Pre-exposure rabies prophylaxis: a systematic review

Abstract Objective To review the safety and immunogenicity of pre-exposure rabies prophylaxis (including accelerated schedules, co-administration with other vaccines and booster doses), its cost–effectiveness and recommendations for use, particularly in high-risk settings. Methods We searched the PubMed, Centre for Agriculture and Biosciences International, Cochrane Library and Web of Science databases for papers on pre-exposure rabies prophylaxis published between 2007 and 29 January 2016. We reviewed field data from pre-exposure prophylaxis campaigns in Peru and the Philippines. Findings Pre-exposure rabies prophylaxis was safe and immunogenic in children and adults, also when co-administered with routine childhood vaccinations and the Japanese encephalitis vaccine. The evidence available indicates that shorter regimens and regimens involving fewer doses are safe and immunogenic and that booster intervals could be extended up to 10 years. The few studies on cost suggest that, at current vaccine and delivery costs, pre-exposure prophylaxis campaigns would not be cost-effective in most situations. Although pre-exposure prophylaxis has been advocated for high-risk populations, only Peru and the Philippines have implemented appropriate national programmes. In the future, accelerated regimens and novel vaccines could simplify delivery and increase affordability. Conclusion Pre-exposure rabies prophylaxis is safe and immunogenic and should be considered: (i) where access to postexposure prophylaxis is limited or delayed; (ii) where the risk of exposure is high and may go unrecognized; and (iii) where controlling rabies in the animal reservoir is difficult. Pre-exposure prophylaxis should not distract from canine vaccination efforts, provision of postexposure prophylaxis or education to increase rabies awareness in local communities.


Introduction
Rabies is a preventable yet fatal disease that is responsible for approximately 59 000 deaths each year. 1 However, widespread underreporting of rabies cases means that the actual number of deaths is likely to be higher. Poor and rural populations are disproportionately affected, with the majority of deaths occurring in children younger than 15 years in Asia and Africa. 2 Ninety-nine per cent of human rabies cases result from dog bites and, once symptoms begin, the disease is almost invariably fatal. 3 Human rabies is preventable through canine vaccination to eliminate rabies at its source or by administering rabies vaccines and immunoglobulin following bites, scratches or saliva exposure from suspected rabid mammals (i.e. postexposure prophylaxis). 4 Another preventive strategy is pre-exposure prophylaxis, which involves giving a series of intramuscular or intradermal injections of rabies vaccine to prime the immune system. This enables fast recall of memory immune responses once a person is re-exposed to the virus. 4 Moreover, people who have received pre-exposure prophylaxis require fewer doses of postexposure rabies vaccine and can be treated without rabies immunoglobulin, which is costly and difficult to procure. 4 Although preventing rabies in dogs is the most cost-effective way of preventing human rabies deaths, pre-exposure prophylaxis is valuable for people at a high disease risk, 5 particularly in areas where controlling disease in the animal reservoir is difficult or has not been implemented and in areas where access to postexposure prophylaxis and rabies immunoglobulin is unreliable or nonexistent. National pre-exposure prophylaxis programmes for high-risk populations have been implemented in Peru and the Philippines. 6,7 In 2010, a World Health Organization (WHO) position paper on rabies vaccines called for studies on the feasibility, cost-effectiveness and long-term impact of incorporating vaccines derived from cell culture or embryonated eggs into immunization programmes for children where canine rabies is a public health problem. 5 The paper also made recommendations on pre-exposure prophylaxis regimens and on the frequency of booster vaccinations and serological surveillance for at-risk individuals, such as veterinarians. The aim of this study was to review the scientific literature published between 2007 and 2016, as well as field data, to assess the current use and cost-effectiveness of pre-exposure rabies prophylaxis (excluding travel vaccines), particularly in children and in high-risk settings, in the context of recommendations made in the 2010 WHO rabies vaccine position paper on pre-exposure prophylaxis and booster vaccine administration.
Objective To review the safety and immunogenicity of pre-exposure rabies prophylaxis (including accelerated schedules, co-administration with other vaccines and booster doses), its cost-effectiveness and recommendations for use, particularly in high-risk settings. Methods We searched the PubMed, Centre for Agriculture and Biosciences International, Cochrane Library and Web of Science databases for papers on pre-exposure rabies prophylaxis published between 2007 and 29 January 2016. We reviewed field data from pre-exposure prophylaxis campaigns in Peru and the Philippines. Findings Pre-exposure rabies prophylaxis was safe and immunogenic in children and adults, also when co-administered with routine childhood vaccinations and the Japanese encephalitis vaccine. The evidence available indicates that shorter regimens and regimens involving fewer doses are safe and immunogenic and that booster intervals could be extended up to 10 years. The few studies on cost suggest that, at current vaccine and delivery costs, pre-exposure prophylaxis campaigns would not be cost-effective in most situations. Although preexposure prophylaxis has been advocated for high-risk populations, only Peru and the Philippines have implemented appropriate national programmes. In the future, accelerated regimens and novel vaccines could simplify delivery and increase affordability. Conclusion Pre-exposure rabies prophylaxis is safe and immunogenic and should be considered: (i) where access to postexposure prophylaxis is limited or delayed; (ii) where the risk of exposure is high and may go unrecognized; and (iii) where controlling rabies in the animal reservoir is difficult. Pre-exposure prophylaxis should not distract from canine vaccination efforts, provision of postexposure prophylaxis or education to increase rabies awareness in local communities.

Literature search
Our literature review was intended as an update of the review of the evidence on pre-exposure prophylaxis carried out for the 2010 WHO rabies vaccine position paper. Our search was conducted according to preferred reporting items for systematic reviews and meta-analyses guidelines. 8 We searched the PubMed, Centre for Agriculture and Biosciences International, Cochrane Library and Web of Science databases for papers on pre-exposure rabies prophylaxis published between 2007 and 29 January 2016 ( Fig. 1) using the search string: "rabies" AND "pre-exposure" AND ("prophylaxis" OR "vaccin*"). We started at 2007 to include studies published after completion of the review for the WHO position paper. Additional references were obtained from citations in relevant publications. We excluded studies that assessed: (i) postexposure prophylaxis only; or (ii) pre-exposure prophylaxis either occupationally or in travellers.
As we considered the safety and immunogenicity of WHO-recommended vaccination regimens for pre-exposure prophylaxis to be well established, we also excluded papers that confirmed the efficacy of these regimens, unless they specifically assessed the safety and immunogenicity of pre-exposure prophylaxis in children or given in combination with other vaccines. We included any type of study, in any language and from any country that assessed: (i) pre-exposure prophylaxis in children; (ii) the cost-effectiveness of pre-exposure prophylaxis; (iii) accelerated or revised pre-exposure prophylaxis regimens; or (iv) booster vaccination recommendations. Studies that assessed the cost of pre-exposure prophylaxis and its use in children were included regardless of publication date.

Field data
We reviewed field data from completed and ongoing pre-exposure prophylaxis campaigns in Peru and the Philippines. In Peru, the campaigns targeted people living in remote areas who were at risk of contracting rabies from vampire bats, whereas in the Philippines they targeted children at risk of dog-transmitted rabies.

Peru
In Peru, vampire bats are a common source of rabies: the life-time risk of a bat bite in rural Amazon basin populations is reported to be 41 to 88%. 9,10 Outbreak reports and responses are delayed by the remoteness of these populations and controlling rabies in the bat reservoir is challenging. 11 Following a rabies outbreak in 2011, Peru began a mass pre-exposure prophylaxis vaccination campaign that targeted people in Condorcanqui and Bagua Provinces at a high risk of rabies from vampire bats. 6 The risk was regarded as high in these provinces because: (i) bat bites were common; (ii) there was evidence of rabies in circulation; (iii) housing conditions increased vulnerability; (iv) protective measures among the population were lacking; (v) tools for vector control were lacking; and (vi) the remote location of villages delayed health service responses. The campaign involved administering three intramuscular doses of human diploid cell or purified Vero cell vaccine on days 0, 7 and 28. Villages were prioritized to receive the intervention by classifying their epidemiological risk using the following variables: (i) the endemicity of bat rabies; (ii) the number of human rabies cases within the previous 6 months; (iii) the number of livestock rabies cases within the previous 6 months; (iv) the frequency of vampire bat bites (where there was bite surveillance); (v) history of postexposure prophylaxis; and (vi) important, recent ecological changes, such as an increase in the population, a change in the feeding habits of vampire bats, illegal mining or deforestation. Among villages with a history of postexposure prophylaxis interventions, priority was given to those in which the intervention took place more than 1 year previously, those where a low percentage of the population had received postexposure prophylaxis and those close to the site of a recent outbreak or of documented circulation of the rabies virus.

Philippines
In the Philippines, pre-exposure rabies prophylaxis is recommended as an additional intervention for high-risk  individuals, such as children and people at occupational exposure. In 2007, the Philippine Government implemented a Department of Health recommendation that free routine pre-exposure prophylaxis should be provided for school children aged 5 to 14 years who are living in high-risk areas. 7,12 To be included an area had to have: (i) an incidence of human and canine rabies above the national average; (ii) an incidence of animal bites above the national average; (iii) no or low canine vaccination coverage, which was defined as less than 30% coverage of the estimated dog population; and (iv) limited access to postexposure prophylaxis, for example, due to geographical isolation, inadequate treatment facilities or poverty. Schoolchildren were targeted because almost 50% of all rabies exposure in the Philippines occurs in children younger than 15 years. Child deaths due to rabies are associated with poverty and, where postexposure prophylaxis is available, with limited or delayed access to health services. The rationale for pre-exposure prophylaxis was that it: (i) may protect children who do not receive postexposure prophylaxis, for example, after unremarked exposure (i.e. if their antibody titre at exposure is ≥ 0.5 IU/mL); (ii) may protect patients when postexposure prophylaxis is delayed; (iii) accelerates antibody responses to postexposure prophylaxis; and (iv) reduces the cost of postexposure prophylaxis by removing the need for rabies immunoglobulin and reducing the number of postexposure prophylaxis doses required from 8 to 2 ( Table 1). The pre-exposure prophylaxis schedule consisted of administering three intradermal doses of purified Vero cell or chick embryo cell vaccine on days 0, 7 and 28.

Results
The systematic review of the literature identified 31 publications on preexposure rabies prophylaxis that met inclusion criteria (Table 2).

Literature search
The search identified 11 studies on the safety and immunogenicity of preexposure prophylaxis in children aged 2 months to 15 years, including two published before 2007 (Table 3). All found it safe and immunogenic in both infants and children. Three found it safe and immunogenic for up to 5 years when given in combination with other childhood vaccines such as those against Japanese encephalitis, diphtheria, tetanus, pertussis and poliomyelitis (both oral and inactivated vaccines). 27,34,40 High-risk settings

Literature search
Pre-exposure prophylaxis programmes for high-risk populations, and especially children, were strongly recommended in reports of expert meetings on rabies and child health in Asia and the Middle East. 6,13,19 In India, the Academy of Pediatrics called for its inclusion in the immunization schedule for high-risk children younger than 18 years. 39

Peru
In 2011, pre-exposure prophylaxis was administered in 286 localities in the Amazonas Region: 86% were in Condorcanqui Province and 14% were in Bagua Province. In total, 13 986 people were immunized. In these areas, the number of rabies deaths dropped from 13 in 2010 and 20 in 2011 to zero child deaths and only two adult deaths (both had refused vaccination) in 2012. 6 Fig. 2 shows the number of human rabies deaths in Bagua and Condorcanqui Provinces between 1975 and 2015. The number of reported bat bites, which is used as a surrogate for rabies exposure, decreased between 2010 and 2013 (Table 4)

Philippines
By April 2010, the routine pre-exposure prophylaxis immunization programme had achieved an average coverage of 47.25% in the target population: 21 637 high-risk children in 31 schools in seven regions were immunized (unpublished data, 2010). In the town of Cabusao, 188 schoolchildren received at least one vaccine dose (i.e. 86% of those eligible) and 90% of the 188 completed the pre-exposure prophylaxis regimen. Subsequently, 3.5% received postexposure prophylaxis within 3 years following suspected exposure. The programme was stopped in 2011 because a large increase in rabies exposure led to a vaccine shortage and priority was given to the immunization of people involved in canine vaccination campaigns. Preexposure prophylaxis of schoolchildren was planned to restart in 2016.

Literature search
Few recent studies have assessed the cost-effectiveness of pre-exposure rabies prophylaxis. One study estimated the annual global direct cost of administering postexposure rabies prophylaxis at 1.7 billion United States dollars (US$), plus an additional US$ 1.3 billion in lost income. 1 In a cost assessment of pre-exposure prophylaxis, researchers showed that it would be cost-neutral if 1% of children were exposed to rabies each year and if the price of the vaccine did not exceed US$ 1.32 per dose, once the cost of postexposure prophylaxis boosters required after exposure was taken into account. 30 The acceptable vaccine cost increased in proportion to the incidence of rabies. In a Thai study, the estimated cost of pre-exposure prophylaxis for children ranged from US$ 2.00 to 7.25 per child depending on the schedule and vaccine used: there was an additional cost of US$ 18.00 to 23.50 per child if postexposure prophylaxis was required later. 17 Pre-exposure prophylaxis became cost-comparable to the least expensive postexposure schedule (i.e. intradermal immunization without rabies immunoglobulin) when the annual risk of a dog bite was approximately 23%. If equine or human rabies immunoglobulin was used with postexposure vaccines, pre-exposure prophylaxis was cost-comparable when the annual risk of a dog bite was 7% or 3%, respectively. As over 30% of Thai children had been bitten by a dog by the age of 15 years, it was estimated that the actual incidence of dog bites in the population of central  23 Journal article 2008 Thailand PCECV, PVRV Adults Khawplod et al. 24 Journal article 2012 Thailand PCECV, PVRV Adults Khawplod et al. 25 Journal article 2007 Thailand PCECV, PVRV Adults Lang et al. 26 Journal article 1997 Viet Nam PVRV Children Lang et al. 27 Journal  33 Journal article 2011 Australia HDCV Children and adults Pengsaa et al. 34 Journal article 2009 Thailand PCECV combined with Japanese encephalitis vaccine Children Ravish et al. 35 Journal article 2013 India PCECV Children Shanbag et al. 36 Journal article 2008 India PVRV, PCECV Children Strady et al. 37 Journal article 2009 France HDCV, PVRV Children Sudarshan et al. 38 Journal article 2011 Worldwide N/A Children and adults Vashishtha et al. 39 Journal article 2014 India N/A Children Vien et al. 40 Journal Thailand was only 2.3% per year. Consequently, pre-exposure prophylaxis with currently licensed vaccines would not be cost-effective in this setting.

Peru
The cost of the mass pre-exposure prophylaxis campaign was estimated to be US$ 4 111 000, of which US$ 3 560 000 was the cost of the vaccine. The average cost per immunized person was US$ 69. By assuming that the risk of rabies was constant (i.e. the rabies virus remained in circulation and the risk of a bat bite was unchanged) and that, each year, rabies caused 20 deaths per 50 000 people in Condorcanqui Province without preexposure prophylaxis, we estimated the cost of pre-exposure prophylaxis to be US$ 205 000 per life saved after the first year. 6

Philippines
Pre-exposure prophylaxis with three doses of purified Vero cell or chick embryo cell vaccine was estimated to cost US$ 4.77 per patient (unpublished data, 2015; (Table 5). For a patient weighing between 26 and 50 kg, pre-exposure prophylaxis reduced the cost of postexposure prophylaxis by up to 38% following category-II exposure (i.e. "nibbling of uncovered skin, minor scratches or abrasions without bleeding") 5 and by up to 85% following category-III exposure (i.e. "single or multiple transdermal bites or scratches, contamination of mucous membranes with saliva from licks, licks on broken skin, exposures to bats"), 5 after the cost of pre-exposure prophylaxis was taken into account.

Accelerated or revised regimens
Nine studies investigated the safety and immunogenicity of an accelerated or revised pre-exposure prophylaxis regimen (Table 6; available at http://www.who.  23,24 or two visits 41 elicited an adequate antibody titre of 0.5 IU/mL or higher for up to 1 year, 23-25,28,33 even when given in combination with Japanese encephalitis vaccine. 20 Adequate titres were observed in people who received a total dose of at least 2 IU of intradermal pre-exposure prophylaxis. 15 Factors associated with an inadequate antibody titre included: (i) a period of more than 21 days between the first and third doses; (ii) male sex; (iii) vaccine type or manufacturer; and (iv) a body mass index of 25 kg/m 2 or higher. 14

Booster vaccinations
Four studies investigated recommendations on booster vaccines (Table 7; available at http://www.who.int/bulletin/volume/95/03/16-173039). They concluded that: (i) the interval between booster vaccinations could be extended by up to 10 years; 16 (ii) serological surveillance or booster vaccination after 1 year is advisable for people in high-risk occupations; 29 (iii) serological testing after the third intramuscular or intradermal pre-exposure prophylaxis dose is unnecessary; 18 and (iv) healthy subjects may not require postexposure prophylaxis boosters on re-exposure to rabies for up to 3 months after preexposure or previous postexposure prophylaxis. 38

Discussion
Several studies demonstrated that preexposure rabies prophylaxis was safe and immunogenic in children and could be co-administered with other childhood vaccines. 21,22,27,32,[34][35][36][37]40 In addition, it could be given with the Japanese en-cephalitis vaccine in both adults and children. In most African countries pre-exposure rabies prophylaxis is unlikely to be included in the expanded programme on immunization because of competing priorities and because postexposure prophylaxis would still be required following suspected contact. Nevertheless, expert consultations advocate vaccination for people in remote, high-risk areas 6,19,36 and national preexposure prophylaxis programmes have been implemented in Peru and the Philippines. 6,7 In Peru, the programme was successful in preventing child deaths due to bat rabies in high-risk areas, which demonstrates the value of targeted preexposure prophylaxis in places where controlling disease in the animal reservoir is challenging. 11 Although it can be difficult for individuals to recall the date of pre-exposure prophylaxis, this does not undermine its usefulness for saving human lives in situations where exposure is uncertain or there is limited access to biologicals. Vaccination certificates are often treasured and kept safe and, in Peru, the identification and recording of vaccinated individuals has improved nationally.
Pre-exposure rabies prophylaxis is also associated with cost savings because fewer postexposure vaccinations and no rabies immunoglobulin are required following suspected exposure. However, the few studies that assessed costs suggest that community vaccination at current vaccine and delivery costs would not be cost-effective in most situations. 1,17,30,42,43 Preliminary studies on accelerated or revised regimens indicate that 1-week or even single-day regimens may be as effective as the recommended 3-to 4-week regimen: shorter treatment and fewer doses would make treatment simpler and less expensive.
The development of a more immunogenic rabies vaccine that provides life-long immunological memory with a single dose and that can be preserved at ambient temperatures, thereby eliminating the need for a cold chain, would make pre-exposure prophylaxis simpler and more cost-effective. The ideal vaccine would induce an antibody titre that remained above 0.5 IU/mL for decades and would protect people who fail to receive prompt booster immunization following exposure. In animal studies, attempts have been made to increase the current vaccine's immunogenicity using adjuvants, 44   Although pre-exposure prophylaxis using currently available biologicals may not be cost-effective in general, we believe it could be beneficial in: (i) remote communities where access to postexposure prophylaxis and rabies immuno-globulin is often delayed or nonexistent; (ii) situations in which the risk of exposure is high and may go unrecognized, for example, in young children or people exposed occupationally, such as veterinarians; or (iii) places where controlling rabies in the animal reservoir is difficult and the risk of human exposure is high, such as in the Amazon basin where bat rabies is endemic. It is important that staff involved in canine rabies control receive pre-exposure prophylaxis because of their higher risk of exposure. Currently, serological surveillance and booster vaccinations are recommended only for people at an occupational risk. 5 Controlling canine rabies remains the cornerstone of preventing human rabies deaths. Pre-exposure prophylaxis should not distract from canine vaccination efforts, the provision of postexposure prophylaxis and education to raise local awareness of rabies. In high-risk areas, pre-exposure prophylaxis should be included in the expanded programme on immunization in children from 1 year of age, 5 followed by a booster after 1 year. Vaccination should be documented with a certificate and any available medical records should be updated. Targeted, mass campaigns in remote, high-incidence areas should be considered to provide protection for both children and adults and travel recommendations should be provided for newcomers. Accelerated vaccination regimes and novel vaccines that provide life-long immunity with a single dose and are stable at ambient temperatures would make pre-exposure prophylaxis more cost-effective and easier to implement. ■ Program of the Philippine Department of Health.
Pre-exposure rabies prophylaxis Jocelyn A Kessels et al.