Clinical Spectrum of Oral Secondary Syphilis in HIV-Infected Patients

Background. Oral lesions may constitute the first clinical manifestation in secondary syphilis, but detailed descriptions in HIV-infected individuals are scarce. Objective. To describe the clinical characteristics of oral secondary syphilis in HIV-infected patients and its relevance in the early diagnosis of syphilis. Methods. Twenty HIV/AIDS adult subjects with oral secondary syphilis lesions presenting at two HIV/AIDS referral centers in Mexico City (2003–2011) are described. An oral examination was performed by specialists in oral pathology and medicine; when possible, a punch biopsy was done, and Warthin-Starry stain and immunohistochemistry were completed. Intraoral herpes virus infection and erythematous candidosis were ruled out by cytological analysis. Diagnosis of oral syphilis was confirmed with positive nontreponemal test (VDRL), and, if possible, fluorescent treponemal antibody test. Results. Twenty male patients (median age 31.5, 21–59 years) with oral secondary syphilis lesions were included. Oral lesions were the first clinical sign of syphilis in 16 (80%) cases. Mucous patch was the most common oral manifestation (17, 85.5%), followed by shallow ulcers (2, 10%) and macular lesions (1, 5%). Conclusions. Due to the recent rise in HIV-syphilis coinfection, dental and medical practitioners should consider secondary syphilis in the differential diagnosis of oral lesions, particularly in HIV-infected patients.


Background
In the USA, since 2001 a resurgence in syphilis incidence, especially among males who have sex with men (MSM), has been observed [1]. By 2004, more than a half of new cases reported of primary and secondary syphilis were estimated to occur in MSM, with a high rate of HIV coinfection [2,3]. During 2007-2008, the total number of cases of syphilis reported to the CDC increased 13.1% [4].
e increasing incidence of syphilis reported in many studies in the last decade, especially among MSM, is clearly a marker for higher risk behavior in this population and raises concerns for a parallel increase in HIV transmission [5][6][7][8].
It has been suggested that HIV infection modi�es the clinical presentation of syphilis with greater organ involvement, atypical and �orid skin rashes, and more rapid progression to neurosyphilis [9][10][11][12]; consequently, the diagnosis of syphilis in HIV-infected individuals represents a challenge for care providers. A careful sexual exposure history, recognition of clinical signs and symptoms, and interpretation of diagnostic testing are crucial in this context [13].
Considering that the incidence of syphilis is increasing in most countries, particularly in HIV-infected individuals [9] and that oral lesions may constitute the �rst or most �orid clinical manifestation of the disease, the aim of the present case series is to describe the clinical spectrum of oral secondary syphilis in HIV-infected patients and its relevance in the early diagnosis of systemic syphilis.

Material and Methods
We report a series of 20 consecutive adult subjects with con�rmed HIV infection who presented oral secondary syphilis lesions. Patients attended during the period from 2003 to 2011 two HIV/AIDS referral centers in Mexico City: the Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán (INCMNSZ), a third level hospital, and the Clínica Especializada Condesa (CEC), a primary care center dedicated to the health care of HIV/AIDS patients who do not have social security. During the study period, between both institutions, approximately 2,000 active patients were seen.
Specialists in oral pathology and medicine, at the same AIDS clinic visit in both referral centers, consecutively evaluated all HIV-infected patients. During these visits, a structured oral examination was conducted by two of the authors (V. R.-Amador and G. A.-Saavedra). Syphilis-related oral lesions were classi�ed according to previous criteria [35,36] as follows.
(i) Macular lesions: �at-to-slightly raised, �rm, red lesions with particular predilection of the hard palate. (ii) Papular lesions: red, raised, �rm round nodules with a grey center that may ulcerate, usually located on the buccal mucosa or commissures. (iii) Mucous patches: slightly raised and covered by a grayish, white pseudomembrane, surrounded by erythema. Lesions appear mainly on the so palate and pillars, tongue, and vestibular mucosa. (iv) Shallow ulcers: oval erosions or shallow ulcers of about one cm in diameter, covered by a grey mucoid exudate with an erythematous border.
In order to rule out other oral lesions, lesions suspected as oral syphilis were further evaluated with cytological smear. In macular and papular lesions, it was necessary to consider in the differential diagnosis oral candidosis (OC), particularly the erythematous type. us, a cytological smear was taken, and OC was excluded by the absence of Candida sp. in periodic acid Schiff (PAS) stained smears and/or the lack of an antifungal treatment response.
Likewise, in cases of shallow ulcerations located on keratinized oral mucosa, the diagnosis of intraepithelial human herpesvirus infection was ruled out by the absence of virusinfected cells in cytologic smears stained with Papanicolaou, and/or a lack of clinical response to systemic antiviral therapy with acyclovir.
A punch biopsy from oral lesions was performed in six patients who gave their consent to biopsy. For hematoxylin and eosin (H&E) stain, 5 m sections were cut from formalin-�xed and paraffin-embedded tissue samples. e silver nitrate-based staining method (Warthin-Starry) was used to identify spirochetes.
Immunohistochemistry was performed using a standard avidin-biotin peroxidase complex technique. Sections were stained with a primary polyclonal antibody against Treponema pallidum; antigen-binding sites were visualized through the peroxidase system (Biocare Medical, 4040 Pike Lane, Concord, CA 94520, USA, rabbit polyclonal CAT CP 135B).
In all patients diagnosis of syphilis was con�rmed with a positive nontreponemal test at any titer (Venereal Disease Research Laboratory (VDRL) and if possible, a con�rmatory �uorescent treponemal antibody test (FTA-ABS).
At diagnosis, medical records were reviewed for information about demographic and clinical characteristics which included risk behavior for HIV transmission, clinical stage [37], tobacco and alcohol use, CD4 + lymphocyte counts, plasma HIV-RNA level, and current antiretroviral therapy.
Description of variables was done using the Statistical Program for Social Sciences (SPSS) package, with a 95% con�dence level. e prevalence of clinical and laboratory characteristics in patients was reported as percentages.

Results
Twenty male adult patients, median age of 31.5 (range 21-59) years with oral secondary syphilis lesions, were identi�ed during the study period, 10 at the AIDS Clinic of the INCMNSZ and 10 at the CEC. e clinical characteristics are shown in Table 1; all patients were MSM, of whom 14 (70%) were receiving HAART, with a median time of use of 32.5 (1-134) months. Of the six patients without HAART, �ve were naive to antiretrovirals and one had discontinued HAART because of intolerance. e 20 cases were con�rmed with positive serologic tests (VDRL in all, FTA-ABS in 15). It is important to mention that CEC is a free governmental HIV/AIDS clinic, so, because of �nancial issues, FTA-ABS was not performed in �ve patients; however, in three of them, besides a positive VDRL result, a biopsy led us to con�rm the diagnosis. In the other two patients, in addition to a positive VDRL test, the clinical picture was characterized by oral mucous patches which is considered the most typical oral manifestation of secondary syphilis.
In 16 (80%) cases, oral lesions were either the �rst or most �orid clinical sign that led to the diagnosis of secondary syphilis. In the remaining four patients (20%), oral manifestations were part of a systemic clinical picture already diagnosed as secondary syphilis by the treating clinician. Eleven of the 20 patients (55%) had mucocutaneous involvement documented in the medical records. Skin manifestations included disseminated maculopapular rash ( = 5), hyper pigmented macules/erythematous plaques on palms and soles ( = 3), eyebrows/eyelashes alopecia ( = 2), and scalp alopecia (1).
e most common clinical manifestation of oral secondary syphilis was mucous patch in 17 cases (85.5%), the so palate and pillars being the most frequent sites . Mucous patches appeared as white slightly raised plaques on an erythematous base with a serpentine and white/reddish well de�ned outline (Figure 1). Two patients presented shallow ulcers (10%), and, in one (5%), macular lesions were the most �orid oral sign.
Ten (50%) of the 20 patients showed more than one type of syphilis-related oral lesion, concurrent mucous patches and papular lesions being the most frequent combination in four patients, followed by mucous patches and macular lesions in three, and shallow ulcers joint with mucous patches in two and with macular lesions in one (Table 2).
We found four cases (20%) of papular syphilitic lesions in our population, all accompanied with mucous patches and located on dorsal tongue. Macular syphilitic lesions were identi�ed on so and hard palate (3), dorsum of tongue (1), and labial mucosa (1). In papular and macular cases, the clinical presentation of the oral lesions mimicked erythematous candidosis (Figure 2(a)); the absence of Candida sp. hyphae in the PAS smear, and the lack of response to topical or systemic antifungal treatment, allowed us to rule out erythematous candidosis.
One case of mucous patch on the lateral portion of the tongue resembled hairy leukoplakia (Figure 2(b)), and in another one, lesions in both sides of the tongue seemed lichen planus. In both cases a biopsy was taken and based in the histopathological features; the �nal diagnosis of oral syphilis was con�rmed. In the patients who presented shallow ulcers on the palate (cases 1 and 13), the diagnosis of intraoral herpesvirus infection was ruled out as described in the Material and Methods section (Figure 3).
A biopsy of the oral lesions was taken in six patients who agreed with the procedure; biopsies were processed and stained with H&E and the silver nitrate-based staining method (Warthin-Starry). Microscopically, the lamina propria showed a diffuse (four out of six cases) and a perivascular (three out of six) lymphoplasmacytic in�ammatory in�ltrate. Occasional eosinophils and neutrophils were also present within the in�ltrate and in the epithelium, forming microabscesses. In one case an interface lichenoid pattern was seen. e six oral biopsies examined showed psoriasiform and spongiotic changes in the epithelium; only in one of these cases, a pseudoepitheliomatous hyperplasia was present. Obliteration of the vessels was evident in all cases. Warthin-Starry stain detected spirochetes in the epithelium in two of the six biopsies. Immunohistochemistry for T. pallidum revealed numerous spirochetes within keratinocytes, free in the stroma of the lamina propria and within the vessel walls in examined biopsies (Figure 4).

Discussion
In this paper we report twenty cases of oral secondary syphilis in HIV-infected patients evaluated in two AIDS referral centers in Mexico City, from 2003 to 2011. In most of these cases (80%) the oral lesions were the key clinical �nding that led to the diagnosis of secondary syphilis; in the remaining cases, oral �ndings were additional clinical signs to a wellestablished clinical picture.
All cases in this series were males who have sex with men, as it has been stated by several authors in most countries, who have found in recent years an increase in syphilis among MSM [38][39][40][41][42]. In addition, almost half of our patients were young, less than 30 years (45%), in agreement with previous reports [39,42]. In relation with CD4 + cell levels, it was found that only 7 (36.8%) of 19 patients had less than 350 cells/mm 3 and 6 (30%) had a VL higher than 4 log 10 copies/mL at the time of syphilis diagnosis. Different studies in HIV-infected patients with syphilis [10,[43][44][45][46][47] have reported a transient decrease in CD4 + cell count and an increase in VL.
It is important to highlight that during the study period no cases of oral primary or tertiary syphilis were observed. As it has been informed in HIV-infected patients [9,11,12], a higher rate of secondary syphilis is a common clinical �nding� oral lesions are usually seen in secondary disease in patients with HIV [12].
Regardless of HIV status, the best recognized and characterized oral manifestation of secondary syphilis is the mucous patch, as shown in this case series in which 85.5% of the patients showed the typical slightly raised plaques, mainly located on the palate and so pillars, which are the common sites described [35,50].
Although the clinical criteria for mucous patches lesions usually involve slightly elevated plaques and super�cial ulcers covered by a gray or white pseudomembrane (shallow ulcers) [36], we decided to consider shallow ulcers separately from mucous patches due to two main reasons: the frequent �nding of isolated shallow ulcers in our population and the remarkable clinical differences between both types of lesions. us, a clinical description for shallow ulcers was added and de�ned independently from mucous patches as oval erosions or shallow ulcers of about 1 cm in diameter, covered by a grey mucous exudate with an erythematous border.
In addition to mucous patches [14,17,48], shallow ulcers [14,17], papular [15,16] and macular lesions [19] have also been described in HIV-infected patients with oral syphilis. As described by Ortega et al. [14,17] we found four cases of shallow ulcers on palate, tongue, and labial mucosa. Also, as other authors have described [19,36], palate was the favorite site for macular syphilis lesions that were found in �ve cases. On the other hand, we found four (20%) papular cases, all accompanied with mucous patches and located on dorsal tongue, as it has been described by others [15,16]. e term papular refers to red, raised, �rm round nodules with a grey center that may ulcerate [36]; in this respect, we found that nodular lesions were covered by a depapillated surface, as described by Dalmau et al. [15] and Baniandrés-Rodríguez et al. [16].
e diagnosis of oral secondary syphilis lesions, called the great imitator [24,51], represents an important clinical challenge. In our report, erythematous candidosis was considered as the differential diagnosis in nine of macular and papular cases, and herpes simplex infection in two patients with shallow ulcers.
Oral lesions mimicking hairy leukoplakia (HL) or lichen planus (LP) were also observed in one patient each. It is important to explain that in order to con�rm the diagnosis of oral syphilis, besides the positive serology test, a biopsy and histopathological analysis were performed to rule out HL and LP. Hairy leukoplakia and leukoplakia-like plaque have been considered in the differential diagnosis of recent case reports of secondary syphilis in HIV seronegative patients [24,28,31]. Remarkably, secondary syphilis mimicking pemphigus vulgaris-like oral lesions has been illustrated in a recent case report [22]. e de�nitive diagnosis of syphilis is based on clinical data and complemented with serological and pathological studies. It is important to emphasize that all cases were con�rmed with a positive serological test and when possible with histopathological analysis. Demonstration of tissue spirochetes using silver stain (Warthin-Starry) can be difficult, nonspeci�c, and time consuming, whereas immunohistochemical technique to detect T. pallidum in tissue sections has proven to be sensitive and speci�c and may be very useful in formalin-�xed paraffin wax embedded tissues [52,53].
e present study has certain limitations that need to be taken into account when considering the diagnosis of oral secondary syphilis: �rst, not all patients gave their consent to the biopsy procedure, and second, there is a lack of con�rmatory FTA-ABS tests in �ve patients of one of the referral centers. However, all �ve patients had positive VDRL tests, also three had con�rmatory histopathological studies and two presented oral mucous patches, which are considered the most typical oral manifestations of secondary syphilis. Also, in doubtful cases a biopsy was done.
Although clinical features (particularly in cases of mucous patches) are highly suggestive of secondary syphilis, VDRL is a nonspeci�c blood test, so the de�nitive diagnosis must be con�rmed by tissue sample and�or FTA-ABS test. However, it must be considered in agreement with some authors [12,13,54] that in secondary syphilis, VDRL is 100% sensitive and that the interpretation of treponemal and nontreponemal serological tests should be equivalent in HIVinfected and non-HIV-infected individuals. Also, in relation with sexual behavior only information about sex preference (MSM or heterosexual) was collected.

Conclusions
Oral manifestations are frequent and protean in HIV-infected patients; lesions caused by syphilis may be easily mistaken for other common entities in these patients, so it is important to emphasize the role of a thorough oral examination in its early diagnosis.
Since syphilis is a treatable disease with a signi�cant potential for serious complications if not treated, a missed diagnosis can have serious consequences for the patient. In routine care, especially in developing countries, accurate diagnostic tools are not always available, so physicians should be conscious that oral lesions could be the only evident clinical manifestation of a common but complex disease like syphilis.
Clinicians should be aware of the variable manifestations of the disease and the need to suspect it in the presence of oral lesions. Based on the present case series, dental and medical practitioners should consider secondary syphilis in the differential diagnosis of white, ulcerative, popular, and nodular oral lesions, particularly in HIV-infected patients.
�on��ct of �nterests e authors declare that they have no con�ict of interests. ey do not have any �nancial and personal relationships with other people or organizations that could inappropriately in�uence their work.