Urticaria and angioedema affect at least 20% of the population at some time1 and can be classified according to their clinical characteristics, triggers, or associations with various diseases.2–4

The characteristic lesion of urticaria is the wheal, an edematous pink to red plaque with a pale center. Wheals vary in size, are evanescent, and can coalesce into larger round or archiform lesions.2

The classic histopathologic descriptions of urticaria note the presence of dermal edema, vasodilation, and a perivascular infiltrate of mononuclear cells. The infiltrate consists mainly of lymphocytes, together with neutrophils and eosinophils in varying numbers, usually arranged between the dermal collagen bundles.3,5 To date, however, histopathologic descriptions have varied considerably, owing to the small number of cases that have included histopathology and to the difficulty of establishing the duration of the biopsied lesions; other confounding factors may be involved as well.6

The presence of neutrophils in biopsy specimens of urticarial lesions was first described in Dermatopathology, a 1967 textbook by Montgomery and coauthors, and later by Jones et al7 in a 1983 article in the British Journal of Dermatology (cited in Winkelmann and Reizner8). As early as 1985 Peters and Winkelmann9 posited the existence of a distinct diagnostic entity they called neutrophilic urticaria, characterized by the appearance of wheals in which histology showed neutrophils in the dermal venules and venular walls and no sign of leukocytoclasia, hemorrhage, or fibrinoid necrosis. In Winkelmann and Reizner's8 subsequent retrospective study, they found a diffuse dermal neutrophilic infiltrate in 23 (8.7%) cases and the presence of neutrophils lining or invading the venular walls in another 19 (7.1%) cases. Thus, neutrophilic infiltrates were found in as many as 15.8% of the 256 cases of chronic urticaria. As a result, the histopathologic concept of neutrophilic urticaria was broadened.

In later histopathologic studies of urticaria, higher numbers of neutrophils were found in the dermis of patients with acute urticaria compared to other types and in the reticular dermis of those with pressure urticaria.6 Furthermore, there seems to be a direct relationship between neutrophil and eosinophil counts.6

When Winkelmann et al.5 initially studied the clinical characteristics and comorbidities of 8 patients, no conclusive data emerged with respect to abnormalities in complete blood counts or proteins; nor were there patterns with respect to indicators of autoimmune disease. More recent articles have suggested that urticaria with a neutrophilic infiltrate, or neutrophilic urticaria, might be characterized by shorter flare-ups and by an increase in acute phase reactants. Interleukin-3 and tumor necrosis factor-α seem to be implicated in the pathogenesis of this condition,10 although few studies have focused on how its characteristics differ from those of urticaria with an infiltrate that is not neutrophilic.

More recently, in 2009, Kieffer et al.11 defined a subtype of neutrophilic urticaria characterized by nonedematous lesions lasting up to 48hours. The distribution of the infiltrate and the scant evidence of edema, however, led the authors to consider this subtype to be more similar to acute neutrophilic dermatosis. The authors identified 9 patients with this type of urticaria, which they termed neutrophilic urticarial dermatitis. These patients had interstitial infiltrates of neutrophils distributed either individually or arranged linearly; no vascular damage was observed. The authors concluded that this subtype may be particularly associated with rheumatic disease.

The histopathologic differential diagnosis of neutrophilic urticaria requires mainly consideration of urticarial vasculitis. In theory, they can be distinguished because of the presence of vascular damage in urticarial vasculitis as evidenced by areas of postcapillary fibrinoid necrosis (in venules).1 Nonetheless, rigorous diagnostic criteria have not been applied in most studies of urticarial vasculitis, and in a published series of patients with this disease, fibrin deposition was detected in samples from only 9.9% of patients,12 many of whom presented a neutrophilic infiltrate with only slight or no vascular abnormalities.1 Additionally, in some published cases of hypocomplementemic urticarial vasculitis, clinical features and histologic findings in the urticarial lesions completely overlapped with those of neutrophilic urticaria.13 In a prospective study of lesions clinically suggestive of urticarial vasculitis, it was observed that there was in fact no vasculitis in the majority of cases and that a finding of a predominance of polymorphonuclear cells on histology was associated with the presence of fibrin, leukocytoclasia, low complement levels, and anemia.12 The problem of the overlap between the 2 diagnoses was pointed out by Winkelmann et al.5 in their initial description of neutrophilic urticaria, in which they stated that its peculiar histopathology might be suggestive of an underlying vasculitis. This hypothesis was not subsequently explored, however. To complicate matters further, the true incidence of leukocytoclastic vasculitis is unknown; nor do we understand its prognostic significance in other types of urticaria, such as solar14,15 or cold urticaria.16–18

In short, neutrophilic urticaria is a clinical entity that partially overlaps with urticarial vasculitis. Much about the condition remains poorly understood, including its true incidence and its association with a higher incidence of systemic conditions, particularly of rheumatic disease.

The primary objective of our study was to compare the prevalence of rheumatic disease in 2 groups of patients: those with neutrophilic urticaria (defined as urticaria with a neutrophilic infiltrate) and those with urticaria with a lymphocytic infiltrate that is not predominantly neutrophilic.

Our secondary objectives were to compare the epidemiologic and histopathologic data for neutrophilic urticaria to all other urticarias that do not fall into that category, as well as to findings previously described in the literature.

We carried out a descriptive, retrospective study using the database of our hospital's pathology department for the period between January 1, 1999 and June 28, 2009 to locate the cases of urticaria for which the results of skin lesion biopsies were available.

Our inclusion and exclusion criteria are shown in Table 1.

We studied 84 patients for whom we had complete access to both clinical and histopathologic data and in whom we could rule out the possibility of other diseases.

Biopsy specimens were fixed in 10% buffered formaldehyde for at least 24hours and were routinely processed into 5-μm sections and stained with hematoxylin-eosin. Each specimen was examined at different magnifications (up to ×400) and evaluated independently by 2 specialists, a pathologist and a dermatologist (J.F.F. and M.L.V.). Any disagreements were settled with the help of another dermatologist (J.S.P.). The histopathologic aspects analyzed in each case are shown in Table 2.

The biopsy that was most representative of the findings for each was classified into one of the following groups:

• 1

  Conventional urticaria with an infiltrate that was not neutrophilic (control group)

• 2

  Urticaria with a neutrophilic inflammatory infiltrate. This group consists of several subtypes: a) diffuse dermal neutrophilia, b) neutrophilic venulitis, c) mixed (diffuse dermal neutrophilia plus neutrophilic venulitis), and d) urticaria in which the neutrophils were mainly intravascular.

The definitions used for each of these groups and subgroups are specified in Table 3 and are illustrated with prototypical examples (Figs. 1–4).

Figure 1.

Conventional urticaria with an infiltrate that was not neutrophilic: normal epidermis, dermis with edema, mild to moderate perivascular lymphocytic infiltrates, vasodilation (hematoxylin-eosin, original magnification ×100).

(0.27MB).

Figure 2.

Interstitial dermal neutrophilia: normal epidermis, severe edema in the dermis, interstitial infiltrate, dilated blood vessels (hematoxylin-eosin, original magnification ×100).

(0.31MB).

Figure 3.

Neutrophilic venulitis. A, normal epidermis and dermis, inflammatory infiltrate in the vessel wall (hematoxylin-eosin, original magnification ×40). B, presence of polymorphonuclear neutrophils and eosinophils in a vessel wall (hematoxylin-eosin, original magnification ×200).

(0.36MB).

Figure 4.

Predominance of polymorphonuclear cells inside the vessel. Mild to moderate vasodilation with lumen completely occupied by neutrophils (hematoxylin-eosin, original magnification ×100).

(0.29MB).

The following information was collected from patients’ medical histories: age, sex, personal history, smoking history, long-term medication, date of first visit, time since onset of disease, frequency and site of flare-ups, (type of lesion, duration, presence or absence of residual lesion after resolution, and symptoms), associated systemic symptoms, hemoglobin values, total white blood cell and differential counts, blood sugar, transaminase levels, erythrocyte sedimentation rate, presence of antibodies, complement abnormalities, parasites in stool, and serum levels of IgE. Examples of our patients’ lesions are shown in Fig. 5.

Figure 5.

A, multiple partially confluent wheals with a circinate configuration on the back of a woman with acute urticaria. B, erythematous, edematous plaques that are well-defined, oval in shape, somewhat infiltrated, with no epidermal involvement.

(0.21MB).

Statistical analysis of the data was carried out using the SPSS statistical package, version 18. Statistical significance testing was performed using G-Stat software, version 2.0.

The primary objective of our study was to determine whether or not there was a greater prevalence of rheumatic disease among patients with neutrophilic urticaria. However, we failed to find any such association with neutrophilic urticaria in general or with any of its histologic subtypes. According to the literature, the estimated difference in the prevalence of rheumatic disease between patients with neutrophilic urticaria or another type is at least as high as 30% to 50%. The power of our study, with a total of 84 patients, to detect differences not due to chance was approximately 80%.

In our series 14.6% of patients with neutrophilic urticaria and 22.2% of those with conventional urticaria (without such an infiltrate) had rheumatic disease. In contrast, Kieffer et al.11 found rheumatic disease in up to 77.7% of their patients. In their article they defined neutrophilic urticarial dermatosis as marked by the presence of erythematous nonpruritic lesions lasting less than 48hours and histologically characterized by considerable interstitial and perivascular involvement, with neutrophils distributed between the collagen bundles either individually or in lines. Curiously, although the presence of edema was one of their exclusion criteria, edema could be seen in their microphotographs. After a critical reading of their article, we have come to the conclusion that the entity that we have included in the category of diffuse dermal neutrophilia overlaps, at least histopathologically, with the one they called neutrophilic urticarial dermatosis. Moreover, in 59% of our patients with diffuse dermal neutrophilia, the clinical lesions were not evanescent. Our results, therefore, are not consistent with those of Kieffer et al. and lend support to the earlier hypothesis that there is no association between neutrophilic urticaria and other diseases.5 However, our study is not directly comparable to theirs, as our criteria did not exclude patients with histopathologically edematous lesions or require the presence of recurrent flare-ups.

Apart from mentioning the overlap between neutrophilic urticaria and urticarial vasculitis that we have referred to above, current reviews do not consider neutrophilic urticaria to be a well defined subtype of urticaria.2 We thus had to define the criteria and histopathologic patterns needed to classify the biopsies of the urticarial lesions in our study. The inflammatory infiltrates found in biopsies of urticarial lesions are indications of an underlying dynamic process. In a study of 22 patients with chronic idiopathic urticaria from whom 2 biopsies were obtained (one of a lesion that appeared less than 4hours earlier and another of one that appeared more than 12hours earlier), Sabroe et al.19 demonstrated this dynamic process, observing a predominance of polymorphonuclear cells in blood and in vessel walls in the more recent lesions and a tendency towards interstitial distribution in older lesions. In our study we identified a group of patients with neutrophilic infiltrates both in the dermis and in some vessel walls; these results support the theory of the dynamic nature of the inflammatory infiltrate of urticarial wheals.8

In our series 57.1% of patients had a neutrophilic pattern of urticaria. Of these, 17.9% were cases of diffuse dermal neutrophilia. In the series of Winkelmann and Reizner,8 only 15.8% of cases had neutrophilic urticaria and 8.7% of these showed a dermal inifiltrate with a diffuse or interstitial distribution. In 26% of their patients they observed a perivascular concentration of neutrophils,8 as we did in the majority of ours. Toppe et al10 found a predominantly neutrophilic pattern in more cases of clinically acute urticaria (55%) than in chronic urticaria (18%). Neutrophils were found inside vessels or in vessel walls in 29.1% of their biopsies. As mentioned above, the study of Sabroe et al.19 suggested that the distribution depended on the age of the lesion.19 While in our study the exact age of the wheal before biopsy was unknown, our hospital tends to select recent lesions for biopsy and this could account for the higher percentage of neutrophilic urticaria in our series. Furthermore, although our inclusion criteria did not take clinical data into account, the urticarial lesions biopsied in our hospital were those with clinical findings suggestive of neutrophilic vasculitis and, to a lesser extent, clinically atypical cases that raised doubts as to diagnosis or that responded poorly to treatment. Our finding that a neutrophilic pattern was more frequent in patients with chronic urticaria is consistent with the increase in dermal neutrophils found in acute urticaria compared to other types described by Toppe et al.6

Winkelmann et al.5 considered that the rare presence of leukocytoclasia, the absence of fibrinoid necrosis, and the scant presence or total absence of red blood cell extravasation in the histologic sections they reviewed could serve to differentiate cases of neutrophilic urticaria from those of leukocytoclastic vasculitis or urticarial vasculitis. In the light of our histologic findings, however, this possibility seems somewhat doubtful, as we found varying degrees of vascular damage. Dense perivascular infiltrates with leukocytoclasia and, more rarely, small points of fibrinoid necrosis can be found in the biopsies of up to 52% of patients with urticaria.20 Clear vascular damage, with destruction of the vessel walls but without fibrinoid necrosis, was found in only 6.3% of our patients with neutrophilic urticaria. With less rigid criteria for vascular damage, up to 60.4% of those cases could have been considered urticarial vasculitis. In contrast, no vascular abnormalities were found when the infiltrate was not neutrophilic. Our study suggests that the disparity in the frequency of vascular damage found in the various series described in the literature might be due to the inclusion of differing percentages of patients with neutrophilic urticaria.

Blood tests showed only 1 significant difference between the 2 groups: white blood cell counts in peripheral blood were higher in patients with neutrophilic urticaria, a finding described previously.

The main limitation of our study, resulting from its retrospective design, was the lack of homogeneity in the clinical and laboratory data and in the duration of follow-up. Consequently, we were unable to analyze some of the data. Moreover, in the absence of reliable information concerning the age of the wheals, we were unable to study possible differences in the composition of the inflammatory infiltrates and their development over time. The lack of this information and of generally accepted criteria for neutrophilic urticaria may explain why we could not rule out the possibility that in some of the cases included in our study the differences found between the 2 groups might be attributable to having biopsied lesions at different points in their development, making it more difficult to interpret the data.

The frequency of a neutrophilic inflammatory infiltrate in biopsy specimens of urticarial lesions is high (57.1%), especially in patients with acute urticaria. None of the different histologic patterns found in neutrophilic urticaria seems to be associated with a greater prevalence of rheumatic disease compared to urticaria with infiltrates that are not neutrophilic.

Neutrophilic urticaria could be a reaction pattern that is intermediate between urticaria with an infiltrate that is not neutrophilic and urticarial vasculitis with fibrinoid necrosis. To determine whether neutrophilic urticaria and urticarial vasculitis are 2 different entities, however, would require a prospective study, given the frequent presence of vascular damage in patients with neutrophilic urticaria and its overlap with urticarial vasculitis.

The authors declare that they have no conflicts of interests.