The challenge in developing a dengue vaccine is that it must protect against DHF/DSS while not making the population susceptible to DHF/DSS. Such a task poses a most difficult problem to workers in the vaccine development field.
Hotta (1957), Sabin and Schlesinger (1945), Schlesinger et al. (1956), and Wisseman et al. (1963) demonstrated the ease and reproducibility of selecting an attenuated live dengue virus vaccine by serial passage in the brains of suckling mice.
More recently, dengue attenuation strategies have used clonal selection of viruses grown in tissue culture to pick variants from wild virus populations. These variants (dengue 1, 2, and 4) demonstrated temperature sensitivity, reduced suckling mouse neurovirulence, reduced rhesus monkey viremia, and somewhat lower antibody responses in infected monkeys, attributes thought to indicate attenuation for humans (Bancroft et al., 1984; Eckels et al., 1984; Halstead et al., 1984a,b,c,d). When tested in humans, however, the variants either were overattenuated, yielding unacceptably low seroconversion rate in susceptibles (dengue 2 and 4), or were underattenuated, producing symptoms in human volunteers (dengue 1 and 4). A dengue 2 vaccine serially passaged 50 times in primary dog kidney cells has been shown to be acceptably immunogenic and nonreactogenic in 10 adult volunteers tested in Thailand (Halstead, personal communication, 1985). However, it shares with all other vaccine strains limited growth potential in suitable mammalian cells, and thus the maximum titer of the vaccine probably will be about 105.
Any dengue vaccine program must have as major components ongoing epidemiological studies of risk factors, surveillance for DHF/DSS cases among vaccinees, and fundamental immunological studies designed to elucidate the mechanisms of DHF/DSS. A coherent understanding of risk factors is essential for the successful development, testing, and use of dengue vaccines. Current knowledge indicates that use of nonpersisting antigens would be extremely dangerous in situations in which dengue viruses continue to circulate in human populations. The existence of a jungle dengue cycle clearly means that the disappearance of one or more dengue viruses from circulation would be transient. As long as an efficient urban vector is established, dengue transmission must be expected at any time.
Unless virological risk factors for dengue shock syndrome can be authoritatively determined for residents of all countries in DHF/DSS endemic areas, a tetravalent vaccine appears to be the best option available. Recent studies from Thailand, however, suggest that only secondary dengue 2 infections in children result in DSS, allowing for the possibility of a monovalent dengue 2 vaccine to protect against severe disease and death (Sangkawibha et al., 1984). Studies in monkeys show that neutralizing antibodies can be developed to all four components of a tetravalent vaccine (Halstead and Palumbo, 1973). The success of the combined measles, mumps, and rubella vaccines demonstrates that multiple live viruses can be inoculated in man without interference.