Bordetella pertussis (Pertussis) and Other Bordetella Species

The Pre-Vaccine and Early DTwP Era

During the pre-vaccine era in the US, pertussis was the leading cause of death from communicable disease among children younger than 14 years. During 1934−1943, an average of 200,752 cases and 4034 pertussis-related deaths were reported annually. During the pre-vaccine era in the US, the peak incidence of pertussis was in children 1–5 years of age, with individuals >10 years and infants <1 year each accounting for less than 15% of cases. ^, Subclinical boosting reinfections and higher transplacental protection, respectively, likely were responsible for these relatively spared age groups. Widespread use of a 4-dose series of wPV (DTwPs) in the US beginning in the late 1940s was responsible for a >100-fold decline in incidence from 110 (1922−1948) to 0.5 cases per 100,000 (1981). The 1010 cases and 7 deaths reported in 1976 in the US are the all-time lowest recorded.

The Late DTwP Era

Incidence of pertussis increased in industrialized countries in the mid-1980s and 1990s related to many factors, including use of less potent vaccines, partial control of pertussis reducing frequent boosting reinfections, waning DTwP immunity, lower DTwP coverage, ^, and interruption of universal DTwP vaccination. Much of this decline in vaccine uptake was due to the safety concerns associated with the reactogenic whole cell pertussis vaccine. Despite high DTwP immunization rates in children <7 years in the US, Australia, and several other industrialized countries, pertussis occurred with increasing peaks of the uninterrupted 3- to 5-year cycles and disproportionately in adolescents, adults, and very young infants ( Fig. 162.2 ; see also Fig. 162.1 ). Additionally, new diagnostic methods (PCR and serology) increased the awareness of pertussis and thus more active surveillance likely increased the number of reported cases. In two prospective studies using these comprehensive laboratory methods, the annual incidences of prolonged undifferentiated cough illness due to pertussis in adolescents and adults were estimated to be 370 and 450 per 100,000 population, respectively. ^, In other studies, 12%–32% of adolescents and young adults with prolonged cough illnesses were confirmed to have pertussis. ^, By 1993, 44% of pertussis cases reported to the Centers for Disease Control and Prevention (CDC) occurred in children <1 year of age.

The DTaP/Tdap Era

Because of local and systemic reactogenicity of DTwP, acellular vaccines (aPVs, DTaPs) were developed. DTaP replaced DTwP in the US for the reinforcing fourth and fifth doses in 1992 and for all doses in 1997. Other industrialized countries (Canada, Australia, most European countries, Japan) also changed to aP vaccines, while 64% of countries worldwide continue to use wPVs. The rise in number of cases that began in the DTwP era continued. By 2004, in the US, the 25,827 cases reported was the highest number since 1959 (see Fig. 162.1 ), and 34% of all US cases occurred in adolescents 11–18 years of age (incidence ∼30 per 100,000). Tetanus toxoid, reduced-content diphtheria toxoid, and acellular pertussis antigen vaccine (Tdap) was recommended universally in 2005 in the US at 11–12 years of age. With 53% of US adolescents 13–18 years vaccinated by 2009, there was an approximately 50% reduction in pertussis in the age group of 11–19 years. Several other countries with similar heightened cases in adolescents and adults, including Australia, Canada, Finland, France, and Germany, introduced a Tdap booster with good results. However, a profound shift in epidemiology after 2006 in the US and in many other countries using acellular vaccines from infancy occurred. More than 27,000 cases were reported in the US in 2010, 2012, and 2014. Between 2011 and 2012, every state in the US except California reported an increase in pertussis, and the case count of 48,277 nationally in 2012 was higher than in the early 1950s. In Europe, 48,446 pertussis cases were reported to the European Surveillance System by 30 EU/EEA countries in 2016. Infants <2 months had the highest and most briskly rising age-related risk for pertussis in Europe and the US (US incidence ∼200 per 100,000). From 2001 through 2010, 189 deaths occurred among 27,995 reported infant cases in the US, a case fatality ratio of 6.8/1000 infant cases. This trend was reversed in the US and other countries with the implementation of Tdap vaccination during pregnancy ^{, ,} (see Fig 162.2 ). In some Nordic European countries (Finland, Norway, and Sweden) the pertussis problem has been less severe and these countries have not adapted Tdap vaccinations during pregnancy.

Durability of protection after aPVs—both DTaP and Tdap— is less than following wPVs. Vaccine effectiveness immediately after 5 doses of DTaP is 75%–89% ^, but declines each year thereafter, associated with rapidly falling antibody levels to pertussis toxin, with odds of acquiring pertussis increasing 42% per year in one study. After the last receipt of DTaP at 4–6 years of age, vaccine effectiveness in children 8–12 years of age has been estimated to be only 24%, and one modeling study, assuming 85% initial protection following DTaP, estimated that only 10% of children would be protected 8.5 years later. Similarly, rapid waning of protection after Tdap occurs, with vaccine effectiveness within the first year of >70% but only 34% by 2–4 years. ^, The abrupt return to susceptibility of immunized children aged 7–10 years since the mid-2000s, then of children aged 13–14 years in 2012 and 14–16 years in 2014, as seen in the US, likely is due to rapidly waning protection in the aging cohort of children who received aPV exclusively. ^{, , ,} The critical importance in protection afforded by receiving remotely a priming dose of DTwP was noted dramatically in a pertussis epidemic in Australia (see the section on Immunity). In Nordic European countries (Finland, Norway, and Sweden), pertussis has been relatively well controlled with the active use of DTaP and Tdap boosters both in adolescents and in adults. Pertussis boosters are recommended universally at age 25 years in Finland and for decennial boosters in Norwegian adults. In Finland, social services and healthcare workers who care primarily for children <12 months of age are required by law to receive Tdap. In 2015 the World Health Organization (WHO) concluded that resurgence of pertussis had occurred in 4 of 19 countries, all of which had transitioned to a complete aPV primary immunization series. WHO recommended that countries using DTwP vaccines continue their use for primary immunization.

Pathogen and Pathophysiology

Bordetella are tiny, gram-negative coccobacilli that grow aerobically on starch-blood agar or completely synthetic media with nicotinamide supplemented for growth, amino acids for energy, and charcoal or cyclodextrin resin to absorb fatty acids and other inhibitory substances. All Bordetella spp. share a high degree of DNA homology among virulence genes. However, only B. pertussis expresses pertussis toxin.

Pulsed-field gel electrophoresis (PFGE) of chromosomal DNA, multilocus variable number tandem repeat analysis, multiantigen sequence typing, and whole genome sequencing of B. pertussis isolates from epidemiologic studies show dominance of certain strains and relatedness of epidemiologically linked cases but diversity of strains over time and location with little evidence of geographic clustering. Whole-genome sequencing is the gold standard for strain differentiation and has enlightened the evolution of B. pertussis from a common ancestor similar to B. bronchiseptica but is expensive to perform. Methodology using a single nucleotide primer extension to detect polymorphisms offers a simpler, highly discriminatory alternative. New genetic methods also have detected that in mice in vivo, approximately 30% of all B. pertussis genes were differentially expressed, compared with organisms grown in vitro, and several novel potential vaccines antigens were expressed exclusively in vivo. These kinds of studies will be important for testing new vaccine candidates, especially with the establishment of the human challenge model (see later), and for gaining an increased understanding of the human immune response in pertussis.

The effect of pertussis vaccination on B. pertussis circulation has permitted adaptation, under selective pressure and evolution of circulating strains able to survive in the vaccine-induced waning immunity. The aPVs were developed to include presumed virulence factors of B. pertussis strains. Current aPVs include chemically or genetically inactivated pertussis toxin (PT) and chemically inactivated filamentous hemagglutinin (FHA) and pertactin (PRN) with or without fimbrial (FIM) antigens. Divergence of circulating B. pertussis strains from vaccine strains began during the DTwP era, with genetic adaptations occurring through allelic changes of surface proteins, including PT, PRN, and FIM, and regulation of toxin expression through the toxin promotor gene ( ptxP ). The PT coding subunit ptxA is the most immunogenic of the group, but allelic variants exist. Most vaccines in the US and Europe were derived from strains expressing ptxA2 and ptxA4 ; most currently circulating strains globally have the ptxA1 allele. ^{, , ,} Multiple alleles are described for the toxin promoter ( ptxP ) but ptxP ⁻ , ptxP1 , ptxP2, and ptxP3 are predominant. The ptxP3 largely has replaced the ptxP1 allele worldwide and is associated with 1.6 times greater toxin production compared with ptxP1 strains. The ptxP3 allele also may confer selective advantage of enhanced transmission by delaying an effective immune response and by evading complement-associated opsonophagocytosis through an autotransporter protein called Vag8. The two currently licensed US DTaP vaccines are ptxP1 derived ( Fig. 162.3 ). There have been adaptations of B. pertussis globally, in countries using whole cell vaccine exclusively and where acellular vaccines are used, although vaccine and circulating strains vary by country. B. pertussis circulating in European countries that use aPVs and in the US show molecular typing profiles of vaccine-divergent alleles prn2-ptxA1-ptxP3 almost exclusively. ^{, , , , ,} Allelic changes were not associated with increased notification of pertussis in the US, but in the Netherlands increased cases were temporally associated with upregulation of PT through the ptxP3 promotor mutation. In Finland the use of aPVs has not been associated with major outbreaks even when new strains have appeared. Widespread use of the adolescent and adult Tdap boosters appears to have induced a large population level of protection. ^{, ,} Fimbrial alleles show more variability, with fim3 predominating and, in the US, transitioning from fim3A to fim3B only in the past decade during exclusive use of acellular vaccines.

Since the exclusive use of aPVs beginning in the late 1990s, PRN-deficient strains have emerged in many countries, including the US, Australia, Canada, Japan, France, Finland, and Spain. ^, PRN-deficient B. pertussis strains were first reported in the US among 11 of 12 isolates from infant cases in Philadelphia in 2011 and 2012 ; the earliest PRN-deficient isolate in the Philadelphia collection (2007 onward) was from 2008. A CDC study of isolates since 1935 showed the first archived isolate of a PRN-deficient strain from 1994, with reappearance only in 2010 and rapid increase since. PRN-deficient B. pertussis has become predominant in outbreaks, such as in Washington State in 2012, as well as in sporadic cases. PRN-deficient strains increased rapidly in Australia from 2008 to 2012, with continued expansion of PRN-deficient ptxP3 strains of multiple lineages since. In Ontario, Canada a PRN-deficient strain was first noted in 2011, with its prevalence increasing to 71% in 2016 and decreasing to 46% in 2017. After predominance of PRN-deficient strains in Japan following the introduction of aPVs, prevalence has decreased, associated with the absence of PRN in 2 of 3 current vaccines. In a retrospective study of 753 pertussis cases with isolates studied from 8 states in the US during 2011 and 2013, patients who had received at least one dose of pertussis vaccine had an adjusted odds ratio (aOR) of 2.2 (95% confidence interval [CI], 1.3−4) for having a PRN-deficient isolate, suggesting possible selective advantage. Unlike other allelic changes, PRN-deficient strains have emerged only in countries using aPVs and demonstrate multiple different gene mutations, suggesting that a number of recombinant events occurred independently and did not just expand on a single clone globally. ^{, , ,} It is noteworthy that PRN-induced antibodies persist longer than do PT-induced antibodies. Mutations in PRN genes likely create an advantage for B. pertussis in the waning phase of immunity when PT immunity has decreased, more adapting to the waning phase rather than totally “escaping” vaccine-induced immunity.

Although PRN is a B. pertussis virulence factor, deficient organisms retain virulence in the mouse ^{, ,} and primate models. In a mouse model, Hegerle found that PRN-deficient isolates resist antibody-mediated cytotoxicity in vitro and that acellular pertussis immunization may provide a fitness advantage for sustained infection. ^, Infants with pertussis due to PRN-deficient and PRN-sufficient B. pertussis strains manifest similar illnesses, with no apparent difference in virulence. ^{, ,}

Bordetella organisms elaborate a number of biologically active substances. The roles of each in pathogenesis and protection have been reviewed by Hewlett ( Table 162.1 ).

PT causes lymphocytosis in experimental animals by preventing their migration from the circulating blood pool. The injection of PT intravenously into adult humans, however, was reported not to be associated with lymphocytosis. Severe or fatal pertussis in infants has been correlated with degree of lymphocytosis, a likely manifestation of toxin activity. ^, There is no evidence of neurotoxic effects of B. pertussis . Although a hydrogen peroxide–detoxified monocomponent PT vaccine protected children in the short term against severe pertussis, and maternal immunization of the pregnant baboon with a monocomponent PT vaccine protects B. pertussis -challenged young offspring from disease, multiple lines of evidence support a central but not singular role for PT in disease.

Transmission

Pertussis is highly contagious, with attack rates of >90% in susceptible individuals exposed at close range to aerosol droplets generated by coughing or sneezing. After intense exposure, as in households and childcare facilities, the rate of subclinical infection or mild disease exceeded 50% in fully DTwP immunized or naturally immune individuals. ^, The role of asymptomatically infected people in transmission is unsettled. When carefully sought, a symptomatic source case (especially a sibling or caretaker) can be found for most infants. ^, Chronic carriage is not documented. Vaccination reduces transmissibility of B. pertussis, even when vaccinated individuals develop symptomatic infection. The baboon model of B. pertussis infection of Warfel, Merkel and colleagues has led to a better understanding of transmission, colonization dynamics, disease, short-term immunity, and vaccine effects. These valuable experiments have been reviewed and are summarized here. A B. pertussis challenge of an infection-naïve baboon leads to dense and prolonged colonization, disease, transmission, and then protection (at least in the short term) from colonization and disease upon re-challenge. Immunization with the acellular vaccine is associated with protection from disease but permits dense and prolonged B. pertussis colonization, which is transmissible. Immunization with whole-cell vaccine leads to an intermediary state, with protection from disease and dense colonization that is more rapidly cleared than following receipt of acellular vaccine but that does persist for a number of days.

Immunity

Neither disease nor DTwP, DTaP, or Tdap provides complete or lifelong immunity against reinfection or disease. There is no generally accepted laboratory correlate of protective immunity. Although vaccine-induced antibody responses were the focus of clinical efficacy and licensing studies of acellular vaccines in the 1990s, the resurgence of pertussis and the knowledge that acellular vaccines do not limit colonization and transmission have stimulated study of the complex immunologic response to B. pertussis following infection, vaccination, and re-vaccination. Innate and adaptive immunologic responses, systemic and local antibody, and cell-mediated immunity likely play important roles and can be more comprehensively measured with current immunologic techniques. ^{, ,} Human, baboon, and mouse studies support a critical role of antibody-independent cell-mediated immune responses in long-term protection and elimination of NP colonization. Priming (the initial challenge) by infection or wPV polarizes a more Th1/Th17 CD4 ⁺ T-lymphocyte cellular response, whereas the acellular vaccine induces a more predominant Th2 polarized response with a lesser Th17 and Th1 response. ^, Natural infection and whole cell vaccines induce antibodies of IgG1, IgG2 and IgG3 subclasses (typical of Th1), whereas aPs evoke IgG1 and IgG4 antibodies (typical of Th2). Th17 response also is associated with clearance of extracellular organisms from mucosal surfaces. Brisk B-lymphocyte–related antibody responses follow infection and both wPV and aPV vaccination, with more rapid waning after aPV. Antibody response following aPVs is limited to those 3–5 antigens included in the vaccines when compared with >3000 antigens included in the wPVs. The initial molecular programming of priming is maintained upon re-exposure to aPVs or B.pertussis . Th1 and Th17 responses lead to more durable protection and the cellular response can be boosted, whereas Th2 response wanes rapidly and fails to boost broadly. In what is known as linked epitope suppression, memory B lymphocytes of aPV-primed individuals, because of their high numbers and affinity for receptors, may out-compete naïve B lymphocytes for access to broader or novel B. pertussis epitopes. The observation of enhanced susceptibility to pertussis during an epidemic in Australia among individuals who remotely had initial receipt of an aPV versus wPV ( Fig. 162.4 ) and inferior antibody responses following infection or Tdap boosting associated with remote priming with aPV support these concepts and the permanence of polarization. ^{, ,}

Experimental animal studies (mice, rabbits, guinea pigs, newborn piglets) and now a challenge model of baboons have increased the understanding of pathogenesis and transmission of pertussis, but questions in immunology remain. A human challenge model was developed in Southampton (UK) to study colonization and immunology. Among challenged healthy study volunteers prescreened to have undetectable or low levels of serum IgG against pertussis toxin, intranasal inoculum of 10 ⁵ colony forming unit of live B. pertussis led to colonization in 80%. The challenge was safe, only minor respiratory symptoms were reported, and all the subjects received azithromycin at day 14 after inoculum. Anti-PT antibodies developed in 9 of the 19 colonized subjects. Nasal wash was more sensitive to detect colonization than nasal swabs. This human colonization model should provide valuable information about both mucosal and systemic immunity and the potential for the development and evaluation of future vaccines and therapies.