Involvement of the Female Genital Tract in Pemphigus Vulgaris

Malik, Mohsin BSc1; Ahmed, A Razzaque MD1

doi: 10.1097/01.AOG.0000185258.74177.91
Original Research

BACKGROUND: Involvement of the genital tract in women with pemphigus vulgaris has rarely been reported. We report 34 patients with pemphigus vulgaris who have involvement of the vulva, vagina, or both.

METHODS: We reviewed the clinical data on 34 women treated at a tertiary dermatology referral center from July 1988 to June 2005.

RESULTS: Of 34 patients, 21 had labial involvement, 3 had vaginal involvement, and 10 had both labial and vaginal involvement. The mean age of onset of pemphigus vulgaris was 49 years. All patients had pemphigus vulgaris involvement at multiple other sites, and all had other mucous membrane involvement, most commonly oral (97%). Recurrent episodes of genital pemphigus vulgaris were noted in 35%, with a mean of 6 recurrences (range 1–11). In all the patients the pemphigus vulgaris was controlled with systemic and local therapy. Long-term follow-up for a mean of 76 months (range 9–204) indicated that no long-term sequela occurred due to genital involvement.

CONCLUSION: Involvement of the female genital tract in pemphigus vulgaris is rare. Vulvar lesions occur more commonly than vaginal lesions. With appropriate topical and systemic therapy, patients can have full recovery with no sequelae.


Involvement of the female genital tract in pemphigus vulgaris is rare and generally occurs in the setting of severe disease, but patients can have full recovery without sequelae.

From the 1Department of Medicine, New England Baptist Hospital, Boston, Massachusetts.

Corresponding author: A. Razzaque Ahmed, MD, New England Baptist Hospital, 70 Parker Hill Ave, Suite 208, Boston, MA 02120; e-mail:

Article Outline

Pemphigus vulgaris is a chronic, autoimmune, blistering disease of the skin and mucous membranes. The reported incidence is 0.1–3.2 cases per 100,000 individuals per year, with a higher incidence in individuals of Ashkenazi Jewish descent.1 Males and females are affected equally. Pemphigus vulgaris had a grave prognosis before the advent of systemic corticosteroids, with a mortality rate of about 90%.2 With the use of systemic corticosteroids and immunosuppressive agents, the mortality rate is less than 10%.2

The typical lesions of pemphigus vulgaris are flaccid blisters that rupture easily and leave denuded, painful surfaces. The clinical profile may range from limited and localized involvement of the skin or mucous membrane to extensive involvement at multiple sites. Although any portion of the skin or mucous membranes may be involved, the majority of pemphigus vulgaris begins in the oral cavity. Most patients have involvement of some mucosal surface during the course of disease.3

The diagnosis of pemphigus vulgaris is made on the basis of clinical appearance, histology, and immunofluorescence studies. Histology demonstrates loss of cell-cell adhesion (acantholysis) above the basal layer of the epidermis. Immunoglobulin G is detected on keratinocyte cell surfaces by direct immunofluorescence in nearly all patients, although sera from pemphigus vulgaris patients contains antibodies that bind to keratinocyte cell surface antigens on indirect immunofluorescence. The pathogenic antibodies are directed against the keratinocyte cell surface molecules desmoglein 1 and 3. An immunoblot assay or enzyme-linked immunosorbent assay may also be used to detect antibodies.

Previous reports have described pemphigus vulgaris involving the esophagus, larynx, nasal mucosa, nails, and penis.4–7 However, data on involvement of the female genital tract in pemphigus vulgaris has been limited to a small number of case reports.8–16 This retrospective study presents data on the frequency of labial and vaginal involvement in 34 patients with pemphigus vulgaris, the duration of disease, responses to therapy, and details of their long-term clinical follow-up.

Back to Top | Article Outline


We reviewed the clinical data for 34 patients treated at a tertiary dermatology referral center in Boston, Massachusetts. In all patients the diagnosis of pemphigus vulgaris was based on clinical presentation, established by histology, and confirmed by direct immunofluorescence of perilesional tissue and indirect immunofluorescence study of their sera. Representative histology and indirect immunofluorescence of vaginal pemphigus vulgaris are shown in Figure 1. Cultures from genital lesions for bacteria, fungi, and herpes virus infection were done on all patients. Biopsies for diagnostic purposes were performed on the 12 patients who had genital lesions on initial presentation.

Patients were seen between July 1988 and June 2005. The following information was collected on each patient: 1) age at onset of pemphigus vulgaris, 2) participation in vaginal sexual intercourse, 3) time to onset of genital disease, 4) total duration of genital disease, 5) severity and symptoms of genital disease, 6) number of relapses if any, 7) distribution of genital lesions, 8) systemic treatment, 9) nongenital involvement, 10) length of remission of pemphigus vulgaris, and 11) total length of follow-up. Remission is defined as the absence of clinical disease after the discontinuation of all systemic therapy.

Severity was rated on an objective scale as follows: Mild Two lesions or fewer, total surface involved less that 5 × 5 mm, and asymptomatic; Moderate: More that two but fewer than seven lesions, total surface involved less than 10 × 10 mm, symptomatic without affecting activities of daily living or quality of life; Severe: Seven lesions or greater, total surface involved greater than 10 × 10 mm, symptomatic and affecting activities of daily living (ADL) or quality of life.

The distribution of lesions in each patient was entered onto a computerized anatomic composite (Photoshop 7.0.1, Adobe Systems Inc., San Jose, CA). This was done to facilitate the determination of frequency of distribution of genital lesions.

Local therapy consisted of topical care (soaks and topical corticosteroids) and corticosteroid injections. If lesions seemed infected when patients were initially evaluated, cultures for bacteria or viruses were done, and appropriate systemic antibiotic or antiviral therapy was started immediately. In addition, topical therapy with anti-infectious agents (mupirocin [Bactroban, GlaxoSmithKline Research, Triangle Park, NC], clotrimazole, and acyclovir) was initiated and local corticosteroid therapy was temporarily discontinued.

Patients were advised about topical care based on the extent of disease. Patients with extensive, widespread cutaneous disease were advised to soak in a bath tub. Patients with limited cutaneous disease who did not need a bath tub were advised to use a Sitz bath. The soaking was done in lukewarm water to which aluminum acetate powder (Domeboro, Bayer Corporation, Morristown, NJ) and chlorhexidine gluconate (Hibiclens, Regent Medical, Norcross, GA) were added. These additives were introduced into water used for soaks to dry oozing, weeping lesions and prevent infections. After soaking the patients were advised to pat dry and then apply topical corticosteroids. Low-potency creams (ie, triamcinolone acetonide 0.1%,) were used for patients taking systemic corticosteroids. High-potency creams (ie, clobetasol propionate 0.05%) were prescribed for patients not taking systemic corticosteroids.

Patients were instructed to soak 3 times daily when the lesions were severe. As lesions improved, the frequency of soaking was decreased. Some patients with multiple lesions or lesions near the urethra experienced significant pain and burning upon micturition. These patients were advised to use a can with a long spout (like those used in gardening) and pour water on the urethral area while micturating. The aim was to dilute the urine and thus reduce symptoms of pain or discomfort. Sometimes underwear adhered to vulvar lesions, particularly to oozing lesions. In such instances patients were advised to cover lesions with nonadhesive dressing (TELFA dressing, The Kendall Company, Mansfield, MA) before wearing undergarments. Patients were advised to avoid sexual activity, because it may be accompanied by trauma. If unavoidable, then significant and excessive lubrication was recommended.

Sublesional corticosteroid injections were used in the following conditions: 1) lesions not healing after aggressive topical therapy for 6–8 weeks; 2) multiple sites in the genital area; 3) lesions significantly affecting quality of life or ADL and causing significant pain, burning, or both upon micturition; 4) lesions not healing despite optimal systemic corticosteroid therapy. In most patients a dose of triamcinolone acetonide 15 mg/mL was used for the first injection. This concentration was achieved by diluting 1 ml of 40 mg/mL triamcinolone acetonide with 2.5 mL of commercially available lidocaine 2% with epinephrine 1:100,000. This dilution provides significant anesthesia and also causes vasoconstriction, which allows the injected corticosteroid to stay in the diseased tissue for longer periods. Before injection, the area was anesthetized by soaking gauze in lidocaine 2% with epinephrine 1:100,000 and leaving it over the lesions for 5–7 minutes. We avoided spray anesthetics because the force of the spray could enlarge the lesions, like a Nikolsky’s sign. Injections were sublesional and not intralesional. The purpose of the injection was to deliver the corticosteroid as close to the site of pathology as possible. Intradermal injections are likely to go beyond the epidermis and into the dermis or subcutaneous tissue. We inserted a 3/8"-long, 26-gauge needle 3–4 mm from the periphery of the lesion with the beveled edge of the needle facing the epidermis and slowly released the corticosteroid once the needle was below the lesion. Observation of blanching as the injection entered the tissue was used as an indication that the corticosteroid had been delivered to the correct depth and was not deeper into the dermis than needed. The injections were repeated after 2 weeks if complete reepithelialization had not occurred. Lesions showing no response were given repeat injections at the same dose or at an increased dose of 20 mg/mL. The concentration of subsequent injections was reduced to 10 or 7.5 mg/mL when lesions showed healing. In any one episode of injections, the total quantity of triamcinolone did not exceed 25 mg. Whenever practical, we tried to follow injections with aggressive topical therapy.

All the patients were initially treated with prednisone in doses from 40–120 mg/d (mean 70 mg). In addition, a steroid-sparing immunosuppressive agent was added. In 10 patients this combination helped control the disease. In 4 patients this combination was ineffective and was abandoned. Those patients were treated with intravenous immunoglobulin (IVIG).

We hypothesized that patients who were sexually active would have a longer duration of genital disease compared with those not sexually active. An unpaired t test was used to compare difference between means (InStat, GraphPad Software Inc., San Diego, CA). Institutional review board approval for the study was obtained from the Center for Blood Research.

Back to Top | Article Outline


The results are summarized in Table 1. The mean age of onset of pemphigus vulgaris in these 34 female patients was 49 years (range 18–80). Thirty-three patients were white, and 6 of these were Jewish. One patient was South Asian.

Cultures for bacteria, fungi, and herpes virus were negative and did not grow any pathogenic organisms. Genital involvement was present at the initial disease presentation in 12 patients. The mean length of time between initial disease presentation and genital involvement was 21 months (range 0–180). Twenty patients engaged in vaginal sexual intercourse during the follow-up period. All patients had involvement of pemphigus vulgaris at other mucosal sites in addition to genital involvement. Thirty-three (97%) had oral involvement. Twenty-four (71%) had cutaneous involvement in addition to mucosal involvement, whereas 10 (29%) had mucosal involvement only. Representative clinical lesions are shown in Figure 2. Genital involvement included vulvar and vaginal sites in 10 patients (29%), vulva only in 21 (62%), and vagina only in 3 (9%). Urethral and cervical involvement was not observed in any patient. Based on the computerized anatomic composite analysis, the most common site for labial involvement in this cohort was the inferior portion of the labia minora (Fig. 3). The frequency of the site of involvement decreases from the vaginal opening to the periphery of the vulva. The most common site for vaginal involvement in this cohort was the distal one-third of the vagina (Fig. 4). The mean severity rating was 1.8 (range 1–3).

The symptoms reported by the patients were burning and pain on micturition, discomfort or pain on intercourse (among those sexually active), adherence of undergarments to lesions, and foul odor. The degree of complaints varied among the patients and varied on a day-to-day basis in the same patient. Nine patients (26%) had significant complaints, were genuinely disturbed by their symptoms, and their ADLs and quality of life were affected. Eleven patients (32%) reported these symptoms, but their ADLs were not hindered or obstructed. Fourteen patients (41%) did not have significant symptoms associated with their lesions. The histology of the lesions from the 12 patients, on hematoxylin and eosin stain, demonstrated intraepidermal vesicle with acantholysis and an intact basal layer (or "tombstone" appearance). The dermis had a variable and mixed cellular infiltrate.

Twenty-six patients received systemic corticosteroids, 18 received dapsone, 13 received azathioprine, and 6 received cyclosporine. Other systemic agents were less frequently administered. Twenty-four patients had severe disease unresponsive to conventional therapy or had serious adverse effects to these therapies, and were consequently treated with IVIG according to a published protocol.17 All patients received topical therapy. Patients with moderate and severe disease were treated with sublesional injection of triamcinolone acetonide. In 65% of the lesions injected, 1 injection was sufficient to facilitate healing. In 25% 2 or 3 injections were required for healing. In 10% more than 3 repeated injections were necessary for healing. No postinjection scarring or atrophy was observed on long-term follow-up. Twelve patients had recurrence of genital disease after the initial episode, with a mean of 6 recurrences (range 1–11).

The median duration of genital disease in all patients was 12 weeks (mean 97, range 3–646). Twenty patients reported engaging in vaginal sexual intercourse, and this group had a median duration of genital disease of 19 (mean 125, range: 6–646) weeks. Fourteen patients reported that they did not engage in vaginal sexual intercourse, and this group had a median duration of genital disease of 10 (mean 34, range: 3–217) weeks. Because of small group-specific sample sizes a Wilcoxon rank-sum test was used to compare these 2 groups. This difference in total duration of vaginal disease was not statistically significant between patients who did and did not report engaging in vaginal intercourse (P = .116). The mean length of remission in all patients was 36 months (range 5–156). The mean duration of follow-up was 76 months (range 9–204). Examination at last follow-up visit indicated that there were no long-term sequela due to genital disease.

Back to Top | Article Outline


Previous authors reported labial, vaginal, as well as cervical involvement in pemphigus vulgaris patients.8–16 The patients described in the literature tended to have involvement of pemphigus vulgaris at multiple other sites, especially the oral mucosa. Batta et al8 suggested that genital involvement generally occurs in the setting of extensive involvement of pemphigus vulgaris at other sites. All patients in our cohort had pemphigus vulgaris at multiple other sites. All had mucosal involvement, most commonly oral (97%). Although the mean amount of time between the onset of pemphigus vulgaris and the start of genital disease was 21 months, 12 patients had genital symptoms at initial presentation, and total of 82% had onset of genital disease within 2 years. We cannot definitely explain why labial or vaginal involvement would tend to occur relatively early in the course of this chronic disease. Sison-Fonacier and Bystryn18 reported regional variations in the distribution of pemphigus antigen, with higher concentrations in the buccal mucosa, face, scalp, and axilla compared with other sites. This distribution mirrored the frequency of clinical lesions. Vaginal mucosa or labia were not specifically evaluated by these authors, although “groin” samples had lower concentrations of pemphigus antigen. Given that all our patients had multiple mucosal involvement, the distribution of desmoglein 3 in the vagina may be similar to that at other mucosal sites. The distribution of desmoglein 3 in the labia could conceivably also be higher than in surrounding areas in the perineum.

Other reports of genital involvement in pemphigus vulgaris ranged from patients with extensive pemphigus vulgaris at other sites to subclinical involvement of the vagina only. However, most patients in our cohort had severe pemphigus vulgaris based on extensive involvement and the need for multiple immunosuppressive agents. The severity of pemphigus vulgaris is further evidenced by the 24 patients who were unresponsive to corticosteroids and immunosuppressive agents and had to be treated with IVIG for control of pemphigus vulgaris. Thus, the authors believe that genital involvement occurs more frequently with severe pemphigus vulgaris.

Other authors have not reported long-term follow-up or recurrent genital pemphigus vulgaris. Although the majority of this cohort only had a single episode of genital involvement, 34% had recurrent disease. As would be expected, recurrent genital disease did cause a significantly longer duration of genital disease compared with those with a single episode.

Diagnostic biopsies were performed on 12 patients who presented on initial examination with genital lesions. Confirmational genital biopsies were not done on other patients because they had biopsies from other sites to confirm the diagnosis. The gynecologist who evaluated these patients did not see the necessity to do confirmational biopsies because genital involvement was considered an integral portion of the total clinical presentation. In light of extensive clinical disease at other sites, it was highly unlikely that the genital disease could have another diagnosis. Further indirect evidence that these were pemphigus lesions comes from the observation that treatment of pemphigus vulgaris improved or cleared these lesions.

In the opinion of the authors, the universal improvement of genital disease, without any sequelae, can in part be attributed to the importance of topical care and the use of sublesional corticosteroid injections. Although vulvovaginal lesions can clear with systemic therapy, high-dose corticosteroids plus an immunosuppressive agent are required for clearing. In the authors’ experience, topical care is often neglected or not adequately addressed in the management of most pemphigus vulgaris patients. Our patients were strongly encouraged to use topical care at least twice or thrice daily and regularly until lesions healed. Although our data do not show a statistically significant difference between sexually active and inactive patients, the sample size is small and may not be adequate to show a difference. Hence, we prefer to advise patients to avoid sexual activity, because trauma is capable of inducing new lesions and preventing existing lesions from healing.

In the cohort of pemphigus vulgaris patients seen at our institution, the prevalence of female genital involvement is 35%. However, we do not think that this would be the prevalence in the overall female pemphigus vulgaris patient population The reason is that we are a referral center and therefore treat patients with extremely severe disease or disease that is difficult to manage. In general, most pemphigus vulgaris patients do not have genital examinations routinely. Hence, the incidence or prevalence has not been documented. The patients were observed partly because of a desire to do thorough examinations and partly because of patient complaints. It is possible that the lack of description of genital involvement in female pemphigus vulgaris patients is due to underreporting or because patients are uncomfortable discussing it. Lesions in the oral cavity are frequent, because it is traumatized daily in the process of speaking and eating. The genital tract does not undergo a similar repetitive use on a frequent basis and may be less traumatized than other skin and mucosal sites in pemphigus vulgaris. Because it can rapidly improve with high-dose systemic corticosteroid therapy, it is possible that many patients may not have mentioned its involvement to their physician. The authors recommend that a gynecologist should be consulted for any female patient suspected to have genital involvement with pemphigus vulgaris.

Back to Top | Article Outline


1. Pisanti S, Sharav Y, Kaufman E, Posner LN. Pemphigus vulgaris: incidence in Jews of different ethnic groups, according to age, sex, and initial lesion. Oral Surg Oral Med Oral Pathol 1974;38:382–7.
2. Robinson JC, Lozada-Nur F, Frieden I. Oral pemphigus vulgaris: a review of the literature and a report on the management of 12 cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1997;84:349–55.
3. Scott JE, Ahmed AR. The blistering diseases. Med Clin North Am 1998;82:1239–83.
4. Palleschi GM, Cipollini EM, Lotti T. Development of oesophageal involvement in a subject with pemphigus vulgaris: a case report and review of the literature. J Eur Acad Dermatol Venereol 2002;16:405–8.
5. Hale EK, Bystryn JC. Laryngeal and nasal involvement in pemphigus vulgaris. J Am Acad Dermatol 2001;44:609–11.
6. Sami N, Ahmed AR. Penile pemphigus. Arch Dermatol 2001;137:756–8.
7. Engineer L, Norton LA, Ahmed AR. Nail involvement in pemphigus vulgaris. J Am Acad Dermatol 2000;43:529–35.
8. Batta K, Munday PE, Tatnall FM. Pemphigus vulgaris localized to the vagina presenting as chronic vaginal discharge. Br J Dermatol 1999;140:945–7.
9. Chan E, Thakur A, Farid L, Lessin S, Uberti-Benz M, James W. Pemphigus vulgaris of the cervix and upper vaginal vault: a cause of atypical papanicolaou smears. Arch Dermatol 1998;134:1485–6.
10. Dvoretsky PM, Bonfiglio TA, Patten SF Jr., Helmkamp BF. Pemphigus vulgaris and microinvasive squamous-cell carcinoma of the uterine cervix. Acta Cytol 1985;29:403–10.
11. Friedman D, Haim S, Paldi E. Refractory involvement of cervix uteri in a case of pemphigus vulgaris. Am J Obstet Gynecol 1971;110:1023–4.
12. Kaufman RH, Watts JM, Gardner HL. Pemphigus vulgaris: genital involvement. Report of two cases. Obstet Gynecol 1969;33:264–6.
13. Lonsdale RN, Gibbs S. Pemphigus vulgaris with involvement of the cervix. Br J Dermatol 1998;138:363–5.
14. Marren P, Wojnarowska F, Venning V, Wilson C, Nayar M. Vulvar involvement in autoimmune bullous diseases. J Reprod Med 1993;38:101–7.
15. Mikhail GR, Drukker BH, Chow C. Pemphigus vulgaris involving the cervix uteri. Arch Dermatol 1967;95:496–8.
16. Zosmer A, Kogan S, Frumkin A, Dgani R, Lifschitz-Mercer B. Unsuspected involvement of the female genitalia in pemphigus vulgaris. Eur J Obstet Gynecol Reprod Biol 1992;47:260–3.
17. Ahmed AR, Dahl MV. Consensus statement on the use of intravenous immunoglobulin therapy in the treatment of autoimmune mucocutaneous blistering diseases. Arch Dermatol 2003;139:1051–9.
18. Sison-Fonacier L, Bystryn JC. Regional variations in antigenic properties of skin: a possible cause for disease-specific distribution of skin lesions. J Exp Med 1986;164:2125–30.
© 2005 The American College of Obstetricians and Gynecologists