1. The varied appearance of clinical lesions can be explained by vessels at different levels of the dermis and subcutis being affected (Figs 1- 7). If capillaries in the papillary dermis are occluded, extravasation of erythrocytes will appear in the papillary dermis and the upper part of the reticular dermis. Depending on the amount of hemorrhage, lesions resulting will be purpuric macules and patches as well as hemorrhagic blisters. If postcapillary venules of the superficial plexus are occluded in number, infarction of a zone in the dermis results, which may be seen clinically as grey necrosis of the epidermis, sometimes as blister, and later as erosion or ulceration. If venules in the dermis are affected with deposits of fibrin to an extent that perfusion is impaired markedly, extravasation of erythrocytes will appear along the vessels, producing macules clinically that resemble the pattern formed by venules, i.e., a ramified shape or, eventually, a net. When hemorrhage is recent, the netlike macules will appear purpuric, and when they are longstanding, they appear brown as a consequence of hemosiderin being deposited in the dermis within macrophages. When vessels in the superficial veins of the subcutaneous septa are affected, the clinical lesion will be a nodule rather than a patch.

2. LV may be accompanied by macules and patches forming a netlike pattern, and that pattern has been called in the past livedo racemosa. It should be mentioned in passing that the pattern of livedo racemosa can be encountered in two different pathophysiological events, (1) when postcapillary venules are affected by fibrin deposits themselves, e.g., in LV, or (2) where an artery is affected by fibrin deposits, and the impaired arterial perfusion leads to hypoxemia and to livid discoloration along the net formed by postcapillary venules, e.g., in Sneddon syndrome. In contrast, no association exists between LV and livedo reticularis (except by mere coincidence), which is a pale blue macule or a patch, with a netlike shape consequent to temporary vasospasms of arteries. In livedo reticularis but not in livedo racemosa macules resolve on warming the skin. Besides LV, livedo racemosa may also be encountered in embolic and thrombotic events of various causes such as cholesterol emboli, calciphylaxis, lupus erythematosus, periarteritis nodosa, but also, at times, in an intravascular lymphoma. The presence of livedo racemosa is not requisite for the diagnosis of LV; sometimes, hemorrhagic patches, blisters, or nodules may be unaccompanied by macules in a branched pattern, depending on the localization and the number of thrombi in venules in the dermis.

3. Thrombi in LV may affect venules of the superficial and deep plexus and in the superficial subcutaneous septa as well as capillaries in the papillary dermis. Marked involvement of venules is always accompanied by stasis in capillaries, and still later, also in arterioles. The oxygenation of tissue is diminished markedly when the blood flow in venules is impaired and infarction of tissue occurs, similar to that produced by emboli in arterioles. That LV seems to be primarily a disease of postcapillary venules can be concluded from the fact that it rarely affects the distal acra, a site that usually is affected primarily by diseases involving arteries and arterioles.

4. The infiltrate in LV ranges from sparse to dense, and it is usually mixed in composition. The presence together of fibrin in the walls of vessels and/or thrombi in the lumina together with a perivascular dermatitis, even though sparse at times, is enough for LV to qualify as a vasculitis according to a strictly morphologic definition of the term. Whether the infiltrate is the reason for the deposits of fibrin and the formation of thrombi or whether fibrin and thrombi are calling forth the infiltrate of inflammatory cells cannot be determined through a microscope. The situation of LV is akin to that of leukocytoclastic vasculitis, in which the infiltrate of neutrophils accompanied by deposits of fibrin in the vessel walls is sufficient for the disease to qualify as a vasculitis morphologically. That circulating complexes of antigens and IgA-antibodies are the most common reason for the findings morphologic of leukocytoclastic vasculitis, i.e., chemotaxis of neutrophils and leukocytoclasia of neutrophils, to evolve, cannot be seen through conventional microscope.

5. The pain of LV is the pain of ischemia. Whenever necrosis of skin occurs because of a lack of oxygenation, that infarction is extraordinarily painful. Stasis in the venous system may cause hypoxemia of the skin as well as impaired arterial perfusion.

6. Even though lesions of LV are often restricted to the legs, the disease is a systemic disease. The legs only serve as a "locus minorus resistentiae" because the hydrostatic pressure acts as a triggering factor for lesions to become manifest clinically. The symmetry of lesions, almost always both legs being affected concurrently, also militates in favor of a systemic rather than a localized disease process. The matter of etiology of LV will be addressed in detail in Part III of this series of articles.

7. Contradictory results in investigations by direct immunofluorescence in regard to deposits of antibodies and complement must be interpreted in the context of findings morphologic being very consistent, namely, fibrin in the walls of vessels and/or thrombi in the lumina together with a perivascular dermatitis, even though sparse at times. If that pattern of changes morphologic is found in association with various deposits of antibodies and/or complement factors, the reason most likely is that a variety of different circumstances can produce the same morphologic changes. Once again, the situation is similar to that of leukocytoclastic vasculitis, in which medications, bacterial antigens, autoantibodies, and other factors may act as causative agent for the disease to become manifest clinically.

In sum, LV is not a specific disease, but it is a distinctive clinicopathologic pattern of changes that may occur in a variety of different circumstances. The unifying features of this pattern are the presence of fibrin in the walls of vessels and/or thrombi in the lumina together with a perivascular dermatitis, in capillaries of the papillary dermis, venules on all levels of the reticular dermis, and sometimes also including those situated in the upper part of the subcutaneous fat. Lesions clinical vary according to the position in the dermis of the vessels affected. Macules may assume a ramified pattern when perfusion is reduced markedly in venules in the dermis and hemorrhage occurs along the vessels. Lesions of LV are painful because stasis in venules is followed by hypoxemia of tissue and infarction.

Results of direct immunofluorescence as well as findings histopathologic suggest that a variety of different circumstances can produce the same morphologic changes, which are considered to be diagnostic of LV. In the next, and last, part of this series of articles we will analyze all diseases that have been said to be associated with LV and try to identify features in common among them.