Rabies remains an important global health concern, with an estimated global annual human death of 25 to 159 thousand cases (1). Most of these cases occur in African and Asian developing countries, with 60% of the global disability adjusted live years lost due to rabies occurring in Asia (1). Human rabies cases in developing countries are mostly transmitted by domestic dogs. Although they are mainly kept free-roaming, dogs in developing countries have a function, as for example in Indonesia where they are often kept as guardians for properties and to chase away wild animals, particularly monkeys and wild pigs (2, 3). Vaccination of dogs has been repeatedly shown to be the most effective strategy to control rabies in a sustainable way. However, diverse challenges, from political to organisational, have been identified to reach a high enough vaccination coverage in dog populations (4). As a consequence, the annual vaccination coverage is most often lower than the vaccination coverage of 70% recommended by WHO for preventing the circulation of rabies virus among dog population, as for example 53% that was reached in Flores Island, Indonesia over the period 2000–2011 (5).

In addition to a high coverage of mass dog vaccination, providing high-quality dog vaccine is key for the success of canine rabies elimination programs (6). High quality and effective rabies vaccine exist on the market, with laboratory trials documenting 93% seroconversion in vaccinated dogs after a single dose (7). However, development of immunity under field conditions might be hampered by several factors. Identifying them, helps to inform effective rabies vaccination campaigns.

Several studies have been conducted in European countries to identify risk factors of development of antibodies in dogs after vaccination (8–13). Age and breed are amongst the most commonly identified factors correlated with the lack of detectable immunity production after vaccination in dog population (9–12). For example, Keneddy et al. studied the antibody response of 10,483 dogs tested at the Veterinary laboratories Agency, Weybridge UK and found that adult dogs aged 1–7 years were more likely to have an antibody titre >0.5 International Units per ml serum (IU/ml) than young dogs (<1 year) (9). Berndtsson et al. investigated factors associated with the development of antibody titre of 6,789 vaccinated dogs in Sweden and found that larger breeds were at higher risk of having antibody titre <0.5 IU/ml (12). However, most of the above studies were conducted in developed countries for the purpose of international pet travel. Field studies to identify risk factors of development of antibodies in free-roaming domestic dogs in developing countries are limited (14, 15). In a randomised controlled study conducted in Tanzanian dogs, Lugelo et al. found that there was a positive correlation between body condition score (BCS) and seroconversion in which those dogs with good BCS were more likely to seroconvert than dogs with poor BCS (14). In contrast, Morters et al. investigated risk factors of antibody response of the free-roaming domestic dogs after vaccination in South Africa and found that the vast majority of the dogs seroconverted 30 days after vaccination, regardless of the health status of vaccinated dogs (15).

The majority of free-roaming dogs in Indonesia, as in most of developing countries, are aged <1 year and have a poor BCS as a consequence of insufficient nutrition uptake or parasitic diseases (16). The influence of BCS on the losing antibody titre has been reported in the previous studies (17). In a prospective study, Wera et al. found that the dogs with low BCS tend to more rapidly lose detectable antibody titres after vaccination than dogs with high BCS (17). However, to which extent the dogs develop binding antibodies after vaccination has not yet been analysed in these populations. In this study, we first investigated the level of and associated factors for the presence of binding antibodies in 130 healthy owned free roaming dogs from Flores Island, Indonesia. Some of them have previously been vaccinated, and the vaccination status was based on owners' reports. This study approach was previously used in the literature (10, 12). However, the low level of antibodies in the investigated dogs does not fully represent the development of immunity, because dogs may have already lost detectable binding antibodies after vaccination, although cellular immunity may still be present (18, 19). Therefore, we secondly investigated binding antibody titres and identified factors associated with the development of antibodies within 30 days after vaccination among a subsample of dogs that had a binding antibody titre <0.5 Equivalent Unit per ml serum (EU/ml) at the day of vaccination.

The success of a parenteral mass vaccination campaign in the field depends on many factors. These include dog owners' behaviour, infrastructure, the sustainability of vaccination campaign programs, and immunity coverage after vaccination. A good balance of these factors could contribute to effective prevention of rabies both in dogs and humans. For example, 2 years mass dog vaccination campaigns in Bali using long-acting vaccines with a coverage of >70% in 2010 and 2011 have led to the reduction of dog and human rabies cases by ~90% (5). Similarly, dog and human rabies incidence in Latin America and Caribbean has successfully decreased by 90% due to sustainable efforts in mass dog vaccination campaigns using a high quality of rabies vaccines (22). A high quality of rabies vaccines is expected to guarantee protection in dogs from rabies, resulting in a high immunity coverage. Vaccination coverage of 70% after sustained annual vaccination is necessary to interrupt the rabies virus circulation among dogs and is thus more cost-effective than PEP alone for preventing human rabies (23, 24). In Flores Island, mass dog vaccinations are carried out annually, targeting all dogs older than 3 months, achieving vaccination coverage around 53% on average (5). This low vaccination coverage was unable to eliminate rabies in Flores Island (25, 26). Furthermore, we earlier reported that the loss of detectable immunity is substantial within 1 year after vaccination in dogs on Flores Island, indicating that vaccination campaign on annual basis may be insufficient (17). Currently, the development of binding antibodies after vaccination has not been investigated yet in these populations, which was the objective of this study.

The present study highlighted the proportion of dogs with antibody titre ≥0.5 EU/ml increased from 30% (39/130) at D0 before vaccination to 86.8% (79/91) at D30 (30 days after vaccination). Increasing the proportion of dogs with antibody titre ≥0.5 EU/ml 1 month after vaccination is well-documented in the literature and findings lay in the range of what was found in the present study (7, 14, 27). Lugelo et al. reported that the proportion of dogs having a titre higher than 0.5 IU/ml increased from 4% (D0) to 85% (D30) (14). Similarly, Minke et al. studied the antibody titre in a group of 30 laboratory dogs vaccinated with Rabisin rabies vaccine and found that the proportion of dogs with a titre ≥0.5 IU/ml increased sharply from 0% at D0 to 93% at D30 post-vaccination (7). Furthermore, Kallel et al. studied the immune response of 1,000 vaccinated domestic dogs under field condition in Tunisia and found that the proportion of dogs with a protective antibody titres at D0 and D30 was 30 and 91%, respectively (27). These results indicate that the vaccine used in the present study was as effective as in comparable studies in producing protective antibody against rabies in the field.

Although development of immunity has already been studied in other contexts, it is of interest to investigate the effectiveness of internationally recognised vaccines in field condition, particularly in areas with limited animal health infrastructure and services, like Flores Island. Moreover, the socio-economic status of dog owners differs between countries, leading to varying availability of resources for ideal dog keeping practises. The results of the present study showed that around 13.2% (12/91) of vaccinated dogs have not produced a sufficient level to be protected against rabies despite the use of an international commercialised rabies vaccine well-known for its efficacy in inducing a strong humoral response (7). Failure to produce a strong humoral response 30 days after vaccination (D30) was also previously found in laboratory experiment (7) and under field conditions (14, 28). Although virus challenge studies would be needed to definitively prove the lack of protection against rabies, for the development of immunity, measuring antibody titres are a useful tool (18, 19). In an experimental study, Auber et al. provided evidence that animals with neutralising antibody titres at the time of challenge towards rabies virus were better protected than those without neutralising antibodies (18). Therefore, these 135% of dogs (n = 12) are very likely to be not protected at D30, because of their lack of reaction towards the vaccine.

In our study, failure to produce a strong humoral response was linked to low BCS. The multivariable logistic regression analysis results indicated that the main determinant of antibody development at D30 was the BCS. The results are consistent with a field study in Tanzania, in which 412 free-roaming domestic dogs following single dose of rabies vaccination, demonstrated the significant association between BCS and seroconversion (14). Similarly, Wera et al. have reported significantly higher proportion of antibody titres >0.5 EU/ml in dogs with good BCS vs. poor BCS at 90, 180, and 270 days after rabies vaccination (17). These findings suggest that the present study provides consistent evidence for BCS influencing the development and presence of binding antibody titres following rabies vaccination.

This observed association between presence of antibody titres and BCS may be attributed to the health status of dogs at the time of vaccination because other dog characteristics such as age, sex and breed, the location they live (urban vs. rural), and management practises (feeding, origin and role of the dog) did not show any significant relation to the presence of antibody titres. Dogs with poor BCS are malnourished and most frequently found with high burden of endo and ecto-parasites (own unpublished data from West Timor, Indonesia). This is most likely due to the fact that dogs in developing countries are free roaming day and night and therefore encounter more dogs and potentially contaminated environment (2, 15, 16). They have, thus, a high potential to be infected by parasites. This further depresses the immune system and consequently reduce antibody production (29). Booster vaccination focusing on dogs with low BCS and educational campaigns to improve knowledge on dog management practise, promote responsible dog ownership including good feeding and health practises, is expected to improve BCS of the dogs, which would enhance producing immunity after vaccination. We thus argue that investing into improved dog keeping practises leading to an increase of dog health and BCS, could highly improve the efficiency of rabies vaccination campaigns in the future.

In our study, we exclude obviously sick dogs because another part of the project is aimed at investigating normal dog roaming behaviour (3, 16, 30). Investigating the effect of the dog's health status in addition to the effect of BCS on the development of immunity after vaccination in the field should be of interest for future studies.

The present study found that 46.8% of dogs that had a history of vaccination within 12 months before D0 had evidence of anti-rabies antibodies at D0. Having a low level of antibodies (<0.5 EU/ml) does not always indicate poor reaction towards vaccination, because it may be that the dogs have developed immunity after the vaccination in the past, but had a decay in the detectable antibodies until blood sampling. Low coverage of antibody titres >0.5 EU/ml after the mass vaccination against rabies in dog populations was also reported by other authors (27, 31, 32). Tepsumethanon et al. described that 42% of vaccinated dogs in Thailand failed to present rabies antibody titres >0.5 IU/ml 360 days after vaccination (31). Similarly, Cliquet et al. reported low immunity coverage (36%) 1 year after vaccination in a dog population in Tunisia (32). These findings suggest that a large number of dogs are potentially susceptible to rabies 1 year after the vaccination campaign, although cellular immunity that is not measured by the antibodies may still be present (18, 19). It is noteworthy that in our study, the majority of dogs with antibody titres <0.5 EU/ml at D0 are dogs having no history of vaccination before D0 (63.7%; 58/91), which was found to be a significant factor for the D0 titre. Therefore, booster vaccination (for example, 6 months after first vaccination) targeting on dogs without previous vaccination is crucial to increase the immune response between yearly campaign and eventually eliminate rabies both in dogs and humans.

The history of vaccination in the present study was mainly based on the dog owners' declaration. This approach is also reported by other authors (33–35) and may lead to an over- or underestimation of the proportion of vaccinated dogs, as dog owners could misclassify the dog's vaccination status. Dog vaccination campaigns in Flores Island are carried out on a yearly basis by public veterinary offices without providing vaccination certificate for dog owners. Despite this limitation, the results of this study illustrate the link of history of vaccination and the presence of a strong immune response in owned free-roaming domestic dogs in Flores Island. Finally, our study revealed that 14.7% (n = 10) of dogs with either no previous vaccination or more than 12 months before D0, had antibody titres ≥0.5 EU/ml at D0. According to the statement of the owners, most of these dogs were previously vaccinated more than 12 months before D0. However, for four (3% of the entire study population of 130 dogs) dogs, owners reported that the dogs were unvaccinated during their life. Although it may be that the statement of the owners was wrong (for example due to lack of memory or change of the dog's owner over time), natural immunity due to exposure to rabies virus not leading to fatal disease can be the reason for this observation (36, 37).

Our findings showed a high immune response in owned free-roaming domestic dogs vaccinated under field conditions in Indonesia. History of vaccination and good BCS were significantly associated with the presence of rabies antibody at D0 and the development of antibodies at D30. Given the significant association with BCS, an educational campaign focusing on dog management practise is expected to improve the health status and BCS of the dogs, which would enhance producing detectable immunity after vaccination.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The animal study was reviewed and approved by Animal Ethics Commission of the Veterinary Medicine Faculty, Nusa Cendana University (Protocol KEH/FKH/NPEH/2019/009). Written informed consent for participation was not obtained from the owners because dog owners in Flores were not familiar with written informed consent.

EW, CW, SD, PMB, and MMS conceived the study. EW, CW, PMB, and MMS collected the field data. EW, CW, SD, and PMB conducted statistical analyses. EW wrote the original draft of the manuscript. CW, SD, PMB, and MMS contributed to reviewing the manuscript. All authors read and approved the version of the manuscript to be published.

The study was funded by the Kupang State Agricultural Polytechnic (Politeknik Pertanian Negeri Kupang) in the form of a grant awarded to EW and PMB (Project Nr. 1/P2M/DIPA.042.01.2.40.1014/2019), the Albert Heim Foundation (Project Nr. 132), and the Wolfermann-Nägeli Foundation (Nr. 2018/28) in the form of a grant awarded to CW and SD.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.