Assessment of vaccine efficacy by challenge studies in man.

Introduction When standardizing and controlling influenza and other vaccines it is clearly essential to have exact information on the antigens it contains and the amount of each. Methods are available to deal with some of the difficulties which arise, such as the differing antigenicity of vaccines put up with different adjuvants. However, there may be a need for further information which any number of laboratory tests cannot provide. One might need to know, for example, whether the antigen evokes immunity against the virus in circulation in the community and whether it will induce a primary or a secondary (booster) type of response; this clearly depends not only on the vaccine but also on the antigens to which the vaccinee has been exposed and the imprint they have left on his immune system, a complex matter which at the moment cannot be mimicked in animals used in control laboratories, although this may yet become possible (Potter et al., 1973). Further, the same antigens may be administered by a number of different routes, and the response by these may not be exactly predicted by laboratory tests because of difference in the physical state of the antigens and their mode of administration-a good example is the administration of killed vaccine by spray or in a fluorocarbon propellant. In this case circulating or local antibodies may be a poor guide to the resistance to infection induced in the respiratory tract. Then again, the effectiveness of a live vaccine obviously depends not only on its antigenic composition but on how and where it replicates in the respiratory tract. The ferret may be of some value here but can be only a rough guide to what happens in man. Finally, the really important index of the effectiveness of all vaccines is the degree to which they render the respiratory tract resistant to infection with the current serotype of virus. This is particularly obvious when one is comparing vaccines which are compounded in different ways or administered by different routes antibody responses may not be comparable either qualitatively or quantitatively and the question at issue is not which produces the best antibody of a certain sort or in a certain compartment but 'what is the net effect on the resistance of the host to infection?'. In the past the only way to get such answers was to conduct field trials and for this one requires large numbers of volunteers and vaccine, a substantial field organization, and good fortune in that the results will not be obtained unless there is a substantial epidemic at a suitable interval after the vaccine is given. The net result is that only relatively simple questions can be asked in such trials, such as 'Does this vaccine protect?'; it is very difficult to obtain quantitative comparisons of effectiveness which will withstand critical statistical analysis. It would appear that challenging vaccinated volunteers with live virus can overcome many of these problems. There are ethical difficulties in giving an unmodified virus-one cannot exclude the risk of serious disease or possible permanent sequelae, and if volunteers are not isolated there is the chance of infecting contacts or even starting a local epidemic; if they are isolated there is the problem of keeping enough individuals in isolation long enough to get a sufficient number of results. It is now possible to obtain reliably attenuated influenza viruses, which will soon, it is hoped, be licensed as live vaccines. These can be used for challenge as well as immunization. Admittedly the virus is modified, but 107 EID50 represents a relatively small number of human infectious doses and the administration by nasal drops or spray is, in a general way, similar to natural exposure, which is presumably by a droplet which lands on the nasal or lower respiratory epithelium. Clearly it can be used in non-isolated volunteers without worries about the occurrence of disease or epidemics.