THE LANCET been removed. This approach might not have been so popular if the early work on maintained epithelial linings in lethal asthma had been consulted. Caution in interpretation of modern biopsy data on denudation would also have been appropriate in view of the evident risks of denudation artifacts. 5 In-vivo physiological studies showing unimpeded airway absorption barriers are evidence against the presence of wide areas of denudation. Indeed, in the 1930s9 an abnormally low rate of airway absorption was demonstrated in allergic rhinitis. Although this finding has now been corroborated in both asthma10 and rhinitis, 11 the popular notion of the opposite—an increased mucosal absorption in these diseases—may prevail as a potentially false9–12 research standard. Mucosal absorption and mucosal exudation (of plasma) are largely independent processes. 12 Hence, nothing can be concluded about absorption perviousness merely because asthma is an exudative disease. 12 In-vivo airway findings that reconcile the features of epithelial shedding and maintained mucosal tightness are now emerging. The repair may thus be a very quick process, even after denudation has occurred: all bordering epithelial cells will promptly dedifferentiate and migrate, at high speeds, to cover the basement membrane. 5 This invivo repair occurs in a “dynamic gel” of plasma-derived fibrinous proteins and activated leucocytes. 5 It is possible that the abundance of shed epithelial cells, leucocytes, and plasma proteins seen in sputum from people with asthma since Berkart’s time3 reflects many epithelial injury-repair sites and their coverings of exudate gels. However, studies of such mucosal features in human airways and in animal models of asthma are still incomplete. 80 years ago, Marchand8 made detailed observations of the mucosal changes in asthma. Instead of regarding the epithelium as a cell type that was lacking in asthma, he pioneered the tantalising thought that the epithelium may release active factors controlling exudation and leucocyte activities of the mucosa. This speculation pre-empts

In the present era of very advanced molecular and cellular biomedical research1 we still cannot adequately define several processes of the asthmatic mucosa, whether it is the activation and actions of eosinophils, the epithelial injuryrepair processes and their sequelae, vascular exudative aspects, or the antiasthma actions of steroids; we still lack essential knowledge about important in-vivo events. 2 The deficit of satisfactory descriptions of mucosal inflammatory processes in asthma is remarkable because the occurrence of these processes in this disease was indicated long ago. It is not only during the past decade or so that asthma has been regarded as an inflammatory airway disease, although published work may occasionally give that impression. Indeed several aspects that are central to our notion of asthmatic inflammation date back to astute observations of the 19th century.