A 38-year-old HIV-1-infected African American woman presented in September 2009 with several weeks of a painful vulvar ulcer. Her HIV disease was diagnosed in 2003 and she had been on HAART (lamivudine, zidovudine and atazanavir) since 2006, although with poor compliance. After another lapse in treatment for 1 year, she resumed therapy in 2008 with a prompt improvement in immune status 2 months later (viral load drop from >1,000,000 to 390 copies/ml and concomitant rise in CD4 count from 3 to 64/mm3). Her HAART regimen was changed to emtricitabine, tenofovir, atazanavir and ritonavir, and approximately 8 months later (11 months after resuming HAART) she developed the painful erosion on the inner left labia minora. Her viral load was still detectable at 240 copies/ml and her CD4 count had risen to 194/mm3. She had a history of genital herpes simplex and was on valacyclovir 500 mg twice-daily prophylaxis up to this time with no clinical symptoms. Lesional culture did not detect HSV but she experienced some symptomatic relief with empiric increase in her valacyclovir dosing to thrice daily. However, the erosion persisted and by 6 months later had enlarged to involve both labia with painful ulceration and yellowish adherent exudates (fig. ). HAART therapy was again changed to abacavir, lamivudine, atazanavir and ritonavir due to nausea. On this regimen, viral activity dropped to <50 copies/ml with a rise in CD4 count to 265/mm3. Again HSV was not detected on culture so a biopsy was obtained to confirm the suspicion and exclude the possibility of fixed drug reaction, erosive lichen planus, syphilis, or malignancy. The result was non-diagnostic, revealing diffuse mixed dermatitis with neutrophils, plasma cells, and numerous eosinophils. Immunohistochemistry did not detect HSV-1 or -2, and special stains for spirochetes were negative. The rapid plasma reagin was non-reactive as well. Supportive care with triamcinolone 0.1% ointment and topical lidocaine were started, and valacyclovir was continued. With spread of the ulceration to the perineal region 2 months later, another skin biopsy was performed showing similar findings (fig. ), but the presence of HSV was confirmed by immunohistochemistry (fig. ). IgG antibodies to HSV-1/-2 were detected at high titer in the serum confirming past exposure to HSV. There was no evidence of treponemal organisms and the rapid plasma reagin was again negative, no cytomegalovirus was detected by immunohistochemistry or PCR, and direct immunofluorescence was negative.
Vulvar ulceration with adherent fibrinous exudates and scalloped borders.
Fig. 2 Histopathology from the edge of the ulceration shows epidermal hyperplasia, mild papillary dermal edema, fibrosis in the upper lamina propria, and a diffuse, deeply extending mixed infiltrate of lymphocytes, neutrophils, and numerous eosinophils. Scattered (more ...)
Immunohistochemistry staining showing positive keratinocytes containing HSV-1/-2 antigens.
Acyclovir-resistant HSV was assumed and intravenous foscarnet was initiated at 40 mg/kg twice daily. After 3 weeks, her vulvar lesions had completely resolved and her perineum was greatly improved. Topical cidofovir 1% gel was then started twice daily with prompt worsening, so she resumed foscarnet for another 4 weeks. Ulcerations persisted again, so her dose of foscarnet was increased to 60 mg/kg twice daily for an additional 3 weeks. At follow-up, she showed improvement; however, she still had small persistent erosions. Her total CD4 count was still low at 168/mm3. She then resumed topical cidofovir and had slow but gradual improvement over the next 9 months. Valacyclovir prophylaxis was added and by 5 months later she was almost completely healed and comfortable. Her total CD4 count was now at 330/mm3. She continues to maintain control with this regimen.