The limited success rate of treatment of AGA with hair growth promoters or modulators of androgen metabolism means that further pathogenic pathways may be taken into account. The implication of microscopic follicular inflammation in the pathogenesis of AGA has recently emerged from several independent studies (Jaworsky et al., 1992; Mahe´ et al., 2000; Whiting, 1993). An early study referred to an inflammatory infiltrate of activated T cells and macrophages in the upper third of the hair follicles, associated with an enlargement of the follicular dermal-sheath composed of collagen bundles (perifollicular fibrosis), in regions of actively progressing alopecia (Jaworsky et al., 1992). Horizontal section studies of scalp biopsies indicated that the perifollicular fibrosis is generally mild, consisting of loose, concentric layers of collagen that must be distinguished from cicatricial alopecia (Whiting, 1993). The term ‘microinflammation’ has been proposed, because the process involves a slow, subtle, and indolent course, in contrast to the inflammatory and destructive process in the classical inflammatory scarring alopecias (Mahe´ et al., 2000). The significance of these findings has remained controversial. However, morphometric studies in patients with male pattern AGA treated with minoxidil showed that 55% of those with microinflammation had regrowth in response to treatment, in comparison to 77% in those patients without inflammation and fibrosis (Whiting, 1993).Inflammatory phenomena: An important question is how the inflammatory reaction pattern is generated around the individual hair follicle. Inflammation is regarded as a multistep process which may start from a primary event. The observation of a perifollicular infiltrate in the upper follicle near the infundibulum suggests that the primary causal event for the triggering of inflammation might occur near the infundibulum (Mahe´ et al., 2000). On the basis of this localization and the microbial colonization of the follicular infundibulum with Propionibacterium sp., Staphylococcus sp., Malassezia sp., or other members of the transient flora, one could speculate that microbial toxins or antigens could be involved in the generation of the inflammatory response. The production of porphyrins by Propionibacterium sp. in the pilosebaceous duct has also been considered to be a possible cofactor of this initial pro-inflammatory stress (Mahe´ et al., 2000). Alternatively, keratinocytes themselves may respond to chemical stress from irritants, pollutants, and UV irradiation, by producing radical oxygen species and nitric oxide, and by releasing intracellularly stored IL-1a. This pro-inflammatory cytokine by itself has been shown to inhibit the growth of isolated hair follicles in culture (Philpott et al., 1996). Moreover, adjacent keratinocytes, which express receptors for IL-1, start to engage the transcription of IL-1 responsive genes: mRNA coding for IL-1b, TNFa, and IL-1a, and for specific chemokine genes, such as IL-8, and monocyte chemoattractant protein-1 (MCP-1) and MCP-3, themselves mediators for the recruitment of neutrophils and macrophages, have been shown to be upregulated in the epithelial compartment of the human hair follicle (Mahe´ et al., 2000). Besides, adjacent fibroblasts are also fully equipped to respond to such a pro-inflammatory signal. The upregulation of adhesion molecules for blood-borne cells in the capillary endothelia, together with the chemokine gradient, drive the transendothelial migration of inflammatory cells, which include neutrophils through the action of IL-8, T cells and Langerhans cells at least in part through the action of MCP-1. After processing of localized antigen, Langerhans cells, or alternatively keratinocytes, which may also have antigen presenting capabilities, could then present antigen to newly infiltrating T lymphocytes and induce T cell proliferation. The antigens are selectively destroyed by infiltrating macrophages, or natural killer cells.