Chapter 31: Haemophilus and Bordetella

Haemophilus are among the smallest of bacteria. The curved ends of the short (1.0-1.5 μm) bacilli make many appear nearly round; hence the term coccobacilli (Figure 31–1). The cell wall has a structure similar to that of other gram-negative bacteria. The most virulent strains of H influenzae have a polysaccharide capsule, but other species of Haemophilus are not encapsulated.

Cultivation of Haemophilus (Greek haema: blood and philos: loving) species requires the use of culture media enriched with blood or blood products for optimal growth. This requirement reflects bacterial need for exogenous hematin and/or nicotinamide adenine dinucleotide (NAD). These growth factors, also termed X factor (hematin) and V factor (NAD), are present in erythrocytes. In culture media, optimal concentrations are not available unless the red blood cells are lysed by gentle heat (chocolate agar) or added separately as a supplement. Although erythrocytes are the only convenient source of hematin, sufficient amounts of NAD may be provided by certain other bacteria and yeasts. This is responsible for the “satellite phenomenon,” in which colonies of Haemophilus have been observed to grow only in the vicinity of a colony of Staphylococcus aureus. The several species of Haemophilus are defined by their requirement for hematin and/or NAD, dependence on CO₂, and other growth-related characteristics (Table 31–1). Species of Haemophilus other than H influenzae (eg, H parainfluenzae, associated with some cases of endocarditis) have the same biology described later for the nonencapsulated strains of H influenzae.

BACTERIOLOGY

Haemophilus that meets the species requirements for H influenzae may or may not have a capsule. Those that do are divided into six serotypes, a through f, based on the capsular polysaccharide antigen. The type b capsule comprises a polymer of ribose, ribitol, and phosphate, called polyribitol phosphate (PRP). These surface polysaccharides are strongly associated with virulence, particularly in H influenzae type b (Hib). The surface of H influenzae features pili and an outer membrane similar to the structure of other gram-negative bacteria. The outer membrane includes proteins (HMW1, HMW2), lipopolysaccharide (LPS), and lipooligosaccharides (LOS). The nonencapsulated, and thus nontypable, H influenzae (NTHi) can be classified by various typing schemes based on outer membrane proteins and other factors. H influenzae produces no known exotoxins.

HAEMOPHILUS INFLUENZAE DISEASE

EPIDEMIOLOGY

Haemophilus influenzae is a strictly human pathogen and has no known animal or environmental sources. It can be found in the nasopharyngeal flora of 20% to 80% of healthy persons, depending on age, season, and other factors. Most of these are NTHi, but encapsulated strains (including Hib) are not rare. Spread is by respiratory droplets, as with streptococci. Before the introduction of effective vaccines, approximately 1 in every 200 children developed invasive disease by the age of 5 years; meningitis was the most common invasive form and most often attacked those under 2 years of age. Cases of epiglottitis and pneumonia tended to peak in the 2- to 5-year age group. More than 90% of these cases were due to a single serotype, Hib.

The introduction of universal immunization with the Hib protein conjugate vaccine (see Prevention) has reduced invasive disease rates by 99%. Most of the cases in immunized populations are now caused either by serotypes other than b or nonencapsulated strains. Evidence suggests a steady increase in infections worldwide due to nonencapsulated strains, primarily targeting perinatal infants, young children, and the elderly. And as before, in countries and populations unable to afford the vaccine, Hib disease continues.

At one point in time, H influenzae that caused meningitis was believed to be an isolated endogenous infection, but reports of outbreaks in closed populations and careful epidemiologic studies of secondary spread in families have changed this view. The risk of serious infection for unimmunized children younger than 4 years of age living with an index case is more than 500-fold than for unexposed children. This risk indicates a need for protection of susceptible contacts (see Prevention).

PATHOGENESIS

Invasive Disease

For unknown reasons, H influenzae strains commonly found in the flora of the nasopharynx occasionally invade deeper tissues. Bacteremia then enables spread to the central nervous system and metastatic infections at distant sites, such as bones and joints (Figure 31–2). These events seem to take place within a short period (less than 3 days) after an encounter with a new virulent strain. Systemic spread is typical only for encapsulated H influenzae strains, and more than 90% of invasive strains exhibit type b capsule. Even among Hib strains there are distinct clones, which account for approximately 80% of all invasive disease worldwide.

Attachment to respiratory epithelial cells is mediated by pili and outer membrane proteins. Evidence suggests that this is a complex regulatory cascade, coordinating capsular biosynthesis and adherence factors that act cooperatively in establishing the microbe within susceptible hosts. H influenzae can be seen to invade between the cells of the respiratory epithelium (Figure 31–3), and for a time resides between and below them. Once past the mucosal barrier, the antiphagocytic capsule confers resistance to C3b deposition in the same manner as it does with other encapsulated bacteria. As with the pathogenic Neisseria, there is evidence that H influenzae LOS may provide an antiphagocytic effect by binding host components such as sialic acid. Outer membrane LOS is toxic to ciliated respiratory cells, and when circulating in the bloodstream produces all the features of endotoxemia.

Localized Disease

NTHi produce disease under circumstances in which they are entrapped at a luminal site adjacent to the normal respiratory flora, such as the middle ear, sinuses, or bronchi (Figure 31–2). This is usually associated with some compromise of normal clearing mechanisms, such as that caused by a viral infection or structural damage. NTHi attach to bronchial epithelial cells and laminin using pili, OMPs, and other proteins. Consistent with their relative prevalence in the respiratory tract, NTHi account for more than 90% of localized H influenzae disease, particularly otitis media, sinusitis, and exacerbations of chronic bronchitis.

IMMUNITY

Immunity to Hib infections has long been associated with the presence of anticapsular (PRP) antibodies, which are bactericidal in the presence of complement. The infant is usually protected by passively acquired maternal antibody for the first few months of life. Thereafter, actively acquired antibody increases with age; it is present in the serum of most children by 10 years of age. The peak incidence of Hib infections in unimmunized populations occurs at 6 to 18 months of age, when serum antibody is least likely to be present. This inverse relationship between infection and serum antibody is similar to that for Neisseria meningitidis (see Figure 30–4). The major difference is that substantial immune protection is provided by antibody directed against a single serotype (Hib) rather than the multiple types of other encapsulated bacteria, such as N meningitidis and S pneumoniae. Thus, systemic H influenzae infections (meningitis, epiglottitis, cellulitis) are rare in adults, but where such infections develop, the immunologic deficit is typically the same as that with meningococci—lack of type-specific circulating antibody.

Like other polysaccharides, Hib PRP behaves as a T-cell–independent antigen, and antibody responses to immunization are poor in children younger than 18 months of age. Significant secondary responses from boosters are not elicited. The conjugation of PRP to protein dramatically improved immunogenicity by eliciting T-cell–dependent responses while preserving the specificity for PRP, and this maneuver represented a significant breakthrough in vaccine immunology in the 1980s.

HAEMOPHILUS INFLUENZAE DISEASE: CLINICAL ASPECTS

MANIFESTATIONS

Of the major acute Hib infections, meningitis accounts for just over 50% of cases; the remaining cases involve pneumonia, epiglottitis, septicemia, cellulitis, and septic arthritis. Localized infections can be caused by encapsulated strains including Hib, but most are caused by NTHi.

Meningitis

Hib meningitis follows the same pattern as other causes of acute purulent bacterial meningitis. The initial signs and symptoms may be those of an upper respiratory infection, such as pharyngitis, sinusitis, or otitis media; whether these represent a predisposing viral infection or early invasion by the organism is not known. Just as often, meningitis is preceded by vague malaise, lethargy, irritability, and fever. Mortality is 3% to 6% despite appropriate therapy, and roughly one-third of all survivors have significant neurologic sequelae.

Acute Epiglottitis

Acute epiglottitis is a dramatic infection in which the inflamed epiglottis and surrounding tissues obstruct the airway. Hib is one of several causes. Onset is sudden, with fever, sore throat, hoarseness, an often muffled cough, and rapid progression to severe prostration within 24 hours. Affected children have air hunger, inspiratory stridor, and retraction of the soft tissues of the chest with each inspiration. The hallmark of the disease is an inflamed, swollen, cherry-red epiglottis that protrudes into the airway (Figure 31–4) and can be visualized on lateral X-rays. As with meningitis, this infection must be treated as a medical emergency, with primary emphasis on maintenance of an airway (tracheostomy or endotracheal intubation) and antimicrobial therapy. Clinical maneuvers such as direct examination or attempting to take a throat swab may trigger acute obstruction and fatal laryngospasm.

Cellulitis and Arthritis

A tender, reddish-blue swelling in the cheek or periorbital areas is the usual presentation of Hib cellulitis. This picture may follow an upper respiratory infection or otitis media; fever and a moderately toxic state are usually present. Joint infection begins with fever, irritability, and local signs of inflammation, often in a single large joint. Haemophilus arthritis is occasionally the cause of a more subtle set of findings, in which fever occurs without clear clinical evidence of joint involvement. Bacteremia is often present in both cellulitis and arthritis.

Other Infections

Haemophilus influenzae is an important cause of conjunctivitis, otitis media, and acute and chronic sinusitis. It is also one of several common respiratory organisms that can cause and exacerbate chronic bronchitis. Most of these infections are caused by NTHi strains and remain localized without bacteremia. Disease may be acute or chronic, depending on the anatomic site and underlying pathology. For example, otitis media is acute and painful because of the small, closed space involved, but after antimicrobial therapy and reopening of the eustachian tube, the condition usually clears without sequelae.

The association of H influenzae with chronic bronchitis is more complex. There is evidence to suggest that H influenzae and other bacteria play a role in inflammatory exacerbations, but a direct cause-and-effect relationship has been difficult to prove. The underlying cause of the bronchitis is usually related to chronic damage resulting from factors such as smoking. Haemophilus pneumonia may be caused by either encapsulated or nonencapsulated organisms. Encapsulated strains have been observed to produce a disease much like pneumococcal pneumonia; however, NTHi strains may also produce pneumonia, particularly in patients with chronic bronchitis. The closely related H parainfluenzae belongs to the so-called HACEK group of fastidious gram-negative bacteria (Haemophilus, Aggregatibacter, Cardiobacterium, Eikenella, Kingella kingae) that are known to produce up to 3% of all infective endocarditis cases, typically in patients with prosthetic valves or underlying heart disease.

DIAGNOSIS

The combination of clinical findings and a typical Gram smear are sufficient to make a presumptive diagnosis of Haemophilus infection. The tiny cells are usually of uniform shape except in cerebrospinal fluid, in which some may be elongated to several times their usual length (Figure 31–1). The diagnosis is usually confirmed by isolation of the organism from the site of infection or from the blood. Blood cultures are particularly useful in systemic H influenzae infections because it is often difficult to obtain an adequate specimen directly from the site of infection. Bacteriologically, small coccobacillary gram-negative rods that grow on chocolate agar but not blood agar strongly suggest Haemophilus. Confirmation and speciation depend on demonstration of the requirement for hematin (X factor) and/or NAD (V factor) and/or biochemical tests. Serotyping is unnecessary for clinical purposes, but important in epidemiologic and vaccine studies.

TREATMENT

All forms of H influenzae disease were effectively treated with ampicillin until the 1970s, when resistance emerged, in a pattern similar to that of Neisseria gonorrhoeae (see Chapter 30). The major mechanism is production of a β-lactamase identical with that found in Escherichia coli. The frequency of β-lactamase–producing strains varies between 5% and 50% in different geographic areas, with rates of 20% to 30% appearing in recent North American isolate collections. More recently, ampicillin resistance has emerged in β-lactamase–negative strains due to alterations in the transpeptidase site of the penicillin-binding protein, PBP3. Current practice is to start empiric therapy with a third-generation cephalosporin (eg, ceftriaxone, cefotaxime), which can be changed to ampicillin if susceptibility testing indicates that the infecting strain is susceptible.

PREVENTION

Purified PRP vaccines became available in 1985; however, owing to the typically poor immune response of infants to polysaccharide antigens, their use was limited to children 24 months of age and older. Because immunization at this age failed to protect those most susceptible to Hib invasive disease, a new vaccine strategy was needed to include improved stimulation of T-cell–dependent immune responses in infants. To achieve this, the first protein conjugate vaccines were developed by linking PRP to proteins derived from bacteria (diphtheria toxoid, N meningitidis outer membrane protein). The first PRP–protein conjugate vaccines were licensed in 1989; by late 1990, they were recommended for universal immunization in children beginning at 2 months of age. As illustrated in Figure 31–5, the impact has been dramatic and sustained to the present time. This 99% reduction in what was once one of the most feared diseases of childhood is one of the greatest achievements in medical history. Fortunately, the decline in Hib has not been accompanied by compensatory rise in the numbers of non-b cases or in the other causes of acute purulent meningitis. An unexpected concomitant finding has been a dramatic drop in H influenzae colonization rates in immunized populations. Under the direction of the World Health Organization, government and philanthropic efforts like those of the Bill and Melinda Gates Foundation are underway to implement Hib immunization of children throughout the world.

As with N meningitidis, rifampin chemoprophylaxis is indicated for unimmunized close contacts. This includes children and adults when there is a child in the family who has not had a full course of the Hib conjugate vaccine.

Haemophilus ducreyi causes chancroid, a common cause of genital ulcer that has been found in Africa, Southeast Asia, India, and Latin America. Occasional outbreaks in North America have most often been associated with the exchange of sex for drugs or money. The typical lesion is a tender papule on the genitalia that develops into a painful ulcer with sharp margins (Figure 31–6). Satellite lesions may develop by autoinfection, and regional lymphadenitis is common. The incubation period is usually short (2-5 days). The lack of induration around the ulcer has caused the primary lesion to be called “soft chancre” to distinguish it from the primary syphilitic chancre, which is typically indurated and painless. The presence of open genital sores due to H ducreyi greatly enhances the risk of transmission of HIV by providing a portal of entry and/or by the recruitment of CD4+ cells to the site. This may contribute to the heterosexual spread of HIV on the African continent, where chancroid is common. However, determination of the true global incidence of chancroid has been obviated by widespread practice of syndromic management for bacterial genital ulcer disease—that is, empiric treatment with agents effective against syphilis and chancroid. Haemophilus ducreyi has also recently been identified as a causative agent of nongenital cutaneous ulcers in children in tropical regions where yaws is endemic (eg, Papua New Guinea, the Solomon Islands).

Candidate H ducreyi virulence factors include pili and an outer membrane protein (DsrA), which mediates attachment to epithelial cells and resistance to complement-mediated killing. In the lesion, H ducreyi localizes with neutrophils and macrophages but remains extracellular. There is evidence to suggest that the organism may gain an advantage by secreting antiphagocytic proteins and by resisting antimicrobial peptides that are part of the innate immune response. Host immunity may be dampened by the action of cytolethal distending toxin on T cells.

The specific diagnosis of H ducreyi infection is difficult. Although the organism grows on chocolate agar, it does so slowly, and other organisms in the genital flora are apt to overgrow the plates. Incorporating antibiotics (usually vancomycin) in the agar overcomes this problem, but few laboratories in the United States have this medium on hand. Preferred treatments for chancroid include single doses of either azithromycin or ceftriaxone; alternative agents include multiple dose regimens of ciprofloxacin or erythromycin. Condoms are effective in blocking transmission.

The genus Bordetella contains seven species. Bordetella pertussis is by far the most important because it is the cause of classic pertussis (whooping cough). Nucleic acid homology and other analyses indicate that B parapertussis and B bronchiseptica are almost similar enough to B pertussis to be considered variants of the same species. Bordetella parapertussis occasionally causes a disease similar to, but milder than, pertussis and has appeared together with B pertussis in outbreaks. This is probably because it does not produce pertussis toxin even though it has a silent copy of the toxin gene. The remainder of this section focuses on B pertussis.

BACTERIOLOGY

GROWTH AND STRUCTURE

Bordetella pertussis is a tiny (0.5-1.0 μm), gram-negative coccobacillus morphologically similar to Haemophilus. Growth requires a special medium with nutritional supplements (nicotinamide), additives (charcoal) to neutralize the inhibitory effect of the compounds in standard bacteriologic media, and antibiotics to inhibit other respiratory flora. Under the best conditions, growth is still slow, requiring 3 to 7 days for isolation. The organism is also very susceptible to environmental changes and survives only briefly outside the human respiratory tract.

The cell wall of B pertussis has the structure typical of gram-negative bacteria, although the outer membrane lipopolysaccharide differs significantly in structure and biologic activity from that of the Enterobacteriaceae. The surface exhibits a rod-like protein called the filamentous hemagglutinin (FHA) because of its ability to bind to and agglutinate erythrocytes. FHA has strong adherence qualities, based on domains in its structure that interact with an amino acid sequence present in host integrins, epithelial cells, and macrophages. FHA also stimulates cytokine release and interferes with T_H1 immune responses. The organism surface also contains other adhesive structures including pili and an outer membrane protein called pertactin.

EXTRACELLULAR PRODUCTS

Pertussis Toxin

Pertussis toxin (PT) is the major virulence factor of B pertussis. It is an A-B toxin produced from a single operon as an enzymatic subunit and five binding subunits that are assembled into the complete toxin on the bacterial surface. The binding subunits mediate attachment of the toxin to carbohydrate moieties on the host cell surface. The enzymatic subunit is then internalized and ADP-ribosylates a G-protein that affects adenylate cyclase activity. Unlike cholera toxin, which keeps cyclase activity switched on, pertussis toxin freezes the opposite side of the regulatory circuit and cripples the capacity of the host cell to inactivate cyclase activity. Multiple intracellular signaling pathways are disrupted by this G-protein modification. Among the results of this action are lymphocytosis, insulinemia, and histamine sensitization.

Other Toxins

Another potent toxin, a pore-forming adenylate cyclase (AC), enters host cells and catalyzes the conversion of host cell ATP to cyclic AMP at levels far above what can be achieved by normal mechanisms. This activity interferes with cellular signaling, chemotaxis, superoxide generation, and function of immune effector cells, including PMNs, lymphocytes, macrophages, and dendritic cells. AC can also induce programmed cell death (apoptosis). Tracheal cytotoxin (TCT) is a monomer of B pertussis peptidoglycan generated during cell wall synthesis. The fragments are released into the environment by multiplying bacterial cells because B pertussis lacks mechanisms present in other bacteria for recycling these monomers. Tracheal cytotoxin is directly toxic to ciliated tracheal epithelial cells causing their extrusion from the mucosa and eventual death. There is little or no effect on the nonciliated cells.

PERTUSSIS (WHOOPING COUGH)

EPIDEMIOLOGY

Pertussis is a major health problem worldwide, with an estimated 50 million cases and 300 000 deaths annually. More than 90% of the cases are in developing nations and most of the deaths are among infants. B pertussis is spread by airborne droplet nuclei and remains localized to the tracheobronchial tree. It is highly contagious, infecting more than 90% of exposed susceptible persons. Secondary spread in families, schools, and hospitals is rapid. Sporadic epidemics occur, but there is no strong seasonal pattern. B pertussis is a strictly human pathogen. It is not found in animals and survives poorly in the environment. Asymptomatic carriers are rare except in outbreak situations. The introduction of immunization in the 1940s produced a dramatic reduction in disease, but outbreaks persisted in 3- to 5-year cycles. Large outbreaks occurred in populations where the immunization rates fell as a result of concerns about febrile reactions to the original pertussis vaccine.

Immunization also produced a change in the age distribution of the residual cases. Previously a disease of toddlers and young children, pertussis began to appear in infants and adults beginning in late adolescence. This is believed to be due to the relatively short duration (10-12 years) of immunity provided by the vaccine. These adults are susceptible if exposed but usually have a milder form of the disease, which is often not recognized as pertussis. These unwitting adults then are the major source for outbreaks in highly susceptible populations, such as infants. In the preimmunization era, newborns were usually protected by maternal transplacental IgG stimulated by the almost universal exposure to B pertussis in the general population. In an immunized population with waning immunity, this antibody has frequently dropped below protective levels by the childbearing years. In a cruel twist, infants have the most severe form of the disease. More than 70% of fatal cases occur in children younger than 1 year of age. These problems appear to have worsened with the switch to an acellular vaccine whose protection is of even shorter duration (See Prevention).

PATHOGENESIS

When introduced into the respiratory tract, B pertussis has a remarkable tropism for ciliated bronchial epithelium attaching to the cilia themselves. This adherence is mediated by FHA, pili, pertactin, and the binding subunits of PT. Once attached, the bacteria immobilize the cilia and begin a sequence in which the ciliated cells are progressively destroyed and extruded from the epithelial border (Figures 31–7 and 31–8). This local injury is caused primarily by the action of tracheal cytotoxin. This toxicity eventually produces an epithelium devoid of the ciliary blanket, needed to move foreign matter away from the lower airways. Persistent coughing is the clinical correlate of this ciliary defect. Although considerable local inflammation and exudate are produced in the bronchi, B pertussis does not directly invade the cells of the respiratory tract or spread to deeper tissue sites.

Virulence Factors

In addition to the local effects on bronchial epithelium, other virulence factors of B pertussis contribute to the disease in diverse ways. The combined action of PT and AC on neutrophils, macrophages, and lymphocytes creates paralysis and even death of these crucial effector cells of the immune system. Many of the systemic manifestations of the disease, such as lymphocytosis, histamine sensitization, and insulin secretion, are due to the action of circulating PT absorbed at the primary infection site. The specific biologic effect depends on how disruption of G-protein regulation by PT is manifested by the host cell type that the toxin reaches. Pertussis is the result of a well-orchestrated delivery by B pertussis of toxic and adhesive factors to host cells at local and distant sites to produce a disease that persists for many weeks.

Genetic Regulation of Pathogenicity

How B pertussis deploys its repertoire of virulence genes is a model for the control of bacterial pathogenicity. B pertussis regulates the synthesis of PT, AC, FHA, pili, and many other genes through genetic loci that control the expression of at least 20 unlinked chromosomal genes at the transcriptional level. Expression is modulated in a two-component system by changes in specific environmental parameters, including temperature. The induction of virulence factors in B pertussis is sequential, with expression of adhesins (FHA and pili) preceding expression of factors involved in tissue injury (PT, AC). The finely honed responses of B pertussis virulence factors to changes in temperature and ionic conditions presumably play a role in the pathogenesis of infection and help the organism adapt in a stepwise fashion to the diverse local conditions throughout the human respiratory tract. Details of the genetic mechanisms involved are discussed in Chapter 22 and illustrated in Figure 22–8.

IMMUNITY

Although IgG antibodies are produced to PT, pili, and pertactin during the course of natural infection and by immunization, they are not long-lasting, and their role in immunity is not well understood. Although naturally acquired immunity is not lifelong, second attacks (when recognized) tend to be mild.

PERTUSSIS: CLINICAL ASPECTS

MANIFESTATIONS

After an incubation period of 7 to 10 days, pertussis follows a prolonged course consisting of three overlapping stages: (1) catarrhal, (2) paroxysmal, and (3) convalescent. In the catarrhal stage, the primary feature is profuse mucoid rhinorrhea, which persists for 1 to 2 weeks. Nonspecific findings such as fever, malaise, sneezing, and anorexia may also be present. The disease is most communicable at this stage because large numbers of organisms are present in the nasopharynx and the mucoid secretions.

The appearance of a persistent cough marks the transition from the catarrhal to the paroxysmal coughing stage. At this time, episodes of paroxysmal coughing occur up to 50 times a day for 2 to 4 weeks. The characteristic inspiratory whoop follows a series of coughs as air is rapidly drawn in through the narrowed glottis. Vomiting frequently follows the whoop. The combination of mucoid secretions, whooping cough, and vomiting produces a miserable, exhausted child barely able to breathe. Apnea may follow such episodes, particularly in infants. Marked lymphocytosis reaches its peak at this time, with absolute lymphocyte counts of up to 40 000/mm³.

During the 3- to 4-week convalescent stage, the frequency and severity of paroxysmal coughing and other features of the disease gradually fade. Partially immune persons and infants younger than 6 months of age may not show all the typical features of pertussis. Some evolution through the three stages is usually seen, but paroxysmal coughing and lymphocytosis may be absent.

The most common complication of pertussis is pneumonia caused by a superinfecting organism such as Streptococcus pneumoniae. Atelectasis is also common but may be recognized only by radiologic examination. Other complications, including convulsions and subconjunctival or even intracerebral bleeding, are related to the venous pressure effects of the paroxysmal coughing and the anoxia produced by inadequate ventilation and apneic spells.

DIAGNOSIS

A clinical diagnosis of pertussis is best confirmed by detection of B pertussis in nasopharyngeal secretions or swabs. Throat swabs are not suitable because the cilia to which the organism attaches are not found there. Specimens collected early in the course of disease (during the catarrhal or early paroxysmal stage) provide the greatest chance of successful isolation. Unfortunately, the diagnosis is frequently not considered until paroxysmal coughing has been present for some time, and the number of organisms has decreased significantly. Usually, the nasopharyngeal specimens are plated onto a special charcoal blood agar medium made selective by the addition of a cephalosporin; this allows the slow-growing B pertussis to be isolated in the presence of more rapidly growing members of the normal upper respiratory flora. The characteristic colonies appear after 3 to 7 days of incubation and look like tiny drops of mercury. Immunologic methods (agglutination, immunofluorescence) are required for specific identification.

A direct immunofluorescent antibody (DFA) technique has been successfully applied to nasopharyngeal smears for rapid diagnosis of pertussis. DFA is particularly helpful in pertussis because of the many days required for culture results. Nucleic acid amplification tests are now replacing both culture and DFA as they have proven to be more timely and sensitive than the classic methods. However, culture confirmation should be established before declaring an epidemic. Serologic tests are widely used for epidemiologic studies but not diagnosis of individual clinical cases.

TREATMENT

Once the paroxysmal coughing stage has been reached, the treatment of pertussis is primarily supportive. Antimicrobial therapy is useful at earlier stages and for limiting the spread to other susceptible individuals. Of a number of antimicrobial agents active in vitro against B pertussis, macrolides are preferred for both treatment and prophylaxis. Erythromycin has the greatest clinical experience, but azithromycin and clarithromycin are equally effective.

PREVENTION

Active immunization is the primary method of preventing pertussis. The original vaccine, which produced a dramatic reduction in disease, was prepared from inactivated whole cell suspensions and given together with diphtheria and tetanus toxoids as DTP. The undoubted efficacy of this vaccine was colored by a high rate of side effects due to the crude nature of the whole cell preparation. These included local inflammation, fever and, rarely, febrile seizures. Although permanent neurologic sequelae were never convincingly linked to pertussis immunization, some argued that the vaccine was worse than the disease. This led to the development of acellular vaccines containing virulence factors purified from inactivated whole cell preparations. The impact of these vaccines is shown in Figure 31–9.

The multiple acellular vaccine products have different combinations of virulence factors. All contain PT and FHA, and some add pertactin or pili (vaccine manufacturers use the term fimbriae). In combination with diphtheria and tetanus toxoids, the acellular vaccine has now replaced the whole cell DTP as DTaP (“a” for acellular). This vaccine is now recommended for the full primary immunization series (at 2, 4, and 6 months) and boosters (at 15-18 months, 4-6 years). The safety and efficacy of these vaccines have now been extensively evaluated. All have dramatically less frequent side effects compared with the whole cell preparations, but their efficacy is increasingly in question. In the United States, major pertussis outbreaks in 2005, 2010, and 2012 have been traced to vaccine failures in fully immunized adolescents and even preadolescent children. Clearly, the acellular vaccine does not provide immunity for as long as the product it replaced (Figure 31–10). The concern for transmission to newborns has led to a strategy called cocooning in which all family members are newly immunized or boosted before the baby comes home. There appears to be no going back to the whole cell vaccine, but adjustments in booster schedules and vaccine formulation are ahead.