Immunopathology of schistosomiasis mansoni in mice and men.

Padraic Fallon recently presented an excellent summary of cytokine responses and immunopathology in infected mice and humans1. We wish to take issue with several points, including Fallon’s implication that morbidity and death in immunodeficient mice caused by type 1 cytokine responses is relevant to human schistosomiasis. In this model, hepatocellular toxicity1,2 is apparently caused by Schistosoma mansoni egg antigens that are neutralized in immunologically intact mice2,3. Fallon et al. recently presented a detailed description of this response2. In considering the varied causes of morbidity and mortality in humans and mice (Table 1) this toxicity, termed ‘cytokine shock’ in the table, has no apparent counterpart in infected humans. Symmers’ pipestem periportal fibrosis is the characteristic hepatic lesion of severe schistosomiasis mansoni4,5. As noted by Fallon, Henderson et al. described liver fibrosis in mice that was strikingly similar morphologically to the human disease6. Modulatory anti-egg idiotypes were absent from both mice and humans with Symmers’ fibrosis but were present in those without Symmers’ fibrosis7,8. Limited data from mice and humans suggest that T helper 1 (Th1) pathways are involved in the pathogenesis of Symmers’ fibrosis9,10 although another study found that mice producing idiotypes stimulating interferon g (IFN-g) were those that did not develop Symmers’ fibrosis7. In our studies in patients with Symmers’ fibrosis we found IFN-g production only when interleukin 10 (IL-10) was neutralized11. These findings suggest, in contrast to the interpretation of Fallon, that type 1 responses are not necessarily linked to hepatosplenic disease11. Vaccination of mice before infection with S. mansoni eggs and IL-12 results in a predominantly Th1 response, compared with the Th2 response in unvaccinated mice. Fibrosis is markedly reduced, granuloma size is decreased, and ‘cytokine shock’ is not seen12. Present information is inadequate to indicate whether such immune deviation towards a Th1-type reaction might be helpful or inadvisable in humans. In mice, our findings indicate that both extreme Th2 and Th1 immune polarization is potentially harmful and that Th2 polarization augments fibrosis13. The Th1 reaction associated with cytokine shock in mice, in our opinion, is not relevant. Conceivably, cytokine shock might develop in patients with immunodepression such as that associated with AIDS, but the very low levels of infection in humans compared with those in mice would suggest that egg-mediated cytokine shock should seldom, if ever, occur even in these patients. S. japonicum and S. haematobium infections differ markedly from S. mansoni infections in both humans and mice. Notably, the egg-mediated hepatotoxicity does not occur in S. haematobiumor S. japonicuminfected T-cell depleted mice14,15 or in S. japonicum-infected nude or severe combined immunodeficiency (SCID) mice16. Thus, this unusual pathology might be restricted to S. mansoni infection in immunodeficient mice. In summary, our areas of agreement with Fallon are much greater than the areas of disagreement. We particularly consider that, interesting as it is in its own right, hepatotoxicity in immunodeficient mice lacks relevance to human schistosomiasis. Better definition of the pathogenesis of both acute toxemic disease and chronic schistosomiasis in humans remains a high priority.