Pertussis: Microbiology, Disease, Treatment, and Prevention

In patients infected with B. pertussis, the incubation period usually lasts 7 to 10 days, but incubation periods as long as 4 weeks have been observed (277). Although pertussis most often presents clinically as whooping cough, the range of signs and symptoms at presentation can vary from the youngest infants to adolescents (Table 3). Infants and children manifest a range of symptoms from asymptomatic or mild upper respiratory tract disease to severe, persistent, and progressive coughing that continues for weeks or months. The classic symptom of pertussis, from which the original name “whooping cough” is derived, consists of violent and rapid coughing with rapid expulsion of air from the lungs, in which the patient is then forced to inhale with a loud “whooping” sound. In infants and young children, the clinical course of pertussis progresses through three consecutive stages after infection and incubation: the catarrhal, paroxysmal, and convalescent stages. Each stage lasts approximately 1 to 3 weeks, and the patients typically do not fully recover for 2 to 3 months (155, 272, 278, 279).

In the first stage, the catarrhal stage, infants and children often present with normal body temperature or low-grade fever, malaise, sore throat, nasal congestion, rhinorrhea, lacrimation, sneezing, and mild progressive dry cough. Although coryza, pharyngitis, and nocturnal cough paroxysms are commonly reported, there are limited systematically collected data for these signs. In the catarrhal stage, the diagnosis of pertussis is frequently overlooked by clinicians because these signs and symptoms mimic those in other viral infections associated with rhinoviruses, coronaviruses, and influenza viruses and perhaps because they expect the vaccinated child to be protected from pertussis infection (280, 281). At this stage of illness, parents frequently bring children to ambulatory care clinics and emergency departments. In these outpatient settings, clinicians need to maintain a high index of suspicion for pertussis in order to provide appropriate referrals to nearby health facilities that are equipped to deliver emergency supportive care and confirmatory diagnostic testing for pertussis. Delays in diagnosis and treatment of pertussis facilitate its transmission to household contacts, friends, schoolmates, and family members (145, 233, 282).

In the second stage of illness, the paroxysmal stage, patients experience bouts of intense and violent coughing (5 to 10 coughs/paroxysms) that last several minutes and are associated with cyanosis, eye proptosis, tongue protrusion, salivation, thick oral mucus production, lacrimation, and engorgement of neck veins (163). The classic sign of pertussis, the inspiratory whoops, manifests at this stage (283). Coughing paroxysms may result from the effects of toxin or from the hardening of mucus that becomes difficult to dislodge from the trachea, bronchi, or bronchioles. Paroxysms can be triggered by stimuli such as crying, laughing, and eating (284). Such paroxysms often occur at night and increase in frequency during the first 1 to 2 weeks of this stage through weeks 2 to 3, with a gradual decline thereafter. In the paroxysmal stage, patients might also experience vomiting (posttussive vomiting), fatigue, and respiratory exhaustion.

In the third stage, convalescence, coughing paroxysms recede in frequency, duration, and severity. However, a mild, chronic, nonparoxysmal cough can last up to 6 weeks. In children who have had the classic type of pertussis, cough paroxysm patterns can recur if the child contracts another viral infection (145, 272, 278). By this stage, patients will have already received appropriate antibiotic and supportive therapy.

Adolescents and adults present with signs and symptoms similar to those in infants and children. In general, however, most adolescents and adults have symptoms that are milder than those of infants and children; consequently, pertussis in adolescents and adults can escape detection by clinicians (233, 285 – 287). Although classic pertussis presentation can still occur in many patients, pertussis without the classic paroxysms is the cause of up to one-third of illnesses with prolonged cough in this age group (219, 288). Therefore, pertussis should be highly suspected as a cause of cough that persists for more than 3 weeks, regardless of whether it is paroxysmal (288). In adolescents and adults, cough duration ranges from 3 to 8 weeks (167, 217, 219, 232, 273, 289). Studies suggested that the clinical presentation of pertussis is influenced by age, gender, infecting species (e.g., B. pertussis versus B. parapertussis), infectious dose, and time from last pertussis vaccination (144, 145, 218, 290). Symptomatic reinfections are common in adolescents and adults (273, 291). In a pertussis vaccine efficacy trial of 246 adults conducted between 1991 and 1994, 64 subjects were identified with clinical and laboratory-confirmed pertussis, of whom 26% had a history of pertussis infection (234). In older adults, the classic whooping cough can be dramatic, as demonstrated by a recent case study of pertussis in a hospitalized adult (292). Also, in older adults, the clinical presentation of pertussis might be influenced by preexisting immunity to B. pertussis as a result of previous exposure or immunization. Adults who had been recently immunized or had a previous pertussis infection tended to have less dramatic presentation, particularly those lacking comorbid medical conditions (144, 145, 218, 290).

In patients of all ages, delayed clinical recognition of pertussis is more likely to result in clinical complications and sequelae (Table 4). Overall, young infants are at highest risk for severe outcomes, such as respiratory failure and death (189, 191, 278, 280, 290, 293 – 311). In the United States, it has been estimated that more than half of infants infected with pertussis require hospitalization, and about one in 100 infants die as a result of pertussis (272, 312). Infants hospitalized for pertussis have been shown to present with apnea, pneumonia, and convulsions (313). In a study conducted in California, among 17 infants less than 3 months of age who were admitted to pediatric intensive care units with severe pertussis, six infants were diagnosed with pulmonary hypertension and four of those six died (314).

In infants and young children, pertussis can be accompanied by coinfection with other respiratory pathogens, particularly respiratory syncytial virus (RSV) (315). In one study, RSV was detected as a coinfection in 33% (11/33) of infants admitted for pertussis to either pediatric intensive care units or general pediatrics wards in London hospitals (251). In 2014, one research team found viral coinfections in 11 of 14 confirmed cases and in six of seven suspected cases of pertussis (316). In that study, influenza A virus, RSV B, rhinovirus, and bocavirus were the most common pathogens found. In several other studies, adenoviruses were isolated from children with pertussis (317 – 320). Coinfection with multiple or mixed pathogens, including bacterial and viral pathogens, has been reported in pertussis patients. In some cases, pertussis infection may be followed by synergistic superimposed bacterial and viral infections. These coinfecting pathogens include parainfluenza virus, RSV, Mycoplasma pneumoniae, adenovirus, and influenza A and B viruses (317, 320 – 323). Selected reports also identified B. pertussis coinfections with helminthic and parasitic pathogens, including Plasmodium falciparum and Fasciola hepatica (324, 325). However, rhinovirus and adenovirus may be found to colonize the upper respiratory tract in asymptomatic infants and children.

In adults, the complications of pertussis most commonly include insomnia, apnea, weight loss, urinary incontinence, syncope, and rib fracture. Less common manifestations include pneumonia, otitis media, and, rarely, death (190, 273, 312, 313, 326, 327, 329). Up to 12% of adults with pertussis require hospitalization (327). From 1990 to 2004, five pertussis-associated deaths in adults were reported to the U.S. Centers for Disease Control and Prevention (CDC) (327). Although reported deaths are infrequent (possibly due to underascertainment), these adults (aged 49 to 82 years) suffered from comorbid conditions, including diabetes, multiple sclerosis with asthma, multiple myeloma (requiring immunosuppressive therapy), myelofibrosis, and chronic obstructive pulmonary disease. In 1998, De Serres et al. reviewed clinical pertussis cases in 280 adolescents and 384 adults who were seen in five public health units in Quebec, Canada (273). They found complications such as sinusitis, pneumonia, otitis media, urinary incontinence, weight loss, rib fracture, and syncope. Six percent of adults who were aged 50 years and older were hospitalized (mean length of stay, 17 days). In this Quebec study, 18% of all adult patients reported previous pertussis infection. Older adults with pertussis have also had uncommon nonrespiratory complications, including encephalopathy, intracranial hemorrhage, stroke, carotid artery dissection, pneumomediastinum, and herniated lumbar discs (313, 326, 330, 331).

The selection of antimicrobial agent should be driven based on the following considerations: age of the patient, drug-related adverse events or interactions, tolerability, medication regimen adherence, and cost. A limited number of antibiotic agents are available for pertussis management, including PEP in individuals who come in contact with clinical cases of pertussis. Close contacts include household members, caretakers, health care workers (HCWs), or any person who has had face-to-face exposure within three to four feet of a symptomatic patient or has shared confined space for 1 h or more or has had direct contact with the respiratory, oral, or nasal secretions of a symptomatic pertussis patient. PEP is recommended for all close contacts regardless of vaccination status (278, 332 – 335). The decision to administer PEP is based on many factors, including the severity of disease, duration and burden of exposure, and immunocompetence of the exposed contact (278).

The usefulness of antimicrobials to treat pertussis when given late in the course of disease is controversial, because the role of therapy is to lessen pertussis disease severity and duration. The use of antimicrobial therapy late in the course of disease is less likely to prevent pertussis symptoms than use early in the course of disease (278). Antimicrobial agents are recommended to clear the nasopharynx of B. pertussis and to prevent the spread of the infection (332, 336, 337). It was estimated that 80% to 90% of patients with untreated pertussis will clear B. pertussis from the nasopharynx within 3 to 4 weeks after the onset of cough (338). However, infants with pertussis who were not treated or vaccinated showed positive culture for more than 6 weeks (339). Table 5 summarizes groups of antimicrobials that have been used for pertussis treatment and PEP, along with recommended dosages across different age groups and potential adverse effects or contraindications. Neither the Clinical and Laboratory Standards Institute (CLSI) nor the European Committee on Susceptibility Testing (EUCAST) has established antimicrobial susceptibility cutoffs for treatment of B. pertussis (340, 341). Thus, clinical laboratories do not routinely perform susceptibility testing for antimicrobial agents used for B. pertussis treatment or PEP.

The effectiveness of antibiotic therapy depends on the stage of pertussis in which therapy is initiated. The first 3 weeks (catarrhal stage) are the optimal time for antibiotics to ameliorate the symptoms of pertussis and eradicate B. pertussis. Unfortunately, treatment is rarely given early enough to impact the course of the disease. For this reason, the U.S. CDC encourages clinicians to start antibiotic treatment based on their clinical judgment and even before laboratory results are known (342). No proven treatments exist that reduce disease severity and frequency of symptoms during the paroxysmal and convalescent stages of the disease (343). However, antimicrobials are still recommended to reduce B. pertussis transmission and to render patients noninfectious (272).

Macrolide antibiotics (e.g., erythromycin, clarithromycin, or azithromycin) have been effective and constitute the mainstay of treatment for patients with pertussis as well as for PEP (278). Although erythromycin has been recommended for treatment of pertussis or PEP, undesirable adverse events have led to poor medication adherence and an increase in prescribing of newer macrolides (i.e., clarithromycin and azithromycin) (278). A relatively long duration of treatment or PEP (i.e., two completed weeks) has been suggested to prevent relapse (278). In 2007, Altunaiji et al. conducted a Cochrane review of 13 clinical trials, which involved more than 2,000 participants, to assess the risks and benefits of antimicrobial treatment of, and PEP against, whooping cough in children and adults (332). They found that short-term antibiotics (azithromycin for 3 to 5 days or clarithromycin or erythromycin for 7 days) were as effective as long-term erythromycin (for 10 to 14 days) to eradicate B. pertussis from the nasopharynx.

Among the three most widely prescribed macrolides, azithromycin is the only recommended antimicrobial for neonates (<1 month old) (278). Erythromycin has been linked with infantile hypertrophic pyloric stenosis (IHPS) in neonates, including those indirectly receiving erythromycin through breast milk (278, 344, 345). In one cohort study, infants receiving erythromycin prophylaxis were found to have greater risk of IHPS (7/157 among those exposed versus 0/125 among those with no erythromycin exposure) (278). In a large retrospective study, Mahon et al. evaluated 14,876 infants treated with erythromycin (oral or ophthalmic administration) as well as mothers who received erythromycin during the third trimester to evaluate the risk of IHPS (346). Young infants (especially those less than 2 weeks of age) who received systemic erythromycin had a higher risk of IHPS than those given ophthalmic erythromycin. No clear association was observed among mothers who received macrolides and IHPS. More recently, in 2014, Danish investigators found an increased risk of IHPS when macrolides were used to treat young infants or when mothers used macrolides within the first 2 weeks after birth (347).

Although azithromycin is recommended for neonates because it is well tolerated and has a high safety profile (278), there may be an association between azithromycin and IHPS. In 2007, Morrison described 7-week-old premature triplets who were hospitalized because of clinical manifestations that were consistent with pertussis (348). All had been treated with azithromycin. Two of the infants were subsequently diagnosed with IHPS and treated surgically.

For children and adults, azithromycin, clarithromycin, or erythromycin may be used (278, 342). Azithromycin and clarithromycin allow for less frequent dosing, which typically results in better adherence to therapy. In addition, azithromycin has a very limited or no effect on the cytochrome P450 system, and more drug-drug interactions are expected with erythromycin and clarithromycin, where they can both inhibit cytochrome P450 (349).

Immunity to pertussis, acquired either from natural infection or through vaccination, is not lifelong. It has been estimated that natural pertussis infection yields 3.5 to 30 years of protection (383 – 386); the estimated protection obtained from the whole-cell pertussis vaccine is 5 to 14 years (387 – 393), and that from the acellular vaccine is 4 to 7 years (241, 390, 394 – 397). As a result of waning immunity over time, adolescents and adults are susceptible to infection with B. pertussis. The severity of their disease appears to be strongly linked to the time since previous vaccination or illness due to B. pertussis (215, 398).

Worldwide, it has been observed that many adolescent and adult patients with confirmed pertussis had a history of pertussis vaccination, had contracted pertussis during early childhood, or both (218, 399, 400). In 2006, Ward et al. conducted a double-blind randomized clinical trial to assess the rates of B. pertussis infection among adolescents and adults (401). In the control group, which did not receive acellular pertussis vaccine (∼1,390 participants), 0.4% to 2.7% showed evidence of pertussis infection during 11 months of follow-up. B. pertussis infection was based on immunoglobulin G (IgG) or immunoglobulin A (IgA) antibodies to PT, PRN, FHA, and FIM. Two percent of nonvaccinated participants had 4-fold-higher antibody titers for one of the following B. pertussis virulence factors: PT, pertactin, FHA, FIM, PT plus any other antibody, pertactin plus PT, pertactin plus FIM, or FHA plus FIM. They concluded that asymptomatic infection was approximately five times more common than reported. In 2010, to estimate the incidence of pertussis, Kretzschmar et al. analyzed serological data from five European countries using two different approaches (267). The first approach consisted of cross-sectional surveys of IgG PT antibody titers combined with longitudinal information about the distribution of amplitude and decay rate of titers in a back-calculation approach. The second approach used age-dependent contact matrices and cross-sectional surveys of IgG PT to estimate a next-generation matrix for pertussis transmission among age groups. Using the first approach, the annual seroincidence of pertussis among adolescents and adults ranged from one to six percent, and when the second approach was used, the pertussis incidence ranged from one to four percent. Another seroepidemiological study was conducted in a highly vaccinated population (>93%) in Israel to estimate the magnitude of B. pertussis infections (402). Approximately 2,000 serum samples were analyzed using standardized methods for IgG PT. The estimated seroincidence rate was ∼2,500 per 100,000 population (≥3 years of age) for the year 2000, compared to an incidence of reported pertussis of 5.6 per 100,000 for the same year, indicating that asymptomatic natural B. pertussis infection is far more common than clinical illness.

In several countries (particularly low- and middle-income countries), persistently low pertussis vaccine coverage has resulted in endemic levels of pertussis accompanied by large outbreaks. In the late 1990s, international health organizations, foundations, and bilateral aid agencies recognized that a global cooperation effort to strengthen immunization programs was needed. From this effort, the Global Alliance for Vaccines and Immunization (GAVI) emerged to become the leading international body to provide national immunization programs with access to funds that enable them to build an immunization infrastructure and improve pertussis (and other) vaccine coverage (434). Despite these efforts, the WHO released a global vaccination coverage report in 2014 indicating that third-dose coverage of diphtheria-tetanus-pertussis (DTP3) vaccine was less than 50% in several countries, including Chad, Central African Republic, Equatorial Guinea, Somalia, South Sudan, and Syria (159). Somewhat higher DTP3 vaccine coverage of 50% to 79% has been reported in Afghanistan, Benin, Democratic Republic of Congo, India, Madagascar, Mali, Myanmar, Nepal, Niger, Nigeria, Pakistan, Papua New Guinea, South Africa, Zambia, and Zimbabwe (159). Even as countries initiate new vaccine introduction to prevent pneumococcal diseases or rotavirus diarrhea, several countries need additional resources and/or technical assistant to achieve DTP3 coverage rates above 80%.

Pertussis outbreaks have been also reported in North America, Europe, and Australia (182, 195, 202, 203, 254, 264, 435 – 438). In contrast, few data are available from Latin American, Asian, and African countries compared with Europe and the United States (174, 439 – 446). In the United States, reported outbreaks that have been investigated by local and state health departments have affected diverse populations, including school children, military personnel, and hospitalized patients (245, 446 – 453). During the past decade, several pertussis outbreaks have been associated with pockets of underimmunization of children (454 – 456). With continued reports of pertussis outbreaks, there have been ongoing demands placed on city and state health departments to carry out timely outbreak investigations and implement effective control measures. Such control measures vary by country and jurisdiction due to differences in local immunization policies (457, 458). In the United States, control measures to limit the duration of a pertussis outbreak and reduce transmission in outbreak settings focus on providing PEP with antibiotics and ensuring that unvaccinated or undervaccinated individuals are appropriately covered (278, 333, 459, 460). For this purpose, the antibiotic of choice depends on a number of factors, including local drug availability, cost, and drug allergies (278).

In suspected pertussis outbreaks, control measures begin by first confirming pertussis among affected individuals. To confirm the illness, clinical specimens (e.g., nasopharyngeal swabs) are collected and tested. According to the U.S. CDC, diagnosis of pertussis in outbreak situations is best done by culture (333). Given the potential for false-positive results associated with PCR, the U.S. CDC advises exercising caution in using PCR during outbreak investigations. Pulsed-field gel electrophoresis (PFGE) is currently the most widely used method to track transmission patterns in an outbreak and characterize B. pertussis strains (462). Newer PFGE methods offer the potential for improved standardization and comparison of B. pertussis strains detected in different laboratories (333, 462). Recently, a new strain-typing technique called single nucleotide primer extension (SNPeX) was developed and showed promise as a flexible and time-efficient way to study outbreaks of B. pertussis and other pathogens (464). SNPeX analyzes fluorescently labeled DNA fragments with high-throughput capacity, a high degree of discrimination, and adaptability for detecting strains that may evolve and vary genetically within outbreaks or from one outbreak to another.

To facilitate the identification of cases in an outbreak, active surveillance must begin with systematic specimen collection and expedited transport of the specimens to laboratories that have the capacity to perform pertussis testing in a standardized fashion (e.g., state health laboratories). The U.S. CDC guidelines for the control of pertussis outbreaks state that confirmation of the pertussis outbreak requires the presence of one or more culture-confirmed pertussis cases (333). Once the diagnosis of the index case(s) and etiology of the outbreak are confirmed, attention then focuses on identifying suspected cases and expanding testing to exposed individuals with suspected pertussis (333). Probable and confirmed cases should be reported rapidly; any delay in reporting will result in more secondary cases and more difficulty in containing the infection (333). Previous studies suggest that reporting delays, often seen during outbreaks, may facilitate pertussis transmission (465, 466). In 2004, Jajosky and Groseclose reviewed a few studies that described reporting timeliness in National Notifiable Disease Surveillance System data (466). They found that the median national reporting delay, based on date of disease onset, was 40 days for pertussis. Marinovic et al. developed a model to evaluate reporting timelines for several communicable diseases, including pertussis (467). They found that the delay in reporting should be reduced by at least 5 weeks to substantially reduce the spread of the infection.

In outbreak settings, ongoing communication with families, parents, schools, teachers, institutional administrators, hospitals, faith-based organizations, and local health departments is essential for efficient and effective outbreak control. Communication with and among these individuals and groups will facilitate ongoing identification of suspected cases of pertussis, and those individuals can be offered confirmatory laboratory testing. In addition, through contact with members of the community and community leaders, education to increase awareness of pertussis and its risk to members in the community can be disseminated through word of mouth, social media, radio, TV, and print media.

If Jules Bordet were alive today, he surely would take some pride in the millions of lives saved in the past several decades through the use of pertussis vaccines. At the same time, he might also argue that additional work remains to be done to eliminate the disease he once observed in his own children (571). The resurgence of pertussis and changes in strain characteristics suggest that some B. pertussis strains may now evade the immune response elicited by existing vaccines. Fortunately, new tools from multiple disciplines are now available to us provide a rich arsenal to guard against the risk of global complacency in the control of pertussis. Research leads from microbiology, public health, and vaccinology highlight at least six avenues of research in pertussis (Table 8). While animal models for pertussis have been instrumental in increasing our understanding of pertussis pathogenesis, multidisciplinary collaborations will be essential to identify host-pathogen and other factors that drive evolution of B. pertussis strains circulating in populations around the world (572 – 574). Such studies may identify environmental, ecological, and human-related factors (e.g., biofilms, use of antibiotics, vaccines, immunity, or behaviors) that are associated with B. pertussis disease and strain emergence (133, 135). Vaccines with improved adverse event profiles and that provide higher efficacy will provide protection to vulnerable populations, including neonates, very young infants, and older adults (575). In these populations, research to understand B. pertussis virulence factors associated with vaccine-associated adverse events may take advantage of immunosignature analysis (576). Immunosignatures in different age groups will also aid the understanding of factors driving humoral and cell-mediated immunity among infants, children, and adults with naturally acquired pertussis and following pertussis vaccination (577). Identification of novel adjuvants to use with existing or new pertussis vaccines has the potential to augment immune responses, including cell-mediated immunity, across a wide range of ages (578). Comparative studies using microbiome analysis among vaccinees and pertussis cases as well as studies to examine the role of probiotic supplementation with vaccination may yield new strategies for prevention or mitigation of disease severity (579). In the fields of public health and clinical medicine, electronic health information systems are prevalent in areas where pertussis outbreaks have been reported (580). Existing information technology and new telecommunication systems may be leveraged to integrate case finding and communication among health care providers and public health agencies with diagnostic testing and guidance for prophylaxis and treatment that facilitates early termination of pertussis outbreaks (581). Such new-generation disease surveillance systems may be adaptable to mobile platforms, particularly in low- and middle-income countries with limited hardwire broadband capacity, and may benefit control of pertussis and many other diseases (582 – 586). Strengthened surveillance for pertussis cases among vaccinees and nonvaccinees will help shed light on the emergence and distribution of vaccine-induced escape strains of B. pertussis that may emerge in different parts of the world (510, 511). A global network, perhaps coordinated through WHO, would collect strains and provide access to strains for collaborating laboratories to perform genome sequencing that characterizes gene mutations. Decreasing costs of whole-genome sequencing now make this genetic analysis feasible. Because vaccine-induced escape strains may be relatively rare, alerting national reference laboratories to collect and store such strains will accelerate genomic analysis.

Further research is also needed to understand changes in pertussis epidemiology that may result from changes in the mucosal environment, biofilms, and bacterial colonization in the nasopharynx and other mucosal sites induced by other vaccines. Studies to understand the role of maternal immunization and its influence on the immunological response to other vaccines in neonates and older infants will help guide improved vaccine design and vaccine schedules. In the area of maternal immunization, additional clinical, epidemiological, and pathological studies are needed to understand the role and significant of conditions (e.g., chorioamnionitis) observed in association with use of Tdap vaccination. In addition, research to better understand the role of the innate immune system, including potential for pertussis adaptation, may also shed light on avenues for combating the ability of B. pertussis to inhibit complement activation mediated by molecules such as the Vag8 autotransporter (587, 588).

Finally, further development and evaluation of dynamic models will help identify conditions under which a pertussis reemergence might be predicted and will aid in evaluating the potential role of booster doses, novel vaccines for neonates and other ages, and newly adjuvanted vaccines to prevent pertussis reemergence (589). New models will benefit from access to empirical data on the genetic characteristics of emerging pertussis strains that evade current vaccines or the human immune response to natural infection.

B. pertussis remains a pathogen of global importance despite having been a recognized cause of whooping cough for one hundred years. Although high vaccination coverage among infants has been achieved in many countries, pertussis cases and deaths are reported annually among children and adults. In existing surveillance systems, underreporting, particularly among adolescents and adults, remains an ongoing multifactorial challenge due to factors including low awareness of pertussis among health care professionals, lack of standardized and efficient diagnostic assays, and weak surveillance infrastructure in many countries. At present, neither currently available vaccines nor previous B. pertussis infection can provide long-lasting protection against later infection. Multipronged approaches to develop new vaccines or enhance the currently available vaccines (e.g., development of more durable and less reactogenic vaccines) will likely require global initiatives aimed at creating new-generation pertussis vaccines through public-private partnerships. The rapid improvement of laboratory research methods in the fields of genomics, immunology, and bioinformatics heralds a new era in research on pertussis that will allow for integration of disciplines and sharing of global research tools such as B. pertussis isolates. The next decade promises new opportunities to understand the evolution of pertussis virulence of strains around the world. This knowledge combined with a deepened understanding of humoral, innate, and cell-mediated immunity to pertussis has potential for guiding new opportunities for treatment and prevention of pertussis across the life span.

We are indebted to Kathryn M. Edwards for her comments on the manuscript and critical revisions during manuscript preparation. We also thank Lisa Maroski for her editorial review of the manuscript.

Paul E. Kilgore is currently funded for research supported through the Global Outcomes of Merck, Inc., and the Pfizer Independent Grants for Learning and Change Program. Heinz-Josef Schmitt declares that he is an employee of Pfizer, Inc.