts on the subject of pertussis in a workshop to discuss the recommendations on laboratory diagnosis and surveillance from the national consensus conference. One of the action plans from this workshop was to set up a Working Group to discuss implementing a national strain characterization program ^{Footnote 64}. Due to competing priorities, this plan was put on hold, but in view of global resurgence of pertussis, the concern of antibiotic resistance and circulation of strains not expressing vaccine antigens, it may be time to re-examine this plan and put it into action.

Limitations

Limitations of this study include not sending this survey to frontline laboratories (including private medical laboratories), hospital laboratories and regional public health laboratories. Therefore, important laboratories may have been missed that may be providing bacteriological cultures for the laboratory diagnosis of pertussis, and also the overall PCR technology being offered for detection of pertussis, including the popular molecular diagnostic platform like the BioFire Respiratory Panel [RP]2.1-EZ for simultaneous detection and identification of 19 respiratory viral and bacterial pathogens including B. pertussis and B. parapertussis. To reach out to all the frontline medical and hospital laboratories would have been a big undertaking. Nevertheless, six regional laboratories were contacted through one provincial public health laboratory to gather frontline information for comparison. None of these six responding laboratories reported providing bacteriological culture for pertussis; two laboratories use BioFire for the detection of B. pertussis and B. parapertussis (including one of these two laboratories also having an LDT) and one laboratory uses a LDT that targets only IS481. None of these six laboratories provided any testing to identify, type or determine antibiotic susceptibility for B. pertussis. This is likely because, at least in some provinces, testing for B. pertussis may be regarded as specialized testing and therefore perceived as best handled at the provincial public health reference laboratories.

Conclusion

This survey has reviewed the current situation of pertussis in Canada and elsewhere globally, summarized the laboratory diagnostic procedures used in the Canadian Public Health Laboratory Network, identified some gaps in the national surveillance system and made recommendations to eliminate gaps identified. While PCR assays to detect B. pertussis is widely available across the country, culture capability may be more limited to some larger provinces. Routine strain typing that can inform strain characteristics such as expression of vaccine antigens, genetic polymorphisms that may affect a mismatch between the vaccine strains and current circulating strains and antimicrobial resistance data are currently lacking. Maintaining some bacteriological culture and strain characterization capability is essential for understanding effects of vaccine induced immune pressure on B. pertussis. A system to collect data representative of national distribution of strain types (including antimicrobial resistance) is essential to understand the evolving nature of pertussis and to prepare for potential need of newer vaccine strain. A sentinel surveillance system including collection of strain typing data coupled with epidemiological information is recommended for a comprehensive understanding of the current epidemiology of pertussis in Canada. Moving forward, strain collection and characterization with antibiotic susceptibility monitoring should be included in a sentinel surveillance system to understand the evolving nature of B. pertussis under national infant, adolescent and maternal pertussis immunization programs.

Authors' statement

CM — Writing–original draft, writing–review & editing, investigation, data curation
JG — Resources, writing–review & editing, methodology
GT — Resources, writing–review & editing, methodology
JM — Resources, writing–review & editing, methodology
PVC — Resources, writing–review & editing, methodology
JK — Resources, writing–review & editing, methodology
BL — Resources, writing–review & editing, methodology
TH — Resources, writing–review & editing, methodology
GD — Resources, writing–review & editing, methodology
LJ — Resources, writing–review & editing, methodology
HP — Resources, writing–review & editing, methodology
RT — Conceptualization, methodology, writing–original draft, writing–review & editing, investigation, supervision

Competing interests

None declared.

ORCID numbers

Jessica Minion — 0000-0002-8863-5697
Julianne Kus — 0000-0001-6033-7244
Todd Hatchette — 0000-0002-5377-2528
Raymond Tsang — 0000-0003-1140-402X

Acknowledgements

The authors thank the provincial and territorial laboratories for providing their responses to this survey. We also thank Gabriella DeAngelis for her assistance with results tabulation.

Funding

None.

Appendix

List 1: Questionnaire on laboratory diagnosis of Bordetella pertussis infection

1. Does your laboratory (province) perform bacterial culture of Bordetella pertussis and other Bordetella species? Yes; No
2. a) Describe the PCR method (and PCR targets) your laboratory (province) performs for detection of pertussis?
   b) Is the assay able to differentiate between B. pertussis, B. parapertussis, and B. holmesii?
   c) If your laboratory is performing qPCR: Yes; No; what is the Ct value used to determine a positive result (presence of pertussis)?
3. a) Subsequent to culture, what method do you use to identify it as B. pertussis and not other Bordetella species?
   b) Does your laboratory (province) perform phenotypic and/or molecular typing: Yes; No
   c) Does your laboratory test for antibiotic susceptibility? Yes; No
   1. If yes, by what method: disk diffusion ( ); E-test ( ); micro-broth dilution method for MIC ( ); agar dilution method for MIC ( ).
   d) How long do you keep your positive cultures? Months; Years

• Previous Page
• Table of contents
• Next page