In contrast, we feel that this conclusion is suspect. Human leukocyte antigen association studies4
suggest that susceptibility or resistance to the development of sarcoidosis may relate to carriage of particular class II alleles, and therefore the presence of CWDFs in an equal proportion of healthy control subjects does not necessarily exclude an etiologic role in sarcoidosis. Moreover, the timing of blood culture may have had a critical bearing on the number of positive yields. Sarcoidosis is characterized by granulomatous inflammation whose function is a presumed “walling-off” of offending antigens. It is possible, therefore, that bacteremia only occurs very early in the disease course before granulomatous structures are fully developed. Mycobacteremia may also not occur in all patients and, perhaps, is confined to only those with multisystem involvement. An important aspect of interpretation of the data from this study should therefore include a correlation of the number of positive blood culture findings with time from disease onset, not necessarily diagnosis, and the presence of extrapulmonary manifestations. If no such correlation exists, we would concur that an etiologic role for these organisms could not be supported and that tissue-based approaches are likely to the most productive way forward.