The genus Bordetella includes three primary human pathogens: Bordetella bronchiseptica, B. pertussis, and B. parapertussis. B. bronchiseptica is reviewed in Chapter 25 because it grows on MacConkey agar. Although B. parapertussis also can grow on MacConkey agar, it is discussed with B. pertussis in this chapter for two reasons: B. pertussis and B. parapertussis both cause human upper respiratory tract infections, with almost identical symptoms, epidemiology, and therapeutic management; and optimal recovery of both organisms from respiratory specimens requires the addition of blood and/or other suitable factors to culture media. Additional Bordetella species may cause rare asymptomatic infections in immunocompromised patients; these include B. hinzii, B. holmesii, B. petrii, and B. trematum. (See the chapter cross-references in the preceding table for information on organisms not discussed in this chapter.)

General Characteristics

General features of Bordetella spp. other than B. pertussis and B. parapertussis are summarized in Chapter 25. In contrast to B. bronchiseptica, B. pertussis and B. parapertussis are nonmotile and infect only humans. In the evolutionary process, these exclusive human pathogens have a close genetic relationship. They remain separate species based on their differences in pathogenesis and host range.

Epidemiology and Pathogenesis

Epidemiology

Before the introduction of the vaccine (and currently in nonimmunized populations), pertussis (whooping cough) periodically became an epidemic disease that cycled approximately every 2 to 5 years. Transmission occurs person to person through inhalation of respiratory droplets. Humans are the only known reservoir.

Pertussis is a highly contagious, acute infection of the upper respiratory tract caused primarily by B. pertussis and less commonly by B. parapertussis. The latter agent generally has a less severe clinical presentation both in duration of symptoms and in the percentage of identified cases. Recently, B. holmesii was reported to cause a pertussis-like illness, but little is known about the biology, virulence mechanisms, and pathogenic significance. Pertussis was first described in the sixteenth century and occurs worldwide, totaling about 48.5 million cases annually. Although the incidence has decreased significantly since vaccination became widespread, outbreaks of pertussis occur periodically. B. pertussis infections appear to be endemic in adults and adolescents, most likely because of waning vaccine-induced immunity; these infections may serve as the source of the epidemic cycles involving unvaccinated or partially immunized infants and children.

Pathogenesis

B. pertussis, the primary pathogen of whooping cough, uses several mechanisms to overcome the immune defenses of healthy individuals. The mechanisms are complex and involve the interplay of several virulence factors (Table 37-1). Some factors help establish infection; others are toxigenic to the host; and still others override specific components of the host’s mucosal defense system. For example, when B. pertussis reaches the host’s respiratory tract, its surface adhesins attach to respiratory ciliated epithelial cells and paralyze the beating cilia by producing a tracheal cytotoxin. A major virulence factor, pertussis toxin (PT), is produced by the attached organism. PT enters the bloodstream, subsequently binding to specific receptors on host cells. After binding, PT disrupts several host cell functions, such as initiation of host cell translation; inability of host cells to receive signals from the environment causes a generalized toxicity. The center membrane of B. pertussis blocks access of the host’s lysozyme to the bacterial cell wall via its outer membrane. B. pertussis and B. parapertussis share a nearly identical virulence control system encoded by the bvgAS locus that is responsive to variation in environmental conditions. Because of this very complex system, Bordetella organisms appear to be able to alter phenotypic expression, enhancing transmission, colonization, and survival.

TABLE 37-1

Major Virulence Determinants of Bordetella pertussis

Function	Factor/Structure
Adhesion (auto transporters)	Fimbriae (FIM), types 2 and 3: Serotype-specific agglutinins for colonization of respiratory mucosa. Filamentous hemagglutinin (FHA): Mediates adhesion to the ciliated upper respiratory tract Pertactin (PRN): Mediates eukaryotic cell binding and is highly immunogenic. Tracheal colonization factor Brk A *
Toxicity	Pertussis toxin (encoded by the ptx gene, an A/B toxin related to cholera toxin): Induces lymphocytosis and suppresses chemotaxis and oxidative responses in neutrophils and macrophages Adenylate cyclase toxin: Hemolyzes red cells and activates cyclic adenosine monophosphate, thereby inactivating several types of host immune cells Dermonecrotic toxin (exact role unknown) Tracheal cytotoxin (ciliary dysfunction and damage) Endotoxin (lipopolysaccharide) Type III secretion ^†
Overcome host defenses	Outer membrane: Inhibits host lysozyme Siderophore production: Prevents host lactoferrin and transferrin from limiting iron

^*Plays a role in pathogenesis by conferring serum resistance.

^†This type of secretion allows Bordetella organisms to transport proteins directly into host cells; it is required for persistent tracheal colonization.

Spectrum of Disease

Several factors influence the clinical manifestations of B. pertussis (Box 37-1). Classic pertussis is usually a disease of children and can be divided into three symptomatic stages: catarrhal, paroxysmal, and convalescent. During the catarrhal stage, symptoms are the same as for a mild cold with a runny nose and mild cough; this stage may last several weeks. Episodes of severe and violent coughing increase in number, marking the beginning of the paroxysmal stage. As many as 15 to 25 paroxysmal coughing episodes can occur in 24 hours; these are associated with vomiting and with “whooping,” the result of air rapidly inspired into the lungs past the swollen glottis. Lymphocytosis occurs, although typically the patient has no fever and no signs and symptoms of systemic illness. This stage may last 1 to 4 weeks.

Box 37-1 Factors Known to Affect the Clinical Manifestation of Bordetella pertussis Infection

• Patient’s age

• Previous immunization or infection

• Presence of passively acquired antibody

• Antibiotic treatment

In addition to classic pertussis, B. pertussis can cause mild illness and asymptomatic infection, primarily in household contacts and in a number of unvaccinated and previously vaccinated children. Since the 1990s, a shift in the age distribution of pertussis cases to adolescence and adults has been observed in highly vaccinated populations. Adults and adolescents are now recognized as a reservoir for transmitting infection to vulnerable infants. Among these immunized individuals, a prolonged cough may be the only manifestation of pertussis; a scratchy throat, other pharyngeal symptoms, and episodes of sweating commonly occur in adults with pertussis. A number of studies have documented that 13% to 32% of adolescents and adults with an illness involving a cough of 6 days’ duration or longer have serologic and/or culture evidence of B. pertussis infection.

Other Bordetella species have been associated with infection in immunocompromised patients. B. bronchiseptica, B. holmesii, and B. hinzii produce a pertussis-like respiratory illness. B. trematum has been isolated from individuals working with poultry, and B. ansorpii has been associated with septicemia.

Laboratory Diagnosis

Specimen Collection, Transport, and Processing

Confirming the diagnosis of pertussis is challenging. Culture, which is most sensitive early in the illness, has been the traditional diagnostic standard for pertussis and shows nearly 100% specificity but varied sensitivity. Organisms may become undetectable by culture 2 weeks after the start of paroxysms. Nasopharyngeal aspirates or a nasopharyngeal swab (calcium-alginate or Dacron on a wire handle) are acceptable specimens, because B. pertussis colonizes the ciliated epithelial cells of upper respiratory tract. Calcium-alginate swabs with aluminum shafts are not recommended for PCR, because they may inhibit the polymerase enzyme in PCR detection. In addition, cotton swabs may be inhibitory to specimen growth and are not recommended. Specimens obtained from the throat, sputum, or anterior nose are unacceptable, because these sites are not lined with ciliated epithelium. For collection, the swab is bent to conform to the nasal passage and held against the posterior aspect of the nasopharynx. If coughing does not occur, another swab is inserted into the other nostril to initiate the cough. The swab is left in place during the entire cough, removed, and immediately inoculated onto a selective medium at the bedside (Table 37-2).

TABLE 37-2

Examples of Selective Media for Primary Isolation of B. pertussis and B. parapertussis

Agar Media	Description
Bordet-Gengou	Potato infusion agar with glycerol and sheep blood with methicillin or cephalexin* (short shelf-life)
Modified Jones-Kendrick charcoal	Charcoal agar with yeast extract, starch, and 40 µg cephalexin (2- to 3-month shelf-life but inferior to Regan-Lowe agar)
Regan-Lowe^†	Charcoal agar with 10% horse blood and cephalexin (4- to 8-week shelf-life)
Stainer-Scholte	Synthetic agar lacking blood products