Study participant characteristics

A total of 90 participants over 65 years of age were included, with 30 received the trivalent XBB.1.5 vaccine, 30 received the bivalent Omicron XBB vaccine, and 30 received the tetravalent XBB.1 vaccine. The sex, age, body mass index, and smoking status distributions were comparable across all three vaccine groups (Supplementary Table 1). Approximately half of the participants reported chronic medical conditions, with hypertension being the most common condition, accounting for 59% of the reported conditions. The percentages of underlying medical conditions among the participants in the three cohorts were comparable. Eighty-eight of 90 participants completed the full course of primary vaccination, 69 of whom received their first booster vaccination, and 19 of whom received a second booster vaccination before the BA.5/BF.7 wave that occurred in late 2022 in China. All participants reported an infection or breakthrough infection during the BA.5/BF.7 wave, and more than half of them experienced an additional infection during the XBB/EG.5.1 wave in 2023. Detailed information on the study participants’ characteristics is displayed in Supplementary Table 1.

Of 90 participants vaccinated, two experienced systemic adverse effects following vaccination (Supplementary Table 1). One participant who was immunized with the trivalent XBB.1.5 vaccine experienced a general fever of 37.6 °C, and another who received the bivalent Omicron XBB vaccine experienced slight loss of appetite, fatigue, and nausea. Both participants recovered within 2–3 days after vaccination. No participant experienced serious adverse events. Detailed information on the three vaccines is summarized in Supplementary Table 2.

Neutralizing antibodies elicited by three XBB-containing vaccines

To assess the serum neutralizing antibody titer induced by the trivalent XBB.1.5 vaccine, bivalent Omicron XBB vaccine, and tetravalent XBB.1 vaccine, we employed a pseudovirus neutralization assay to measure neutralization titers against ancestral D614G and variants BA.5, XBB.1.5, JN.1, KP.2, and KP.2. For the trivalent XBB.1.5 vaccine, all the serum samples had neutralization titers >30 against D614G, BA.5, KP.2, and KP.3 before vaccination, whereas 96.7% (29) and 93.3% (28) of the 30 serum samples had neutralization titers >30 against XBB.1.5 and JN.1 variants, respectively (Fig. 1a). After booster immunization, all seronegative (titer < 30) serum samples were seroconverted against variants XBB.1.5 and JN.1. Compared with the GMTs before vaccination the geometric mean titers (GMTs) of D614G, BA.5, XBB.1.5, JN.1, KP.2, and KP.3 were boosted 3.7-, 18.8-, 13.8-, 11.2-, 7.4-, and 4.2-fold, respectively (Fig. 1a). For the bivalent Omicron XBB vaccine, all the serum samples had neutralization titers >30 against D614G and BA.5, whereas 96.7% (29), 76.7% (23), 96.7% (29), and 93.3% (28) of the 30 serum samples had titers >30 against the XBB.1.5, JN.1, KP.2, and KP.3 variants, respectively (Fig. 1b). After immunization, all seronegative (titer < 30) serum samples seroconverted against variants XBB.1.5, JN.1, KP.2, and KP.3. The neutralization titers against D614G, BA.5, XBB.1.5, JN.1, KP.2, and KP.3 were boosted 2.4-, 5.7-, 5.2-, 7.0-, 3.2-, and 3.6-fold, respectively, in the GMTs compared with the GMTs before vaccination (Fig. 1b). For the tetravalent XBB.1 vaccine, a similar pattern to that of the trivalent XBB.1.5 vaccine was observed for the GMTs against the tested pseudoviruses, with 6.1-, 20.4-, 18.0-, 10.4-, 6.2-, and 6.5-fold increase of the GMTs compared with the GMTs before vaccination (Fig. 1c).

Fig. 1

Neutralizing antibody titers before and after immunization with three XBB-containing vaccines. a–c Neutralization of D614G, BA.5, XBB.1.5, JN.1, KP.2, and KP.3 pseudoviruses by individual-matched serum obtained before or after vaccination with the trivalent XBB.1.5 vaccine (a, n = 30), bivalent Omicron XBB vaccine (b, n = 30), or tetravalent XBB.1 vaccine (c, n = 30). Sera were collected before vaccination (“before”) and 3 weeks after vaccination (“after”). Each dot represents the 50% neutralization titer (NT50) for an individual, and a line connects the NT50 values for the same individual before and after vaccination. The horizontal dotted line in the neutralization assay indicates a limit of detection of 30, with serum samples demonstrating neutralization below 30 plotted as 10. The geometric mean titers (GMTs) of the neutralizing antibodies and the percentages of individuals with NT50 values above the limit of detection are presented on top of each group. Statistical analyses were performed using a two-tailed Wilcoxon matched-pairs signed-rank test to compare neutralizing antibody titers before and after vaccination. A p-value <0.05 was considered statistically significant, and only significant differences are displayed in the figure, with the fold change in the GMT denoted in brackets

Next, we analyzed variant-specific neutralization titers stratified before and after immunization with three XBB-containing vaccines. We observed a similar neutralization pattern for the tested pseudoviruses for individuals in the three vaccine groups before vaccination. First, neutralization titers against D614G, BA.5, and XBB.1.5 were comparable (Fig. 2a–c). Second, the neutralization titer against JN.1 was the lowest. Third, KP.2 and KP.3 showed comparable neutralization escape to the JN.1 variant. After immunization, a clearer pattern was observed in which neutralizing antibody titers against JN.1, KP.2, and KP.3 variants were significantly lower than neutralization titers against the D614G, BA.5, or XBB.1.5. The neutralization titer against BA.5 in the trivalent XBB.1.5 vaccine booster group was significantly greater than the neutralizing antibody titer against D614G, and the neutralization titer against BA.5 in the tetravalent XBB.1 vaccine group was higher than the neutralization titer against XBB.1.5. Notably, there were no significant differences in neutralization titers against the JN.1, KP.2, and KP.3 variants. (Fig. 2a–c).

Fig. 2

Neutralization of the SARS-CoV-2 lineage before and after immunization with three XBB-containing vaccines. a–c Comparison of SARS-CoV-2 lineage-specific neutralization titers against the indicated pseudoviruses before and after trivalent XBB.1.5 vaccine (a), bivalent Omicron XBB vaccine (b), and tetravalent XBB.1 vaccine (c) immunization. Each dot represents the 50% neutralization titer (NT50) for an individual. The horizontal dotted line in the neutralization assay indicates a limit of detection of 30, with serum samples demonstrating neutralization below 30 plotted as 10. The geometric mean titers (GMTs) of the neutralizing antibodies and the percentages of individuals with NT50 values above the limit of detection are indicated below each group. Statistical analyses were performed using a two-tailed Friedman test with a false discovery rate. A p-value <0.05 was considered statistically significant, and only significant differences are displayed in the figure, with the fold change in the GMT denoted in brackets

Stronger antibody response induced by trivalent XBB.1.5 and tetravalent XBB.1 vaccines

Next, we compared neutralization titers against the tested pseudoviruses among the three cohorts. We observed that individuals from the three cohorts before immunization had comparable neutralization titers against D614G, BA.5, XBB.1.5, JN.1, and KP.2 (Fig. 3a). However, individuals from the trivalent XBB.1.5 cohort before immunization had even higher neutralizing antibody titers against KP.3 (Fig. 3a). After vaccination, individuals in the trivalent XBB.1.5 vaccine and tetravalent XBB.1 vaccine cohorts exhibited increased neutralization titers against BA.5, XBB.1.5, and JN.1 (Fig. 3b). In contrast, individuals from the trivalent XBB.1.5 cohort produced higher neutralizing antibody titers against the variants KP.2 and KP.3, and individuals from the tetravalent XBB.1 vaccine cohort produced higher neutralizing antibody titers against D614G (Fig. 3b). Overall, recipients of trivalent XBB.1.5 and tetravalent XBB.1 likely produced similar antibody titers against the indicated pseudoviruses but were significantly higher than those of bivalent Omicron XBB vaccine recipients.

Fig. 3

The trivalent XBB.1.5 vaccine booster elicited increased neutralizing antibody titers. a, b Comparison of neutralizing titers against the indicated pseudoviruses in serum collected from individuals before (a) and after (b) administration of the trivalent XBB.1.5 vaccine, bivalent Omicron XBB vaccine, or tetravalent XBB.1 vaccine. Each dot represents the 50% neutralization titer (NT50) for an individual. The horizontal dotted line in the neutralization assay reflects a limit of detection of 30, with serum samples exhibiting neutralization below 30 represented as 10. The geometric mean titers (GMTs) of the neutralizing antibodies are displayed on top of each group. The bar represents the GMTs and 95% confidence intervals. The dotted line indicates the limit of detection of the NT50. Statistical analyses were performed using the Kruskal‒Wallis test with the false discovery rate method for three-group comparisons of GMTs. A p-value <0.05 was considered statistically significant, and only significant differences are displayed in the figure

Neutralizing antibody titers in individuals with underlying medical conditions

Considering that half of the participants reported underlying medical conditions, we compared antibody titers between participants with and without underlying medical conditions. The results showed that antibody titers against the tested pseudoviruses in individuals without chronic medical conditions were similar to those in individuals with chronic medical conditions, both before and after immunization (Fig. 4a, b). A further sub-analysis comparing neutralization titers between participants with hypertension and those without medical conditions revealed no statistically significant differences (Supplementary Fig. 1).

Fig. 4

Neutralizing antibody responses in individuals with or without medical conditions before and after immunization with three XBB-containing vaccines. a–d Comparison of the 50% neutralization titer (NT50) against the indicated pseudoviruses in individuals with (n = 13) and without (n = 17) medical conditions before receiving the trivalent XBB.1.5 booster (a), in individuals with (n = 20) and without (n = 10) medical conditions before receiving the bivalent Omicron XBB vaccine (b), in individuals with (n = 11) and without (n = 19) medical conditions before receiving the tetravalent XBB.1 vaccine (c), and in all pooled sera (d). e–h NT50 of individuals with and without medical conditions after immunization with the trivalent XBB.1.5 booster (e), bivalent Omicron XBB vaccine (f), tetravalent XBB.1 vaccine (g), or all pooled sera after immunization (h). Each dot represents the NT50 for an individual. The horizontal dotted line in the neutralization assay reflects a limit of detection of 30, with serum samples exhibiting neutralization below 30 represented as 10. The bar represents the GMTs and 95% confidence intervals. Statistical analyses were performed using the Wilcoxon rank-sum test for group comparisons of GMTs between individuals with and without chronic conditions. A p-value <0.05 was considered statistically significant, and only significant differences are displayed in the figure

Antigenic cartography

By employing pooled and separate serum neutralization data from all three vaccine cohorts, antigenic maps were created to quantify and illustrate the antigenic disparities between ancestral D614G and the tested variants (Fig. 5a–d). The map shows that the sera from the three cohorts before vaccination substantially overlapped and was centered around D614G, and the sera from the three cohorts after immunization shifted toward the BA.5 and XBB.1.5 variants (Fig. 5a). The JN.1, KP.2, and KP.3 variants were clustered together, showing greater antigenic distinction from D614G than from the XBB.1.5 variant (Fig. 5a). Specifically, the antigenic distances between D614G and the XBB.1.5 and JN.1 variants after the administration of the three XBB-containing vaccines indicated a significant boost in antibody potency and breadth (Fig. 5b, c). However, the antigenic distances between D614G and KP.2 or KP.3 were not shorter after immunization than before vaccination (Fig. 5b, c), suggesting that XBB-containing vaccines boost the antibody potency and breadth to KP.2 and KP.3 but are limited, which is consistent with a mean 5- and 4-fold increase in GMTs being also observed after immunization (Fig. 1). Taken together, the JN.1, KP.2, and KP.3 variants exhibited similar antigenic and distant properties.

Fig. 5

Antigenic map of serum virus neutralization data. a–d Antigenic maps were constructed using the Racmacs program (1.1.4) for neutralization titers against the indicated pseudoviruses from all cohorts (a), individuals before and after receiving the trivalent XBB.1.5 vaccine (b), bivalent Omicron XBB vaccine (c), and tetravalent XBB.1 vaccine (d). The circles represent the indicated variants, whereas the squares denote individual serum samples. The x- and y-axes depict antigenic units (AUs), with each grid corresponding to a 2-fold serum dilution of the neutralization titer. One square on the grid corresponds to one AU squared. Arrows between D614G and selected variants are annotated with the distance between those variants in AUs