MANUSCRIPT CITATION

Madhi SA, Anderson AS, Absalon J, Radley D, Simon R, Jongihlati B, Strehlau R, van Niekerk AM, Izu A, Naidoo N, Kwatra G, Ramsamy Y, Said M, Jones S, Jose L, Fairlie L, Barnabas SL, Newton R, Munson S, Jefferies Z, Pavliakova D, Silmon de Monerri NC, Gomme E, Perez JL, Scott DA, Gruber WC, Jansen KU. Potential for Maternally Administered Vaccine for Infant Group B Streptococcus. N Engl J Med. 2023 Jul 20;389(3):215-227. doi: 10.1056/NEJMoa2116045. PMID: 37467497.

REVIEWED BY

Dustin Daniel Flannery, DO, MSCE
Assistant Professor of Pediatrics
Children’s Hospital of Philadelphia
Email: flanneryd@chop.edu

Karen Marie Puopolo, MD, PhD
Professor of Pediatrics
Children’s Hospital of Philadelphia

TYPE OF INVESTIGATION

Prevention/Treatment –Safety and Immunogenicity

QUESTION 

In healthy pregnant women 18 to 40 years of age (P) does administration of a single dose of GBS6 vaccine (5µg, 10µg, or 20µg, with and without aluminum phosphate) (I) compared with placebo (C) result in fetal transfer of antibodies associated with a reduced risk of invasive GBS disease without adverse events (O) through 12 months after delivery (T)?

METHODS

Design: This was a phase 2, randomized, placebo-controlled trial (vaccine trial), and a parallel case-control, longitudinal, observational cohort study (seroepidemiologic study)

Allocation: Pregnant women were randomly assigned in 1:1:1 ratio to receive vaccine with aluminum phosphate, vaccine without aluminum phosphate, or placebo. Sentinel cohorts were used to assess safety to allow progression to the next higher dose.

Blinding: Trial staff who dispensed and administered vaccine or placebo were not blinded; all other trial personnel and participants were blinded

Follow-up period: For maternal participants, solicited local and systemic events were monitored for 7 days after infection, unsolicited adverse events for 1 month after injection, and obstetrical complications through 12 months after delivery. For infant participants, unsolicited adverse events were monitored for 6 weeks after delivery, and serious adverse events were monitored through 12 months after delivery.

Setting: Three clinical research centers in South Africa

Participants:

• Vaccine trial: Healthy women who were 18 to 40 years old and pregnant between 27 0/7 and 35 6/7 weeks’ gestation on the day of injection from March 2019 to June 2020 with an uncomplicated, singleton pregnancy. Further detailed inclusion and exclusion criteria are listed in the study protocol.
• Seroepidemiologic study: Pregnant women attending for antenatal care at one of the participating antenatal-clinics or delivery hospitals. Neither antenatal culture-based GBS screening nor intrapartum antibiotic prophylaxis for GBS prevention were standard of care at these sites. Further detailed inclusion and exclusion criteria are listed in the study protocol.

Intervention: A single intramuscular injection of GBS6 vaccine (6 formulations: 5µg, 10µg, or 20µg, with and without aluminum phosphate) or placebo (saline).

Outcomes:

• Vaccine trial: Maternal and infant safety and immunogenicity
• Seroepidemiologic study: Risk of invasive GBS disease as a function of distribution of anti-capsular polysaccharide (CPS) immunoglobulin G (IgG) concentrations

Analysis and Sample Size:

• Vaccine trial: There was no formal statistical hypothesis. The sample size was determined by the likelihood of adverse events being observed. Analyses for safety and immunogenicity were descriptive. Point estimates and 95% Cis were determined for each group using the Clopper and Pearson method. Serotype-specific anti-CPS IgG concentrations were log-transformed geometric mean concentrations were used. There was no imputation for missing serologic results, no adjustments for multiplicity, and values less than the lower limit of detection were assigned as 0.5 this value for statistical purposes.
• Seroepidemiologic study: Mother-newborn dyads were prospectively enrolled and monitored for invasive GBS disease in the 89 days after birth. Cord-blood serum anti-CPS IgG concentrations were measured from cases (infants with invasive GBS) and controls (infants without invasive GBS born to mothers with GBS colonization), and infant serum anti-CPS IgG was measured at the time of infection for infants with invasive disease. Vaginal and rectal swabs were obtained at delivery from a subgroup of pregnant women and case patients were matched to controls with vaginal colonization by infant invasive serotype and gestational age.
• Serologic assay: Maternal serum samples were obtained prior to and after injection at various times; cord blood or infant serum samples were obtained at delivery. For both the vaccine trial and parallel seroepidemiologic study, a quantitative direct immunoassay (Luminex) was used to determine anti-CPS IgG concentrations for serotypes Ia, Ib, II, III, IV, and V. A Bayesian model was fitted for antibody concentrations among cases and controls.

Patient Follow-up:

• Vaccine trial: There were 494 pregnant women were screened and 360 (73%) were enrolled; 40 pregnant women received vaccine in each of the 6 vaccine formulation subgroups (5µg, 10µg, or 20µg, with and without aluminum phosphate) and 120 women were assigned to placebo. One pregnancy in the vaccine group (5µg with aluminum phosphate) resulted in stillbirth and two women in the placebo group withdrew from the study. Safety analyses included data from 360 maternal participants and 357 infants.
• Seroepidemiologic study: There were 24,053 pregnant women who were prospectively screened and 125 infants with invasive GBS disease were retrospectively screened for enrollment within 72 hours after confirmed illness; there were 77 cases and 250 controls included in the final analysis. There were 2950 women included in the subgroup analysis of rectovaginal GBS colonization.

MAIN RESULTS

Seroepidemiology study: Twenty-six percent of the selected subgroup who underwent vaginorectal screening were found to be colonized with GBS. There were 28 cases of invasive GBS disease from the prospectively followed group, and an additional 90 cases of invasive GBS disease were enrolled retrospectively upon presentation. There were 77 cases where a cord blood or infant serum sample, invasive isolate, and matched controls were available (20 from the prospective cohort and 57 from the retrospective cohort); the most common infecting isolates were serotype III (58%) and serotype Ia (23%); all other serotypes accounted for 6% or less. The overall incidence of early-onset and late-onset GBS disease was 0.72 and 0.83 per 1000 live births, respectively, and were higher among births at <34 weeks’ gestation as compared to ≥34 weeks’ gestation. Among births at <34 weeks’ gestation, the incidence of late-onset disease was nearly double that of early-onset disease (3.05 vs 1.53 per 1000 live births).

Determination of serocorrelate for infant protection: Due to the available sample size, investigators only determined the risk-concentration relationships for serotypes Ia and III. The assay was cross-standardized to allow for comparison between serotypes, and risk-concentration curves were compared between cases and controls for serotype Ia, serotype III, and the other four serotypes combined. An anti-CPS IgG concentration of 0.184 to 0.827µg/mL was determined to be the threshold associated with 75% to 95% risk reduction.

Vaccine trial:

• Safety: Maternal adverse events were similar at all time points between the two groups, except for pain at the injection site which was more common among women who received vaccine with the aluminum phosphate component. The one stillbirth in the vaccine group was determined by investigators to be unrelated to the vaccine. Infant adverse events were also similar between the two groups, and no infant adverse event was determined to be related to the vaccine or placebo. There were three infant deaths (one in the vaccine group and two in the placebo group); all were determined by investigators to be unrelated to the vaccine or placebo.
• Immunogenicity: All vaccine formulations led to maternal anti-CPS IgG antibody production responses compared to placebo. There was variation in antibody response by serotype, and a higher vaccine dose (20µg) without the aluminum phosphate component led to higher antibody concentrations. Placental antibody transfer ratios of 0.4 to 1.3 were observed for the different serotypes across vaccine formulations. The proportions of infants with anti-CPS IgG greater than the proposed lower limit for disease protection are shown in the Table. For the higher vaccine dose, 57 to 97% of infants had anti-CPS IgG concentrations about the 75% risk reduction serocorrelate of protection (0.184µg/mL).

Table: Proportions of infants with anti-CPS IgG greater than the proposed lower limit for disease protection

Serotype	% of infants with anti-CPS IgG > proposed lower limit for disease protection*
Ia	89-97
Ib	49-71
II	97-100
III	72-83
IV	85-97
V	36-57

*ranges represent the results from vaccination with the lowest (5 mcg) to highest (20 mcg) CPS content preparations

CONCLUSION

Maternal antibody responses to all six GBS serotypes were induced and transferred to the newborns at protective threshold levels. No safety concerns with the GBS6 vaccine were identified. Formulations with aluminum phosphate were more immunogenic but also associated with more local reactions. A single (not serotype-specific) protective threshold of 0.184µg/mL of anti-CPS IgG may be associated with a 75% risk reduction for invasive infant GBS disease.

COMMENTARY

Group B Streptococcus (GBS) causes approximately 150,000 stillbirths and infant deaths annually worldwide.¹ Infants who are exposed to maternal GBS colonization are at risk of sepsis and meningitis, which when not fatal, are associated with significant life-long morbidity.² Accordingly, the development of a maternal GBS vaccine is a public health priority.¹ A vaccine could help prevent GBS-associated morbidities and mortality. The results of this phase 2, randomized, placebo-controlled trial provide not only evidence that Pfizer’s GBS hexavalent vaccine (GBS6) is safe and immunogenic; the parallel seroepidemiology data also provide an evidence-based serocorrelate of protection that can be utilized in future studies.³

 

In some regions, universal GBS screening and targeted intrapartum antibiotic prophylaxis (IAP) have led to a substantial decline in cases of early-onset disease, though several limitations serve as the foundation for maternal GBS vaccine development.⁴ First, in many regions, IAP is not standard of care. Second, early-onset disease still persists in regions utilizing IAP due to limitations in antenatal screening. Third, a substantial portion of early-onset disease occurs among preterm infants, who may not benefit from late pregnancy screening and IAP. Fourth, IAP does not impact late-onset disease. Lastly, there are evolving concerns about the impact of IAP on maternal and neonatal dysbiosis, and selection pressure leading to colonization and infection with antibiotic-resistant organisms.⁵

 

The investigators provide reassuring data that GBS6 is safe. Maternal and infant adverse events were similar between groups and no serious adverse events were attributed to the vaccine. The vaccine was immunogenic in pregnant women and placental antibody transfer, known to be less robust for GBS compared with other pathogens, was achieved at concentrations that correlate with proposed threshold concentrations for protection against invasive infant disease. Importantly, investigators used a cross-standardized immunoassay and propose a single protective threshold that could be utilized in ongoing and future GBS vaccine research.

 

The study had several limitations. The study was conducted in a setting that does not use IAP, as the impact of IAP on infant infection could cloud the picture when assessing a serologic threshold of protection. Serologic estimates from low- and middle-income countries may overestimate thresholds for high-income countries. Demographics of the study cohort may limit generalizability: enrolled pregnant persons were generally young, multiparous, of Black-African race/ethnicity, and roughly one third had HIV infection. The estimated serocorrelate of protection should also be viewed with caution. Acknowledging the remarkable effort represented by this study, the serocorrelate modeling was based on measurements from only 77 cases: when broken out by 6 serotypes and early- and late-onset disease, only type III infection was represented by >10 cases. The authors emphasize that the reported serocorrelate must be better defined in future studies with different populations. Such studies are needed to address the effect of pre-existing, naturally-acquired serotype-specific antibody on vaccine response, as well as maternal morbidities that may affect placental antibody transfer. Finally, a larger set of case-control data encompassing all common serotypes and both early- and late-onset disease is needed to generate robust estimates for serotype-specific serocorrelates of protection.

 

REFERENCES

1. Immunization, Vaccines and Biologicals. https://www.who.int/teams/immunization-vaccines-and-biologicals/diseases/group-b-streptococcus-(gbs). Accessed August 24, 2023.

2. Nakwa FL, Lala SG, Madhi SA, Dangor Z. Neurodevelopmental Impairment at 1 Year of Age in Infants With Previous Invasive Group B Streptococcal Sepsis and Meningitis. Pediatr Infect Dis J. 2020;39(9):794-798. doi:10.1097/INF.0000000000002695

3. Madhi SA, Anderson AS, Absalon J, et al. Potential for Maternally Administered Vaccine for Infant Group B Streptococcus. N Engl J Med. 2023;389(3):215-227. doi:10.1056/NEJMOA2116045/SUPPL_FILE/NEJMOA2116045_DATA-SHARING.PDF

4. Puopolo KM, Lynfield R, Cummings JJ. Management of infants at risk for group B streptococcal disease. Pediatrics. 2019;144(2). doi:10.1542/peds.2019-1881

5. Nogacka A, Salazar N, Suárez M, et al. Impact of intrapartum antimicrobial prophylaxis upon the intestinal microbiota and the prevalence of antibiotic resistance genes in vaginally delivered full-term neonates. Microbiome. 2017;5(1). doi:10.1186/S40168-017-0313-3

 

 

CONFLICTS OF INTEREST

The authors report no conflicts of interest.

FUNDING

The authors report no funding for this work.