[A histological and immunohistological study of vascular and inflammatory changes in rosacea].

BACKGROUND

Rosacea has a number of pathophysiological components, chief of which are vascular abnormalities and inflammation. The morphology of the dilated vessels in rosacea may indicate an increase in the number and size of lymphatic vessels. We carried out a histological and an immunohistological study to quantify these abnormalities in rosacea and compared them with those seen in lupus erythematosus.

METHODS

We reviewed all cases of rosacea analysed over a 4-year period. Ultimately, we only included 86 cases in which the diagnosis could be confirmed by a dermatologist based upon histopathological correlation and follow-up. All biopsies were reviewed for histopathological features, and 25 of these were compared with 25 facial biopsies in documented cases of lupus erythematosus, using standard staining followed by immunohistochemical analysis with anti-CD3, CD4, CD8 and CD20 (lymphocytic) antibodies, anti-CD68 (histiocytic) antibodies, anti-CD31 (endothelial cell) antibodies and anti-D2-40 (podoplanin, a marker for lymphatic endothelial cells) antibodies.

RESULTS

In 88% of cases of rosacea, large superficial dermal vessels of geometrical or bizarre configuration were noted, and turgescent cells and dermal edema were frequently seen. Over 75% of cases involved Demodex, including erythemato-telangiectatic subtypes. The rosacea included a mean 15 vessels/mm(2), eight of which expressed D2-40; six were greater than 30μm in diameter (mean: 103μm; maximum: 400μm), with only two of these being D2-40+. The lupus erythematosus biopsies exhibited a mean 15 vessels/mm(2), nine of which expressed D2-40; four measured over 30μm in diameter (mean: 59μm; maximum: 100μm), of which two were D2-40+. The vessels measuring over 100μm were only seen in rosacea, and notable actinic elastosis was associated in 80% of these cases. No Demodex was seen in the lupus cases. The lymphocytic infiltration consisted mainly of CD4+ T cells in both groups, but was chiefly sub-epidermal in lupus, occasionally masking the small vessels of the superficial dermis.

CONCLUSIONS

Rosacea is characterised by large, dilated, anfractuous capillaries, which are both larger and more numerous than in lupus, although there is no difference in dermal vascular density between the two diseases. Contrary to what their form may suggest, these dilated vessels are not lymphatic. D2-40+ vessels (lymphatic), which are flatter, are found in both lupus and rosacea. The association of large telangiectasias with actinic elastosis may indicate a causative role of exposure to UV radiation. These vessels likely exhibit increased permeability, resulting in dermal edema. Inflammation is consistently present, even in the early forms, strongly suggesting a dual inflammatory and vascular mechanism.