Immunity to Pertussis

During the past 3 decades, great strides have been made in relation to the molecular microbiology of Bordetella pertussis (the cause of pertussis) and the epidemiology of infection and disease caused by this pathogen [1]. Specific occurrences during this period include (1) the development of acellular pertussis component diphtheria-tetanus–acellular pertussis (DTaP and Tdap) vaccines, which are less reactogenic, because they contain virtually no endotoxin, and (2) the realization that B. pertussis infection and illness involves all age groups and not just children. The findings of the study by Edelman et al. [2] in this issue of the journal support and extend the findings of previous studies of antibody titers over time following immunization or infection [3, 4]. In addition, to our knowledge, Edelman and colleagues present, for the first time, evidence of persistent cell-mediated immunity to the 3 vaccine proteins (pertussis toxin [PT], filamentous hemagglutinin, and pertactin [PRN]) that the subjects had been exposed to 5 years previously. Presently, it seems worthwhile to review the epidemiology of B. pertussis infection and disease and examine what the antibody measurements and CMI values mean with regard to protection against B. pertussis infection and disease. In the prevaccine era, observed pertussis was an epidemic disease with cyclic peaks occurring every 2–5 years [5, 6]. There was an average of 157 cases per 100,000 persons per year. More than 93% of these cases were noted in persons aged <10 years, and ∼10% were noted in infants. Universal infant and early childhood immunization reduced the overall rate of reported pertussis to <1 case per 100,000 persons per year during the early 1970s. At this time, although there were only 1000–2000 cases per year, ∼50% of these cases were noted in infants. Since 1984, there has been a modest increase in reported cases of pertussis, which peaked in 2004 at 25,827 cases (rate, 8.9 cases per 100,000 persons per year), and the majority of reported cases are now seen in adolescents and adults. Contrary to popular belief, it is my opinion that this increase in adolescent and adult cases of pertussis is not a new phenomenon. Pertussis in adults was reported in Germany during the early 1990s, when pertussis was epidemic because of the lack of immunization, and adult pertussis was recognized by the pertussis experts in the prevaccine era in the United States [7–9]. During the past 15 years, a number of studies have contributed to our understanding of the epidemiology of B. pertussis infection [1, 5, 6]. Three types of studies have been performed: (1) studies of prolonged cough illnesses in adolescents and adults, (2) studies of rates of infection by noting increases in antibody titer in serum samples collected at ⩾2 time points, and (3) studies of prolonged cough illnesses in defined populations. The breakthrough that allowed these studies to reveal useful data was the development of ELISA with high precision to measure IgG and IgA antibodies to PT [10]. Approximately 13%–20% of prolonged cough illnesses in adolescents and adults are due to B. pertussis infection [1, 5]. The rate of yearly infection in adolescents and adults is 1%–6.7%, and the yearly rate of cough illness due to B. pertussis infection is 0.37%–1.5% [1]. These data suggest that there are 800,000–3.3 million cases of pertussis in adolescents and adults in the United States each year [1, 5, 6]. The overall frequency of specific clinical manifestations of pertussis in adolescents and adults has not been clearly delineated. A segment of the case patients experience typical illness, with paroxysmal cough and whoop [1, 11]. The majority of cases are unrecognized as pertussis and are diagnosed as bronchitis or upper respiratory infection [1, 10]. It should also be noted that the rates of reported pertussis are 40–160-fold less common than actual illness rates and that asymptomatic infections are 4–22 times more common than symptomatic infections [6]. Today, symptomatic adolescents and adults are the major source of infection in infants in whom disease severity and mortality are the greatest [1, 5, 6, 12]. Although it is frequently stated that there are no serologic correlates of immunity associated with pertussis, I believe that this view should not be accepted. Immunologic correlates of both cellular and humoral immunity have been extensively studied in mouse challenge experiments, and these studies, in the broad sense, have suggested that antibody and cellular factors to many different individual B. pertussis proteins are protective [1]. However, the mouse assay system is not representative of infection and disease in people. Studies performed among children >50 years ago found that protection from pertussis correlated with high titers of agglutinating antibodies (agglutinins) [1, 13]. We now know that the antigens (agglutinogens) to which antibody was directed and that caused agglutination were fimbrial protein serotypes 2 and 3, PRN, and lipopolysaccharides. Although 9 extensive vaccine efficacy trials were performed during the 1980s and 1990s, only 2 of these were performed in a manner in which serologic correlates of immunity could be evaluated [13, 14]. Cell-mediated immunity and its correlation with protection have not been studied in people. In the 2 separate, household-contact substudies nested within large cohort studies, the serologic correlate findings were similar [13, 14]. In both of these studies, it was found that relatively modest levels of antibody to PRN were most important for protection and that antibodies to PT and fimbrial proteins were also important. However, because many vaccine failures had high antibody titers to PRN and to PT and/or fimbrial proteins, it is clear that there is still much to be learned about protection against B. pertussis infection and disease. Despite the concern noted above, the findings reported by Edelman et al. [2] suggest that booster immunizations of adolescents and adults with Tdap vaccine could be successful for reducing the spread of B. pertussis infection among the population. The prolonged prevalence of antibody to PRN noted in the study [2], as well as that noted by Le et al. [3], suggests that an every-10-year Tdap booster program will be successful if it can be universally implemented. It is time for those who care for adults to learn from pediatrics that universal immunization works and, if implemented, would do much to control the spread of pertussis in the United States [1, 5, 15]. Potential conflict of interest.J.D.C. has been a consultant to GlaxoSmithKline and Sanofi Pasteur and is on the Speakers' Bureau of GlaxoSmithKline and Sanofi Pasteur.