Skin Diseases in the Era of Highly Active Antiretroviral Therapy: A Retrospective Study of 534 Patients

Abstract

Background:

With the increasing number of HIV-infected patients receiving highly active antiretroviral therapy (HAART), the shift in their dermatologic profile becomes less characteristic of AIDS-defining illnesses.

Methods:

Retrospective review of mucocutaneous pathology among patients seen at HIV-Dermatology Clinic from January 2009 to December 2013.

Results:

Among 534 patients, there were 68.4% males and 31.6% females, with 8.7-year average duration of infection; 82.8% were receiving HAART. Kaposi sarcoma was the only relatively frequent AIDS-defining disease. Fungal and viral infections were common, with human papilloma virus (HPV) as the most frequent overall. Benign and premalignant tumors were associated with HAART and CD4 >200/mm³ (P < .05). Psoriasis was prevalent among patients without HAART (P < .05). Prurigo was associated with lower CD4 count (P < .001).

Conclusion:

Patients receiving HAART are faced with chronic skin problems such as benign and premalignant tumors, and HPV infection adds to their neoplastic predisposition. Further research is recommended to develop protocols for treating psoriasis and screening for HPV-associated neoplasia among patients.

Keywords

skin diseases HIV HAART

Introduction

Patients living with HIV infection are usually afflicted with skin diseases, and their variability and severity often reflect their level of immune deficiency. Ranging from infections, inflammations, and neoplasms, these can lead physicians to the initial diagnosis of HIV infection or can manifest the various stages of its disease evolution. In addition to this is the problem of drug reactions and iatrogenic effects as well as the reactivation of certain diseases alongside immune reconstitution while receiving highly active antiretroviral therapy (HAART).¹

The epidemiologic profile of dermatologic illnesses in relation to HIV varies between countries. This is mainly affected by economic and political factors pertaining to the availability of HAART, as well as the risk-taking behavior of patients. In Portugal, a few years after the recognition of HIV (1985-1991), the mucocutaneous manifestations were dominated by AIDS-defining conditions, such as oropharyngeal candidiasis and Kaposi sarcoma, as well as the aggravation of more common conditions, such as herpes simplex virus (HSV) infection, herpes zoster, dermatophytosis, seborrheic dermatitis, and drug-related skin disorders.² However, for the past 5 years, there has been a decreasing trend in terms of the number of new HIV-infected cases but with an increasing number of people living with the infection alongside the increasing number of patients receiving HAART.³ These observations result to a less number of opportunistic infections, but instead there is the emergence of medical problems that were less common before⁴ and a shift to more chronic forms of illnesses.

This study aims to have a descriptive measure of the current epidemiology of mucocutaneous pathology among HIV-positive patients, based on the referrals to the dermatology clinic of a tertiary hospital in Lisbon, Portugal, from 2009 to 2013. By understanding the trend and identifying shifts in the epidemiologic burden of disease, the authors hope to identify areas that can be bases for future research in assessing the economic burden of HIV in the field of dermatology and hence aid in the formulation of new policies and planning of resource allocation.⁵

Materials and Methods

Clinical Data

We accessed data from the electronic records of all HIV-infected patients on their first consult at the Specialized HIV Clinic of the Dermatology Service of Hospital de Santa Maria during the period of January 2009 to December 2013. These include in-hospital referrals as well as those referred from other hospitals in Greater Lisbon area. Dermatologic diagnoses were based on clinical data and, if needed, laboratory and pathologic correlations as in the case of malignant neoplasms. The medical ethics board of the hospital approved the study.

Statistical Analysis

The primary outcome variable was the frequency rate of identified skin diseases among the study participants, classified according to those receiving and not receiving HAART. We also determined the mean CD4 count for each diagnosis. The secondary outcome variables were the significant correlation of each diagnosis with the patient’s HAART status and the significant difference in the mean CD4 count among the diagnoses.

We analyzed the data statistically using IBM Statistical Package for the Social Sciences (SPSS) version 22. We used summary statistics, where N is the total number of HIV-infected patients included in the study. For the analysis of the secondary outcome variables, we used Pearson χ² test (P < .05) to check for correlations between categorical data. For parametric data, we used independent samples t test at the same significance level.

Results

Epidemiologic Profile

We initially assessed a total of 564 patients, of which 5% did not have CD4 count available on record. In the end, we included a total of 534 patients, of which 68.4% were males and 31.6% were females. They were mostly Caucasians with 18% non-Caucasians. The mean age at the time of dermatologic consult was 44.6 years (standard deviation [SD] = 11.9), whereas the mean age at initial HIV diagnosis was 36 years (SD = 11.9), both values with no significant difference between the genders. The average duration of HIV infection was 8.7 years (SD = 6.4). The primary mode of HIV transmission was heterosexual, with 18% being men having sex with men (MSM). History of intravenous (IV) drug use was identified in 21.8%, of which 96.6% were Portuguese and 98.3% were heterosexuals. Also, the majority of the study population had HIV-1, with 3.9% HIV-2, of which 76% were non-Caucasians.

In this study, around 82.8% were receiving HAART at the time of their dermatologic consult. Almost all (96.4%) of them were compliant, with no significant difference between genders. The 3.4% of those who were not compliant were all heterosexuals. In this case, noncompliance was significantly correlated with the history of IV drug use (P < .05).

The mean CD4 count for those receiving HAART was 522/mm³ (SD = 340.85), whereas for those not receiving HAART was 457/mm³ (SD = 284.74), with no significant difference between the 2 groups. Those receiving HAART had a significantly longer duration of HIV infection (9.6 years, SD = 6.2) versus those not receiving HAART (4.1 years, SD = 5.4; P < .001).

Dermatologic Diagnoses Leading to HIV Testing

Table 1 shows the 12 cases in which HIV was diagnosed by a dermatologist. Syphilis was the most common diagnosis that had prompted dermatologists to do HIV testing in this study. These patients were 76.9% Caucasians, with an almost equal proportion of heterosexuals and MSM.

Table 1.

Most Common Dermatologic Diagnoses Leading to HIV Diagnosis.

Dermatologic Diagnosis That Led to HIV Diagnosis	Percentage (N = 12)
Syphilis	58.7% (7)
HPV, anogenital	30% (4)
Kaposi sarcoma	8.3% (1)

Most Common Dermatologic Diagnoses among HIV-Infected Patients

Around 21% of the study population had more than 1 dermatologic diagnosis at the time of the consult and these were tallied accordingly. Table 2 shows a list of the most common dermatologic diagnoses encountered in the study, classified according to those receiving HAART and not. As mentioned earlier, there was no significant difference between the mean CD4 of the 2 groups in general. However, when the dermatological diagnosis was analyzed individually, there was significantly lower mean CD4 level observed among patients with prurigo (303/mm³, SD = 207, P < .001), Kaposi sarcoma (296/mm³, SD = 252, P < .05), and molluscum contagiosum (325/mm³, SD = 191, P < .05) when compared to the overall mean CD4 count. The observed frequencies of these disease entities had, nonetheless, no correlation with their HAART status. Prurigo was also significantly associated with a shorter mean duration of HIV infection, that is, 4.9 years (SD = 4.95, P < .05). On the other hand, the frequency of psoriasis and syphilis was found to have a significant correlation with negative HAART status (P < .05), although their mean CD4 count was not significantly different from the average in this study.

Table 2.

Most Common Dermatologic Diagnoses among HIV-Positive Patients Classified by HAART Status.

Dermatologic Diagnoses	Mean CD4 count/mm³ (±Standard Deviation)^a	% (Frequency; with HAART), N = 442^b	% (Frequency; without HAART), N = 90^b
HPV, anogenital	506 (±276)	8.1% (36)	13.3% (12)
Seborrheic dermatitis	560 (±412)	7.9% (35)	8.9% (8)
HPV, nongenital warts	512 (±288)	7.9% (35)	6.7% (6)
Cutaneous xerosis	471 (±259)	5.9% (26)	3.3% (3)
Dermatophytosis	522 (±421)	5.4% (24)	5.6% (5)
Psoriasis	436 (±322)	5.4% (24)^c	12.2% (11)^c
Onychomycosis	523 (±224)	5.2% (23)	4.4% (4)
Eczema	592 (±372)	4.1% (18)	4.4% (4)
Folliculitis	586 (±444)	3.6% (16)	4.4% (4)
Seborrheic keratosis	604 (±370)	3.6% (16)	0
Prurigo	303 (±207)^d	3.4% (15)	4.4% (4)
Rosacea	573 (±372)	3.2% (14)	3.3% (3)
Kaposi sarcoma	296 (±252)^d	3.2% (14)	2.2% (2)
Melanocytic nevus and other benign nevus types	483 (±241)	2.9% (13)	1.1% (1)
Syphilis	490 (±291)	2.7% (12)^c	8.8% (8)^c
Fibromas	804 (±290)	2.3% (10)	0
Molluscum contagiosum	325 (±191)^d	2.3% (10)	2.2% (2)
Herpes simplex	380 (±205)	2.3% (10)	0
Pityriasis versicolor	481 (±172)	1.8% (8)	1.1% (1)
Trichilemmal cyst	627 (±361)	1.8% (8)	2.2% (2)
Acne	522 (±271)	1.6% (7)	1.1% (1)
Epidermoid cyst	835 (±548)	1.6% (7)	2.2% (2)
Postinflammatory pigmentation	297 (±216)	1.6% (7)	1.1% (1)
Actinic keratosis	501 (±237)	1.4% (6)	0

Abbreviation: HAART, highly active antiretroviral therapy.

^aOverall mean CD4 count = 514 (±335), where n = total frequency of diagnoses.

^bPertains to n (number) of patients in that group and used as the denominator in that column. But as some patients had more than 1 diagnosis, the total frequency of dermatologic diagnoses may add up to more than 100%.

^cSignificantly different when compared to mean CD4 of the rest of the diagnoses at P < .05 (for prurigo, P < .001).

^dSignificantly different between those taking HAART and not taking HAART at P < .05.

It is also noteworthy that 17% of the study population had CD4 count below 200/mm³, 80.2% of which were receiving HAART. The most common diagnoses, in this case, were psoriasis (9.1%), seborrheic dermatitis (9.1%), dermatophytosis (7.3%), prurigo (5.5%), nongenital human papilloma virus (HPV) warts (5.5%), anogenital HPV warts (4.5%), Kaposi sarcoma (4.5%), and syphilis (4.5%).

Finally, when the diagnoses were grouped etiologically (ie, viral, fungal, or bacterial infections; benign and premalignant tumors; malignant neoplasms; and other inflammatory conditions) and then divided according to (1) patients’ HAART status and (2) CD4 cutoff level of 200/mm³, it was found that the group of benign and premalignant tumors was significantly more common among those receiving HAART (21% versus 10%, P < .05) as well as those with CD4 count >200/mm³ (22% versus 5%, P < .05).

Discussion

The demographic profile of the study population is compatible with that of the World Health Organization’s report in 2011, except for the higher rate of IV drug use in this study (21.8% versus the 7.0% reported rate) and lower rate of MSM (18% versus reported rate of 27.6%).³ Nevertheless, it was reported that the overall trend for the country is decreasing in terms of HIV infection due to injecting drug use, as opposed to transmission via MSM which has been increasing throughout the years.⁶ In addition, more recent studies show an older age profile of patients (51 years)⁴ when compared to that of the pre-HAART era (20-40 years).^2,7,8 There has also been an increasing proportion of female patients, similar to the trend described in India⁹ and the United States.¹⁰ This has been attributed to increased awareness among females, leading to voluntary testing and detection.⁹ In the case of Portugal, it may be partly due to the immigration of patients from former colonies in Africa. In fact, while there were 32% females overall, there was an equal proportion between genders among the non-Caucasian group. This group also represented majority of the HIV-2 infections in this study, although only 3.9% of the total population.

Since the introduction of HAART, treatment facilities in the country have increased considerably from 2004 to 2009,³ as reflected by the high rate of treatment in the study population. Those who were not receiving HAART were more recently diagnosed (4.1 years) than the ones receiving treatment (9.6 years). Nonetheless, both groups in general had a relatively good immune status at the time of the dermatologic consult, as reflected by the profile of mucocutaneous pathologies in Table 2. In fact, Kaposi sarcoma was the only relatively frequent AIDS-defining condition. Oral candidiasis, which has been associated with low CD4 count or negative HAART status,^11,12 was not common in this report.

In a study during the pre-HAART period wherein 11% of HIV-infected cases were diagnosed initially by a dermatologist, Kaposi sarcoma was the most common diagnosis that had led to HIV testing; this was followed by perianal condylomas, other sexually transmitted infections, seborrheic dermatitis, and HSV.⁷ In the current study, 2.4% of the HIV-infected cases were diagnosed by a dermatologist; syphilis has become the most common indication to test for HIV, followed still by anogenital HPV. Syphilis was also the second most common diagnosis among MSM. These findings highlight the role of this disease as one of the most common causes of genital ulcers, and hence in the spread of HIV infection, especially in Western Europe and the United States.¹³ As such, the importance of identifying HIV infection among patients with syphilis can never be undermined. Conversely, when the immune deficiency is severe, syphilitic infection among HIV-infected patients can have an altered course. It may have negative serologic tests in spite of active disease, more severe clinical picture, more rapid progression to late stage, recurrence despite adequate treatment, and difficulty in controlling active infection with the recommended dose of penicillin.¹⁴

Some fungal and viral infections have continued to be common ever since the pre-HAART period.⁷ Included in the former are dermatophytosis. For the latter, there are HSV, molluscum contagiosum, and HPV. The first 2 viral infections have been associated with aggravation of AIDS.⁷ In this study, patients with molluscum contagiosum had significantly lower mean CD4 levels, as was observed in previous studies.^9,10 On the other hand, anogenital warts are not only important in raising suspicions for HIV but are also a chronic problem among patients living with the infection. They were also the most common finding among MSM in this study. Known for their carcinogenic potential, some risks of non-AIDS-defining malignancies, including anal carcinoma, have already been identified.⁴ These can be attributed to worsening immune status, direct action of HIV on various cellular mechanisms, coinfection with other oncogenic viruses, and additional carcinogenic potential of some antiretroviral drugs.⁴ As such, it is advised to do patient monitoring that includes serial physical examinations of the anogenital area, colposcopy or proctoscopy, cervical and anal HPV determination, and cytology, followed by histologic confirmation when indicated.¹

In the current study, no other malignant skin tumor has been frequently observed yet aside from Kaposi sarcoma, but a significant number of benign and premalignant tumors was associated with CD4 count >200/mm³ and HAART when grouped and analyzed etiologically. A similar finding was observed in a study in the United States,¹⁰ wherein nevus, epidermoid cyst, seborrheic keratosis, actinic lentigo, and acrochordon were significantly associated with CD4 count >200/mm³. The latter was also associated with HAART. This now raises the question of whether the dermatologic disease profile among HIV-infected patients receiving HAART approaches that of the population without the infection.

In addition, the etiologies mentioned earlier are the inflammatory conditions, among which psoriasis warrants attention. Together with seborrheic dermatitis, it was the most common dermatologic disease among patients whose CD4 counts were below 200/mm³. It is reported that psoriasis presents late with increasing immunodysfunction, as it creates an environment that is conducive to developing the disease among individuals with susceptible genotypic constitution.¹⁵ In this study, psoriasis was the second most frequent diagnosis among patients not receiving HAART and, in fact, was significantly associated with negative HAART status. Treatment of HIV-associated psoriasis can be challenging, as there is still no standard guideline for this specific group of patients. Although some cases have associated the use of HAART with clinical improvement in psoriasis,¹⁶ the general principles of management of psoriasis and treating HIV apply, that is, to start HAART at a CD4 count <350/mm³.¹⁵ This scenario should entail further multidisciplinary discussion to effect a treatment plan that weighs the risk–benefit ratio in each patient.¹⁵

Another important skin problem is prurigo, especially among patients who had relatively more recent diagnosis of HIV as well as those with poor immune status. It does not have a single underlying etiology but is consistently associated with low CD4 count.^10,12 Among the Chinese¹² and Africans,¹⁷ it has been linked to arthropod bites and, at times, to poor economic conditions. Eczema¹⁷ and intractable pruritus due to immune dysregulation¹⁰ are also possible causes. The HAART is said to effectively improve skin lesions, but patients who are already receiving therapy and still have persistent lesions often show antiviral treatment failure.¹²

To this end, this study had some limitations considering that an unknown proportion of HIV-infected patients might have opted to be seen in private dermatology clinics. Also, since we only had access to the current CD4 count at the time of the referral and majority of the sample were receiving HAART, we were not able to study the relationship between nadir and some dermatologic diseases. However, in Portugal, all HIV-infected patients receive HAART through government hospitals and clinics, and as our dermatology clinic is a referral center for a number of hospitals in the Greater Lisbon area, our sample can be indicative of the most common dermatologic conditions in this immunosuppressed population.

Conclusion

The current profile of mucocutaneous pathologies from our study population in Portugal is similar to other European series¹⁸ and is reflective of a country wherein there is good coverage of HAART among people living with HIV/AIDS. As HAART-treated patients now tend to follow a chronic course of infection, the usual AIDS-defining illnesses have become uncommon, and there is an increase in cases of benign and premalignant tumors and an increasing risk of HPV-related, non-AIDS-defining cancers.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Rieger

Chen

Cockerell

. Cutaneous manifestations of HIV infection. In: Bolognia

Jorizzo

Schaffer

, eds. Dermatology. 3rd ed. St Louis, MO: Elsevier; 2012:1285–1302.

Lamarão

Eliseu

Castro

. Patologia cutâneo-mucosa na infecção pelo Virus de Imunodeficiência Humana. Acta Méd Port 1994; 7: 141–148.

World Health Organization. Key facts on HIV epidemic in Portugal and progress in 2011. 2012. http://www.euro.who.int/__data/assets/pdf_file/0009/191088/Portugal-HIVAIDS-Country-Profile-2011-revision-2012-final.pdf. Accessed December 29, 2017.

Carmo

. Infecção VIH e neoplasias não definidoras de SIDA: experiência de um centro. Acta Méd Port. 2014;27(2):181–190.

Honeycutt

Hoerger

Hardee

Brown

Smith

. RTI International. An Assessment of the State of the Art for Measuring Burden of Illness. Washington, DC: US Department of Health and Human Services; 2011. http://aspe.hhs.gov/sp/reports/2011/BurdenOfIllness/index.pdf. Accessed December 29, 2017.

Carvalho

. A situação da epidemiologia em portugal: o que sabemos e o que precisamos saber? Portugal: Virus de Imunodeficiência Humana (VIH); 2011. http://slideplayer.com.br/slide/1767652/. Accessed December 29, 2017.

Brandão

Cachão

Carvalho

. Manifestações Cutâneas da SIDA: Revisão da Casuística (1987-1991) e Aspectos Particulares de Algumas Dermatoses. Portugal: Reuniões da Sociedade Portuguesa de Dermatologia e Venereologia (SPDV); 1992.

Monteiro

Guimarães

Albuquerque e Sousa

. Três anos de SIDA: Experiência do hospital curry cabral com as infecções pelo VIH (1985-1988). Acta Méd Port. 1989;2(6):270–275.

Kore

Kanwar

Vinay

Wanchu

. Pattern of mucocutaneous manifestations in human immunodeficiency virus-positive patients in North India. Indian J Sex Transm Dis. 2013;34(1):19–24.

10.

Zancanaro

McGirt

Mamelak

Nguyen

Martins

. Cutaneous manifestations of HIV in the era of highly active antiretroviral therapy: an institutional urban clinic experience. J Am Acad Dermatol. 2006;54(4):581–588.

11.

Han

Lun

Meng

. Mucocutaneous manifestations of HIV-infected patients in the era of HAART in guangxi zhuang autonomous region, China. J Eur Acad Dermatol Venereol. 2013;27(3):376–382.

12.

Huang

Chen

. Clinical analysis of skin lesions in 796 Chinese HIV-positive patients. Acta Derm Venereol. 2011;91(5):552–556.

13.

Da Costa

Domingues

Castro

Exposto

. Úlceras genitais causadas por infecções sexualmente trasmissíveis: actualização do diagnóstico e terapêuticas, e a sua importância na pandemia do VIH. Acta Méd Port. 2006;19(4):335–342.

14.

Cunha

Pacheco

Rodrigues

Cardoso

Pinto Ferreira

Malhado

. Sífilis em doente com infecção pelo VIH. Acta Méd Port. 1997;10(6-7):503–507.

15.

Morar

Willis-Owen

Maurer

Bunker

. HIV-associated psoriasis: pathogenesis, clinical features, and management. Lancet Infect Dis. 2010;10(7):470–478.

16.

Guerra Rodrigo

. Infecção pelo vírus da imunodeficiência humana: perspectiva dermatológica. Acta Méd Port. 2002;15(4):293–309.

17.

Amerson

Maurer

. Dermatologic manifestations of HIV in Africa. Top HIV Med. 2010;18(1):16–22.

18.

Blanes

Belinchón

Merino

Portilla

Sánchez-Payá

Betlloch

. Current prevalence and characteristics of dermatoses associated with human immunodeficiency virus infection. Actas Dermosifiliogr. 2010;101(8):702–709.