Fluctuations in serogroup B meningococcal vaccine antigens prior to routine MenB vaccination in France

Categories: Disease & Virus

March 26, 2025

Characteristics of meningococcal serogroup B isolates cases during the study period

We have selected all cultured invasive isolates of serogroup B that we were able to recover over almost 5 decades during the period 1975–2022 (n = 1691) that corresponded to all cultured serogroup B isolates from sterile sites (cerebrospinal fluid (CSF), blood and other sterile sites) and for which the year of isolation was available. Isolates were recovered from blood (587; 34.7%) and CSF (1076; 63.6%) in addition to other sterile sites (n = 28; 1.7%).

We divided the isolates into four equal periods of 12 years (1975–1986, 1987–1998–1999–2010 and 2011–2022). However, the isolates were distributed unevenly into the four periods (n = 173, n = 459, n = 47 and n = 1012 respectively). Moreover, the proportion of isolates that were recovered from blood significantly increased from the first period to the fourth period (19.7–44.9%, p = 0.0008). We then screened for additional data (age-group, and sex). Sex was available for 1665 isolates (98.5%). The ratio male/female of 1.1 was obtained for the whole set of isolates and of 1.1, 1.2, 1.1 and 1.0 for the four periods 1975–1986, 1987–1998–1999–2010 and 2011–2022 respectively, with no significant difference (p = 0.27).

Next, we considered age groups distribution (<1 year, 1–4 years, 5–14 years, 15–24 years, 25–44 years, 45–64 years, 65 years and older). These data were available for 1499 cases (88.6%) for the whole period with more missing data from the first period 1975–2012 (age data were available for 73.3% versus 99.4% for the periods 1975–2012 and 2013–2022 respectively). The overall distribution of cases according to age groups is depicted in Fig. 1 that suggested different distributions between the four periods with a shift to older age groups during the 2011–2022 period.

Clonal complexes (CC) distribution

We performed Illumina sequencing for the whole set of isolates and whole genome sequencing (WGS) data were used to extract typing data. The overall distribution is depicted in Fig. 2 according to the major hyperinvasive clonal complexes (CC8, CC11, CC32, CC41/44 and CC269)²⁴. A significantly different distribution was observed between the four periods (p < 0.0001) with more isolates of CC8, CC11 during the first two periods (1975–1986 and 1987–1998) when compared to the last two periods (1999–2010 and 2011–2022) (Fig. 2). The isolates of CC8 have not been detected since 2005 and were absent during the last period 2011–2022. The CC11 isolates were also more frequent during the first two periods. As those isolates were usually reported to belong to serogroups C and W, we compared their genetic relationships to isolates of serogroups C and W. The gene-by-gene approach of the core-genome MLST (cgMLST v2) was used as described in the Methods section. The B:CC11 isolates were distributed into several clades on the tree with the C:CC11 and W:CC11. One clade corresponded to a highly linked B:CC11 isolates that were all detected in 1988 (Supplementary Fig. 1).

Isolates belonging to CC32, CC41/44 and CC269 showed fluctuations over the four periods and the B:CC32 isolates seem to increase in France since the second period 1987–1998 (Fig. 2).

It is worthy to note that during the last 3 years, 2020–2022, that were marked by the COVID-19 pandemic, CC32 and CC41/44 remained the most prevalent CCs as for the entire second period 2011–2022 (41% and 19% versus 34.0% and 20.5%, respectively).

Characterisation of fHbp peptides

We next analysed the alleles of fhbp gene among the 1691 isolates and deduced the corresponding amino acid sequences. fHbp peptides in isolates were diverse and 141 different peptides were detected among the tested NmB isolates with 73 peptides appearing only once and 1 new peptide in an isolate from the year 1975. Two fHbp peptides were the most frequent: peptide 510 and peptide 1 (the peptide included in the 4CMenB vaccine) that were represented by 225 and 224 isolates and accounted for 13.3% and 13.2% respectively of all isolates. Most fHbp peptides belonged to variant 1 (n = 952; 56.3%). Variants 2 and 3 peptides accounted for 558 isolates (33.0%) and 181 (10.7%), respectively. However, this distribution differed significantly between the four periods. The proportion of fHbp variants are shown in Table 1 with significantly higher proportion of isolates expressing fHbp variant 1 over the study periods (increasing from 24.3% for the period 1975–1986 to 65.9% for the period 2011–2022).

Table 1 Distribution of the three variants of factor H-binding protein f(Hbp) (numbers (%) according to the four periods

Individual fHbp peptides also differed between the four periods. The most frequent fHbp (peptide 510) was only present in the period 2011–2022 and was present in 22.0% of all the isolates of the fourth period. These isolates belonged in majority to CC32 (98.7%). Moreover, the second most frequent fHbp peptide (peptide 1) was more frequent during the two periods 1987–1998 and 1999–2010 compared to the two other periods 1975–1986 and 2011–2022. fHbp peptide 14 showed an increasing trend over the four periods of the study (Fig. 3). The most prevalent fHbp peptides 1 and 510 were also plotted over the study period. The increased percentage of peptide 1 from 8,6% during 1975-1986 to 25.9% during 1987–1998 was associated to the increase of CC32 isolates that rose from 9.2% to 30.5% during the same period. While the percentage of peptide 1 decreased gradually after the period 1987–1998 to reach 8.0% during 2011–2022. The peptide 510 emerged and increased remarkably after 2010. CC32 remained stable after 1986, ranging between 27.7% and 34.0% (Supplementary Fig. 2).

fHbp allele 16 (variant 2) was more frequent during the first period 1975–1986 (n = 75; 43.4%) of all the isolates of this period and then decreased gradually to account only for 1.8% (n = 19) of all isolates during the period 2011–2022 (Fig. 3).

Characterisation of NHBA peptides

We next analyzed the alleles of nhba gene among the 1691 isolates and deduced the corresponding amino acid sequences. NHBA peptides in isolates were diverse and 92 different peptides were detected among the tested NmB isolates with 38 peptides appearing only once. One hundred isolates harboured one of 71 new NHBA peptides. Two NHBA peptides were the most frequent: peptide 20 and peptide 29 that were represented by 311 and 244 isolates and accounted for 18.4% and 14.4% respectively of all the 1691 isolates. However, the distribution of these two peptides differed significantly (p < 0.0001) between the four periods. The peptide 20 was more frequent in the period 1975-1986 and peptide 29 was more frequent in the period 2011–2022 (Fig. 4). In fact, the proportions of peptide 20 decreased from 60.7% in 1975–1986 to 9.4% in 2011–2022. Most likely, this decrease was consequent to the drastic decline of CC8 isolates, as 95.5% (106/111) of CC8 isolates expressed NHBA peptide 20. At the opposite, NHBA peptide 29 increased to account for 22.2% of the isolates during the period 2011–2022 and it predominated among CC32. NHBA peptide 29 increased concurrently to fHbp peptide 510 (Supplementary Fig. 2). Indeed, NHBA peptide 29 accounted for 43.1% (221/513) of CC32 isolates, of which 93.7% (207/221) carried fHbp peptide 510. The peptide 2 (that is included in the 4CMenB vaccine) was represented by 113 isolates and accounted for 6.7% of all the 1691 peptides but increased from 1.2% during the period 1975–1986 to 9.5% during the period 2011–2022 (Fig. 4).

Characterisation of PorA peptides

The variable regions 1 (VR1) and variable regions 2 (VR2) of PorA were detected with several hundreds of combinations among the 1691 NmB isolates tested. The VR2 P1.4 corresponding to the PorA in the 4CMenB vaccine was present in 151 isolates (8.9%) but most of these isolates (94.7%) were in the period 2011–2022 (n = 143; 14.1%) and belonged in majority (54.5%) to the ST-41/44 clonal complex.

Characterisation of NadA peptides

The nadA gene was detected in 736 isolates (43.5%) for the whole period of the study (1975–2022). These isolates were distributed all over the four periods. However, the peptide 8 (included in the 4CMenB vaccine) was only present in 123 isolates that were all from the period 1975–1986 (n = 67; 38.7%) and the period 1987–1998 (n = 56; 12.0%) but absent thereafter. Isolates harbouring NadA peptide 8 were mostly of the clonal complex CC8.

MenDeVAR -based prediction coverage by 4CMenB and Bivalent rLP2086 vaccines

MenDeVAR analysis was first used to predict vaccine coverage of the 1691 NmB isolates by both vaccines 4CMenB and bivalent rLP2086. This approach classifies the isolates into four categories: exact match, cross-reactive, non-covered and insufficient data. The first two categories correspond to ‘covered isolates’. The MenDeVAR revealed high proportions of isolates with insufficient data on the alleles encoding the vaccine antigens to allow prediction. For the 4CMenB these isolates accounted for 577 isolates (34.1%) and 423 isolates (25.0%) for the Bivalent rLP2086 vaccine. Consequently, MenDeVAR predicted a coverage of 973 isolates (57.5%) for the 4CMenB vaccine and of 1268 isolates (75.0%) for the Bivalent rLP2086 vaccine of the 1961 total isolates of the study. This coverage fluctuated but not significantly over the four periods between 46.8% and 60.6% for the 4CMenB. This fluctuation was between 63.4% and 81.3% with a significant increasing trend for bivalent rLP2086 vaccine that reached the highest coverage rate of 81.3 during the period 2011–2022 (Fig. 5).

gMATS-based predicted coverage by the 4CMenB vaccine

Given the high proportion of isolates with insufficient data, we next evaluated the coverage of NmB isolates by the 4CMenB using the gMATS approach that classifies the isolates into covered, non-covered or unpredictable. When applied to the 1961 NmB isolates, gMATS predicted 1153 covered isolates, 228 non-covered isolates and 310 unpredictable isolates corresponding to an overall level of coverage of 77.4% (LL-UL 68.2–86.5%). The coverage rates did not differ significantly between the four study periods that were 85.0%, 76.3%, 74.5% and 76.7% for the 1975–1986, 1987–1998, 1999–2010 and 2011–2022 respectively (Table 2). The overall coverage was mainly by one antigen with 62.5% of the covered isolates for the whole study period and fluctuated over the four periods. The overall coverage for the COVID-19 pandemic period alone (2020–2022) was 82.3% (LL-UL 73.7–90.4%) and did not significantly differ from the overall coverage of the whole period. NHBA and fHbp with different combinations (alone or with one or two of two other antigens) contributed to the coverage for 737 isolates (63.9%) and 763 isolates (66.1%) of all covered isolates, respectively. While P1.4 antigen contributed to the coverage of 151 isolates (13.3%) but rarely alone with only 11 isolates that were covered by P1.4 alone (1%) (Fig. 6). It is worth noting that during the periods 1975–1986, 1987–1998, and 1999–2010 NHBA contributed more frequently for the coverage while fHbp contributed more to the coverage during the period 2011–2022. Additionally, PorA antigen, P1.4, also contributed more frequently to the coverage during 2011–2022 (Fig. 6).

Table 2 Coverage levels in percentage by genetic meningococcal antigen typing system (gMATS) according to clonal complexes

**Fig. 6: Proportion of meningococcal group B isolates covered by different combinations of 4CMenB vaccine antigens.**

Clonal complexes-based coverage as predicted by gMATS was calculated for the hyperinvasive CC (CC8, CC11 CC32, CC41/44 and CC269) in addition to isolates belonging to the other CC and isolates that were unassigned to a known CC (UA). Higher coverage rates for isolates belonging to CC8, CC11 and CC32 were observed in comparison to isolates belonging to CC41/44, CC269, other CCs and UA (Table 2). No significant difference was observed in coverage rates between the four periods except for CC41/44 isolates that showed a higher coverage rate during the period 2011–2022 (Table 2).

Age-related distributions of gMATS-based coverage showed slightly variable rates between different age groups with no significant differences between the four periods (Table 3).

Table 3 Coverage levels in percentage by genetic meningococcal antigen typing system (gMATS) according to age groups