Cutaneous Co-infected Cytomegalovirus and Herpes Simplex Virus Perigenital Ulcers in Human Immunodeficiency Virus Patients

Jason Schoenfeld, MD, Jefferson Medical College of Thomas Jefferson University, Philadelphia, Pennsylvania; Sarah Cannon, MD; Kristin Cam, MD; Matthew Keller, MD, Department of Dermatology and Cutaneous Biology, Jefferson Medical College of Thomas Jefferson University, Philadelphia, Pennsylvania

Disclosure: The authors report no relevant conflicts of interest.

There is uncertainty regarding the pathogenic nature of cytomegalovirus in cutaneous lesions co-infected with herpes simplex virus. It is widely believed that herpes simplex virus is the main pathogenic factor in such lesions and that cytomegalovirus plays little if any role. There are, however, isolated case reports that describe cytomegalovirus as an important driving pathogen in such lesions. The authors present two human immunodeficiency virus patients who have cytomegalovirus and herpes simplex virus co-infected perigenital ulcers, one of whom improved on valacyclovir, while the other, who was already on valacyclovir for chronic herpes simplex virus suppression, showed no improvement with a single dose of cidofovir. He only showed rapid improvement when treated with valganciclovir. The latter patient underscores the viewpoint that at least in some cases, cytomegalovirus may be an important driving force behind the formation of such lesions. The authors therefore recommend that clinicians be aware of the possible pathogenic role of cytomegalovirus in these ulcers, and, in nonhealing ulcers, use anti-cytomegalovirus agents to prevent the onset of systemic disease. These results warrant further study of the pathogenesis of cytomegalovirus in co-infected herpes simplex virus ulcers.
(J Clin Aesthet Dermatol. 2013;6(10):41–43.)

Genital ulcers have a broad differential diagnosis, and may arise from infectious or noninfectious causes.[1] Infectious causes include herpes simplex virus (HSV), syphilis (Treponema pallidum), chancroid (Haemophilus ducreyi), granuloma inguinale (donovanosis), lymphogranuloma venerum (Chlamydia trachomatis serotypes L1 ,L2, L3), fungi (i.e., Candida), and secondary bacterial infections.1 Noninfectious causes include trauma, psoriasis, Behcet’s syndrome, Wegener’s granulomatosis, and fixed drug reactions.[1] In the United States, most young, sexually active patients who present with genital ulcers have HSV or syphilis.[1,2]

In immunocompromised patients, additional etiologies, such as acquired immunodeficiency syndrome (AIDS)-related neoplasms (i.e., nodular kaposi sarcoma and B cell non-Hodgkin lymphoma) and CMV co-infection must also be considered in the differential diagnosis.3 This is because the presence of CMV may be a marker of impending systemic infection and is an especially important diagnostic consideration when an ulcer does not respond to initial therapy.[4–8]
Although it is commonly believed that HSV, not CMV, is the true pathogenic causative agent in these lesions, the authors provide evidence that in certain cases, CMV may be an important driving pathogen causing genital ulcerations.

Case Presentations
Case 1. A 38-year-old man with AIDS (CD4 38) was admitted to Thomas Jefferson University Hospitals, Philadelphia, Pennsylvania, with altered mental status and left-sided weakness secondary to central nervous system (CNS) toxoplasmosis. During his stay, dermatology was consulted to evaluate penile lesions. He reported that the lesions were present for “weeks to months,” painful when scratched, but otherwise did not bother him. He denied a prior diagnosis of HSV and believed the lesions were due to tape from the foley catheter. The patient denied any fevers or any other skin lesions. His medications included dexamethasone, levetiracetam, olanzapine, pyrimethamine, sulfadiazine, azithromycin, fluconazole, leukovorin, and enoxaparin. On physical exam, two ulcerations were present on the shaft of the penis, and there was hypopigmentation evident from a prior lesion on the glans. There was shotty, non-tender, left inguinal lymphadenopathy, condyloma on the scrotum, and no perianal lesions. His HIV viral load was 1.17 million, rapid plasma reagin (RPR) was nonreactive, and urine chlamydia/gonorrhea polymerase chain reaction (PCR) was negative. A Tzanck smear showed characteristic multinucleated keratinocytes, which was confirmed by a positive HSV culture. A CMV PCR was also positive at 29,100 deoxyribonucleic acid (DNA) copies/mL. The patient was treated with valacyclovir and put on chronic HSV suppressive therapy with improvement.
Case 2. A 38-year-old man with AIDS (CD4 35) presented to the dermatology clinic with a complaint of worsening “boils” on his buttocks for the last two weeks. He also reported having draining sores and enlarged, painful lymph nodes. The patient reported that he had been seeing a urologist for one year for possible methicillin-resistant Staphylococcus aureus (MRSA) of the penis and taking trimethoprim-sulfamethoxazole for one year with no relief. He was also on highly active antiretroviral therapy (HAART) and chronic valacyclovir prophylaxis. On physical exam, there was a large ulcer with a peripheral hyperkeratotic papule and loss of glans from prior surgery. There were multiple pedunculated, ulcerated, skin-colored papules with purulent drainage in the perianal area. There were several purulent suprapubic ulcerations and bilateral inguinal lymphadenopathy that was painful to palpation. Biopsy of one the suprapubic lesions demonstrated a likely eosinophilic dermatosis with enlarged and pale keratinocytes at the ulcer edges with multinucleation, chromatin margination, and nuclear molding. An immunohistochemical stain for HSV was positive. His HIV viral load was 591. A CMV PCR was also positive at 2144 DNA copies/mL. He was first given a single dose of cidofovir and showed no improvement. He subsequently was given valganciclovir, with rapid improvement of the lesions.

In one of the patients with HSV and CMV co-infected ulcers (Case 1), the lesions resolved with anti-HSV medication (valacyclovir). In the other patient (Case 2), who was already on chronic anti-HSV prophylaxis (valacyclovir), the lesions failed to resolve with cidofovir and only showed rapid improvement when treated with valganciclovir, an anti-CMV medication.

CMV infection in immunocompetent patients is often unnoticed or may manifest as a mononucleosis-like syndrome. Immunocompromised patients may respond to primary or recurrent CMV infections by remaining completely asymptomatic, or, more frequently, by developing disseminated disease.[8] Such infection may be characterized by viremia with fevers, leukopenia, hepatitis, pneumonitis, esophagitis, gastritis, colitis, and retinitis. The mortality rate is high (80–90%).[4,9] Though rare, cutaneous manifestations that have been attributed to CMV include maculopapular rashes, urticarial eruptions, papules, and ulcers, which are frequently localized to the perianal and perigenital areas.[4,7,10–13] Immunohistochemical staining and viral culture are two methods of diagnosing cutaneous CMV infection. To diagnose active viremia, a CMV antigenemia assay or the more sensitive PCR is preferred.[8]
CMV and HSV, both herpes viruses, may co-infect in the same lesion.[8,11,14] Such lesions may be seen in the immunocompromised as cutaneous lesions, often as ulcers.[4,5,11,12,15] The pathogenesis of CMV in co-infected cutaneous lesions is controversial.[4,5,11,12,15,16] Some clinicians have concluded that CMV does not play a significant pathogenic role in these lesions. Dauden et al[5] reflect this belief in three studies of CMV and HSV co-infected ulcers.  They conclude that CMV is often cultured with other infectious agents that have a recognized pathogenic role and that CMV may appear in otherwise healthy skin. Others maintain that CMV does have a pathogenic role, and there are reports of such lesions improving only with anti-CMV medications, such as ganciclovir. Moodley et al[11] describe the case of an HIV patient with a co-infected HSV and CMV vulval ulcer that resolved only when treated with ganciclovir. Garib et al[8] additionally describe three immunocompromised patients with cutaneous co-infected HSV and CMV lesions that improved with ganciclovir and valganciclovir.

Ganciclovir, valganciclovir, foscarnet, and cidofovir have been approved for treatment of CMV.[13] Ganciclovir is a guanosine derivative that has considerably more activity against CMV than acyclovir.[17] After intracellular conversion by a viral phosphotransferase, it is a selective inhibitor of CMV DNA polymerase.[9] In severe infections, ganciclovir is often combined with CMV immune globulin.[9] Valganciclovir is an orally bioavailable prodrug that is rapidly metabolized to ganciclovir in intestinal tissues and the liver. Myelosuppression is the foremost dose-limiting side effect, with neutropenia seen in 15 to 40 percent and thrombocytopenia in 5 to 20 percent.[18] Neutropenia is observed most commonly during the second week of treatment and is usually reversible one week after stopping the drug.[18] CNS side effects may be seen in 5 to 15 percent and range from headaches to coma.[18] Foscarnet inhibits CMV DNA polymerase and is effective against many ganciclovir-resistant strains because it does not require phosphorylation to be active.[9] It causes considerable toxicity, however, including renal dysfunction, genital ulcers, dysuria, nausea, and paresthesias.[9] Cidofovir is a nucleotide analogue that allows intermittent intravenous administration. It can cause severe nephrotoxicity through dose-dependent proximal tubular cell injury.[9]

The patient described in Case 2 who was on chronic valacyclovir prophylaxis and improved only on valganciclovir (and not a single dose of cidofovir) underscores the viewpoint that in at least a subset of these patients, CMV does seem to play a pathogenic role. It is likely that the patient who was only given a single dose of cidofovir did not receive enough of the medication to be active against CMV. While valganciclovir is also active against HSV, it is more effective against CMV than HSV.[17]

While this patient did not develop systemic disease from the CMV, it is postulated that cutaneous CMV lesions may be a marker of impending systemic infection. With that being said, in any immunocompromised patient who presents with perianal or perigenital lesions, it is important to at least consider CMV as a causative agent, especially in lesions not responding to conventional treatment. The authors’ findings warrant further study into the pathogenesis of cutaneous CMV infection in immunocompromised patients.

1.     Roett M, Mayor M, Uduhiri K. Diagnosis and management of genital ulcers. Am Fam Physician. 2012;85:254–262.
2.     Mertz K, Lewis J, Morse S, et al. Etiology of genital ulcers and prevalence of human immunodeficiency virus coinfection in 10 US cities. J Infect Dis. 1998;178:1795–1798.
3.     Grayson W. The HIV-positive skin biopsy. J Clin Pathol. 2008;61:802–817.
4.     Drago F, Aragone M, Lugani C, Rebora A. Cytomegalovirus infection in normal and immunocompromised humans: a review. Dermatology. 2000;200:189–195.
5.     Dauden E, Fernández-Buezo G, Fraga J, et al. Mucocutaneous presence of cytomegalovirus associated with human immunodeficiency virus infection: discussion regarding its pathogenetic role. Arch Dermatol. 2001;137:443–448.
6.     Lee J, Peel R. Concurrent cytomegalovirus and herpes simplex virus infections in skin biopsy specimens from two AIDS patients with fatal CMV infection. Am J Dermatopath. 1989;11:136–143.
7.     Bournerias I, Boisnic S, Patey O, et al. Unusual cutaneous cytomegalovirus involvement in patients with acquired immunodeficiency syndrome. Arch Dermatol. 1989; 125:1243–1246.
8.     Garib G, Hughey L, Elmets C, et al. Atypical presentation of exophytic herpes simplex virus type 2 with concurrent cytomegalovirus infection: a significant pitfall in diagnosis. Am J Dermatopathol. 2012 Apr 24 [Epub ahead of print].
9.     Hirsch M. Cytomegalovirus and human herpes virus types 6, 7, and 8. In: Longo D, Fauci A, Kasper D, et al. Harrison’s Principles of Internal Medicine. 18th ed. 2012. < edu/content.aspx?aID=91518> Accessed on July 10, 2012.
10.     Choi Y, Kim J, Jang K, et al. Characteristics of cutaneous cytomegalovirus infection in non-acquired immune deficiency syndrome, immunocompromised patients. Brit J Dermatol. 2006;155:977–982.
11.     Moodley M, Chetty R, Moodley, J. Vulval cytomegalovirus coexisting with herpes simplex virus in a patient with human immunodeficiency virus infection. BJOG. 2003; 110:1123–1124.
12.    Smith K, Skelton H, James W, Angritt P. Concurrent epidermal involvement of cytomegalovirus and herpes simplex virus in two HIV-infected patients. J Am Acad Dermatol. 1991;25:500–506.
13.     Belazarian L, Lorenzo M, Pace N, et al. Exanthematous viral diseases. In: Wolff K, Goldsmith L, Katz S, et al. Fitzpatrick’s Dermatology in General Medicine. 2008. <    nt.aspx?aID=2997277>. Accessed on July 15, 2012.
14.     Regazi, J, Eversole LR, Barker B, et al. Herpes simplex and cytomegalovirus coinfected oral ulcers in HIV-positive patients. OMPJ. 1996;61:55–62.
15.     Horn T, Hood A. Cytomegalovirus is predictably present in perineal ulcers from immunosuppressed patients. Arch Dermatol. 1990;126:642–644.
16.    Saito Y, Takenaka H, Chin T, et al. Cytomegalovirus in cutaneous ulcers in a patient with adult T cell lymphoma. Acta Derm Venereol. 2006;86:181–182.
17.     Biron K, Stanat S, Sorrell J, et al. Metabolic activation of the nucleoside analog 9-{[2-hydroxy-1-(hydroxymethyl) ethoxy]methyl} guanine in human diploid fibroblasts infected with human cytomegalovirus. Proc Natl Acad Sci USA. 1985;82:2473–2477.
18.     Acosta E, Flexner C. Antiviral Agents (Nonretroviral). In: Brunton L, Chabner B, Knollmann B. Goodman and Gilman’s The Pharmacological Basis of Therapeutics. 12e. 2011. <>. Accessed on November 1,1 2012.