Experimental urethral infection with Neisseria gonorrhoeae

Abstract

Gonorrhea rates and antibiotic resistance are both increasing. Neisseria gonorrhoeae (Ng) is an exclusively human pathogen and is exquisitely adapted to its natural host. Ng can subvert immune responses and undergoes frequent antigenic variation, resulting in limited immunity and protection from reinfection. Previous gonococcal vaccine efforts have been largely unsuccessful, and the last vaccine to be tested in humans was more than 35 years ago. Advancing technologies and the threat of untreatable gonorrhea have fueled renewed pursuit of a vaccine as a long-term sustainable solution for gonorrhea control. Despite the development of a female mouse model of genital gonococcal infection two decades ago, correlates of immunity or protection remain largely unknown, making the gonococcus a challenging vaccine target. The controlled human urethral infection model of gonorrhea (Ng CHIM) has been used to study gonococcal pathogenesis and the basis of anti-gonococcal immunity. Over 200 participants have been inoculated without serious adverse events. The Ng CHIM replicates the early natural course of urethral infection. We are now at an inflexion point to pivot the use of the model for vaccine testing to address the urgency of improved gonorrhea control. Herein we discuss the need for gonorrhea vaccines, and the advantages and limitations of the Ng CHIM in accelerating the development of gonorrhea vaccines.

1. Gonorrhea vaccines are needed to combat increasing incidence and spread of antibiotic resistance.

Neisseria gonorrhoeae (Ng) is a common sexually transmitted bacterial pathogen. All sexually active persons are at risk for gonorrhea. Approximately 90 million gonorrhea cases occur globally each year (Rowley et al., 2019). In the United States, gonorrhea is the second most reported notifiable disease, and incidence has risen over 60% since 2014 (Unemo et al., 2019;CDC, 2019). Ng is a Gram-negative bacterium that typically infects the male and female genital tracts with a spectrum of phenotypes, as described below (reviewed in (Lovett and Duncan, 2018)); gonococci can also infect other mucosal tissues, such as the oropharynx, rectum, and conjunctiva. Infection at multiple mucosal sites is common and in some high-risk populations, oropharyngeal infections exceed the prevalence of genital infections (Chow et al., 2019).

Untreated genital infection can ascend into the upper reproductive tract, including the epididymis and fallopian tubes (McConaghy and Panchal, 2016;Westrom et al., 1992;Wiesenfeld et al., 2012;Westrom, 1994). Ascended infection can have severe long-term sequelae, including pelvic inflammatory disease (PID) in females (Reekie et al., 2018). PID damages the fallopian tubes, and infertility and chronic pelvic pain may ensue, even in the absence of acute signs and symptoms (Wiesenfeld et al., 2012). In males, up to 30% of untreated urethritis cases can progress to epididymitis and result in male infertility (McConaghy and Panchal, 2016). In the pre-antibiotic era, Ng was known to cause chronic infections in males associated with long-term sequelae, like urethral stricturing (Richards, 1946). Urethral structuring remains a common sequela of gonococcal urethritis in middle-to-low-income countries (Fall et al., 2011) where gonorrhea exposure is high and access to adequate Ng detection and treatment is problematic (Rowley et al., 2019;Wi et al., 2017). In 0.5 to 3 percent of patients with gonorrhea, infecting gonococci may spread systemically and can cause fever/septicemia, arthritis, tenosynovitis, endocarditis, or vasculitis (Belkacem et al., 2013). During pregnancy, genital gonorrhea has been linked to obstetric and fetal sequelae, including preterm labor and preterm birth, low birthweight, and congenital infections (Heumann et al., 2017;Liu et al., 2013a) (reviewed in (Waltmann et al., 2021)). Ng can be vertically transmitted to the neonate, and neonatal conjunctival infection can lead to corneal scarring and blindness (Woods, 2005). Asymptomatic infection is common. As many as 50% of cervicovaginal infections and the majority of rectal and oropharyngeal carriage are asymptomatic (Kent et al., 2005). In addition to the potential for development of complicated infections with serious consequences, asymptomatic infections serve as a reservoir for transmission, as these infections are often not recognized or treated. Not only is Ng a source of substantial burden for sexual and reproductive health among infected persons, it also increases transmission of HIV. Symptomatic and asymptomatic genital gonorrhea infection is associated with inflammation in the genital tract, which in turn is a risk factor for Human Immunodeficiency Virus (HIV) acquisition (Masson et al., 2015;Passmore et al., 2016). In addition, an almost 5-fold increased risk of HIV acquisition was observed prospectively in patients when gonorrhea was present (Mlisana et al., 2012). Furthermore, seminal plasma of HIV-seropositive men with urethritis had significantly higher detectable levels of HIV-1 RNA with the highest HIV-1 RNA levels in men with gonorrhea (Cohen et al., 1997).

Gonorrhea control has focused on directly observed single dose antibiotic administration at the time of diagnosis or syndromic management (Grant et al., 2020). Syndromic management is imperfect, due to modest sensitivity (40%–75%) and specificity (54%–76%), particularly for recognition of asymptomatic and extragenital infections (van Gemert et al., 2018;Grant et al., 2020), thus leading to over- or under-treatment (Grant et al., 2020). Overuse of antibiotics exerts selective pressure for development of Ng antimicrobial resistance (Unemo et al., 2017;Wi et al., 2017) and under-treatment of gonorrhea facilitates onward transmission and disease. Effective antimicrobial treatment of gonorrhea is threatened by resistant Ng strains. As a result of gonococcal genetic plasticity, resistance to every first-line antimicrobial used to treat infection has rapidly emerged since antibiotics were first introduced (Unemo and Nicholas, 2012). Ceftriaxone, an extended-spectrum cephalosporin, is the currently recommended option for empirical first-line gonorrhea monotherapy, but decreased susceptibility and resistance have been reported globally (Wi et al., 2017), including in resource-rich countries, like Australia and the United Kingdom (Eyre et al., 2019;Jennison et al., 2019). As a result of rising levels of Ng cephalosporin resistance in the US, CDC treatment guidelines now recommend an increased dose of ceftriaxone monotherapy to treat uncomplicated gonorrhea(St Cyr et al., 2020). In the face of increasing incidence and the spread of antibiotic resistant Ng, an effective gonococcal vaccine is a priority long-term solution for gonorrhea control.

2. Limited protective immune responses to naturally-acquired Ng infection and immune evasion by Ng are major impediments to effective vaccine development.

Most uncomplicated Ng infections are associated with increased risk of recurrent infections, even with the same strain (Fox et al., 1999;Hobbs et al., 1999)and persistent gonococcal infection and reinfections (Hazel et al., 2014;Mauss, 1946;Stupiansky et al., 2011;Handsfield et al., 1974) are common. These data suggest that humans do not develop robust natural protective immunity to Ng infection, at least in settings of low intensity of gonorrhea re-exposure and/or when prompt antibiotic treatment is available and used promptly to treat infection (Handsfield et al., 1974;Hazel et al., 2014). A study by Plummer et. al (1989) of Kenyan female sex workers in an area of high Ng exposure showed that participants were less likely to be reinfected with the same serovar and the longer they engaged in sex work the less frequently they were infected; in the same study, a subgroup of ~30% of participants were never infected during the study, suggesting that these females may be protected from or at a substantially reduced risk of-gonococcal infections and among infected participants, a reduced rate of ascending infection was observed. This led Plummer et. al (1999) to hypothesize that protective immunity may be achieved but may require multiple infections with the same strain. In the absence of antibiotic therapy, there is some evidence that limited immunity to natural infection can develop. For example, in the pre-antibiotic era, symptomatic infections resolved spontaneously over months, although the basis for apparent immune resolution was not studied (Richards, 1946). Before treatment of men with asymptomatic gonorrhea became the standard of care, a prospective study of the natural history of asymptomatic male infection (Handsfield et al., 1974), showed that 2/3 of the participants remained asymptomatic until they were treated between 7-165 days after the initial infection was detected. Of the remaining participants, roughly 1/6 developed urethritis, and the other 1/6 spontaneously cleared the infection (Handsfield et al., 1974). Longitudinal studies of the natural history of infection using retrospective nucleic acid testing for Ng in asymptomatic (untreated) females who provided self-collected vaginal swabs have shown that some females, but not all, clear Ng over an 11-week period without antibiotics (Stupiansky et al., 2011). Taken together, these observations indicate that the human immune system does exert some selective pressure on Ng fueling efforts to develop vaccines to stimulate more robust protective responses than those elicited by natural infection.

3. Persistent gonococcal infections in males and females suggest that the bacteria evade or resist host immune responses.

Ng surface structures, including lipo-oligosaccharide (LOS), opacity (Opa) protein and pilin adhesins, undergo frequent phase and antigenic variation (Simms and Jerse, 2006). As a result, immune responses face an ever-changing array of targets among different Ng strains and even within an individual strain. In addition to alterations of surface antigenic structure, Ng expresses proteins that directly impact immune signaling in the human host. For example, gonococcal Opa proteins bind to host cells via the carcinoembryonic antigen cell adhesion molecule (CEACAM) proteins on human epithelial cells and immune cells (Bos et al., 1997;Chen et al., 1997;Gray-Owen et al., 1997a). Opa-CEACAM1 interaction inhibits T cell receptor signaling through the recruitment of the tyrosine phosphatase SHP-1 (Chen and Gotschlich, 1996;Gray-Owen et al., 1997b). Ng Opa-CEACAM1 interactions also inhibit T cell proliferation and reduce B cell antibody production (Boulton and Gray-Owen, 2002;Pantelic et al., 2005). Ng infection has been shown to typically stimulate adaptive Th17 responses away from protective Th1 (Liu and Russell, 2011). This pathway may be reversed by giving mice anti-TGF-β or IL-10 antibodies or by vaginal administration of microencapsulated IL-12. This was seen to lead to an increase in Th1 and Th2 cells and a humoral memory response (Liu et al., 2014;Liu et al., 2013b). Notably, several candidate vaccines that showed protective efficacy in mice induced Th1 responses (Zhu et al., 2004;Gulati et al., 2013;Liu et al., 2017). Mechanisms of immunosuppression have been reviewed in Jerse et al. (2014). Overall, the presence of Ng immune evasive mechanisms that a gonococcal vaccine must achieve a “better than nature” immune response in order to provide protective efficacy against Ng.

4. History of gonorrhea vaccine development.

Only two gonorrhea vaccines have ever tested in humans as part of field clinical trials, and both were unsuccessful. One of the two field trials, an intradermally delivered two-dose purified pilin subunit vaccine carried out in high-risk US military personnel stationed in Korea (Boslego et al., 1991), was preceded by efficacy testing using the controlled human infection model of urethral Ng infection (Ng CHIM)(Brinton, 1982). Unlike the field trial in which the vaccine was not found to be protective (Boslego et al., 1991), the vaccine indeed showed protection from homologous Ng challenge, with broad antibody response in serum and in genital secretions, including secretory IgA (Brinton, 1982). There were two important confounders to the discrepancy between the real-world data and the Ng CHIM findings: 1) parenteral immunization was used in the controlled infection, whereas the field trial utilized intradermal delivery; and 2) a subsequent trial using the controlled human urethral infection model showed no protection against a heterologous strain expressing antigenically variant Pil (Tramont and Boslego, 1985). The last gonococcal vaccine to be tested in humans was in 1985. Using the Ng CHIM, a vaccine preparation from the outer membranes of a single strain of Ng was given in a randomized, placebo-controlled fashion (unpublished, reviewed in (Rice et al., 2017)). Two to four weeks post-vaccination, male participants were challenged with the homologous Ng strain via intraurethral inoculation. No significant difference in infection after challenge was observed between the two groups, but paradoxically resistance to infection was high in both the active vaccine and the placebo groups: 46% of vaccinees and 36% of placebo recipients resisted infection. As explained in the review by Rice et al. (2017), though the active vaccine preparation vs. the placebo preparation aimed to enrich for the major outer membrane protein Porin (Por), the antigen purification technology available at the time was not sufficiently precise to eliminate contamination with other outer membrane constituents, particularly lipooligosaccharide (LOS) and reduction modifiable protein (Rmp). LOS and Rmp also effected immune responses in vaccines, and anti-Rmp antibodies were later found to be “blocking antibodies”, i.e. the presence of these downregulates effective bactericidal function of antibodies directed against other antigens, such as Por and LOS (Rice et al., 1986). In effect, anti-Rmp antibodies antagonized those elicited by Por and LOS, and protection in vaccinees was only observed in retrospective analyses when stratified by the ratio of the concentration of Por and LOS antibodies to that of Rmp antibody. None of the antibody levels alone correlated with protection against challenge. The importance of this study with respect to the use of a human model of gonorrhea infection to test vaccine efficacy is two-fold: 1) it is feasible with a relatively small number of participants (this study enrolled 63 males) to determine the protective effects of a gonorrhea vaccine; and 2) inter-individual variation in protection and putative correlates of protection can be studied with such a system of uniform exposure because it minimizes confounders. After the disappointing results of the field trials and human challenge studies, enthusiasm for a gonorrhea vaccine waned. With the expensive chimpanzee model of infection being the only animal model available at the time and the rapidly emerging HIV crisis of the 1980s, efforts to develop vaccines were further stymied.

5. Evidence for vaccine-induced immunity.

In addition to the data supporting the existence of naturally-acquired immunity, albeit limited, evidence of vaccine-induced immunity also exists. Data from the pilin vaccine field trial has shown that protective responses can develop as long as the reinfecting strain does not differ from the original strain (McChesney et al., 1982). When chimpanzees vaccinated with a killed whole-cell gonorrhea vaccine, analogous to the ones used in the first human trials, were challenged with the homologous Ng strain, a 1000-fold larger inoculum was needed to establish infection (Arko et al., 1976); notably, some animals cleared the infection spontaneously after a few weeks of persistent infection (Arko et al., 1976). The same animals were not protected when re-challenged 2 years after the primary infection, thus durable protection was not observed (Kraus et al., 1975).

Perhaps the most convincing evidence of vaccine-induced immunity has come from mass immunization campaigns with Neiserria meningitidis (Nm) outer membrane vesicle (OMV) vaccines. Data from multiple observational studies that investigated the community effects of a currently FDA-approved and commercially available vaccine for Nm serogroup B, a pathogen closely related to Ng, indicate that mass vaccination against group B Nm may have cross-protective effects against Ng (Paynter et al., 2019;Petousis-Harris et al., 2017). This hypothesis is yet to be formally tested, though our group are planning to do so using the Ng CHIM. Nm B-OMV vaccines (Naess et al., 1999;Peeters et al., 1996)^, containing several proteins and lipoproteins with homology to Ng antigens, have been proven safe and have been used in mass vaccination campaigns in the setting of NmB epidemics in several countries, including Cuba, Norway, Canada and New Zealand (de Moraes et al., 1992;Holst et al., 2009;Nokleby et al., 2007). While STI education campaigns could have contributed to decreased Ng rates in these settings, at least one retrospective case control study in New Zealand found that vaccinated individuals compared to unvaccinated individuals demonstrated a decreased odds of gonococcal infection, but not of Chlamydia trachomatis infection (Ct, another sexually transmitted bacterial pathogen that causes urethritis and cervicitis) (Petousis-Harris et al., 2017). The data from New Zealand also suggest that after controlling for confounders, the Nm-OMV-based vaccine offered 31% protection against Ng infection (95% confidence interval 21%–39%). Since then, it has been demonstrated by us and others that 4C-MenB vaccinees recognize gonococcal antigens (Semchenko et al., 2019;Leduc et al., 2020).We have also shown in collaboration with researchers at the Uniformed Services University of Health Sciences that 4C-MenB could induce protective immune responses to Ng in mice (Leduc et al., 2020): mice vaccinated with 4C-MenB substantially accelerated Ng clearance and reduced vaginal Ng burdens compared to adjuvant only-treated animals (Leduc et al., 2020), supporting the ecologic data in humans that suggest Nm-OMV and 4C-MenB can induce immune protection against Ng and that gonococcal vaccines are biologically feasible.

6. Ng is an exclusive human pathogen and limited preclinical animal models have slowed vaccine development.

In nature, gonococci infect only humans; there are no non-human animal or environmental reservoirs. Many Ng interactions with its host related to adherence, immune evasion, and nutrient acquisition are restricted to recognition of human factors. Without other natural hosts, it has been difficult to establish animal models of infection that replicate human disease, and to evaluate immune responses and protection from infection induced by experimental vaccines. Until 1999, when the estradiol-treated female mouse model was developed (Jerse, 1999), the only non-human model of infection was the chimpanzee (Arko et al., 1976). Biomedical experimentation on chimpanzees is no longer considered ethically acceptable, and the chimpanzee model of gonococcal infection is no longer available to support vaccine development. Other non-human primate models have been explored but establishing Ng infection or colonization in monkeys has met with limited success (D. Patton, personal communication). Perhaps the best characterized animal model is the estradiol-treated female mouse model of vaginal gonococcal colonization, established in 1999 and refined over the last two decades (reviewed in (Raterman and Jerse, 2019;Jerse et al., 2011)). Several limitations in this murine model pose challenges to its use in vaccine development. Estradiol treatment prolongs proestrous phase necessary for gonococcal colonization, but also likely impacts host innate and adaptive immunologic responses during the infection. Genital tract infection in mice is also transient. The mouse also lacks host-restricted factors such as human transferrin and CEACAMs, which are potential targets of Ng virulence factors that are potentially exploitable as vaccine antigens. Despite these limitations, the lower genital tract model of gonorrhea in female mice has been pivotal in identifying potential vaccine candidates (Gulati et al., 2013;Liu et al., 2017;Plante et al., 2000;Gulati et al., 2019) and in studying immune evasion mechanisms of the gonococcus (Liu et al., 2014;Packiam et al., 2010;Packiam et al., 2012). Murine infection models are especially important to identify features of infection and potential protection in the female genital tract, in light of the restriction of experimental human infection to males. Recent murine models employing transgenic mice that express human cellular adhesion factors or complement components may eventually circumvent host-restriction limitations (Li et al., 2011;Rice et al., 2017).

7. Male human urethral challenge is safe, provides a relevant model for studies of Ng pathogenesis, and will advance gonococcal vaccine development.

Urethral gonococcal infection in men usually produces a mild illness, and the risk of serious complications is remote based on the large number of natural infections, especially with prompt treatment of infections. In the absence of robust animal models, investigators at the Walter Reed Army Institute of Research and at the University of North Carolina at Chapel Hill developed a controlled human infection model (CHIM) of male urethral gonorrhea in the 1980s (Hobbs et al., 2011). In accordance with the ethical principles set forth in the Declaration of Helsinki and the US Code of Federal Regulations for the Protection of Human Subjects, CHIM studies with Ng are subject to rigorous safety and ethical review, and all participants provide written informed consent. Participation is limited to men; potential complications from ascendant gonococcal infection in women pose an unacceptable safety risk. To initiate experimental infection, a suspension of Ng is delivered to the anterior urethra using a small, smooth plastic catheter. Participants are monitored daily and receive 100% effective antibiotic therapy promptly when they develop symptoms of urethritis. Experimental infection of male volunteers with Ng reproduces clinical features of naturally-acquired gonococcal infection; infected participants develop urethritis between 1 and 5 days after inoculation. Several hundred individuals have participated in experimental infection studies without reported complications or unexpected adverse events (Hobbs et al., 2011). Comparisons between experimental infections with wild-type and isogenic mutants of Ng and competitive infection studies have proved useful for testing the importance of putative gonococcal virulence factors for urethral infection in men (Cornelissen et al., 1998;Hobbs et al., 2013). The model can assess both microbiologic outcomes (i.e., do participants acquire infection with recoverable Ng?) and disease outcomes (i.e., do participants develop urethritis after inoculation?). Thus, the Ng CHIM has the potential to advance gonococcal vaccine development in several ways: pathogenesis studies to identify vaccine targets; vaccine or natural protection studies to identify correlates of protection, and clinical trials with high exposure to Ng for early-stage investigations of vaccine efficacy.

8. Ng CHIM has been used in pathogenesis studies to identify potential vaccine targets.

The Ng CHIM has been used in conjunction with isogenic strain mutants to test the requirement of specific gonococcal factors to human infection. For example, Ng acquires iron during human infection by scavenging transferrin or lactoferrin, via expression of specific transferrin- and lactoferrin-binding proteins. All Ng express a transferrin receptor, and roughly half of circulating isolates also express a lactoferrin receptor. Cornelissen and colleagues demonstrated that a Ng mutant lacking both transferrin- and lactoferrin-binding capacity was rendered essentially non-infectious in human challenge (Cornelissen et al., 1998). This study highlighted the possibility of causing “nutritional immunity” through vaccines that block Ng iron acquisition. Similar studies using isogenic mutants lacking adhesin proteins that were believed to be critical for Ng infection (major structural protein of the pilus, PilE, and the 13 Opa proteins) revealed that gonococci were still infectious in the absence of their expression (Hobbs et al., 2011). These CHIM studies effectively eliminate the idea that the function of these latter proteins could be targeted by vaccine to protect against gonorrhea, though a vaccine that elicited adhesin-specific bactericidal activity might still provide protection. Combined, these experiments highlight the need to understand correlates of protection or mechanisms of immune-mediated clearance in order to develop effective protective vaccines against Ng.

9. Ng CHIM can be used to identify mechanisms of natural immune clearance or correlates of immunity.

In studies of natural Ng acquisition, it is difficult to assess the risk of infection after exposure and to find at risk individuals who have naturally cleared Ng or resisted infection after exposure, because those individuals do not seek medical attention. Participants in human challenge studies have uniform exposure by nature of the study design, and this allows comparison of immune responses, host genotypes, phenotypes, and other factors to be compared between infected and uninfected individuals. In an experiment using rechallenge, Schmidt and colleagues demonstrated that an initial treated infection did not prevent reinfection with the same challenge strain of Ng (Schmidt et al., 2001). The lack of protection in this experiment suggests that a study of naturally-acquired protection may require a challenge in a population of highly Ng-exposed individuals or in the setting of a partially effective vaccine candidate.

10. Passive protection studies using the Ng CHIM suggest that circulating bactericidal antibody alone is insufficient to prevent gonococcal infection.

Bactericidal antibodies are known to correlate with protection from invasive Neisseria meningitidis infection and have often been considered a potential immunologic correlate to tested in early Ng vaccine development. We recently conducted a randomized, controlled trial to test whether infusion of a human monoclonal antibody (hMAb) against bacterial β-(1→6)-linked poly-N-acetyl-d-glucosamine (PNAG) with bactericidal activity against Ng could prevent experimental infection in men. In this study, ten participants were enrolled: 3 received no infusion, and 7 were randomized to 1 of 3 doses of the hMAb (1, 3, or 10 mg/kg). Three to 6 days after infusion (mean = 4 days), participants were inoculated intraurethrally with wild-type Ng strain FA1090 and were followed daily for up to 5 days. One of 3 control participants and 4 of 7 hMAb-infused participants developed urethritis. Further, bacterial load in urine of infused and infected participants was not different compared to those observed in historical controls infected with the same Ng strain (Figure 1). After antibody administration, serum from all infused participants had bactericidal activity titers greater than 1:16 against the FA1090 challenge strain. These results suggested that circulating monoclonal bactericidal antibody may not be sufficient to provide protection from Ng infection in men. Equilibration of the hMAb in urogenital tissues or with effector cells to support antibacterial responses in the setting of antibody infusion has not been studied. Antibody-mediated protection may require polyclonal responses or functionality not manifested shortly after infusion of the monoclonal product.

Figure 1.

Infusion with bactericidal hMAb did not impact recovered bacteria in experimental Ng urethritis.

11. Ng CHIM offers statistical power for initial Ng vaccine efficacy and safety prior to field trials.

Intraurethral challenge with Ng in humans provides an opportunity to test vaccine efficacy against gonococcal infection with higher rates of exposure and infection than are found in even high-risk populations. For a partially effective vaccine, the sample size required for an adequately powered study to detect statistically significant efficacy increases from under 100 for a human challenge study to more than a 1000 in a very high-risk population and more than 25,000 in a population with average risk for Ng infection (Figure 2). The required sample size could allow human challenge to be incorporated into phase 2 safety trials of Ng vaccines as a rapid and cost-effective way to evaluate vaccine performance for the rapid down-selection of promising vaccine candidates prior to implementation of larger and more costly field trials. A potential public health risk presented by field trials of Ng vaccine studies is presented by the possibility that a vaccine could prevent development of urethritis or symptomatic Ng infection without halting asymptomatic infection and potential for transmission of infection. Study of vaccine efficacy using human challenge allows assessment of vaccine candidate’s ability to prevent establishment of microbiologic infection as well as symptomatic infection and could provide reassurance that a vaccine trial will not inadvertently lead to increased transmission of Ng from study participants to intimate contacts.

Figure 2.

Sample size of trial 90% powered to detect vaccine efficacy for given level of vaccine efficacy with p< 0.05 calculated with Fisher Exact test.

12. Limitations of Ng CHIM for evaluation of vaccine efficacy.

By design, the controlled nature of CHIMs limits their generalizability for predicting the efficacy of candidate vaccine against natural exposure. This limitation was actually realized during the development of an Ng Pilin vaccine in the 1980’s. An intradermally delivered two-dose purified pilin subunit vaccine trial was carried out in high-risk US military personnel stationed in Korea (Boslego et al., 1991). The vaccine generated a broad antibody response in both serum and genital secretions and demonstrated protection from infection after experimental challenge with a homologous Ng strain (Brinton, 1982). Despite the proof-of-concept for this pilin vaccine, the vaccine was not protective in a randomized, placebo-controlled, double-blind efficacy trial; among 3123 men in the trial, 108 in the active vaccine group and 102 in the placebo group acquired gonorrhea within two weeks post vaccination (Boslego et al., 1991). The failure of pilin vaccine in this setting after success in human challenge model is largely attributed to then unrecognized, but now well-characterized extensive antigenic variation in Ng pilin structure. Even for vaccine antigens that are not subject to inter-strain variation, antigen variation between strains could limit generalizability of vaccine effectiveness using homologous challenge. It is likely that challenge with multiple Ng strains, alone or in combination, will be required to demonstrate robust vaccine efficacy against heterologous strains. Currently the mechanisms of immune clearance from the cervix, oropharynx and rectal mucosa are unknown. It is possible clearance or protection in the urethra is mediated by different immune responses from the clearance mechanisms than might protect these other sites of infection. Data from a human urethral challenge model showing vaccine efficacy could be compromised in this case and not be predictive of field trial success.

Despite its limitations, advantages detailed above of the Ng CHIM in the natural host of this well-adapted pathogen offer promise for testing neisserial OMV-containing vaccines and other Ng vaccine platforms in development and for identifying correlates of protection to move the field forward.