Epidermodysplasia verruciformis (EV) is an autosomal recessive genodermatosis characterized by disseminated dyschromic and squamous skin macules resembling pityriasis versicolor (PV-like macules) and flat warts, predominant on sun-exposed areas.1 Most cases are caused by invalidating mutations in 2 genes, EVER1 and EVER2 located on chromosome 17q25, resulting in a susceptibility to a group of beta human papillomavirus (β-HPV) or EV-associated HPV types. EVER genes encode transmembrane proteins located in the endoplasmic reticulum and are involved in signal transduction.2 It has been demonstrated that EVER proteins regulate zinc homeostasis by interacting with zinc transporter proteins (ZnT1). This interaction is inhibited by the E5 protein of HPV. It has been suggested that loss-of-function mutations in the EVER genes and the lack of E5 protein in β-HPV result in specific susceptibility and host restriction to these viruses.3 Local innate and adaptive immunity to specific HPV types in the infected keratinocytes is likely to be involved.1 This would account for the transition from an abortive to a productive infection by EV-HPV types, which are usually ubiquitous and commensal defective viruses in normal skin.
The oncogenic HPV-5 and HPV-8 are the most frequent HPV types, identified in 90% of patients with EV.1 Approximately 20 other β-HPV types are known, including HPV-9, 12, 14, 15, 17, 19 through 25, 28, 29, 36, and 47. The main histologic features of the disease are the coexistence of epidermal thickening, loose horny layer with a basket weave–like appearance, and presence of large cells in the spinous and granular layers presenting with a large blue-gray cytoplasm. A perinuclear halo can be seen.
Skin eruption usually begins during childhood and progressively spreads. The lesions remain chronic and refractory. As the causal mutations do not confer susceptibility to oncogenic or nononcogenic mucous HPV, the mucous membranes are generally unharmed. Cutaneous nonmelanoma skin cancers develop in 30% to 50% of the patients, within a mean of 25 years after the onset of benign skin lesions.4 These cancers are mostly Bowen's disease and invasive squamous cell carcinomas, growing slowly on areas exposed to the sunlight. The physiopathologic features of cutaneous oncogenesis is suggestive of the combined effects of the oncogenic potential of HPV-5 and HPV-8, an environment of UV-induced keratinocyte mutations, and a specific genetic background.
Susceptibility to HPV infection has been reported in patients with genetic primary immunodeficiencies (WHIM [warts, hypogammaglobulinemia, infections, and myelokathexis] syndrome and severe combined immune deficiency) and in patients with acquired cell-mediated immunodeficiency (HIV-positive patients and immunosuppressive treatment in organ allograft recipients). These patients are at high risk of developing other HPV-related diseases, such as common warts, anogenital condylomas, and mucosal intraepithelial neoplasia, but EV-like eruptions are rarely observed.
5- 9 We studied a series of 11 cases of EV-like syndrome occurring in HIV-infected patients in order to describe the epidemiologic, clinical, virologic, and immunologic characteristics of these patients. A systematic literature review was carried out.
We studied HIV-positive patients seen between 2003 and 2009 in the Department of Dermatology at Tarnier-Cochin Hospital, Paris, France. All patients with clinically evocative and histologically confirmed EV-like syndrome were included in this study. The histological diagnostic features were a combination of a thickened epidermis and the presence of large cells, with an atypical nucleus and a large blue-gray cytoplasm, in the granular and spinous layers (Figure 1). Data collection was standardized. The following demographic and epidemiologic data were recorded: sex, ethnic origin, phototype, age at diagnosis of HIV infection, age at onset, and diagnosis of EV lesions. We also recorded the following clinical signs: existence of flat warts, PV-like macules, lichenoid lesions, topographic abnormalities, history of cutaneous or mucosal HPV-associated infection (warts, condylomas, bowenoid papulosis) and nonmelanoma skin cancer, lymphoproliferative malignant disorders, and viral hepatitis. Timing of lesion onset with respect to the initiation of HAART, CD4 T-cell count, HIV viral load at EV onset and at the time of this report, and HAART efficacy were also recorded. For cases occurring soon after HAART initiation, CD4 T-cell count and HIV viral load at HAART initiation were noted. Results were expressed as percentages or as medians and interquartile ranges for quantitative variables.
Histologic features of epidermodysplasia verruciformis. Epidermal thickening with presence of large cells, sometimes in nests, in the spinous and granular layers presenting with an atypical nucleus and a large blue-gray cytoplasm.
Fresh frozen skin biopsy specimens from typical EV lesions were collected after obtaining informed consent. The skin samples were cut and mechanically dissociated. DNA was extracted with the QIAmp DNA extraction kit for tissues (Qiagen Inc, Valencia, California), using a slightly modified version of the manufacturer's protocol. Briefly, the lysis period was extended to 18 hours, with shaking at 56°C, and repeated additions of proteinase K (Qiagen Inc): 400 μg of the enzyme was added every 3 hours during the first 6 hours and 800 μg before overnight lysis. DNA was eluted in 200 μL of elution buffer (Macheret; Macherey-Nagel, Duren, Germany). Concentration and purity were verified by the NanoDrop method (NanoDrop Technology, Wilmington, Delaware). DNA quality was checked by amplifying a 90–base pair (bp) fragment of the human apolipoprotein B (HAPB) gene (GenBank NG_011793) with the following primers and probe: HAPBF, 5′-TGAAGGTGGAGGACATCCCTCTA-3′; HAPBR, 5′-CTGGAATTGGCATTTCTGGTAA-3′; and HAPBP, 5′-VIC-CGAGAATCACCCTGCCAGACTTCCGT-TAMRA-3′. Amplification was performed on 200 ng of DNA with each primer (300nM) and a 100nM probe in 50-μL Taqman Master Mix (Applied Biosystems, Courtaboeuf, France). After a 10-minute denaturation step at 95°C, cycling conditions consisted in 50 cycles of 15 s at 95°C and 1 minute at 60°C on an ABI 7500 platform (Applied Biosystems).
We used the consensus primer pair FAP59/62, described by Forslund et al.10 Amplification was performed in a final volume of 50 μL containing 200 ng of sample DNA, 500 nmol/μL of each consensus primer and 1X Ampli Taq Gold buffer (Applied Biosystems). The polymerase chain reaction (PCR) conditions were as follows: 10 minutes for initial denaturation at 94°C, followed by 45 cycles of 30 s at 94°C, 1 minute at 54°C, and 1 minute at 72°C, with a final 5-minute elongation at 72°C. As a positive control, 10 ng of plasmid containing the whole genome of HPV-5 or HPV-8 was included. Water was used as a negative control. The PCR products were analyzed by subjecting 20 μL of amplicon to electrophoresis in a 2% agarose gel. For positive samples, the bands were isolated from the gel with the QIAQuick Gel extraction kit (Qiagen Inc), according to the manufacturer's instructions, and eluted into a final volume of 50 μL. The purified PCR products were subjected to automated nucleotide sequencing in an ABI 3130XL genetic analyzer (Applied Biosystems). DNA sequences were compared with the sequences stored at the National Center for Biotechnology Information, Bethesda, Maryland.
Sequences of the primers used were HPV5F (5′-AATCGCATTTGGCACTGC-3′) and HP5R (5′-TTGCACGGTGACCTCT-3′) for HPV-5, and HP8F (5′-TTACAATGCTGTGACTTGTGC-3′) and HPV8R (5′-CACTACATTCAGCTTCCAAAATACA-3′) for HPV-8. These primers amplified a 720-bp fragment from HPV-5 and an 840-bp DNA fragment from HPV-8. Amplification was carried out in a final volume of 50 μL containing 200 ng of sample DNA, 500 nmol/μL of each specific primer and 1X Ampli Taq Gold buffer (Applied Biosystems). The PCR conditions were as follows: initial denaturation for 10 minutes at 94°C, followed by 50 cycles of 30 s at 94°C, 1 minute at 50°C, and 1 minute at 72°C, with a final 5-minute elongation at 72°C. As a positive control, 10 ng of HPV-5 or HPV-8 plasmid were included. Water was used as negative control. We then subjected 20 μL of the PCR products to electrophoresis in a 2% agarose gel. The specificity of the amplification was checked by sequencing the purified amplicon.
We searched the MEDLINE electronic database for articles reporting acquired EV during HIV infection. The keywords used were “epidermodysplasia verruciformis” and “HIV.” We selected and reviewed articles in which the reported cases fulfilled the same inclusion criteria used in our study. Data were collected in a standardized manner, as previously described for our own series.
Eleven HIV-positive patients fulfilling the inclusion criteria were studied. The main characteristics of these patients are summarized in the Table. There were 6 men and 5 women. Six of the patients were black and originated from an African country. The median age at diagnosis of HIV infection was 27 years (interquartile range [IQR], 22-33 years; range, 20-37 years). The first EV lesions occurred at a median age of 40 years (IQR, 29-43 years; range, 15-45 years). The lesions occurred after HIV diagnosis in 10 cases, with a period of 9 years between HIV diagnosis and lesion onset (IQR, 7-13 years; range, 1-20 years). Patient 1 developed EV lesions before the diagnosis of HIV infection. The median age at diagnosis of EV was 44 years (IQR, 35-46 years; range, 32-50 years). The median delay to EV diagnosis was 4 years (IQR, 3-7 years; range, 0-18 years). At the time of this report, EV lesions have been present for a median of 10 years (IQR, 7-12 years; range, 1-24 years). Median age at the time of this report was 45 years (IQR, 41-50 years; range, 39-51 years). Patients 8 and 9 died at the ages of 45 and 44 years, respectively. Patient 10 was lost to follow-up at the age of 50 years.
Extensive flat warts were observed in all cases, with associated pityriasis-like macules in 5 cases. In 3 cases, the clinical presentation was lichenoid papules resembling lichen planus (patients 1, 3, and 10). The skin lesions affected the upper trunk in all cases and were generalized in 3 cases (patients 1, 2, and 3) (Figure 2). Fresh frozen skin biopsy specimens taken from EV lesions were not available for 3 patients (patients 8, 9, and 10), for whom PCR for HPV typing could not be carried out. In all the cases studied, EV-associated HPV was detected, and the strains identified included oncogenic HPV-5 and/or HPV-8 in 6 cases. Coinfection with HPV within EV lesions was observed in 4 cases. In patient 3, HPV-3, associated with flat warts, was found. Patient 6 had another 2 HPV subtypes (HPV-38 and 65) not usually associated with EV.
Clinical features of epidermodysplasia verruciformis. Clinical presentation includes lichenoid dyschromic papules (A) and disseminated flat warts and extensive pityriasis versicolor–like macules (B and C).
Eight patients (73%) currently had or had a history of another HPV-associated cutaneous or mucosal infection: common palmar and/or plantar warts in 4 cases and anogenital bowenoid papulosis and/or condylomas in 7 cases. Associated oncogenic mucosal HPV subtypes were detected in 3 patients (patients 5, 6, and 11). Three of the women in this series had previously undergone treatment for high-grade cervical intraepithelial neoplasia (patients 4, 10, and 11). In patients for whom familial history was available (patients 3, 4, 5, and 6), we found no history of warts or consanguinity in their parents and siblings.
Three patients (patients 3, 5, and 10) developed nonmelanoma skin cancers: 2 cases of Bowen's disease and 1 case of invasive squamous cell carcinoma. None displayed nodal or visceral metastatic progression of the skin cancers. For these skin cancers, HPV typing was not possible. Four patients had a history of lymphoproliferative disorder (patients 3, 8, 9, and 10), which was the cause of death in patients 8 and 9. These lymphoid tumors were not related to EBV. The first EV lesions occurred at the time of diagnosis of human herpesvirus 8–linked multicentric Castleman disease in patient 8 and of a large B-cell lymphoma in patient 10. Current active viral hepatitis was found in 3 patients.
In all cases, HAART was ineffective on the course of skin lesions. The first EV lesions occurred after the initiation of HAART in 7 cases. The time between HAART initiation and EV onset was less than 2 years in 3 cases (patients 2, 4, and 8) and, in 2 other cases, skin eruption occurred in a context of virologic escape (patients 5 and 6). At the onset of EV lesions, median CD4 T-cell count was 170/μL (IQR, 116/μL-326/μL; range, 2/μL-767/μL), and HIV viral load was undetectable in 3 cases (data not available for patient 1). At the time of this report, the median CD4 T-cell count was 495/μL (IQR, 304/μL-615/μL; range, 78/μL-1040/μL), and HIV viral load was undetectable in 8 cases. Patients 2 and 4 displayed a significant increase in CD4 T-cell count and a decrease in HIV viral load on HAART, coinciding with the development of the first skin symptoms. All the patients received topical treatments without any improvement (tretinoin, imiquimod, urea, fluorouracil, and cryotherapy). Patient 3 reported a partial improvement of EV skin lesions during a course of interferon alfa treatment for hepatitis C virus infection.
We identified 17 articles reporting 25 other cases of acquired EV in patients with HIV infection.11- 27 Most of these cases were published as individual case reports. Even the largest series did not exceed 3 cases. The data for these cases are given in the eTable. Articles reporting other HPV-associated skin eruptions in HIV-positive patients, such as flat warts and common warts28,29 without histological features of EV were not included.
Epidermodysplasia verruciformis–like eruptions have rarely been described in immunodeficient patients, unlike HPV-associated anogenital malignant diseases.5- 9 The EV phenotype and EV-HPV genotypes30- 32 were first reported in kidney transplant recipients and then in hematological malignant diseases,33- 35 iatrogenic immunodeficiency,36,37 and CD4+ T lymphocytopenia.38,39 In HIV-positive patients, HAART resulted in a decrease in the incidence of opportunistic infections, but the prolonged survival was associated to the emergence of cutaneous and mucosal HPV-related diseases, especially anogenital intraepithelial neoplasia.5,6,40 The EV-like syndrome provides an example of emerging widespread HPV infections in HIV-positive patients. The present series appears to be the largest reported with clinical, virologic, and immunologic documentation.
We found HIV-positive men and women equally affected by EV. The high frequency of black patients and the lack of pediatric cases are consistent with our recruitment. The first skin lesions mostly occurred in adult patients, after HIV diagnosis, and following a long period of immunosuppression (median of 9 years) and then were chronic and refractory (median duration of 10 years). The onset of HIV infection was unknown for patient 1 but probably occurred during adolescence in Cameroon, at the time of the first skin EV lesions. Then, this patient should be considered as more likely having an acquired EV related to HIV infection than a classic genetic EV. The clinical lesions resembled those of genetic EV, predominantly on the skin exposed to the sun, in accordance with the role of UV in β-HPV–induced lesions. However, the occasional presence of flat warts in sun-protected skin suggests that UV exposure is not always necessary. The presence of lichenoid papules mimicking lichen planus is a clinical feature that has been rarely reported and may account for the long delay of diagnosis (median of 4 years). Pathological examination and HPV typing of these lesions showed typical characteristics of EV and EV-associated HPV that excluded coincidental skin disease associated with viral hepatitis. For virologic results, we increased the sensitivity of HPV typing by amplifying HPV-5 and HPV-8 sequences specifically, in addition to amplifying FAP59/62 sequences, which might identify only the preferentially amplified genotype. β-HPV was found in all skin biopsy specimens and oncogenic HPV-5 or HPV-8 in 75% of specimens, which is similar to that reported for patients with genetic EV1 and in EV-like disorders.32 Gamma HPV-65 and HPV-38 found in patient 6 had previously been described in genetic EV41 and in an immunocompromised patient.34
Unlike patients with genetic EV who are only susceptible to β-HPV, HIV-positive patients with EV often displayed other HPV-related diseases including mucosal HPV in 64% of cases. Most of these mucosal HPV were oncogenic, highlighting the risk of intraepithelial cervical neoplasia observed in HIV-positive women. This is a strong argument for a clinical follow-up with detailed genital and anal examinations in this HIV-positive population. Skin cancers were a rare event in our series and in the published cases, which differs from patients with genetic EV, who are affected by nonmelanoma skin cancer in 50% of cases. Because the mean interval between the onset of benign EV lesions and the first skin cancer in genetic EV cases is long,4 a prolonged follow-up after the onset of skin involvement is probably required to detect a higher incidence of nonmelanoma skin cancer in this population. First, EV lesions were often concomitant to intense immunodeficiency state, confirming the role of T-cell immune responses in the control of HPV infection.42- 44 We focused on the presence of current or past lymphoproliferative disorder in 36% of the patients in our series, including 2 fatal cases. In 2 patients (patients 8 and 10), immunodeficiency related to lymphoproliferative syndrome may have triggered the skin eruption. These observations demonstrate that acquired EV should be considered a marker of chronic and intense immunosuppression.45
We noticed that preexistent skin lesions were not improved by immune restoration under HAART. Complete and partial responses after the initiation of HAART have been previously documented only twice.14,24 An impairment of CD4 cell function may be involved in viral persistence, suggesting that nadir CD4 T-cell counts should be considered.46 This confirms the inefficiency of HAART reported in other HPV infections.47 Cutaneous and mucosal HPV-induced lesions may reveal a specific defect in the immune response to the HPV family not corrected by immune restoration. In other cases, EV lesions are concomitant with the immune restoration by HAART. In the case reported by Mermet et al19 and in 2 patients in the present series (patients 2 and 4), the chronology of skin eruption and the immunologic and virologic data were consistent with immune reconstitution inflammatory syndrome Patient 8 seems different because his first EV lesions occurred at the start of both multicentric Castleman disease diagnosis and HAART. Like common and flat warts, which do not display the histological features of EV, EV-like eruptions are an example of skin manifestations of immune reconstitution inflammatory syndrome.28,29 The physiopathologic mechanism of such eruptions remains unclear owing to a lack of peripheral and local immunomonitoring data, but we suggest an exacerbated cutaneous inflammatory response against commensal β-HPV in a particular genetic background.
Evidence for the genetic heterogeneity of EV is provided by the lack of EVER mutations in some EV cases, including familial cases. X-linked inheritance has been suggested in some of these cases.48- 50 The low frequency of EV-like syndrome in HIV-positive and other immunocompromised patients are suggestive of additional genetic susceptibility. In our series, familial history of patients was not suggestive of any pattern of inheritance. An EVER2 polymorphism potentially associated with β-HPV infection was previously reported.26 Further genetic explorations are warranted in immunocompromised patients to identify risk factors associated with an EV phenotype.
In conclusion, acquired EV-like eruptions in HIV-positive patients should be considered not only as a marker of specific susceptibility to β-HPV infection and skin cancers, but also as a generalized immunodeficiency state associated with low- or high-risk mucosal HPV infection and lymphoproliferative disorders. Skin lesions often remain refractory, despite immune restoration by HAART, and may reflect a genetic susceptibility.
Correspondence: Simon Jacobelli, MD, Department of Dermatology, Tarnier-Cochin Hospital, 89 rue d’Assas, 75006 Paris, France (email@example.com).
Accepted for Publication: October 25, 2010.
Published Online: January 17, 2011. doi:10.1001/archdermatol.2010.399
Authors Contributions: Drs Jacobelli, Avril, and Dupin had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Jacobelli, Avril, and Dupin. Acquisition of data: Jacobelli, Laude, Carlotti, Rozenberg, Deleuze, Morini, Franck, Gorin, Avril, and Dupin. Analysis and interpretation of data: Jacobelli, Avril, and Dupin. Drafting of the manuscript: Jacobelli and Laude. Critical revision of the manuscript for important intellectual content: Laude, Carlotti, Rozenberg, Deleuze, Morini, Franck, Gorin, Avril, and Dupin. Administrative, technical, and material support: Laude, Carlotti, and Rozenberg. Study supervision: Avril.
Financial Disclosure: None reported.
Additional Contributions: Michel Favre, PhD (Pasteur Institute, Paris, France), provided plasmids containing the whole genomes of HPV-5 and HPV-8.
Jacobelli S, Laude H, Carlotti A, Rozenberg F, Deleuze J, Morini J, Franck N, Gorin I, Avril M, Dupin N. Epidermodysplasia Verruciformis in Human Immunodeficiency Virus–Infected PatientsA Marker of Human Papillomavirus–Related Disorders Not Affected by Antiretroviral Therapy. Arch Dermatol. 2011;147(5):590-596. doi:10.1001/archdermatol.2010.399