Shortcomings of pertussis vaccines: why we need a third generation vaccine

Abstract

First generation whole-cell (wP) and second generation acellular (aP) pertussis vaccines have been highly effective in preventing childhood deaths due to pertussis, yet both vaccines have drawbacks that have limited their long-term usefulness. These include issues of reactogenicity and potency in the case of wP, and limited durability of protection and the potential for selection of escape mutants in the case of aP. Neither vaccine prevents disease in neonatal infants who continue to die from pertussis. A third generation of pertussis vaccines that provides broad, durable protection is needed. In the meantime, countries using wP should continue to do so, while countries using aP need to consider policies and schedules that reduce pertussis transmission to unvaccinated infants. In this respect, maternal vaccination appears to be a promising solution.

Keywords::

• Bordetella pertussis
• efficacy
• immunogenicity
• pertactin
• vaccine

Improved detection methods, heightened awareness and continuing major pertussis epidemics with deaths in infants too young to be vaccinated have led to a growing sense of urgency to improve current pertussis prevention strategies. First-generation whole-cell pertussis (wP) vaccines consist of suspensions of Bordetella pertussis organisms and have been used since the 1940s. Although replaced by second-generation acellular pertussis (aP) vaccines in many industrialized countries, around 74% of the global birth cohorts reside in countries offering wP vaccines in national immunization programs [1].

Potent wP vaccines show good efficacy but suffer several drawbacks: local reactions, fever and other systemic side effects are common after vaccination [2,3]. Concerns that wP caused neurological disease in young children have never been substantiated, but were responsible for substantial decreases in wP vaccine coverage in many countries during the 1970s, stimulating the search for less reactogenic aP [4,5].

Manufacturing processes have changed little since the first wP vaccines were developed early in the twentieth century, and individual wP vaccines vary in immunogenicity, efficacy and the level of impurities (including lipopolysaccharide) contained in them [6]. This variability is not always resolved by pre-release testing, because the currently recommended test (the mouse intracerebral challenge model [Kendrick test] used to test pertussis vaccine potency since the 1940s [7]) does not always guarantee efficacy in humans. Specifications for passing the Kendrick test are not identical worldwide. Furthermore, the test is technically challenging and correct performance and accurate calibration of reference vaccines used in the test are important to the outcome. The widespread use of poorly efficacious vaccines that passed release and potency tests has, in the past, led to decreased pertussis control and local epidemics [8–11]. Clearly, an improved test capable of reflecting wP vaccine efficacy could reduce the risk of future outbreaks [12]. To this end, the mouse intranasal challenge model has been proposed as an alternative, being able to discriminate between vaccines in terms of their protective efficacy and substantially easier to perform [13,14] (see Queenan et al., in this issue). The intranasal challenge model provides clearance curves over time after B. pertussis challenge. Further work to identify reference vaccines and to establish acceptable dose responses to vaccination is needed for acceptance of this model as a qualified release assay.

aP vaccines were developed in the early 1980s and contain between one and five components which may be purified individually or co-purified. Release tests for aP vaccines use manufacturer-specific assays that compare batches, physical chemistry and potency (immunogenicity). Some countries use the murine sensitization test which detects residual pertussis toxin (PT) or a modified version of the Kendrick test. Both of these assays may be useful research tools, but their technical complexity and inter-laboratory variability currently limits their usefulness as release assays. Work continues in this area to identify alternative tests for measuring aP potency.

The use of aP-priming has resulted in a resurgence of pertussis in older children. The first cohorts who received aP-priming have now reached adolescence/early adulthood, allowing assessment of the epidemiology of pertussis disease among fully aP-vaccinated, fully wP-vaccinated or mixed wP/aP-vaccinated individuals. Available data indicate limited durability of the immune response induced by primary and booster vaccination with aP vaccines during childhood, with increased rates of pertussis observed in fully-aP-vaccinated cohorts compared to fully or partially wP-vaccinated cohorts [15–17]. The nature of the first priming dose appears to be important in modulating the immune response with improved durability of protection when the first dose administered is wP [16].

The duration of protection against pertussis may be even shorter when aP is used to boost aP-primed individuals in adolescence. Vaccine effectiveness of aP in preventing pertussis among 13- to 16-year olds who received primary vaccination during the transition period between wP to aP in the USA and an aP booster vaccination at age 11 years was estimated at 47% (95%CI: 19–65) shortly after the booster [18]. aP vaccine efficacy in wP-primed 17- to 19-year olds was 66% (95% CI 30–84), that is, several years after the booster (pertussis was defined as acute cough illness [any duration] associated with B. pertussis isolation or cough illness ≥2 weeks with one of the following: paroxysms/inspiratory ‘whoop,’/post-tussive vomiting without other apparent cause and either confirmed by PCR or epidemiologically linked to a culture-confirmed or PCR-confirmed case). These results contrast with those of the NIH-sponsored Acellular Pertussis Trial in which the efficacy of aP against pertussis (defined as a cough illness ≥5 days with PCR or serological confirmation) in wP-primed adults was 92% (95% CI: 32–99) over a follow-up period of 2.5 years [19]. The discrepancy in vaccine efficacy between aP- and wP-primed individuals could indicate poor priming by aP in infancy, leading to less durable responses to subsequent doses. Alternatively, priming to PT, a major virulence factor present in all aP vaccines, could be misdirected due to chemical (specifically formaldehyde) detoxification processes used during production, which removes up to 80% of surface epitopes. Chemical detoxification reduces immunogenicity of PT and could lead to original antigenic sin (i.e., utilization of immune memory to the PT vaccine epitopes to produce antibodies that are ineffective against a wild-type strain) in response to subsequent doses/exposures [20].

Aside from the problem of limited durability of protection, there have been fears that aP vaccines select for vaccine-resistant pertussis variants. Evaluation of globally representative pertussis strains available from 1920 to 2010 showed that polymorphisms in genes coding for vaccine antigens have occurred continuously [21]. Indeed, most mutations affecting the components of aP vaccines occurred during the wP-vaccination era. Since 2010, B. pertussis strains isolated from countries that use aP have shown multiple alterations to the prn gene, all leading to failure to express pertactin (PRN) [22]. aP vaccination may be driving the selection of PRN-deficient strains [23], but the relative contribution of the number of vaccine components, primary and booster schedule, and coverage are not known. Careful studies are needed to evaluate the implications for aP vaccine efficacy.

Interestingly, researchers in France have identified PRN-negative B. parapertussis strains collected after 2007, which show similar pathogenicity to PRN-positive strains [24], suggesting a link to aP immunization (see Hegerle and Guiso in this issue).

What then is the solution to solve the ongoing problem of pertussis? Despite apparently strong reasons for advocating a return to wP, this is generally viewed as unfeasible in countries where aP is established, primarily because of the increased reactogenicity associated with wP [3]. The efficacy of good wP has remained high despite the emergence of strains with polymorphisms affecting key virulence factors, suggesting that the antigenic diversity of potent wP is sufficient to counteract these changes. An attenuated (i.e., less reactogenic), killed wP will be difficult to achieve because potency and reactogenicity of current wP vaccines are linked. A novel live-attenuated B. pertussis strain known as BPZE1 is being investigated for intranasal administration at birth [25]. wP with chemically extracted lipopolysaccharide has been explored in mice with early evidence of improved tolerability [26]. Third-generation pertussis vaccines that contain a large number of purified components might show more diverse protection and improved resistance to genetic mutations with acceptable reactogenicity. Improvements/additions to current aP vaccines could include changing to genetically detoxified PT and inclusion of additional virulence factors such as adenylate cyclase toxin and lipooligosaccharide conjugates to broaden the immune response [27,28]. New vaccine adjuvants have been developed, but underlying concerns about safety will make their introduction into third-generation pertussis vaccines destined for use in infants challenging. Nevertheless, new adjuvants, antigen delivery systems and routes of administration for pertussis vaccines targeting adolescents and adults warrant investigation. How these will protect infants, however, is unclear.

Until such time that third-generation pertussis vaccines are available, vaccination schedules and policies need close evaluation to maximize protection using currently available vaccines. Consideration has been given to recommending regular booster doses of aP throughout life, ‘cocooning’ of infants by vaccinating their close contacts and vaccination of specific age groups, such as adolescents who have high rates of pertussis disease or neonates in whom pertussis disease is most severe. All of these approaches face challenges, in terms of either implementation and expected achievable coverage or, in the case of neonatal immunization, the impact on subsequent immune responses. A recent study in the USA showed that maternal immunization with Tdap resulted in high anti-pertussis antibodies in their infants that persisted through the first few months of life [29]. Importantly, subsequent responses to Tdap were relatively unimpaired. In September 2012, the UK introduced a temporary pertussis immunization program targeting pregnant women [30] in response to a national pertussis epidemic with rates of pertussis in infants <3 months of age of 240/100,000. At the end of 2013, vaccine uptake among pregnant women was as high as 60% in some regions, the incidence of pertussis in infants decreased to 50/100,000 and there were three infant deaths due to pertussis compared with 14 the previous year [30]. The effectiveness of the UK maternal pertussis program have recently been published [31].

Eradication of pertussis is not realistic in the foreseeable future. Countries using wP should continue to do so based on the available evidence of safety, efficacy and durability of the immune response. Countries using aP need to consider policies and schedules that reduce pertussis morbidity and protect unvaccinated infants who die from pertussis. In this respect, maternal vaccination appears to be an effective solution. In the longer term, improved pertussis control will only be possible using third-generation vaccines that induce high and sustained efficacy, have acceptable reactogenicity and are resistant to escape mutants.

Acknowledgements

The author thanks Joanne Wolter (independent writer) for writing assistance.

Financial & competing interests disclosure

The author is an employee of Crucell. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Writing assistance was utilized in the production of this manuscript and was provided by Joanne Wolter (independent writer on behalf of Crucell).