LL RHEUMATIC fever attacks follow streptococcal infections,' but only A 3 per cent2 or less3 of streptococcal infections are followed by rheumatic fever. Hence the search for differences, in degree and in kind, between the few infections that are followed by rheumatic fever and the many that are not? Among these differences may be one or more of causal importance, i.e., determining the occurrence of rheumatic fever after a streptococcal infection.Lack of an adequate experimental model and reluctance to experiment in man have hampered this search. The only relevant findings have thus far been obtained by studying the natural history of streptococcal infections and of rheumatic fever. The pitfalls of such a purely observational approach are notorious; caution is indicated in interpreting the findings, lest guilt be determined by association.The extensive study of primary attacks of rheumatic fever in the general population has established that rheumatic fever, unlike glomer~lonephritis,~ may follow infection with any of the 50-odd types of Group A streptococcus.' Infections with the same bacterial strain under epidemic conditions will cause rheumatic fever in some, and only in some, of the patients infected.2 Therefore, genetic heterogeneity in the pathogenicity of the parasite is not an adequate explanation of the difference in outcome of the infection. Conversely, genetically identical subjects ( monozygotic twins ) even when reared together. are usually discordant for rheumatic fever.6 Thus, genetic heterogeneity in susceptibility of the host, if at all involved, does not suffice to explain the occurrence of rheumatic fever after only a small minority of streptococcal infections. The mean antibody titer of patients with rheumatic fever is somewhat higher than that of patients with uncomplicated streptococcal sore throat, but extensive overlap of the two groups O C C U~S .~ It has been suggested that repeated, closely spaced streptococcal infections are particularly conducive to rheumatic fever.s The finding of elevated streptococcal antibody titers at the onset of streptococcal pharyngitis in patients who later developed rheumatic fever seemed to support this suggestion.' Other studies, however, failed to confirm this finding.0 The fact is that it is dficult to obtain meaningful data on the remote streptococcal history of patients with first attacks of rheumatic fever. It is doubtful whether the first serum sample obtained from a patient who comes to observation because of symptoms of pharyngitis reflects the pre-infection status or, instead, the early infectious stage with attendant beginning
303-