UC Berkeley

Despite the near disappearance of rheumatic heart disease (RHD) in wealthier nations of

the world, this disease continues to cause substantial morbidity and mortality in poor

countries worldwide. The burden is projected to be particularly important among

residents of urban slums in poor countries. However, the epidemiologic features of RHD

in developing countries are poorly understood. RHD is characterized by damage to

cardiac valves that is the long-term consequence of an immune process initiated by

infection with Streptococcus pyogenes (Group A Streptococcus; GAS). Progression to

RHD often takes decades; it would require decades-long prospective studies to observe

associations between GAS infections and the outcome of RHD in the same subjects.

Therefore, in this dissertation we first attempted to assess the current burden of RHD in

Salvador (Chapter 2), a city where more than half of the population is living in slums, by

conducting a population-based study of operations performed for cardiac valve disease.

We found that a large proportion of valvular surgeries performed in Salvador from 2002-

2005 was for RHD. We then wished to investigate possible reasons for the large burden

of this disease. It is known that RHD is influenced by biological factors of beta-hemolytic

Streptococcus as well as health-care seeking behavior and treatment of streptococcal

infections. Therefore, in the third and fourth chapters of this dissertation we focused on

the biological factors of Streptococcus, focusing our work to address current hypotheses

regarding RHD pathogenesis as found in the literature, such as: 1) the association of

infection or colonization with certain streptococcal strain types and clinical outcome; 2)

the increased risk of RHD with repeated infections with a diverse set of GAS strains, and

3) the possible association of RHD with Streptococcus spp. other than GAS.

In Chapter 3, we compared the genotypes of GAS strains recovered from children aged 3-

15 years of age who live in slum versus non-slum communities. This was done to

determine if there are differences in the strain genotype distributions (as measured by

genotype diversity) in these two populations. Detection of differences in genotype

distributions by community (estimating slum communities to have higher diversity of

genotypes circulating in the population compared to non-slum populations), would

provide preliminary data to support the hypothesis that high GAS genotype diversity in

slums may be associated with the observation of high prevalence of RHD in slum

populations compared to the low prevalence of RHD among non-slum populations.

Furthermore, we investigated two additional species of beta-hemolytic streptococci from

slum and non-slum communities as well, and found an unexpected finding that

colonization with Streptococcus dysgalactiae equisimilis was associated with lower odds

of sore throat in children (Chapter 4). We discuss possible explanations for this finding,

including biological plausibility as well as alternate explanations. While our

observational studies can not define causal associations between epidemiologic features

of beta-hemolytic Streptococcus and the outcome of RHD, they provide preliminary data

that the epidemiologic features of GAS and non-GAS infections in urban slums of

Salvador may be distinct from that in non-slum populations.