Abstract
Pertussis or whooping cough is an acute bacterial infection of the respiratory tract, caused by Bordetella pertussis and, less frequently, by B. parapertussis . It occurs worldwide, affects all age groups, and is most serious in young, unprotected infants. Due to underconsulting, underrecognition, underdiagnosis, and underreporting of cases, incidence rates vary considerably by country and region. Disease characteristics vary by age and underlying immunity, from mild, unspecific cough to typical pertussis, a three-stage illness, comprising catarrhal, paroxysmal, and convalescent phases. In typical cases, cyanosis, neck vein distention, bulging eyes, tongue protrusion, salivation, lacrimation, sweating, and posttussive vomiting of food or viscous mucus occur whereas fever is usually minimal or absent. There is a wide spectrum of complications of pertussis, most of which predominantly occur in young infants: pneumonia, respiratory failure due to hyperleukocytosis, apnea, and encephalopathy being most serious. Specific PCR applied to nasopharyngeal secretions or induced sputum allows rapid diagnosis. Treatment consists of macrolides, which can ameliorate symptoms when given early during the course of disease, and general support and exchange transfusion (to reduce leukocytosis) and extracorporeal membrane oxygenation in serious cases. Active immunization is the most effective prevention, usually with acellular pertussis component vaccines (containing various numbers and quantities of B. pertussis antigens). The primary series should start early in infancy (6–8 weeks of age), followed by life-long boosters to reinforce waning immunity. Recently, immunization in pregnant women has been shown to be most efficient in preventing serious pertussis in infants too young to be protected by active immunization.
Keywords
pertussis, Bordetella pertussis, vaccine
Pertussis (“whooping cough”) is an acute bacterial infection of the respiratory tract. The illness occurs worldwide and affects all age groups, but it is most serious in young, unprotected infants. It is caused by Bordetella pertussis (first described in 1906) and, less frequently, by Bordetella parapertussis (1937). The term pertussis was coined in 1670 and means “violent cough.” The clinical picture was described for the first time in 1640 by Guillaume de Baillou, based on a 1578 epidemic in Paris.
Epidemiology
Pertussis occurs worldwide, and humans are the only host of B. pertussis. Transmission occurs effectively by droplets, with secondary attack rates close to 100% in exposed susceptible individuals. The incubation period is usually 7–10 days but may vary substantially. In a household contact study, secondary cases were noted to have their onset up to 6 weeks after the onset of illness in the primary case, especially when antibiotics were used. Several studies have shown that females and males are equally affected by pertussis in childhood, whereas a female preponderance (55%–69%) is noted in adolescents and adults. Most likely, this is due to more frequent contact of females with young children, from whom they acquire infection. The seasonal pattern of pertussis varies between different geographic locations.
The epidemiology of pertussis caused by B. pertussis infection is incompletely understood. Reasons for this are the lack of a uniform case definition, inconsistent use of diagnostic laboratory tests, variable surveillance systems, and incomplete case ascertainment due to underconsulting, underrecognition, underdiagnosis, and underreporting. Furthermore, epidemiology is greatly influenced by pertussis immunization programs that, with high coverage rates, confer not only individual protection but also herd protection to some extent. In contrast to B. pertussis, B. parapertussis infections appear to occur independent of immunization activities, and their relative frequency varies considerably by geography and time.
Although in most countries pertussis still mainly affects infants and young children, it is increasingly diagnosed in adolescents and adults, who are also important sources of infection in young, unprotected infants in their families and households ( Table 32.1 ). In recent years, a shift from mothers to siblings as the predominant source of B. pertussis transmission to infants has been noted in the United States. Among 1306 cases of pertussis in infants, the source of infection was identified in 569: overall, 35.5% were siblings, 20.6% were mothers, and 10.0% were fathers.
| Country | Investigator (Year) | Study Population | Observation a |
|---|---|---|---|
| United Kingdom | Crowcroft (2003) | 25 infants <5 months of age admitted to ICU because of proven pertussis | Primary case: Parent: N = 11 (44%) Sibling: N = 6 (24%) |
| United States | Bisgard (2004) | 616 infants with proven pertussis | Source discovered in 264 (43%) cases: Parent: N = 123 (47%; 20% of total) Grandparent: N = 22 (8%; 4% of total) Sibling: N = 52 (20%; 8% of total) |
| France | Bonmarin (2007) | 1668 hospitalized infants <6 months of age with proven pertussis | Source discovered in 892 (53%) cases: Parent: N = 491 (55%; 29% of total) Sibling: N = 223 (25%; 13% of total) |
| Multinational | Kowalzik (2007) | 99 infants admitted to ICU because of proven pertussis | ≥1 source ( N = 30) discovered in 24 (24%) cases: Parent: N = 18 (60%; 18% of total) Other adult: N = 6 (20%; 6 % of total) Sibling: N = 5 (17%; 5% of total) |
| Multinational | Wendelboe (2007) | 95 infants <6 months of age admitted to hospital because of proven pertussis | ≥1 source discovered in 44 (46%) cases: Parent: N = 27 (55%; ≈25% of total) Grandparent: N = 3 (6%; ≈3% of total) Sibling: N = 8 (16 %; ≈5% of total) |
| The Netherlands | de Greef (2010) | 201 infants <6 months of age admitted to hospital because of proven pertussis | ≥1 source discovered in 96 (48%) cases: Parent: N = 53 (55%; ≈25% of total) Sibling: N = 39 (41%; ≈19% of total) |
| United States | Skoff et al. (2015) | Cases of pertussis in infants, identified through “enhanced pertussis surveillance” in seven US states | Source identified in 569 (44%) cases: siblings (35.5%), mothers (20.6%), fathers (10.0%), grandparents (7.6%), aunts/uncles (6.5%), and other source, not specified (6.3%) |
a Restricted to household contacts; other sources, if any, were nonhousehold contacts.
A resurgence of cases and a gradual shift toward an increase of pertussis in adolescents and adults has been noted in North America and elsewhere. Whether this increase is caused by waning vaccine immunity and a decreased chance for natural boosters, a consequence of increased awareness and diagnostic tools or both is an ongoing debate. Of note, pertussis in adolescents and adults is frequently atypical, and true numbers of cases are certainly higher than reported. Furthermore, in support of a true change in epidemiology, a rise in fatalities due to pertussis has been observed in infants in the United States during the last decades. Whereas 1.67 deaths per million infants per year were reported in the 1980s, the rate increased to an average of 2.40 in the 1990s. This increased incidence almost exclusively affected infants under 4 months of age.
The crucial role that mass immunization plays in controlling pertussis has been clearly demonstrated in countries such as Japan, England, and Sweden, where infant pertussis vaccination was either discontinued or markedly curtailed as a result of unsubstantiated concerns about vaccine-related adverse events.
Etiology
Bordetella organisms are small, aerobic, gram-negative coccobacilli. Today, the genus comprises 10 different species, of which 9 have been shown to cause respiratory tract illness in humans; these species include B. pertussis and B. parapertussis, the causative agents of whooping cough; B. bronchiseptica and B. holmesii, which can cause variable respiratory symptoms; B. trematum, which has been found in ear and wound infections, and B. petrii, isolated from respiratory tract secretions in patients with cystic fibrosis and from a patient with mastoiditis. However, the overwhelming majority of Bordetella infections are caused by B. pertussis and B. parapertussis.
Pathology/Pathogenesis
The organism is transmitted by aerosol droplets from infected to susceptible humans. After transmission, adhesion of the bacteria to ciliated cells of the upper and lower respiratory tract establishes colonization, followed by multiplication and spread on the epithelium, local mucosal damage, and finally, induction of respiratory symptoms. Invasiveness is extremely rare. Asymptomatic, transient colonization frequently occurs during reinfection in immune individuals. Animal studies suggest that a variety of virulence factors is involved in the various steps of infection of the respiratory tract ( Table 32.2 ). Expression of these factors is regulated in response to environmental changes by BvgAS, a two-component signal transduction system. The precise mechanisms during B. pertussis and B. parapertussis infection in the human host are unknown. Laboratory studies suggest that several factors working in concert allow adherence of the organisms to the epithelium, and filamentous hemagglutinin (FHA) inhibits phagocytosis. Later on, effects caused by adenylate cyclase toxin and pertactin expression allow effective phagocytosis and killing of the bacteria by the host. The popular belief that pertussis is a single-toxin illness caused by pertussis toxin (PT), exclusively produced by B. pertussis, is refutable by the observation that a similar illness results from infection with B. parapertussis, which does not express PT. Further insight into the pathogenesis of Bordetella infection will be gained now that several members of the genus have been sequenced.
| Factor (Gene) | BVG Regulation | Molecule | Major Role | Other Functions | Comments |
|---|---|---|---|---|---|
| Pertussis toxin (PTX) | Yes | “A” protomer and “B” subunits | Toxin and first-line adhesion factor | Causes leukocytosis by lymphocytosis | Precise role in disease unknown |
| Filamentous hemagglutinin (FHA) | Yes | Large, filamentous protein (220 kDa) | Major adhesion; predominantly in trachea | Not known | Need for inclusion in vaccine questionable |
| Fimbriae 2 and 3 (fim2, fim3, fimX) | Yes | Small, filamentous proteins (23 kDa) | Adhesion factor; predominantly in trachea | Agglutinogens; sustain infection | Important stimulator of host’s immune response |
| Pertactin (prn1, prn2, prn3) | Yes | 69-kDa outer membrane protein | Adhesion factor, induces type-specific antibody | Major protective antigen (mouse model) | Important vaccine antigen, used for genotyping |
| Adenylate cyclase (cyaA) | Yes | Protein toxin | Toxin; inhibits phagocytosis by ↑ cAMP | Inhibits chemotaxis and induces apoptosis of macrophages | Candidate for future vaccines! |
| Tracheal cytotoxin (TCT) | No | Peptidoglycan derivative | Toxin; paralyzes mucociliary clearance system | Inhibits DNA synthesis and cell death | Nonimmunogenic → not suitable for vaccine |
| Dermonecrotic toxin (DNT) | Yes | Heat-labile toxin (140 kDa) | Toxin; dermal necrosis and vasoconstriction | Effect only after injection in skin | Role in human disease unknown |
| Tracheal colonization (tcfA) | Yes | Proline-rich protein | Adhesion factor; predominantly in trachea | Not known | C-terminal homology to prn, factor brkA, and vag-8 |
| Bordetella resistance to killing factor (brkA) | Yes | Outer membrane protein (32 kDa) | Adhesion factor | Provides resistance to complement | C-terminal homology to prn, tcfA, and vag-8 |
| Virulence-activated gene 8 (vag-8) | Yes | Outer membrane protein (95 kDa) | Adhesion factor (?) | Not known | C-terminal homology to prn, tcfA, and brkA |
| Lipooligosaccharide (wlb) | Yes | Lipid A and trisaccharide | Presumably required for nasal colonization | Not known | Substantially species-specific structure within Bordetella |
| BVG intermediate-phase (bipA) | Yes | Outer membrane protein (137 KDa) | Transmission (?) and adhesion factor | Not known | First of a new class of Bordetella protein A antigens (“intermediate phase”) |
| Type III secretion system (bsc) | Yes | Several, not yet specified proteins | Secretes effector proteins into host cells | Downregulation of the host immune system | Appears to be functional only in B. bronchiseptica (and some B. parapertussis strains) |
| bteA (bteA) | Yes | Linked to type III secretion system (72 kDa) | Induction of cytotoxicity | Persistent infection (animal model) | Potential vaccine antigen |
The pathology of pertussis has been characterized by studies of B. pertussis infection. It causes inflammation, congestion, and infiltration of the respiratory mucosa with lymphocytes and granulocytes and leads to accumulation of viscous secretions in the lumens of the bronchi, bronchiolar obstruction, and occasional atelectasis. Later in the infection, necrosis of the midzonal and basilar parts of the bronchial epithelium result in necrotizing bronchitis ( Fig. 32.1 ). Subsequently, bronchopneumonia may develop, either caused by B. pertussis itself or by secondary infections with other pathogenic bacteria.
Infection with B. pertussis of a previously naive host results in the production of serum and salivary antibodies against a number of antigens such as PT, FHA, pertactin, and adenylate cyclase toxin. Enzyme-linked immunosorbent assay techniques allow discrimination of class-specific antibodies, with Immunoglobuline G (IgG) being more reliably detectable than Immunoglobulines A, E, and M. In individuals who have been “primed” by B. pertussis infection or immunization, reinfection will elicit a secondary immune response with or without concomitant symptoms. Of note, infection with B. pertussis does not provide lifelong immunity, and apparently no cross protection exists between different species of Bordetella. After natural infection, sustained IgG and IgA serum antibody levels against FHA, pertactin, and—though less—PT have been observed before returning close to baseline values after approximately 5 years. Yet the precise role of serum antibodies in immunity against pertussis is a matter of ongoing debate, and various studies show conflicting results.
Cell-mediated immune responses to B. pertussis have been shown to play an important role in protection against pertussis. In mice, challenge with B. pertussis resulted in a predominant T helper 1 (Th1) cell–mediated immune response followed by complete bacterial clearance. Specific protection could be conferred by adoptive transfer of immune spleen cells into immunosuppressed mice, further underlining the role of Th1 cells. Interestingly, persistent vaccine efficacy has been documented in young children several years after immunization despite significant antibody decline, and it was also preferentially mediated by Th1 cells.
Clinical Features
Symptoms
Typical pertussis is a three-stage illness, comprising catarrhal, paroxysmal, and convalescent phases. The catarrhal stage lasts for about 1–2 weeks and is characterized by flulike symptoms such as coryza, sneezing, lacrimation, conjunctival injection, malaise, and nonspecific cough. It is followed by the paroxysmal stage, which in classical cases is marked by an increase of frequency and severity of coughing, with paroxysms as the most typical feature; characteristically, repetitive series of 5–10 or more hacking spells of cough occur during a single expiration. At the end of a paroxysm, a typical whoop, which is caused by the sudden rush of inspired air through a narrowed glottis, is noted. Paroxysms may occur up to several times per hour, during both day and night, triggered by various stimuli such as eating and drinking and physical or emotional stress. The paroxysmal stage may last from a few days to several weeks, until the convalescent stage is reached; this phase is marked by a decrease in the frequency and severity of coughing spells. However, over a period of several months, similar coughing episodes may again occur, often associated with other respiratory tract infections. Notably, B. pertussis infections present with considerable variability, which primarily depends on age, previous immunization, or infection. Other variables—including the presence of passively acquired antibody (in young infants), degree of exposure to the source of infection, specific bacterial inoculum, host genetic and acquired factors, and genotype of the organism—may contribute to attenuation of symptoms.
The variability of symptoms is exemplified by results from a study of 1860 culture-positive cases in unvaccinated children and adolescents in Germany. In that study, 38% of patients had a total coughing illness duration of 4 weeks or less, 18% had nonparoxysmal cough, 21% did not whoop, and 47% did not have posttussive vomiting. On rare occasions, B. pertussis infection has been found to cause otitis media and to be associated with unilateral hyperlucent lung (MacLeod or Swyer-James syndrome) and the hemolytic uremic syndrome.
Overall symptoms with B. parapertussis infection are similar to those caused by B. pertussis, but illness is usually less severe and of shorter duration. Dual infections of B. pertussis and B. parapertussis have been observed. Occasionally B. bronchiseptica and B. holmesii have been isolated from children with pertussis-like illness. The clinical role of Bordetella species isolated from sputum specimens in patients with cystic fibrosis is currently unknown.
Physical Findings
Cyanosis, neck vein distention, bulging eyes, tongue protrusion, salivation, lacrimation, sweating, and posttussive vomiting of food or viscous mucus may occur with pertussis. Fever is usually minimal or absent. There is a wide spectrum of complications of pertussis, most of which predominantly occur in young infants ( Table 32.3 ). Respiratory complications include bronchopneumonia with or without atelectasis, pulmonary hypertension, and otitis media mainly secondary to other respiratory tract pathogens. Pertussis has also been associated with activation of latent tuberculosis. Additional complications that have been observed as a consequence of pertussis include ulcer of the frenulum of the tongue, epistaxis, melena, subconjunctival hemorrhages, meningoencephalitis, encephalopathy with cerebral seizures, tetanic seizures caused by severe alkalosis as a result of loss of gastric contents due to persistent vomiting, subdural hematomas, spinal epidural hematoma, rupture of the diaphragm, rib fracture, umbilical hernia, inguinal hernia, rectal prolapse, dehydration, syndrome of inappropriate antidiuretic hormone secretion, apnea, and nutritional disturbances. Death secondary to pneumonia, pulmonary hypertension, or sudden death, probably due to severe hypoxemia, occurs mainly in infants, for whom the mortality rate is 0.6%.
