(See the Major Article by Ndeffo-Mbah and Pandey on pages 2026–34, and the Major Article by Kareko et al., on pages 2018–25.) The yellow fever (YF) 17D vaccine is a live attenuated vaccine developed by Max Theiler and colleagues in the 1930s. The vaccine is produced in embryonated chicken eggs using technology that has changed little since the seed-lot system was implemented in the early 1940s. It has proved to be a very safe and effective vaccine and, somewhat surprisingly, is the only vaccine for which a Nobel prize has been awarded (in 1951). Despite the great success story of the vaccine, the mechanism of protective immunity and our understanding of how the vaccine works are very limited. Historically, the vaccine has been administered as a one-dose vaccine with boosters every 10 years to maintain protective immunity. In 2013, the World Health Organization (WHO) convened a working group to assess the need for booster doses, and a recommendation was made that booster doses were not needed for the general population and should be restricted to special populations. The recommendation was endorsed by the WHO Strategic Advisory Group of Experts on Immunization [1] and, subsequently, national organizations, including the US Centers for Disease Control and Prevention’s Advisory Committee on Immunization Practices (ACIP) [2]. This issue of The Journal of Infectious Diseases contains 2 articles that look at very different aspects of YF vaccination. It would be fair to say that not everyone agrees with the booster recommendation, with concerns over waning immunity over time. In one article, Kareko et al [3] provide a well-executed study suggesting that the need for a booster dose at 10 years be reconsidered, at least in nonendemic areas. In addition, there is the WHO Eliminate YF Epidemics (EYE) initiative, which proposes to immunize 1.4 billion people in endemic countries over 10 years [4]. Deciding on priority geographic areas for vaccination campaigns and vaccine coverage levels are critical components of EYE, and the article by Ndeffo-Mbah and Pandey in this issue [5] uses modeling to suggest immunization levels needed in the EYE initiative. Significantly, the proposed level is higher (90%–95%) than that currently being considered by EYE (80%). If the conclusions of both articles were implemented, there would be need for much more vaccine than that available at the present time. Kareko et al studied a cohort in the United States, a nonendemic area, and provide evidence of waning immunity over time, as determined by loss of detection of neutralizing antibodies. Of particular concern is the implication from the data that 20% of vaccinees will lack detectable neutralizing antibodies at 10 years after vaccination. This, in turn, would suggest that travelers should receive a booster dose of vaccine every 10 years. Whereas the implication is straightforward, analyses of the data are not so straightforward. The mechanism of protective immunity of licensed flavivirus vaccines in humans is poorly understood, so we tend to use neutralizing antibodies as a surrogate of protection when in reality it is a correlate of protection. The definitions here are important. A correlate of protection is defined as “an attribute that is statistically associated with an endpoint (without the association necessarily being causal)” whereas a surrogate of protection is defined as “an attribute on the direct causal pathway, ie, the vaccine induces protection via the measured mechanism” [6]. For YF, we use neutralizing antibodies as our “gold standard” for protective immunity, but this is based on studies of passive transfer of human sera to monkeys and protection from wild-type challenge. This was translated into a log neutralization index (LNI; ie, constant antibody varying concentrations of virus) of 0.7 correlated with protection. Today, we use the plaque reduction neutralization test (PRNT; ie, constant virus, varying concentrations of antibody). The LNI and PRNT have never been formally compared using standardized reagents, but this author would estimate that an LNI of 0.7 is equivalent to a 50% PRNT (PRNT50) of around 1 in 40 to 50. Today, either 90% PRNT (PRNT90), 80% PRNT, 60% PRNT, or PRNT50 titers are reported in studies, and as stated by Kareko et al, articles use different neutralization assays, which makes comparisons difficult. Furthermore, most studies do not use standards, so that assays in different laboratories cannot be compared easily. Nonetheless, Kareko et al are to be applauded for their efforts to compare different studies from nonendemic countries. However, an international standard is available for neutralization assays (https://www.who.int/bloodproducts/catalogue/Anim.pdf?ua=1), and researchers are encouraged to use this standard as it enables different studies to be directly compared. Flaviviruses exhibit extensive serological and genetic cross-reactivity that makes identification of infection due to a specific flavivirus difficult. As such, neutralization assays have long been used in arbovirology as the assay of choice to identify a particular flavivirus infection. A PRNT90 titer is used, as this avoids much of the cross-reaction because neutralization is being measured at the bottom of the S-shaped neutralization curve. One would not predict much variation in PRNT90 titers and this was found in the current study. If a PRNT50 were used, one would expect neutralization titers to be higher as it is the midpoint of the S-shaped curve. It should be pointed out that most clinical trials use a PRNT50 assay with a titer of 1 in 10 as a correlate of protection. Despite the variability in the assays, the current study and other articles agree that neutralization titers wane over time. The critical question becomes “is this important?” and, if yes, “when is it important?” These are not easy questions to answer and 2 additional snippets of information make the situation more confusing. First, not all vaccinees have an anamnestic neutralizing antibody response following a booster dose (ie, a 4-fold or greater increase in neutralization titer following the booster dose to that prior to booster vaccination). This would be suggestive of sterilizing immunity, but is not seen in all individuals. Second, given the clinical picture of YF disease, it would be expected that if there is no longer protective immunity, that cases of YF would be reported. However, very few cases of vaccine failure have been recorded, albeit these are in endemic areas and not in travelers. Thus, our understanding of how the vaccine works and whether or not the mechanism is neutralizing antibodies is unclear. This makes interpretation and consideration of implementing changes in the vaccination regimen difficult, but studies such as those by Kareko et al are important as they help fill our gaps in knowledge, in this case in a nonendemic population. As pointed out by the authors plus ACIP and WHO recommendations, the critical issue is whether travelers will visit a YF-endemic area. For many YF-endemic countries, the entire country is not YF endemic and the virus is restricted to a subset of regions/areas and, as such, recommendations for vaccination of travelers should be made based on the area(s) that the traveler will visit and whether or not the virus is found in these areas. In addition to questions regarding waning immunity, there is the need to vaccinate all individuals at risk in endemic countries, and this is investigated in the second article in this issue by Ndeffo-Mbah and Pandey. This is no easy task given that approximately 800 million people are at risk from YF in about 46 countries in sub-Saharan Africa and tropical South America. Clearly, the risk is not the same in all areas, but 2016 was a wake-up call when YF returned to Angola and expanded to the Democratic Republic of Congo (DRC), with infected travelers taking the virus to other countries, including the People’s Republic of China where there were 11 cases due to export of infected individuals from Angola. Fortunately, there were no secondary cases in China, but this was the first time cases had been seen in Asia. The Angola/DRC outbreak was a major milestone in YF vaccination history as the reserve stockpile of 6 million doses of vaccine was exhausted twice in the first half of 2016 and a fractional-dosing strategy had to be implemented to provide vaccination to those who needed it. In brief, fractional dosing involved giving vaccinees one-fifth of a vaccine dose. The premise was that one dose was 0.5 mL and that it would be possible to provide equal volumes of vaccine if a volume of 0.1 mL was given using a tuberculin syringe. YF vaccine recommendations are that each dose of vaccine should contain no less than 1000 international units (IU) and, because of shelf life considerations, a dose usually contains ≥10,000 IU; thus, a fractional dose contained ≥1,000 IU. A vaccination campaign using fractional dosing was undertaken and early data indicate this strategy worked. The Angola/DRC outbreak made it clear that vaccine supply and demand is an issue, plus there is a need to vaccinate all at risk in endemic areas to prevent outbreaks. This gave rise to the WHO EYE initiative [4] that has a bold plan to administer 1.4 billion doses of vaccine in 10 years. Further support for the need for EYE was the need for fractional dosing in Brazil in 2017–2018 when vaccine supplies were again exhausted. Some have suggested that fractional dosing should be used for routine implementation of the vaccine as this would overcome issues of supply and demand of the vaccine, although WHO has stated this strategy is for emergency use only. An important consideration here is that vaccine from 4 producers is used internationally. They are WHO “prequalified” and use 3 different substrains of 17D vaccine (17D-204, 17D-213, and 17DD) that contain different quantities of vaccine in a dose, there being only a minimum quantity of vaccine (1000 IU) and not a maximum quantity of vaccine in one dose. Studies on fractional dosing have only been published with the 17DD substrain vaccine from Bio-Manguinhos in Brazil and, thus, vaccine from this producer has only been used in fractional dosing vaccination campaigns. Current research is investigating all 4 prequalified vaccines in fractional dosing studies, and we await the results. Ndeffo-Mbah and Pandey propose that 90%–95% of the population needs to be immunized to prevent urban YF outbreaks (ie, a human-mosquito-human transmission cycle), and this would significantly add to the number of vaccine doses needed in the WHO EYE initiative, which is planning on 80% population coverage. It should be pointed out that Ndeffo-Mbah and Pandey focused on vaccination to control urban YF, which takes place in cities. Fortunately, urban YF is relatively rare compared to sylvatic YF (monkey-mosquito-monkey cycle) when humans enter areas where the sylvatic cycle is found and mosquitoes bite humans. In addition, urban YF is very rare in tropical South America, with all urban outbreaks since 1928 involving no more than 10 people, and usually the cases cluster. The situation in Africa is more complicated with urban outbreaks more frequent and involving larger numbers of cases, and these usually involve movement of infected individuals from jungle rural areas in to cities, such as happened in Luanda, Angola, in 2016. Thus, the need for 90%–95% vaccine coverage will need to be considered carefully given the ecology of the virus and the risk of urban YF in cities. Can we eliminate YF? Probably not, as the virus is zoonotic and exists in a mosquito-monkey cycle. As such, vaccination is critical for control of the disease and to eliminate YF epidemics, as is being implemented by the WHO EYE initiative. Eliminating outbreaks will make it safer for both those living in endemic areas and, potentially, for travelers to these areas. Overall, the 2 YF articles in this issue provide important data regarding gaps in our knowledge, but there are still many gaps that need to be filled, especially for a legacy vaccine such as YF 17D that was developed before the modern regulatory era. The vaccine has served us very well over the last 80 years, and there is no doubt that without it, YF would probably be the biggest biosafety level 4 pathogen in the world. Financial support. A. D. T. B. is supported by the Gillson Longenbaugh Foundation. Potential conflicts of interest. Author certifies no potential conflicts of interest. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
No takes yet. Share an insight, caveat, or question.
Alan D.T. Barrett (2019) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: