Diffuse HIV-associated seborrheic dermatitis – a case series

Introduction

In patients infected with human immunodeficiency virus (HIV), seborrheic dermatitis (SD) is common (prevalence 20–83%), and has been reported to have distinct clinical and histopathologic features compared to HIV-negative patients.^1–3 Here, we present 20 cases of HIV-associated SD to further delineate its unique clinical and histologic characteristics. Recognition of this variant may aid in early detection and treatment of HIV.

Cases

The 20 cases were seen by the authors in dermatology clinics in the United States, Botswana, and South Africa. All patients were African or African American and HIV-positive, with CD4 cell counts ranging from 25 to 1244 (mean 345) cells/μl; only seven had CD4 cell counts lower than 200 cells/μl. Eleven patients were on antiretroviral therapy (ART). The patients ranged in age from 25 to 62 years (mean 37). Sixteen patients were women (see Table 1).

OPEN IN VIEWER

Table 1.

Demographics.

Case	Gender	Age	CD4	On ART	Biopsy performed
1	M	34	340	Yes	Yes
2	F	62	190	No	No
3	F	Unknown	31	Yes	Yes
4	F	27	310	Yes	No
5	F	27	1244	No	Yes
6	F	46	320	Yes	No
7	M	40	650	Yes	Yes
8	F	27	300	Yes	No
9	F	30	250	Unknown	No
10	M	Unknown	200	No	No
11	F	25	25	Yes	Yes
12	M	46	649	No	Yes
13	M	40	290	Yes	N
14	M	45	401	No	No
15	F	35	280	Yes	No
16	M	46	461	No	No
17	F	29	115	Yes	No
18	M	31	420	No	No
19	M	40	240	Yes	No
20	M	31	190	No	No

Physical findings included exuberant erythema and scale originating within a typical seborrheic distribution on the face and scalp, but also extending to other areas. There was often extensive involvement of the scalp, axillae, groin, and extremities, particularly flexural surfaces. There was marked yellow scaling of the scalp, which was greasier and thicker than typical SD. Multiple patients had bacterial superinfection. Notably, five patients were erythrodermic; three of these had biopsies consistent with SD (see Figure 1). OPEN IN VIEWER

Biopsies were performed in six cases. Four were read by local pathologists as consistent with SD. Slides from two cases were available for review by a US dermatopathologist, which demonstrated epidermal acanthosis, widespread parakeratosis, spongiosis, necrotic keratinocytes, and a mixed infiltrate of plasma cells, lymphocytes, neutrophils, and eosinophils. These features have been described in association with HIV-related SD. ⁴

Importantly, our patients were refractory to standard SD treatments and required prolonged courses of topical and systemic antifungals plus topical or oral corticosteroids. Treatment failure and frequent recurrence were common. Several cases showed improvement following initiation of anti-staphylococcal antibiotics.

Discussion

The distinct appearance and recalcitrance of SD in HIV has been noted previously, with some authors suggesting it should be classified as a distinct entity altogether.^4–7

Clinically, our cases were notable for extensive involvement beyond typical seborrheic areas, including the extremities, groin, and axillae, consistent with other reports.^3,4 The markedly thick, yellow scale on the scalp also appears to be a distinguishing feature, as is the propensity to cause erythroderma, which has not been well described in the literature.

Histologically, lesions of SD in HIV-negative patients demonstrate spongiosis and a lymphohistiocytic infiltrate, as well as follicular plugs with parakeratotic cells and uneven rete ridges. ⁷ Our two cases reviewed by a US dermatopathologist showed necrotic keratinocytes, more widespread parakeratosis, and a superficial perivascular infiltrate of plasma cells. Soeprono et al. ⁴ described these differences in detail. Of note, however, our cases did not demonstrate focal dermal-epidermal obliteration as seen by Soeprono et al.

The pathophysiology of this SD variant is unknown, but one hypothesis involves immune dysregulation leading to alterations in quantity and speciation of fungal growth on the skin. However, there is disagreement in the literature as to whether HIV-associated SD has higher ⁸ or lower ⁵ levels of Pityrosporum yeast compared to SD in HIV-negative patients. The skin of patients with HIV may also have an altered dermal lipid profile^9,10 and produces heat shock proteins not seen in HIV-negative patients with SD or psoriasis. ⁶

The relationship of SD severity to CD4 cell count appears uncertain both in the literature and in our patients. SD can be an early finding in HIV patients with high CD4 cell counts,^11–13 but it is also reported to increase in severity with worsening immunosuppression. ¹ In our cases, it is notable that only seven patients had a CD4 cell count lower than 200 cells/μl, suggesting that the degree of immunosuppression need not be profound. In general, however, those with very low CD4 cell counts had particularly severe disease. The three patients with the lowest CD4 cell counts (25, 31, 115 cells/μl) were all erythrodermic. It is not clear whether ART improves the dermatitis or makes it less refractory to treatment.^13–15 As is typical for HIV-related SD, many of our patients received prolonged treatment with topical and systemic antifungals plus topical or oral corticosteroids as well as oral antibiotics, when needed, to address bacterial superinfection.

In summary, we describe a series of 20 patients with diffuse HIV-associated SD to highlight its severe and distinctive clinical appearance, its notable propensity for erythroderma, and its chronic course that is often refractory to standard SD treatments. We believe this pattern serves as a marker for underlying HIV infection. Because patients may present with a relatively high CD4 cell count, recognition of this disease variant may aid early screening and diagnosis of HIV.