The Two-years Immune Effect Between 0-1-2-months and 0-1-6-months of HBV Vaccination Schedule in Adults

Background: The short-term 0-1-2-months hepatitis B virus (HBV) vaccination schedule was previously proposed in the adult population; however, its long-term immune effect remains unclear. The present study was aimed to investigate 1) the 2-months and 2-year immune effect of HBV vaccination; and 2) compliance rate between 0-1-2-months and 0-1-6-months vaccination schedules in adults. Method: A total of 1281 subjects tested for HBsAg(-) and Hepatitis B surface antibody (anti-HBs)(-) were recruited. Participants from two distant counties were inoculated hepatitis B yeast vaccine for 10ug per dose each time, with 0-1-2-months (n=606) and 0-1-6-months (n=675) vaccination schedule, sequentially followed-up at two months and two years after the 3 rd injection. Results: There was no statistical difference in anti-HBs seroconversion rate between 0-1-2-months and 0-1-6-months vaccination schedule at two months (91.96% vs 89.42%, p=0.229) and two years (81.06% vs. 77.14%, p=0.217). Quantitative anti-HBs level of 0-1-2-months vaccination schedule was not different with 0-1-6-months vaccination schedule at 2 months (anti-HBs 1 ) (342.12 ± 378.42 m IU/ml vs. 392.38 ± 391.96 m IU/ml, p=0.062), but was higher at two years (anti-HBs 2 ) (198.37 ± 286.44 m IU /ml vs. 155.65 ± 271.73 m IU /ml, p=0.048). By subgroup analysis, 0-1-2-months vaccination schedule showed better maintenance (p=0.041) and delayed reinforcement (p=0.019) in comparison to 0-1-6 vaccination schedule. The 0-1-2-months vaccination schedule also increased the 3 rd -time injection completion rate (89.49%


Background
Hepatitis B virus(HBV) infection was estimated that the global prevalence of HBsAg in 2016 was 3.9%. 1 A modelling study 2 estimated HBsAg prevalence in China to be 6·1%, 3 other study reported the prevalence in 2018 was more than 80 million estimated chronic infections. 4 The World Health Organization (WHO) estimates that more than 658,000 individuals die annually from hepatitis B virus (HBV)-related complications, such as fulminant hepatitis, cirrhosis, and liver cancer. 5 The universal infantile HBV vaccination in the national immunization program achieved great success in preventing and controlling HBV infection in the past 20 years. HBsAg positive rate in mainland of China decreased from 14·0% in 1957-89 to 5·4% in 1990-2013. 6 In adults, the immunized population had a much lower prevalence of HBsAg than the un-immunized population. 7 Therefore, hepatitis B vaccine immunization in adults should also be recommended. 2 HBV immunization at 0, 1 and 6 months (0-1-6-months) have been recommended by the WHO and US Centers for Disease Control and Prevention, as well as Chinese National Guidelines on chronic hepatitis B prevention and treatment (2015). However, the WHO-recommended 0-1-6-months vaccination schedule often leads to a lower completion rate of vaccination in adults. [8][9][10] In fact, in China there was a special population called oating population (so-called migrant workers in other countries), accounting for 230 million people per year. Generally, they leave their hometowns to nd jobs in other cities and change their jobs frequently. The oating population of China had an increased risk of sexual transmission of HBV due to having lower education, lower economic income, younger age, and multiple sexual partners. 11 Therefore, previous studies proposed a vaccination schedule at months 0, 1, and 2 (0-1-2-months), resulting in a comparable short-term safety and immunogenicity. 12,13 Moreover, shortening the vaccine schedule time can effectively increase the completion rate of vaccination and even stimulate earlier and faster anti-HBs production. 12,13 Such accelerated immunization schedules ( 0-1-2-months, 0-1-3-months and 0-7-21-days, etc.) were veri ed to have the same short-term immune effect and to increase the completion rate in the general population 14 , vein-injected drug users 15,16 and adults who were refused to receive the second or third dose owing to occupational reasons 17,18 . However, antibody maintenance by immune memory was even more crucial than antibody production in protecting patients from HBV infection. Therefore, the comparative results for the best vaccination schedule could not be obtained from previous short-term follow-up studies.
The long-term immune effect of the accelerated vaccination schedule remains unclear due to the lacking evidence. 12,19 Ren et al recently reported that the same positive seroprotection rate and the quantitative anti-HBs level between 0-1-3-months vaccination and 0-1-6-months schedule in 8 years after the vaccination. 18 However, the author also declared in limitation part that high loss to follow-up (from initial 771 participants to 242 in nal follow-up, 529 participants were lost to follow-up) caused by the oating population and relatively small sample size may in uence the reliability. We proposed a prospectively interventional study in two comparable towns study to investigate the short-term and long-term immune effects, as well as the completion rate between the accelerated vaccination schedule (0-1-2-months) and the standard vaccination schedule (0-1-6-months).

Study design
This was a prospective study to explore the response of different vaccination schedules of hepatitis B on both HBsAg(-) and anti-HBs(-) adults. Randomization were not available due to the different 3rd injection time (participants from small towns would aware of different schedules The Health Planning Committee of Sichuan (grant number 16PG280). Funding body was in charge of the design of the study and collection, analysis, and interpretation of data.
Appropriate training was given to research stuff from the leading investigators. Standardized, questionnaire-based face-to-face interviews were performed after obtaining written consent. The questionnaire collected information including gender, age, height and weight, etc. Each subject was con rmed by screening identi cation card and taking a photo before each vaccine injection. After that, they were offered vaccination through the regular service of township hospitals, where trained medical staff performed vaccinations by intramuscular injection. Vaccinations with 10ug per dose of hepatitis B yeast vaccine (Hualan Biological Vaccine Company, Chengdu, China) in assigned to 0-1-2-months vaccination schedule in Jinfeng and 0-1-6-months vaccination schedule in Longmen.

Follow up
The 1st and 2nd follow-up visits were conducted at 2 months and 2 years after nishing the third vaccine injection. The follow-up time to each subject was informed by phone-calling 3 weeks before visiting. Each subject was con rmed by screening an identi cation card and taking a photo. Every screening card was individualized with the follow-up time of each participant as a memorandum. Both schedules were consistent with the community standard of care at the time.

Serum assay and Blood sample tests
Serum samples were also collected and tested at each follow-up visit in local health stations and community clinics. The anti-HBs levels at 2 months (anti-HBs 1 ) and 2 years (anti-HBs 2 ) after 3rd -time injection vaccination were analyzed for vaccine response. The serum samples were tested for the quanti cation of HBsAg and anti-HBs by electrochemiluminescence immunoassay (Abbott i2000SR, U.S.A) in West China Hospital. The HBsAg level < 0.05 m IU/ml was de ned as negative ones. Anti-HBs level below 10 m IU/ml, 10-100 m IU/ml, 100-1000 m IU/ml and above 1000 m IU/ml were de ned as no response, low response, normal response and high response, respectively. 20 Because anti-HBs ration less than 0.05 m IU/mL were unable to detect and exceeding 1,000 m IU/mL were excluded from further dilution test, a value of 0 m IU/mL and 1000 m IU/mL were assigned to these subjects respectively for quantitative analysis of anti-HBs, for reference as previous anti-HBs geometric mean concentration (GMC) test. 13,21 To explore the reason for different maintenance of Anti-HBs level between two vaccination schedule, the subjects were divided into four different clinical scenarios: 1) "well production and good maintenance" by anti-HBs 1 (at two months) (+) and anti-HBs2(at two years) (+); 2) "well production and poor maintenance" by anti-HBs 1 (+) and anti-HBs 2 (-); 3) "persistent non-production" by anti-HBs1 (-) and anti-HBs 2 (-); and 4) "delayed production", anti-HBs 1 (-) and anti-HBs 2 (+). Furthermore, the "well production and good maintenance" group was further divided into four subgroups: 1) "High production and good maintenance" by high anti-HBs 1 (> 100 m IU/ml) and high anti-HBs 2 (> 100 m IU/ml), 2) "well production and fast decrease" by high anti-HBs 1 (> 100 m IU/ml) and low anti-HBs 2 (10-100 m IU/ml), 3) "relative low production" by low anti-HBs 1 (10-100 m IU/ml) and low anti-HBs 2 (10-100 m IU/ml), and 4) "delayed reinforcement" by low anti-HBs 1 (10-100 m IU/ml) and high anti-HBs 2 (> 100 m IU/ml).

Statistical analysis
Mean values and prevalence of baseline characteristics were calculated. Data were reported as the mean ± standard deviation for normal and median (interquartile range) for non-normal continuous variables (when sample size > 40, mean ± standard deviation was used to express data), while the frequency was used for discrete variables. In the univariate comparisons, we used the Student t-test and ANOVA with Bonferroni adjustments for continuous samples and the chi-square test or Fisher's exact test for the qualitative ones. Non-parametric alternatives (Mann-Whitney U and Kruskal-Wallis tests) were used for non-normal distributions. Logistic regression models were used to estimate adjusted odds ratios (aORs) with our principal outcome of full-time completion rate and 3rd -time injection rate. Covariates were selected for analysis according to their biologically plausible potential to act as confounders or predictors for each outcome. The potential predictors at baseline were as follows: age, gender, BMI, previous hypertension, previous type 2 diabetes mellites (T2DM), abnormal alanine aminotransferase (ALT) level. The collinearity between factors included in the multivariable analyses was checked by using variance in ation factor (VIF) and tolerance ( (Table 1) There were 621 participants completed both 2-month and 2-year follow-up visits. The information obtained from phone-calling indicated that subjects who un nished the three injections or did not showup at follow-up time were mainly owing to occupied working schedules or migrant workers working in distant places. The distribution equally were re-assessed since almost half of participants were lossed followed-up. (Supplementary table 1) 2. Comparison of the short-and long-term immune effects of two vaccination schedules (1)No difference was found in anti-HBs seroconversion between two vaccination schedules Anti-HBs seroconversion rates were no signi cant difference between 0-1-2-months and 0-1-6-months vaccination schedule at short-term seroconversion (at two months) (89.42% vs. 91.96%, p=0.229) ( Figure  1A), as well as at long-term seroconversion rate (81.06% vs. 77.14%, p=0.217) ( Figure 1B). Therefore, a similar immune effect was obtained by 0-1-2-months vaccination schedule, and even an earlier protection was gained at second months rather than at sixth months compared to 0-1-6-months vaccination schedule.  Table 2) Furthermore, among the subjects with successful seroconversion (anti-HBs 1 >10 m IU/ml), no difference was found in the proportion of low response (10 m IU/ml< anti-HBs 1 <100 m IU/ml), normal response (100 m IU/ml< anti-HBs 1 <1000 m IU/ml) and high response (anti-HBs 1 >1000 m IU/ml) between 0-1-2-months and 0-1-6-months vaccination schedule at 2-months followup (p for ANOVA=0.517) ( Figure 1C) However, 0-1-2-months vaccination schedule induced better maintenance of anti-HBs at 2-year followed up, with a higher proportion of normal response (44.73% vs. 32.87%) and high response (7.69% vs. 6.48%), and a lower proportion of low response (7.69 vs. 6.48%) (p for ANOVA=0.010). ( Figure 1D) (3) 0-1-2-months vaccination schedule showed better maintenance and a delayed reinforcement To explore the reason for different maintenance of Anti-HBs level between two vaccination schedules, the subjects were divided into four different clinical scenarios (detailed in methods part). We found that the 0-1-2-months vaccination schedule showed a lower proportion of the "well production and poor maintenance" group than that in 0-1-6-months vaccination (12.56% vs. 18.61%, p=0.041), suggesting the 0-1-2-months vaccination schedule induced better maintenance than 0-1-6 vaccination schedule. (Figure 2) The "well production and good maintenance" group was further divided into four subgroups by qualitative level of anti-HBs. The result showed that the proportion of the 4 th subgroup (low anti-HBs 1 and high anti-HBs 2 ) in the 0-1-2-months vaccination schedule was higher than that in the 0-1-6months vaccination schedule (9.33% vs. 3.51%, p=0.019), suggesting 0-1-2-months vaccination schedule possibly induced a "delayed reinforcement" of antibody production, as having low anti-HBs at initial but high anti-HBs at two years. (Figure 2)

Discussion
In the present study, we found that the short-and long-term anti-HBs seroconversions were not different between 0-1-2-months and 0-1-6-months vaccination procedures. However, 0-1-2-months vaccination schedule showed a higher anti-HBs level at 2-year follow-up by inducing better maintenance and a delayed reinforcement in comparison to the 0-1-6-months vaccination schedule. In the multivariant analysis, no signi cant difference of anti-HBs seroconversion rate was found between 0-1-2-months and 0-1-6-months vaccination schedule at two months and two years. However, the 0-1-2-months vaccination signi cantly increased the 3rd -time injection completion rate. Therefore, In conclusion, the 0-1-2-months vaccination could obtain a similar short-term immune effect, but achieve a better long-term immune memory and a higher completion rate in the adult population.
In a recent systematic review, the different vaccination schedule obtain the similar short-term immune effect, demonstrating that anti-HBs concentrations ≥ 10 m IU/ml were approximately 65.0-85.0% using the 0-1-2-months schedule, approximately 77.0-90.8% using the 0-7-21-day schedule, 87.0% using the 0-2-6-week schedule, and around 79.0% after the 0-14-28-day schedule, respectively. 22 Ren et al 18 also reported that the accelerated schedule (0-1-3-months) and the standard schedule (0-1-6 months) enhanced the long-term immune memory (8 years after) in comparison to (0-1-12-months) schedule, which we found a similar long-term immune effect in the present study (2 years after). Regarding the short-term immune effect, we speculate that the different short-term immune effect may be explained by that the measurement was taken at a different time (4th months for "0-1-3-months", 7th months for "0-1-6-months" and 13th months for "0-1-12-months), since the 6-months and 12-months injection serve as a booster dose, which was known to increase the seroconversion rate. 22 Regarding the long-term immune effect, we further explored that the longer immune memory of shorter interval of 0-1-3-months was induced by a "better maintenance" and "delayed reinforcement" in the present study. Thus, the shorter interval of vaccination schedule provided a stronger re-estimation to the immune system, resulting in a higher anti-HBs level at 2 years after three injections.
The in-developing immune system in children was different from the mature immune system in adults, therefore the vaccination schedule could be modi ed according to a speci c population. Cesare Belloni 23 reported that in children the 0-1-6-months vaccination schedule presented a higher percentage of seroconversion than the 0-1-3-months vaccination schedule, but also pointed out that the reduced response after the 0-1-3-months vaccination schedule was mainly due to the relative immaturity of the immune system in the younger infants. 24 Similarly, the bacterial conjugate vaccines were also recommended for adult to use only one dose at initial and a booster immunization, which is su cient for adults to stimulate memory B cells and to maintain the antibody due to the mature immune system. 25 Thus, the decision of shorter or longer period of vaccination injection was dependent on the mature or immature status of the host immune system. The immature immune system in children was responded differently in 0-months, 1-months and 6-months since the system was still developing; thus the 3rd injection at 6 months in children could induce a better response by a relatively more mature immune system at 6 months than that at 3 months or earlier. However, the mature immune system in the adults was responded similarly at 0 months, 1 month, 3 months or 6 months, thus a shorter period could gain more completion rate.
Our results indicated that the 0-1-2-months vaccination established a better completion rate than the 0-1-6-months vaccination schedule. However, a previous study reported that a 0-1-12-months vaccination schedule may be more suitable for the oating population and 0-1-6-months schedule is recommended for the xed population, both of them had a better completion rate than the 0-1-3-months schedule. 8 The oating population of China was characterized by changing their jobs frequently and returning home annually to celebrate the spring festival after 11 months of working and rest for one month. The different results between the present and previous studies could be explained by calculating the 3 rd− time injection rather than a full-time injection since the 1st and 2nd injections were the same at 0 and 1 months.
There were limitations in the present study. Firstly, there were patients lost follow-up patient in the rst and second follow-up visit. Nevertheless, the lost follow-up was inevitable in follow-up visit, because the secondary outcome of the present study was designed to investigate the completion rate on a oating population. Moveover, we re-assessed the baseline characteristics of followed up patients, resulting a destribustion equally between 2 groups (supplementary table 1). Secondly, information on the participants' status (including smoking, drinking and diabetes) were not recorded (even the HIV coinfection and active HBV were excluded), which might affect antibody response after hepatitis B vaccination. Lastly, a standardized time point for the measurement of anti-HBs antibody levels, to enhance comparability of the immune response between different studies; although we found better maintenance and a delayed reinforcement of anti-HBs by 0-1-2-months vaccination schedule,the insu cient observed time points within 2 years limited us to explore the anti-body production tendency.

Conclusion
The 0-1-2-months vaccination could obtain a similar short-term immune effect, but achieve a better longterm immune memory and a higher completion rate in the adult population. Consent for publication: None, present study was retrospective study design and did not contains speci c personal medical information (video or image) about an identi able living individual. However, the consent of information for each participantets was obtained before the rst vaccination.

Abbreviations
Availability of data and materials: Yes, supplementary data. All data generated or analyzed during this study are included in this published article and its supplementary information les.  Anti-HBs 2 , hepatitis B antibody at 2 years after three vaccination injections.