To the Editor.
—In the April 1991 issue of the Archives, Sachs1 discussed hypersensitivity reactions to streptococcal antigens, which probably play a key role in producing cutaneous cellulitis. As Sachs postulated, the production of cytokines by intraepidermal immunocompetent cells could contribute to the inflammatory symptoms and to the quick clearance of pathogens from the skin. The author concluded that administration of nonsteroidal anti-inflammatory drugs (NSAIDs) could be beneficial by reducing the inflammation. The concept may be correct, but it is far too hypothetical to be generalizable in current dermatologic practice. On the one hand, the effect of therapy with NSAIDs could show a positive reduction of inflammatory symptoms; on the other hand, it could promote bacterial multiplication. In vitro studies supported the hypothesis that NSAIDs induced an impaired host defense against infection because of an alteration of chemotaxis, phagocytosis, and bacterial activity of granulocytes.2 In a retrospective study