Portal de Periódicos da CAPES

About the Histopathology of Erythema Induratum-Nodular Vasculitis

1999; Lippincott Williams & Wilkins; Volume: 21; Issue: 3 Linguagem: Inglês

10.1097/00000372-199906000-00018

1533-0311

Evaristo Sánchez Yus, Pilar Simón,

Autoimmune Bullous Skin Diseases

To the Editor: We have read the interesting paper by Schneider and Jordaan on the histopathology of 20 cases of erythema induratum of Bazin (1); this article was a continuation of another one by Schneider et al. on the detection of Mycobacterium tuberculosis DNA in the same lesions of the same cases which was also published in The American Journal of Dermatopathology(2). The detection of mycobacterial genome using the polymerase chain reaction will probably be a fortunate milestone in the history of erythema induratum of Bazin (EI). The anterior milestone in the history of this condition was, in our opinion, an unfortunate one. We are referring to 1945, when Montgomery, O'Leary, and Barker not only separated nontuberculous from tuberculous cases, but proposed a new name, nodular vasculitis (NV), for the former (3). Seven years before, Pautrier organized a meeting in Strasbourg with the main goal of demonstrating that erythema nodosum (the most frequent tuberculid at that time) was not only seen in children suffering a tuberculous primoinfection but in adults without such a circumstance (4). Fortunately, nobody proposed, then and later, a new name for the nontuberculous cases of erythema nodosum. From 1971, EI and NV begun to be used as two synonymous terms (5,6) and to be considered as a disorder related to several etiologic factors, one of which might be tuberculosis (7,8). The Schneider and Jordaan report was on the histopathology of the cutaneous panniculitis of 20 patients with cutaneous hypersensitivity to purified protein derivative (PPD): "clearance of skin lesions on antituberculous drugs and at least three of the following histopathologic features: septolobular or lobular panniculitis, granulomatous inflammation, primary vasculitis, and significant subcutaneous necrosis (1)." Several questions arise concerning this investigation. The first one is: what about a control group? In other words, which were the histopathologic features of panniculitis of patients who either had less hypersensitivity to PPD or did not clear on antituberculous drugs? Baselga et al. have recently failed to find significant histopathologic differences (except for the presence and degree of histologic necrosis; p = 0.04) between polymerase chain reaction positive and negative cases of EI-NV (9). According to these authors, none of the clinical (including PPD hypersensitivity and response to antituberculous therapy) and histologic variables evaluated could accurately predict the results of polymerase chain reaction amplification (9). From the point of view of histopathologic features, we think that two items of the paper by Schneider and Jordaan (1) must be discussed: those concerning the pattern of panniculitis and concerning the type of vessel involved. In the article, Schneider and Jordaan repeatedly speak of "septolobular" panniculitis: "All biopsies showed septolobular panniculitis (1)." The classification of panniculitis into septal and lobular was first done, to the best of our knowledge, by Reed and Clark in 1971: "The Panniculitides may be grouped into two broad categories depending on whether the fibrous septa or the fat lobules are the principal site of damage." Later, they stated that in septal panniculitis "the septa are thickened, edematous, and show perivascular infiltrate of chronic inflammatory cells. Granulomas, if present, are in the thickened septa and in the fat lobules bordering septa. In lobular panniculitis the inflammatory infiltrate involves the major portion of one or several lobules (10) (italics are ours)." Ackerman also classified panniculitides into septal and lobular, "depending on where most of the inflammatory cells are found" and noticed that these patterns must be "perceived with the scanning objective (6) (italics ours)." In brief, every panniculitis is a mixed (septolobular) panniculitis; that is, both septa and lobules are simultaneously involved by the inflammatory phenomena. Pure septal panniculitis and pure lobular panniculitis are nonexistent patterns. However, in most cases, the inflammatory phenomena mainly involve either the septum and the periseptal area of the lobules or affect all of either one or several lobules and the surrounding septa (11). In the first case, it is said that the panniculitis is mainly, or mostly, septal (septal panniculitis), and in the second one, the panniculitis is mainly, or mostly, lobular (lobular panniculitis). The problem is when the biopsy is small. In such a case, it is often impossible to make a decision about the pattern of the panniculitis because, "it must be perceived with the scanning objective (6)." The paper by Schneider and Jordaan (1) lacks scanning photomicrographs, so we cannot know what a "septolobular" pattern of panniculitis is. The only medium magnification photomicrographs (Fig. 1 and 2) show in each one a very thin septum (at right-bottom in Fig. 1, at left in Fig. 2) (1). If these figures are representative of the septal involvement of their cases, they are examples of almost pure lobular panniculitis. In the previous paper of Schneider et al. regarding the same cases (2), Fig. 3 is a scanning magnification with a typical lobular panniculitis, although its legend says, "septolobular panniculitis (2)." We conclude that what the authors name "septolobular" panniculitis is what is usually named, from 1971, lobular panniculitis.FIG. 1: This schema is a replica of that of Fig. 8-2 in Ackerman's book (6). Here, the "traditional" thin-walled vein, at the right, has been substituted by the thick-walled vein, with a plurifasicular muscle layer, that is usually found in the subcutaneous fat of the areas of the body suffering a higher hydrostatic pressure. A, artery; al, arteriole; vl, venule; V, vein.Concerning the affected vessels in the "primary vasculitis," Schneider and Jordaan say that, "Group I displayed primary neutrophilic vasculitis involving one muscular artery in five cases (Fig. 3) (1)." Such a figure is centered by a muscular vessel whose muscular layer is formed by fascicles of smooth muscle fibers separated by elastic fibers. In 1987, Sánchez Yus et al. demonstrated that this is the typical structure of many subcutaneous veins (12); conversely, the muscular layer of the subcutaneous arteries is always formed by one thick compact fascicle separated from the intima by a thick undulated inner elastic lamina. The venous nature of such a thick-walled vessel could be demonstrated because, in step sections, valves were found in the confluence of smaller veins with the thicker ones (12)(Fig. 1). Other authors (6,13) have also erroneously interpreted such veins as arteries with "fragmented external elastic membranes (6)." In Fig. 5 of the Schneider and Jordaan paper, the vessel on the left is also a vein, although the legend only mentions "...a smaller caliber blood vessel... (1)." Once established that the medium-sized vessels involved in EI-NV are mainly veins, the question remains whether such involvement is either the cause or the consequence of the panniculitis. There is a general agreement that thrombosis and inflammation of a large subcutaneous vein (as in thrombophlebitis migrans) is not enough to activate the sequence of events that may lead to a lobular panniculitis like that seen in EI-NV, but panniculitis surrounding the involved vessel is usually minimal and thrombophlebitis migrans is classified among septal panniculitides (14). Our theory, both in 1987 (12) and now (14), is that simultaneous damage to veins and venules and, perhaps also to arterioles, are a primary phenomenon in the pathogenesis of EI-NV. This interpretation is based on the histopathologic study of small, early nodules, more palpable than visible. In such nodules, an inflamed and thrombosed vein was surrounded by an area of panniculitis with simultaneous thrombosis of its venules (12). Awareness of the existence in several areas of the subcutaneous fat, mainly in those with a higher hydrostatic pressure, of a thick-walled vein (Fig. 1), with a middle layer formed by muscular fascicles separated by elastic fibers, may be an aid for a better knowledge of some clinical, histologic, and pathogenetic problems of panniculitis with large vessel vasculitis. Evaristo Sánchez Yus, M.D. Department of Dermatology; Hospital Clínico San Carlos; Facultad de Medicina; Universidad Complutense; Madrid, Spain Pilar Simón, M.D., Ph.D. Department of Cell Biology; Facultad de Medicina; Universidad Complutense; Madrid, Spain