Haemophilus influenzae Type b Vaccine Immunogenicity in American Indian/Alaska Native Infants

We conducted a phase IV, prospective, 1:1 randomized, open-label, parallel group, noninferiority clinical trial comparing the immunogenicity of standard-of-care Hib-containing vaccine (PedvaxHIB) primary series to a primary series with Vaxelis, as measured by anti-PRP immunoglobulin G (IgG) concentration (US National Library of Medicine ClinicalTrials.gov Identifier: # NCT04978818). We recruited AI/AN infants receiving primary care at 5 Indian Health Service/Tribal Health facilities in Alaska and the Navajo Nation: Alaska Native Medical Center (Anchorage, Alaska), Chinle Comprehensive Health Care Facility (Chinle, Arizona), Gallup Indian Medical Center (Gallup, New Mexico), Northern Navajo Medical Center (Shiprock, New Mexico), and Tséhootsooí Medical Center (Fort Defiance, Arizona). The study was approved by the Johns Hopkins Bloomberg School of Public Health institutional review board (#00011170), the Navajo Nation Human Research Review Board (#20.374), the Fort Defiance Indian Hospital Review Board, the Alaska Area institutional review board (#2020-02-011), the Southcentral Foundation Executive Committee, and the Alaska Native Tribal Health Consortium Human Research Review Committee. Written, informed consent was obtained from the parent or legally authorized representative (LAR) of infant participants before the initiation of any study activities. Participants were randomized immediately before administration of the study vaccine using an allocation sequence generated by an independent statistician in the randomization module within the Research Electronic Data Capture data collection and management system. The allocation sequence was concealed until participants were assigned. Randomization was blocked and stratified by study site. Only the laboratory personnel assessing anti-PRP concentration were blinded to vaccination group. Other study personnel, participants, care providers, and data analysts were not blinded to vaccination group assignment.

We enrolled healthy infants, born at a gestational age of ≥35 weeks, aged 42 to 90 days at the time of the first vaccination, identified as AI/AN by the parent/LAR. Exclusion criteria included a history of receipt of blood products or immunoglobulin products, any confirmed or suspected condition which may have interfered with the evaluation of the immunogenicity results, or acute illness at the time of the first vaccination. See Supplemental Information for full eligibility criteria.

PedvaxHIB (Merck & Co, Inc., Rahway, New Jersey) contains 7.5 µg PRP and 125 µg OMP. Vaxelis (distributed by Merck Sharp & Dohme Corp, a subsidiary of Merck & Co, Inc, Kenilworth, New Jersey; and Sanofi Pasteur Inc, Swiftwater, Pennsylvania) contains 3.0 µg PRP and 50 µg OMP, as well as diphtheria, tetanus, pertussis (acellular, component), poliomyelitis (inactivated), and hepatitis B (recombinant DNA). Vaxelis is licensed and recommended for use as a 3-dose primary series in infants in the United States.^15,19 

Study vaccines were administered intramuscularly at age 2 months (day 1) and 4 months (day 61) for participants in the PedvaxHIB group, and at age 2 months (day 1), 4 months (day 61), and 6 months (day 121) for participants in the Vaxelis group (Table 1). Protocol-defined visit windows are defined in Table 1. Infants participating in the study received their other routine immunizations according to the US vaccination schedule for children.

Given that both PedvaxHIB and Vaxelis are licensed products, we did not actively monitor reactogenicity. All serious adverse events (SAEs) were recorded and reported throughout the entire study period. SAEs were assessed via parental questionnaire and chart review at each study visit. SAEs were assessed for severity and causality by the investigators.

Serum specimens were collected by venipuncture, or heel stick if venipuncture was unsuccessful, at day 1 before the first dose, day 31, day 121, and day 151 (Table 1). Serum anti-PRP IgG was measured using the Binding Site VaccZyme Human Anti-Haemophilus influenzae type b Enzyme Immunoassay Kit. This assay is calibrated against the Lot 1983 standard from the Center for Biologics Evaluation and Research, US Food and Drug Administration. The sensitivity of the assay (0.11 µg/mL) is also the lower limit of detection (LLD), and the measuring range of the assay is 0.11 to 9.0 µg/mL. The instructions for use were followed, with the exception that all wash steps were performed 5 times rather than 3 to allow for testing mildly and moderately hemolyzed sera without dilution. Hemolysis was assessed at the time of testing according to a visual grading chart. Grossly hemolyzed sera were tested after diluting them 1:2 or 1:4. Grossly hemolyzed sera with results below the LLD after dilution were excluded from analysis. Sera with anti-PRP IgG concentrations >9.0 µg/mL were diluted 1:10. Anti-PRP IgG concentrations for all diluted samples were multiplied by the dilution factor for reporting. For calculations, samples with anti-PRP IgG concentrations lower than the LLD were assigned a value of half the LLD (0.055 µg/mL).

The primary objective was to estimate the vaccine-group specific change in anti-PRP IgG geometric mean concentration (GMC) in AI/AN infants from baseline to 30 days after vaccination to determine whether Vaxelis was noninferior to PedvaxHIB. The anti-PRP IgG GMC for children receiving dose 1 of Vaxelis was defined a priori as noninferior if it was within a 1.5-fold margin of the GMC for children receiving dose 1 of PedvaxHIB, corresponding to the lower bound of the 95% confidence interval (CI) of the IgG GMC ratio (Vaxelis to PedvaxHIB) being >0.67.

To facilitate comparisons with previous studies of Hib vaccine immunogenicity, the secondary objectives were to describe the proportion of participants in each group with anti-PRP IgG concentration above the putative correlates of short-term (≥0.15 µg/mL) and long-term (≥1.0 µg/mL) protection at each specimen collection time point, and to describe the proportion of participants in each group with at least a fourfold rise in anti-PRP IgG between baseline and each of the subsequent timepoints.^20–22 

A sample size of 150 infants per group was calculated to achieve 80% power to detect noninferiority, defined as the lower bound of the 95% CI of natural log anti-PRP IgG GMC ratio of Vaxelis to PedvaxHIB ≥0.67, assuming a true ratio of 1. An enrollment target of 330 infants was set to allow for 10% participant attrition.

The total vaccinated cohort (TVC) included all evaluable specimens (specimens from which valid concentration measurements were made) from participants who received at least 1 dose of study vaccine. The according-to-protocol cohort included all evaluable specimens from the TVC adhering to study procedures and intervals between primary doses, as defined in the protocol (Table 1). SAEs were analyzed among the TVC cohort.

Constrained longitudinal data analysis (cLDA), an analytical technique wherein the baseline means are assumed to be equal between groups as a feature of the randomization, was used to evaluate the primary outcome.^23,24  The cLDA approach is tolerant of moderate deviations from the assumed normal data distribution and allows the contribution of data from participants who do not have complete data for all study visits. In the cLDA model, the log anti-PRP IgG concentration was regressed on whether the measure was pre- or postvaccination, and the interaction of this indicator with vaccination group. The main effect of vaccination was omitted because of the assumption of equal means at baseline. We assumed a common unstructured random effects covariance matrix and residual for the 2 groups and unstructured covariance across visits. See Supplemental Information for additional details on cLDA methods.

The proportion of children with anti-PRP IgG above the putative thresholds for short-term protection (≥0.15 μg/mL) and long-term protection (≥1.0 µg/mL) were compared using the χ² test at days 1, 31, 121, and 151.²¹  The proportions of children with at least a fourfold rise in anti-PRP IgG from day 1 to subsequent visits in the Vaxelis group versus the PedvaxHIB group were compared using the χ² test.

Missing or nonevaluable measurements were not replaced. Participants with missing or nonevaluable measurements for a specific time point were excluded from the analysis for that time point only.

Of the 1736 infants screened for eligibility from December 2021 to April 2023, 333 infants were enrolled in the study: 166 were assigned to the PedvaxHIB group and 167 to the Vaxelis group (Fig 1; reasons for decline or exclusion are shown in Supplemental Tables 3 and 4). Participant follow-up concluded in October 2023. All participants were identified as AI/AN by the consenting parent/LAR. The randomization groups were balanced on key characteristics (Supplemental Table 5). Median age at randomization was 60 days (interquartile range: 46–63 days); 47.5% of participants were male. Participant progression through the study and number of evaluable samples obtained per visit are presented in Fig 1. The protocol-specified specimen collection windows and median days between vaccination and specimen collection are described in Supplemental Table 6. No crossover between vaccination groups was observed during study follow-up.

The observed postdose 1 anti-PRP GMC was 0.39 µg/mL (95% CI 0.31–0.50) in the PedvaxHIB group and 0.41 µg/mL (95% CI 0.33–0.52) in the Vaxelis group (Table 2). The postdose 1 GMC ratio of Vaxelis to PedvaxHIB, as modeled by cLDA, was 1.03 (95% CI 0.75–1.41). The lower bound of the 95% CI was greater than the prespecified noninferiority margin of 0.67. Thus, Vaxelis postdose 1 anti-PRP IgG immunogenicity met the noninferiority criterion compared with PedvaxHIB.

The anti-PRP GMCs at days 1, 31, and 121 were similar between the PedvaxHIB and Vaxelis and groups (Table 2, Fig 2). At day 151, the GMC was higher in the Vaxelis group compared with the PedvaxHIB group (GMC ratio 1.78 [95% CI 1.66–2.60; P = .003]).

At day 31, the proportion of participants with anti-PRP IgG concentrations ≥0.15 µg/mL was 71.2% (95% CI 63.3–78.0) in the PedvaxHIB group and 75.7% (95% CI 68.2–81.8) in the Vaxelis group (Fig 3A, Supplemental Table 7). At day 31, the proportions of participants with anti-PRP IgG concentrations ≥1.0 µg/mL were 27.4% (95% CI 20.8–35.2) and 25.0% (95% CI 18.7–32.5) in the PedvaxHIB and Vaxelis groups, respectively (Fig 3B, Supplemental Table 7). The only statistically significant difference between the vaccination groups was in the proportion above 1.0 µg/mL at day 151 (71.8% [95% CI 62.9–79.2] in the PedvaxHIB group and 83.6% [95% CI 76.1–89.1] in the Vaxelis group; P = .03; Fig 3B, Supplemental Table 7). The proportion of participants with a fourfold, or greater, increase in anti-PRP IgG concentration between day 1 and the subsequent time points was similar between groups (Fig 4, Supplemental Table 8).

SAEs occurred in similar numbers between groups (15 SAEs in 12 participants in the PedvaxHIB group and 10 SAEs in 9 participants in the Vaxelis group); none were considered by the investigators to be related to study vaccination. The most common SAE was acute respiratory infection (n = 21).

Previous studies have shown that Vaxelis is safe and highly immunogenic after at least 2 doses in AI/AN children.¹⁸  However, postdose 1 data are crucial for determining if Vaxelis, which contains less Hib antigen than the vaccine currently preferentially recommended for AI/AN infants (PedvaxHIB), provides noninferior protection at the historic age of peak Hib risk for AI/AN infants. We found that the antibody response to the first dose of Vaxelis was noninferior to that after the first dose of PedvaxHIB, thereby supporting inclusion of Vaxelis among the preferred Hib vaccines for AI/AN infants.

The current preferential recommendation for use of PRP-OMP Hib conjugate vaccine in AI/AN children was informed by considerations unique to this population. Given the historic peak of Hib disease at 4 to 8 months of age in AI/AN infants, the use of a vaccine that provides protection early in life is vital.^7,9,10  Studies showing that the majority of infants developed anti-PRP concentrations ≥0.15 µg/mL after the first dose of PRP-OMP vaccine, with little invasive Hib disease observed after the first dose in infancy, suggested that this conjugate vaccine would provide early protection. Indeed, widespread use of PRP-OMP vaccine precipitated a substantial decline in invasive Hib disease among AI/AN infants.^11–14  The importance of PRP-OMP vaccine in this population was underscored in the late 1990s, when use of HbOC (a different Hib conjugate vaccine that was conjugated to diphtheria toxin CRM197), which did not achieve protective antibody concentrations until after the third dose, was associated with an increase in invasive Hib disease in rural AN children.²⁵ 

In the study presented herein, 71.2% of infants in the PedvaxHIB group and 75.7% of infants in the Vaxelis group had anti-PRP concentrations above the putative threshold of short-term protection 30 days after the first dose. These proportions are lower than was observed in studies during the prevaccine era, in which the proportions ranged from 80% to 95%.^12–14,26  Potential reasons for this include:

1. differences in laboratory assays (eg, radioimmunoassays used in the prevaccine era detected both IgG and non-IgG anti-PRP antibodies, whereas the enzyme-linked immunosorbent assay used in this study measures only IgG antibody); and

2. shifts in epidemiology after nearly 3 decades of routine use of Hib vaccines, which has resulted in reduced circulation of Hib and less opportunity for natural boosting.

At day 121 (∼6 months of age, 2 months after the second dose), there was no statistically significant difference in anti-PRP GMC between groups. The proportions of participants with anti-PRP concentrations above the short- and long-term thresholds of protection were similar between groups, consistent with previous studies.^18,22  In a multicenter trial of Vaxelis, 97% and 73% had postdose 2 anti-PRP IgG ≥0.15 μg/mL and ≥1.0 μg/mL, respectively, similar to the 95% and 76% reported herein.^18,27 

At day 151 (∼7 months of age, 3 months after the last primary series dose for the PedvaxHIB group and 1 month after the last primary series dose for the Vaxelis group), the anti-PRP GMC was higher among Vaxelis recipients, likely reflecting the receipt of a third dose of vaccine 1 month before. The proportion of participants with anti-PRP concentrations above the short-term threshold of protection was similar between groups, but the proportion of recipients above the long-term threshold of protection was higher in the Vaxelis group compared with the PedvaxHIB group (84% vs 72%, P = .03). In a prelicensure trial of Vaxelis, 1-month postdose 3, the proportion of infants with anti-PRP IgG ≥1.0 μg/mL was somewhat higher at 92.7% (95% CI 86.7%–96.6%) among a subset of AI/AN infants, as compared with the proportion observed in this study.¹⁸ 

Nearly all infants receiving either vaccine regimen had protective levels of anti-PRP antibodies at day 151. Because Vaxelis recipients receive the final dose of their primary series at 6 months, rather than 4 months, their antibody decay starts at a later age than PedvaxHIB recipients. This study only followed participants through 7 months of age, so we were unable to evaluate longer-term antibody persistence. Among AI/AN children living in the Navajo Nation and White Mountain Apache Tribal lands in the Southwest United States, most invasive Hib disease cases now occur outside of infancy, so providing durable protection is critical.^28–30  Data from previous studies suggest that booster doses after a primary series of Vaxelis are immunogenic. Notably, a heterologous booster (ie, PRP conjugated to a carrier protein other than OMP), such as PRP conjugated to tetanus toxoid, at 12 to 15 months may provide stronger and more durable protection than a homologous booster.^18,31–34  Future work should address whether a heterologous booster dose could better protect AI/AN infants and young children from invasive Hib disease.

We enrolled only children from AN and Navajo Nation communities in this study, so the results may not be generalizable to other settings. In addition, this study assessed immunogenicity, rather than disease outcomes; therefore, we cannot directly compare the efficacies of the vaccines. Because cell-mediated immunity and immunologic memory cannot be measured by antibody concentration, anti-PRP immunogenicity may be an underestimate of the protection conferred by these vaccines.^21,35 

Vaxelis postdose 1 immunogenicity met the prespecified noninferiority criterion compared with PedvaxHIB in AI/AN infants, a population with elevated risk of invasive Hib disease. Including Vaxelis among the preferentially recommended Hib vaccines for this population would increase the options available and provide an opportunity to optimize protection of AI/AN infants, given the known benefits of combination vaccines.

We thank the study participants and their families, our partners at participating Indian Health Service/Tribal Health facilities, as well as the study staff who assisted with this study. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Indian Health Service or the Centers for Disease Control and Prevention.

AI/AN

American Indian/Alaska Native

CI

confidence interval

cLDA

constrained longitudinal data analysis

GMC

geometric mean concentration

Hib

Haemophilus influenzae type b

IgG

immunoglobulin G

LAR

legally authorized representative

LLD

lower limit of detection

PRP-OMP

polyribosylribitol phosphate covalently linked to the outer member protein

SAE

serious adverse event

TVC

total vaccinated cohort