Introduction

Infection with Streptococcus pneumoniae can result in serious illness, with pneumonia and invasive pneumococcal diseases (IPD) such as meningitis and bacteremia the most common manifestations in adults. Pneumococcal vaccines are an effective prophylactic health intervention, reducing disease incidence in countries where pneumococcal vaccination is utilized as part of national immunization programs. Indirect protection has been observed for adults from vaccinated children1,2,3,4, and there is a further direct benefit of adult vaccination5,6,7,8.

The bacterial capsular polysaccharide (CPS) is the major virulence factor of S. pneumoniae, and differences in CPS structure determine the approximately 100 pneumococcal serotypes identified to date9. An immune response raised against the CPS can protect the host via complement-mediated killing, and therefore, pneumococcal vaccines are designed to elicit an immune response to CPS for the serotypes targeted by the vaccine. In general, anti-CPS antibodies are specific to the pneumococcal polysaccharides contained in the vaccine, but vaccination can also induce antibodies to CPS epitopes from a serotype contained in the vaccine that cross-react to other serotypes10. Cross-reactivity, and in some cases cross-protection, has been demonstrated following pneumococcal vaccination for serogroups 6 and 15, among others11,12,13,14,15. A potential limitation is that cross-reactive antibodies can have a lower capacity to opsonize bacteria and, therefore, may not confer cross-protection against disease and/or prevention of nasopharyngeal carriage16,17.

V116 is a 21-valent PCV specifically designed to protect against serotypes associated with the majority of residual IPD among adults ≥65 years of age in countries with established pediatric national immunization programs. V116 contains the pneumococcal polysaccharide antigens 6A and 15A, as well as de-O-acetylated 15B, which is structurally similar to serotype 15C. A pre-specified objective of the V116 development program was to assess cross-reactivity of V116-induced antibodies and OPA for non-vaccine serotypes 15B and 6C. Here, we report a secondary analysis of five V116 phase 3 trials with adult participants that collected immunogenicity data on serogroups 6 and 15. Vaccination with V116 induced functional cross-reactive antibodies to serotypes 15B and 6C, supporting broadened serotype coverage for pneumococcal disease (PD) prevention beyond the serotypes included in the vaccine.

Methods

Study design

Group 1 participants were derived from V116-003 (clinicaltrials.gov NCT05425732)18 and V116-010 (clinicaltrials.gov NCT05569954)19 studies. Participants were assigned to receive either V116 or an active control (PCV20 or PPSV23) on day 1. Group 2 participants included pneumococcal vaccine-naïve adults living with HIV (V116-007, clinicaltrials.gov NCT05393037) and adults with chronic medical conditions that increase the risk of PD (V116-008, clinicaltrials.gov NCT05696080). Participants included in this analysis were assigned to receive either V116 on day 1, followed by placebo 8 weeks later, or PCV15, followed by PPSV23 8 weeks later. Trials V116-007 and V116-008 are each in the process of publication, and the study data have been publicly disclosed. Group 3 participants were derived from the V116-006 (clinicaltrials.gov NCT05420961)20 study which included participants who had previously received a pneumococcal vaccine(s) at least 1 year prior to enrollment. Participants included in this analysis were assigned to receive either V116 or an active control (PCV15 or PPSV23), depending on the participants’ prior pneumococcal vaccination history. For Groups 2 and 3, only participants who received V116 were included in this analysis. All individual studies assessed the safety and tolerability of the study vaccines, and these data can be found in the above-referenced reports or are currently in the process of publication. Written informed consent was obtained from each participant prior to any study procedure. Participants included in this analysis were from a total of 23 countries (Supplementary Table 1) and the IRBs/IECs that approved the studies are found in Supplementary Data 1. Participant inclusion criteria are found in Supplementary Table 2.

Treatment allocation/randomization occurred centrally using an interactive response technology system. Study vaccines were prepared and administered by unblinded, qualified study site personnel who were not involved in subsequent participant assessments or study procedures. An external data monitoring committee conducted periodic safety reviews for the studies. A scientific advisory committee comprised of external and MSD scientists contributed to the development of the protocols and statistical analysis plan. The studies were conducted in accordance with the principles of Good Clinical Practice and approved by the appropriate institutional review boards/ethics committees.

Vaccines and administration

V116 (Merck, Sharp & Dohme LLC, a subsidiary of Merck & Co, Inc., Rahway, NJ, USA [MSD]) is a 21-valent PCV. Each 0.5 mL dose contains 4 µg of each pneumococcal capsular polysaccharide (PnPs): 3, 6A, 7F, 8, 9N, 10A, 11A, 12F, 15A, deOAc15B, 16F, 17F, 19A, 20A, 22F, 23A, 23B, 24F, 31, 33F, and 35B, all individually conjugated to CRM197. The deOAc15B PnPs is referred to as 15C as it is structurally similar to the 15C PnPs and antibodies to 15C were evaluated in each study. PCV20 (Pfizer) and PPSV23 (MSD) are licensed pneumococcal vaccines. Study vaccines were supplied as prefilled syringes, stored at 2–8 °C, and were all administered as 0.5-mL intramuscular dose. Participants were observed for at least 30 minutes post-vaccination for immediate reactions.

Immunogenicity assessments

Blood was collected prior to study vaccination and 30 days following vaccination for measurement of serotype-specific anti-PnPs antibodies. Functional antibodies were measured using a serotype-specific microcolony multiplex opsonophagocytic killing assay (MOPA). Serotype-specific anti-PnPs IgG antibodies were evaluated with a multiplexed electrochemiluminescence assay. Both assays were developed by Merck & Co., Inc., Rahway, NJ, USA21,22. Both assays were adapted to V116 vaccine serotypes and validated for use in the V116 clinical program23. For a brief description of MOPA, heat-inactivated serum was added to serotype-specific Streptococcus pneumoniae to allow for specific binding of antibodies. Rabbit complement was then added to initiate complement deposition on the surface of the bacteria. The prepared bacteria were then incubated with HL60 macrophages, which subsequently phagocytose and kill bacteria bound by specific antibody and complement. A reduction in subsequent bacterial colony growth was used to determine the degree of bactericidal activity in the sample. While adapting the MOPA assay to V116 serotypes, potentially cross-reactive antibodies were placed on separate assay cassettes to avoid competition within each cassette.

Competition assay: in order to test potential cross-reactivity between 6A/6C and 15B/15C, 100 µg/mL of either purified homologous, heterologous, or potentially cross-reactive polysaccharides were pre-incubated with 29 clinical samples before testing in the MOPA assay. The competition assay included participant samples with antibody titers in a range that were sufficiently inhibited by pre-incubation with 100 µg/ml of homologous competitor. In the case of serotype 6A specificity, homologous PnPs were 6A, heterologous PnPs were 15C, and potentially cross-reactive PnPs were 6C. In the case of serotype 15C specificity, homologous PnPs were 15C, heterologous PnPs were 6A, and potentially cross-reactive PnPs were 15B.

For all studies included in this analysis, immunogenicity objectives included evaluation of opsonophagocytic activity (OPA) reported as geometric mean titers (GMTs) at 30 days after vaccination. Evaluation of serotype-specific IgG geometric mean concentrations (GMCs) and the proportion of participants with a ≥4-fold rise in serotype-specific antibodies from before to 30 days following vaccination were additional endpoints.

Statistics and reproducibility

For the continuous endpoints, the within-group 95% confidence intervals (CIs) were obtained by exponentiating the CIs of the mean of the natural log values based on the t-distribution. For the dichotomous endpoints, the within-group 95% CIs were based on the exact binomial method proposed by Clopper and Pearson24. Analyses were conducted on the “per protocol” population, which consisted of all participants without protocol deviations that may substantially affect the results of the immunogenicity endpoints. Immunogenicity analyses were further restricted to samples with sufficient serum volume to perform the assays.

Analysis software

All analyses were performed using SAS software, version 9.4, of the SAS System for Unix (copyright 2012 SAS Institute, Inc).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Results

A total of 4829 participants were included in this secondary analysis, 2911 of whom received V116. For each of the individual studies included in the analysis, >95% of participants completed the study (data not shown). Participants were categorized as Group 1: pneumococcal vaccine-naïve participants ≥50 years of age who received V116, PCV20, or PPSV23; Group 2: pneumococcal vaccine-naïve participants 18-64 years of age with specific stable, chronic medical conditions, including persons living with HIV, that increase risk of PD; and Group 3: participants ≥50 years of age with prior pneumococcal vaccination as adults (Supplementary Table 1). Baseline characteristics are shown by group in Table 1. All participants in Group 2 had specific, stable, chronic medical conditions, and 33.2% and 43.7% of participants in Groups 1 and 3, respectively, had stable, chronic medical conditions with the most common chronic risk condition being diabetes mellitus.

Table 1 Participant characteristics
Full size table

Fig. 1 shows data from a serotype-specific MOPA (see Methods) for serotypes 6A/6C and 15B/15C using homologous, heterologous, and potentially cross-reactive competitors on samples from unvaccinated participants or those who received PPSV23 or V116 in phase 1 studies. For all serotypes, pre-incubation with homologous competitor polysaccharide resulted in high geometric mean fold inhibition while heterologous competitor polysaccharide resulted in low (≤1.4) geometric mean fold inhibition (Fig. 1 and Supplementary Table 3). A proportion of the serotype 6A antibody pool (V116 serotype) was inhibited by pre-incubation with 6C competitor polysaccharide (non-V116 serotype), and a higher amount of inhibition was seen when 6A competed with 6C. Inhibition also occurred with combinations of 15B (non-V116 serotype) and 15 C (V116 serotype) in the assay (Fig. 1 and Supplementary Table 3). These data suggest vaccine-induced antibodies elicited against serotypes 6A and 15C cross-react to serotypes 6C and 15B, respectively.

Fig. 1: MOPA assay specificity among serogroups 6 and 15.
figure 1

Results of competition assays are shown with pre-incubation of homologous, heterologous, and potentially cross-reactive polysaccharides described in the x-axis labels. Geometric mean fold inhibition values are displayed above each bar. n for each group is found in Supplementary Table 3.

Full size image

This analysis further examined cross-reactive antibodies in the pooled population that received V116. Immune responses were assessed in each study by OPA GMTs, IgG GMCs, and the proportions of participants with a ≥4-fold rise in OPA and IgG responses from baseline (before vaccination) to 30 days after vaccination.

Across the 3 groups, V116 induced robust fold change in OPA GMTs to vaccine serotypes 6A and 15C as expected (Supplementary Fig. 1 and Supplementary Data 2). Non-vaccine serotypes 6C and 15B also demonstrated increased proportions of participants with a ≥4-fold rise in OPA responses after V116, in some cases comparable to that which was observed for the vaccine serotype in the same serogroup (Supplementary Fig. 1, Supplementary Fig. 4, and Supplementary Data 2). An increase in OPA GMTs was observed for both serotypes 6A and 6C after vaccination with all study vaccines (Fig. 2 and Supplementary Data 2). OPA GMTs for serotypes 15B and 15C were generally comparable for V116 recipients at day 30, and values for serotype 15B were comparable to participants who received PCV20 or PPSV23, vaccines that contain serotype 15B (Fig. 2 and Supplementary Data 2). Antibody concentrations, as assessed by serotype-specific IgG GMCs, were consistent with the OPA results (Fig. 3 and Supplementary Data 3). In Group 3 (pneumococcal vaccine-experienced participants), baseline antibody levels were higher than in pneumococcal vaccine-naïve participants, and vaccination further increased antibody levels to both vaccine serotypes and cross-reactive serotypes (Fig. 3).

Fig. 2
figure 2

OPA GMTs for pneumococcal serogroups 6 (A) and 15 (B). OPA GMTs were measured for the shown serotypes prior to (baseline) and 30 days following (day 30) vaccination with V116, PCV20, or PPSV23. 95% confidence intervals are shown for each group. n for each group is found in Supplementary Data 2.

Full size image
Fig. 3
figure 3

IgG GMCs for pneumococcal serogroups 6 (A) and 15 (B) with 95% confidence intervals shown. IgG was measured prior to (baseline) and 30 days following (day 30) vaccination with V116, PCV20, or PPSV23. n for each group is found in Supplementary Data 3.

Full size image

Discussion

This report summarizes data from more than 4800 participants examining cross-reactivity of V116-induced antibodies in serogroups 6 and 15. Prior published studies have demonstrated serogroup 6 and/or 15 cross-reactivity in individuals vaccinated with PCVs11,12,13,14, in addition to supporting data from structural and surveillance studies9,25,26,27,28. For V116, the pivotal study in the phase 3 clinical program showed cross-reactive immune responses to serotypes 6C and 15B in participants vaccinated with V11618. This secondary analysis extends the V116 data to a larger and more diverse participant population with a variety of pneumococcal vaccination histories and pre-existing chronic medical conditions. Vaccination with V116 elicited cross-reactive antibodies with comparable opsonophagocytic activity for serotype 15B compared to serotype 15C and, to a lesser extent, to serotype 6C compared to 6A. Participants with a history of pneumococcal vaccination generally had higher antibody levels prior to vaccination, but were boosted to levels comparable to pneumococcal vaccine-naïve participants following V116, suggesting that these findings are generalizable to adults with differing pneumococcal vaccination experiences.

Assessment of functional, bactericidal activity of antibodies is a surrogate of serotype-specific immune protection for pneumococcal vaccines, and thus OPA analyses are important when assessing cross-reactive antibodies since some studies have suggested lower bactericidal activity of cross-reactive antibodies elicited by PCVs compared to vaccine serotypes16,17. The specificity of the MOPA for individual serotypes within serogroups 6 and 15 was supported by homologous and heterologous data from the competition assays. Additionally, there was a positive correlation between the OPA data and total antibody level as assessed by IgG GMCs. It is still unknown whether these cross-reactive antibody responses will translate to cross-protective vaccine effectiveness for V116.

Serotype 15B CPS differs from 15C due to O-acetylation from the presence of a functional wciZ-encoded enzyme. A frameshift in the wciZ gene in serotype 15C renders it non-functional and, therefore, prevents O-acetylation29. Some have suggested that functional antibodies raised against 15B are specific for the O-acetylated polysaccharide and, therefore, have low cross-reactivity to 15C30, while others have reported antibodies raised against 15B from vaccination with PPSV23 have killing activity against 15C29. V116 was designed to elicit anti-serotype 15C antibodies through de-O-acetylation of the 15B CPS; and the modified CPS retains a very small proportion of acetyl groups. In clinical trials, comparable opsonophagocytic killing against native 15B and 15C was observed in V116-vaccinated participants. These data suggest that functional antibodies against 15B are not limited to those that bind to the acetyl group. A measure of cross-reactivity could also be observed in the inverse for PCV20 and PPSV23, which both contain serotype 15B and demonstrated some cross-reactivity to 15C. For serogroup 6, prior published data demonstrated that cross-reactive immune responses to serotype 6C were more robust from vaccines that included both 6A and 6B (PCV13 and PCV15) than the vaccine that included only 6B (PCV7)11,12,14,31. Here, antibodies raised against serotype 6A by V116 displayed at least as much cross-reactivity to serotype 6C as PCV20 (containing serotypes 6A and 6B) and PPSV23 (containing serotype 6B). These data reflect what is understood regarding the structural similarities of serotypes 6A and 6C which only differ genetically in the coding of the wciN glycosyl transferase, while 6B and 6C are less similar32.

Published studies within the V116 clinical program have demonstrated the safety, tolerability, and immunogenicity of V116 in adults18,20,33,34. Surveillance data in the United States showed that the serotypes in V116 account for approximately 85% of IPD in adults ≥50 years of age35. This secondary analysis further demonstrates that cross-reactive antibodies were elicited to serotypes 15B and 6C in V116-vaccinated individuals. Taken together, these data suggest V116 has the potential to meet a significant clinical need in the prevention of adult PD.