zation) differ from those of patients with SSPE onset in 1988-1989 indicate a possible change in the nature of the disease since the introduction of measles vaccine and a concurrent decrease in the incidence of measles. If such a change is confirmed by other studies (and this will be difficult, because there are so few new cases of SSPE), it could indicate a different etiology for current SSPE cases compared with those in the past. It could also merely indicate a change in the time of life at which a child is infected with measles and subsequently develops SSPE (e.g., since the beginning of widespread immunization, perhaps only infants who are too young for immunization are infected with measles virus and only a proportion of these develop SSPE).

It will be difficult to obtain other evidence for a causal relation between measles vaccine and SSPE. First, the number of cases of SSPE in the United States is now so low that detection of even moderately strong associations may be difficult. Second, the period of time between infection with the measles virus and development of SSPE is quite long, and if an association between measles vaccine and SSPE exists, a similarly long latency (perhaps 10 years or more) would be expected. Even if the latencies for the two conditions were different and the difference were moderately large, the difference would be difficult to detect because the range of time from measles infection to SSPE is fairly long and the number of new cases of SSPE is low.

Although application of new scientific methods, such as RNA sequencing, could be used to describe more completely the virus that causes SSPE, the well-known genetic alterations of the virus from wild-type measles virus will confound interpretation of the data and make it unlikely that investigators will be able to determine whether there is an independent association between measles vaccine and the development of SSPE.

There has been some concern as to whether measles vaccine could exacerbate preexisting SSPE (Dodson et al., 1978) and whether a second dose of measles vaccine could more often result in SSPE (Halsey, 1990). After publication of the case report of Dodson et al. (1978) of an 8-year-old boy with SSPE whose condition appeared to have been exacerbated by administration of the measles vaccine, Halsey et al. (1978) reported data suggesting that such a concern was not warranted. The National Registry for Subacute Sclerosing Panencephalitis contained records of nine patients who received attenuated or killed measles vaccine after the onset of SSPE symptoms. Four of the nine patients died an average of 3.6 years after the onset of SSPE symptoms and 2.4 years after vaccination. The remaining five patients on record at that time were still alive an average of 10.5 years after the onset of SSPE symptoms and 9.3 years after vaccination. Halsey and colleagues argued that the variability in the course of SSPE rendered the assertions of Dodson et al. (1978) questionable. The same data set