Study design, participant demographics and safety

All participants received at least two ID doses of purified Vero cell rabies vaccine (PVRV, > 2.5 IU/IM dose; 0.5 mL/ampoule; Speeda®, China) (2 × 0.1 mL) on D0 and D7 for the PrEP regimen (for a total of 0.4 mL) and were categorized into three groups: the participants in the PrEP group received only the PrEP regimen (n = 43). Participants in the 1st booster group received the first ID booster dose the year after PrEP (n = 42), and participants in the 2nd booster group received the second ID booster dose the year after the 1st ID booster dose (n = 19) (Fig. 1). Most of the participants in the three groups were males (58.1, 81.0 and 100.0%), and the median ages at enrolment were 52.0 (IQR 45.0–60.0), 57.5 (IQR 51.2–63.0) and 61.0 (IQR 56.5–65.0) years for the PrEP, the 1st booster and the 2nd booster group, respectively. The most common underlying diseases in all groups were diabetes mellitus, hypertension and hyperlipidaemia (Table 1).

Fig. 1

Schematic diagram of intradermal rabies vaccination and blood collection. D day, ID intradermal, PrEP preexposure prophylaxis

Table 1 Characteristics of the participants

Overall, among participants who reported adverse events following immunization (AEFIs) in the 2nd booster group, 2 (10.5%) reported more than one symptom, a significant decrease when compared with 19 (45.2%) in the PrEP group and 16 (39.0%) in the 1st booster group (Table 2). AEFIs, especially systemic reactions, fever and fatigue, tended to decrease after the 1st and 2nd booster doses. The most common symptoms in the PrEP group were muscle pain (14.0%), fever (11.6%) and itching (9.3%), while those in the 1st booster group were local AEFIs associated with injection site itching (16.7%), swelling (11.9%) and muscle pain (9.5%). However, no severe AEFI was observed in this study.

Table 2 Percentage of participants who reported adverse events at D28 after vaccinationImmunogenicity and booster effectiveness

The anti-rabies antibody levels in the three groups, PrEP and the 1st and 2nd booster groups, on D0, D28 and D180 were detected by CEE-cELISA. The lowest concentration of an anti-rabies antibody that could be detected by the CEE-cELISA was approximately 0.7 EU/mL, which was comparable with the REFFT result at 0.5 IU/mL [11]. On D0 before immunization and D28 after immunization, the median rabies antibody level was not different among the three groups; however, on D180, the median level in PrEP (16.5 EU/mL [95% CI 7.2–24.7 EU/mL) was significantly lower than that in the 1st booster group (25.8 EU/mL [95% CI 18.9–101.2 EU/mL]) and the 2nd booster group (27.0 EU/mL [95% CI 26.0–38.2 EU/mL]) (Fig. 2a).

Fig. 2

Immunogenicity and booster effectiveness. a Rabies neutralizing antibody levels at each time point were measured by CEE-cELISA; each data point represents the median with 95% CI for each individual; the horizontal dashed line indicates 0.7 EU/mL (indicator of adequate vaccination); statistical analysis was performed using Dunn’s multiple comparisons test; b seroconversion rates at each time point; each bar represents the percentage with 95% CI of participants who had an antibody level ≥ 0.7 EU/mL in each group; the white, grey and black dots or bars indicate the PrEP, 1st booster and 2nd booster groups, respectively; statistical analysis was performed using a Chi-square test; *p < 0.05; no asterisk and ns nonsignificant

The persistence of rabies antibodies was first observed on D0. The measurable antibodies on D0 in the 1st and 2nd booster groups persisted approximately one year after PrEP and the 1st booster vaccination. Among the participants who had an adequate antibody level ≥ 0.7 EU/mL on D0 in the 2nd booster group, 63.0% [95% CI 38.3–83.7%] had an adequate antibody level significantly greater than that in the 1st booster group (36.0% [95% CI 21.6–52.0%]). The percentage of participants who had an inadequate antibody level (< 0.7 EU/mL) after PrEP vaccination for one year (D0 of the 1st booster group) was 64.0% [95% CI 48.0–78.4], while that at D28 was 2.0% [95% CI 0.1–12.3%] after the 1st boost. Similarly, the percentage of patients who underwent a 1st boost for one year (D0 of the 2nd booster group) was 37.0% [95% CI 16.3–61.6%] when the percentage at D28 was 5% [95% CI 0.1–26.0%]. These results imply that even if the level of antibody after PrEP or the 1st boost vaccination declines to an inadequate level, the antibody level can be restored after boosting. All groups exhibited a high seroconversion rate (≥ 95%) on D28 and D180 after participants received PrEP or booster doses (Fig. 2b); however, the antibody level in the PrEP group decreased significantly on D180 (Fig. 2a).

To estimate the half-life of the antibodies in response to rabies vaccination, response profiles that showed a decrease in antibody levels over at least two consecutive sampling points in the PreP and 1st booster groups were selected, and one-phase exponential decay curves fitted to the decay phase of the selected response profiles were generated (Additional Fig. 1). The mean half-life of the antibodies in the PrEP group was 118.0 days, whereas that in the 1st booster group was 121.4 days. The half-lives of the responses did not differ significantly between the two groups.

Taken together, the data indicated that at least one ID booster dose was sufficient to increase the adequate antibody concentration after a participant had been vaccinated with the PrEP regimen for one year. Moreover, one ID booster dose was used when measuring the persistence of an adequate rabies antibody level for more than 180 days and to increase the percentage of seroconversion on D0 in the next year.

Pattern of antibody response after the 1st and 2nd booster doses

Among the participants in the 1st and 2nd booster groups, 19 were the same individuals who were enrolled in different years, and the pattern of antibody response to both recombinant rabies virus glycoprotein (rRVGP) and the rabies vaccine was observed by CEE-cELISA and indirect ELISA. The RVGP plays important roles in the early and late phases of virus replication [12]. The patterns of antibody response on D0 and D28 after the 1st and 2nd booster doses were similar and correlated between the CEE-cELISA and indirect ELISA methods (Additional Fig. 2a–c and Additional Fig. 3).

Only one male participant was negative for an antibody response on D28 after the 1st and 2nd booster doses according to CEE-cELISA, but antibodies against both the rRVGP and the rabies vaccine were detectable via indirect ELISA. However, the detectable antibody level of this participant was low on D28 after the 1st booster dose (46.9 AU/mL; median across this group, 114.1 AU/mL [% CI 77.1–148 AU/mL]), and after the 2nd booster dose, a level of 38.6 AU/mL was detected (median across this group, 117.8 AU/mL [% IC: 75.7–170.8 AU/mL]) (Additional Fig. 2a–c). Samples taken on D28 following receipt of the 2nd booster dose confirmed the neutralizing activity of the rabies antibody by the Rapid Fluorescent Foci Inhibition Test (RFFIT) performed at the National Institute of Infectious Diseases, Tokyo, Japan, and the RFFIT data were > 0.5 IU/mL (indicator of adequate vaccination). Thus, this participant was a low responder to the rabies vaccine and was able to produce an adequate antibody response after immunization with one ID booster dose.

Anti-rabies antibody and avidity indices in response to recombinant rabies virus glycoprotein and the rabies vaccine on D28 after vaccination

To determine the strength of the binding interaction between the antigen and the antibody, the avidity index of the anti-rabies antibody was measured. Anti-rabies antibody was observed by indirect ELISA with or without urea treatment. Anti-rabies IgG antibody levels in response to the rRVGP and rabies vaccine on D28 in the three groups were measured via indirect ELISA. The median IgG antibody concentration against the rRVGP significantly increased in the 1st booster (102.8 AU/mL [95% CI 68.8–126.2 AU/mL]) and 2nd booster (117.8 AU/mL [95% CI 75.7–170.8 AU/mL]) groups compared with the PrEP group (15.4 AU/mL [95% CI 7.7–22.5 AU/mL]) (Fig. 3a). Similarly, the median level of IgG antibody to the rabies vaccine was significantly elevated in the 1st booster (51.6 AU/mL [95% CI 40.2–61.8 AU/mL]) and 2nd booster (52.6 AU/mL [95% CI 43.8–87.6 AU/mL]) groups compared with the PrEP group (10.8 AU/mL [95% CI 8.0–19.8 AU/mL]) (Fig. 3b).

Fig. 3

Anti-rabies IgG antibody and avidity index in response to recombinant rabies virus glycoprotein and the rabies vaccine at D28. IgG antibody levels in the a recombinant rabies virus glycoprotein (rRVGP) and b rabies vaccine groups measured by indirect ELISA; percentage of avidity indices in response to the c rRVGP and d rabies vaccines; the samples were selected from the PrEP (n = 17), 1st booster (n = 36) and 2nd booster (n = 19) groups; the bars indicate the medians with 95% CIs, and each dot represents the value of each sample; statistical analysis was performed using Dunn’s multiple comparisons test; ***p < 0.001; **p < 0.01; ns nonsignificant

As shown for the antibody level, the avidity index (expressed in %) increased significantly in the 1st booster group (83.4% [95% CI 81.1–87.4%) and the 2nd booster group (90.5% [95% CI 85.4–92.6%) compared with the PrEP group (40.3% [95% CI 26.5–52.2%) in response to the rRVGP (Fig. 3c). Similarly, for the rabies vaccine, the avidity index was significantly greater in the 1st booster group (78.8% [95% CI 77.7–85.7%) and the 2nd booster group (90.1% [95% CI 84.5–100.0%) than in the PrEP group (26.8% [95% CI 13.7–75.1%) (Fig. 3d). These data indicated that the elevated level and avidity maturation of rabies-specific antibodies occurred during vaccination, especially after the 1st booster.