H5N1 is a type of influenza virus that causes highly infectious, severe respiratory disease in birds, known as avian influenza. The H5N1 bird flu incident in Hong Kong in 1997 was the first known instance of a purely avian virus causing severe human disease and death. Widespread re-emergence of avian influenza occurred in 2003 in Asia and later in Africa, Europe, and the Middle East. Influenza A virus subtype H5N1 (A/H5N1) causes illness in humans and many other animal species. Vietnam is 1 of 6 countries that were identified as A/H5N1 hotspots at that time. From 2003 to 2009, the World Health Organization (WHO) recorded 861 humans infected with H5N1 and 455 deaths, of which 50.4% occurred in Vietnam [1].

Vaccines remain one of our best defense weapons against the influenza virus. In 2006, the WHO and the wider international public health community launched the Global Action Plan for Influenza Vaccines to help ramp up the production capacity of vaccination in 14 countries, including Vietnam [2]. In 2012, the Institute of Vaccines and Medical Biologicals (IVAC) successfully produced an influenza vaccine named IVACFLU-A/H5N1, which constitutes a whole-virion vaccine produced by technology involving chicken embryos, with utilization of egg micronutrients, formalin inactivation, adsorption of aluminum hydroxide, and no preservatives. Evidence from 1,005 Vietnamese people who voluntarily received injections showed that this vaccine produced a high seroconversion rate (75.8%), equivalent to a high level of antibodies after vaccination [3].

Antibody persistence, which is widely defined as the maintenance of seropositivity, is vital for sustained protection [4]. The persistence of vaccine antibodies decreases over time and depends on several factors, such as age, sex, and the vaccine’s components (e.g., adjuvants) [5]. To our knowledge, there is limited evidence on the relationships among factors related to antibody persistence rate over time. IVACFLU-A/H5N1 was documented as effective in Vietnam; however, antibody persistence after vaccination has not been well characterized due to the recent approval of the vaccine. A study in the United States showed that the antibody rate was 76.7% at 6 months after vaccination with an ASO_3A-adjuvanted H5N1 vaccine [6]; this figure decreased to 30.9% and 8.3% after 2 years for vaccine formulations with and without adjuvants, respectively [5]. Other evidence suggested that antibody rates persisted for more than 2 years after H5N1 vaccination [7].

We examined the persistence of antibodies after vaccination and related risk factors in individuals enrolled in the phase II IVACFLU-A/H5N1 vaccine trial in Ninh Hoa, Vietnam, who received a 15-μg dose (2 injections 21 days apart). Our study is one of the first to explore long-term antibody persistence and its associated characteristics in IVACFLU-A/H5N1 vaccinated individuals.

In this study, we found that 84.9% and 52.3% of participants showed persistence of at least a 1/10 level of A/H5N1 antibodies at 30 months and 36 months after complete vaccination, respectively. More than half (57.0%) of participants showed seroconversion at 30 months, with levels at least 4 times higher than baseline. Meanwhile, this proportion at 36 months was 30.2%. The GMTs (average antibody titers) at 30 months and 36 months were 4 times and 2 times higher than baseline, respectively (Table 1).

Table 2 presents the demographic characteristics, personal behaviors, medical history, and exposure risk factors of those who showed persistence of at least a 1/10 level of A/H5N1 antibodies at 30 months and 36 months. The proportions of male with persistence of at least a 1/10 level of antibodies at 30 months and 36 months after completing the full IVACFLU-A/H5N1 vaccination were 73.9% and 43.5%, while these figures in female were 88.9% and 55.6%, respectively. In our study, sex was not significantly associated with the persistence of at least a 1/10 level of antibodies at both 30 months and 36 months after vaccination. Older people (41–65 years) had higher odds of persistence of at least a 1/10 level of A/H5N1 antibodies than younger participants (18–40 years; p=0.02) (using participants’ age at the day of vaccination in 2016). In contrast, people who experienced flu-like symptoms in the past 18 months or between 2 visits had lower odds of persistence of at least a 1/10 level of A/H5N1 antibodies than those who did not (p=0.02).

Our study did not find associations between persistence of at least a 1/10 level of antibodies and personal behaviors (alcohol use and smoking) or exposure risk factors (having a poultry farm, directly contacting poultry, and directly slaughtering and processing poultry).

The overall proportion of participants with persistence of at least a 1/10 level of A/H5N1 antibodies was 84.9% and 52.3% at 30 months and 36 months after IVACFLU-A/H5N1 vaccination, respectively. The GMT of participants at 30 months and 36 months post-vaccination was 22.3 and 12.0, respectively. Antibody persistence was greater in older people (odds ratio [OR], 5.65; 95% confidence interval [CI], 1.06 to 39.87; p=0.02). However, the estimate of antibody persistence after vaccination was lower in people who experienced flu-like symptoms in the past 18 months or between 2 visits (OR, 0.09; 95% CI, 0.01 to 0.81; p=0.02).

We found that most participants (84.9%) showed persistence of at least a 1/10 level of A/H5N1 antibodies, and the proportion of those with seroconversion at 30 months after vaccination was 57.0%. At 36 months after vaccination, the former figure decreased to 52.3% and the latter declined to 30.2% Subjects with a titer of ≥1:10 can be considered seropositive. However, opinions on protective titers for the influenza vaccine are quite diverse. Classic research, supported by more recent studies, has established a correlation between the antibody titer evaluated by the HI assay and protection against influenza virus infection, with an HI titer of 1:40 equivalent to 50% protection against infection [10–12]. Lower titers, such as 1:32 or 1:20, have previously been used for HI and neutralizing antibodies to characterize the immune responses to influenza vaccines [13–15]. Meanwhile, other studies have argued that there is insufficient evidence to deduce that an HI titer of 1:40 predicts 50% protection against influenza virus infection [16,17]. Therefore, further research on IVACFLU-A/H5N1 should be conducted to establish a time point for booster vaccination. Nevertheless, it is noteworthy that these results were obtained at 30 months and 36 months post-vaccination without booster shots, indicating a relatively high persistence of antibodies [18].

Our findings on the GMT (the average antibody titer) persisting over time are remarkable compared to previous studies on other H5N1 vaccines. Comparing to a study by Gillard et al. [18] on ASO_3A-adjuvanted heterologous H5N1 vaccines, the GMT at 36 months post-vaccination was comparable (12.0 vs. 10.5). At the same time, our figure for 30 months was notably higher despite a longer follow-up time (30 vs. 12, 18, 24, and 30 months). In another study on an ASO_3A-adjuvanted A/H5N1 vaccine, the GMT for 12 months following vaccination was reported to be 11.8, which is considerably lower than our results at both 30 months and 36 months [19]. Aside from the physiological condition of the study subjects, differences in the results of immunological indicators between studies might be attributable to variation in vaccine composition, particularly adjuvants. Adjuvants have been proven to play an important role in enhancing immunogenicity through either physical or chemical associations with antigens [20]. Although evidence has demonstrated the safety and effectiveness of ASO₃ in boosting the immune response, aluminum-based adjuvants have remained the “gold standard” for adjuvants due to their long-term success [21]. The results observed in this study suggest that future research should compare the performance of ASO₃ and aluminum-based adjuvants in terms of the persistence of immune responses elicited by the A/H5N1 vaccine.

Statistically significant associations were found in this study between persistence of at least a 1/10 level of A/H5N1 antibodies and older age (OR, 5.65), the experience of flu-like symptoms in the past 18 months or between 2 visits (OR, 0.09). However, notwithstanding previous findings of the persistence of influenza antibodies in the elderly [22,23], there remains no specific evidence or explanation for these findings in the age groups examined in this study (18–40 and 41–65 years old). Meanwhile, the finding of lower antibody persistence in people with flu-like symptoms and the mechanisms underlying this relationship need more investigation.