Group A streptococcal vaccines.

Abstract

IN THIS ISSUE OF JAMA, KOTLOFF AND COLLEAGUES 1 REport on the safety and immunogenicity of a recombinant multivalent M protein group A streptococcal (GAS) vaccine evaluated in 28 healthy adult volunteers. This preliminary report is a positive step in a long journey ahead to develop a GAS vaccine. Numerous attempts to immunize humans against GAS infections date back to the 1930s, when dead GAS intravenously administered to children who had rheumatic fever reportedly decreased recurrence rates. There was no evidence that these immunizations caused or reactivated rheumatic fever. In the 1940s, heat-killed and ultraviolet-killed GAS was administered subcutaneously, repetitively, and in high inocula to hundreds of military recruits with evidence of toxicity but not efficacy. The role of M type-specific antibodies in human immunity was first evaluated in the early 1950s. Infection with homologous M types of GAS occurred 6 times as frequently among adults without M type-specific antibodies to that particular strain compared with those adults with antibody to that type, although type-specific antibodies did not affect tonsillopharyngeal acquisition of GAS. In 1969, the results of vaccination of 23 healthy siblings of patients with rheumatic fever with a partially purified serotype M3 protein vaccine were reported. Increasing amounts of vaccine were administered subcutaneously for 18 to 33 injections. These patients had a total of 18 GAS infections with high antistreptolysin O titer increases while under prospective surveillance. Two children developed definite acute rheumatic fever and 1 child probable acute rheumatic fever after vaccination, adverse events attributed to the vaccine. Several studies in the 1970s using purified M protein vaccine administered subcutaneously or mucosally (intranasal or pharyngeal) found partial protection from homologous challenge and colonization by the mucosal vaccines. However, there was no correlation between protection against infection and M type-specific antibodies. In 1979, Beachey et al described immunization of adult volunteers with a structurally defined polypeptide fragment of GAS type M24 protein. M type-specific antibodies developed and no hostreactive antibodies were detected. These early studies showed that M type-specific serum bactericidal antibodies had little impact on pharyngeal acquisition of homologous M type streptococci. The presence of bactericidal antibodies was associated with decreased rates of symptomatic infection. Mucosal antibodies against M proteins were associated with decreased rates of pharyngeal acquisition of GAS, which were associated with reduced rates of symptomatic infection. In the dose escalation study by Kotloff et al, a hexavalent (6 M types) recombinant fusion protein vaccine was administered to 8 adult volunteers at the lowest dose (50 μg) and to 10 adult volunteers at each of 2 higher doses (100 μg and 200 μg). The vaccination schedules differed with the lowest dose administered on days 0, 28, and 56, while the higher dosages were administered on days 0, 28, and 112. This precludes some comparisons because the vaccination dose and schedule differed among the 3 regimens, with a very small sample size. The medical screening, by necessity, was intense with specific exclusion of individuals who had a history of rheumatic heart disease, cross-reactive antibodies with M protein in heart, renal, or joint tissues, any cardiac abnormality, or a personal family history of rheumatic fever, poststreptococcal glomerulonephritis, or cardiovascular diseases. Importantly, the participants could not have experienced recent GAS infections or have enzymelinked immunosorbent assay antibody present before vaccination to at least 3 of the vaccination components. Because this was a phase 1 study, the power was inadequate to detect adverse effects. None of the participants experienced an acute GAS tonsillopharyngitis infection during follow-up. It appeared that the highest-dose group may have had the lowest bactericidal response. Kotloff et al did not attempt to define an individual responder in the functional assays because they did not know how much of an increase would be necessary to confer protection. The authors did not perform repeated analysis of postimmunization outcomes; they limited their statistical comparison to 2 time points: day 0 before the first dose of vaccine vs 14 days following the last dose of vaccine. They did not compare geometric titers for each component M type because that analysis would have involved multiple comparisons and their sample size was not sufficiently