Extramammary Paget Disease (EMPD): Mohs Indications & Treatment

Extramammary Paget disease (EMPD) is a rare cutaneous adenocarcinoma that arises in apocrine gland-bearing skin outside the breast. First described by Crocker in 1889 as the extramammary counterpart of mammary Paget disease, EMPD accounts for approximately 6.5% of all Paget disease cases. The estimated incidence is approximately 0.1-0.4 per 100,000 person-years, making it a genuinely rare neoplasm that most dermatologists and Mohs surgeons will encounter infrequently during their careers. The vulva is the most common site (65-70% of cases), followed by the perianal region (15-20%), groin/inguinal folds (5-10%), axillae (3-5%), and penile/scrotal skin (3-5%). EMPD predominantly affects postmenopausal women (median age 65-70 years), with a female-to-male ratio of approximately 3-5:1 for vulvar disease. However, the sex distribution varies significantly by anatomical site. Perianal EMPD shows a male predominance, and penile/scrotal EMPD is by definition male-exclusive. There is a higher incidence reported in Asian populations, particularly Japanese and Chinese cohorts, where perianal EMPD is relatively more common compared to Western populations. Importantly, EMPD is classified as either primary (75-85% of cases), representing an in situ apocrine adenocarcinoma arising de novo in the epidermis, or secondary (10-25%), representing cutaneous spread from an underlying visceral malignancy. Most commonly colorectal, urogenital, or gynecological carcinoma. This primary versus secondary distinction is the single most important diagnostic determination in EMPD, as it fundamentally changes workup, management, and prognosis.

EMPD presents as an erythematous, eczematous, well-demarcated plaque on apocrine-bearing skin, often with scaling, crusting, erosion, or a velvety surface texture. Pruritus is the most common symptom, reported by 60-80% of patients, followed by burning, pain, and bleeding. The lesion may be moist or weeping, and can have areas of hyperpigmentation, leukoplakia, or excoriation from chronic scratching. One of the most clinically significant features of EMPD is the extraordinarily long delay between symptom onset and diagnosis. The average time from first symptom to definitive diagnosis is 2-4 years, and delays exceeding 5-10 years are well documented. This diagnostic delay occurs because EMPD closely mimics several common benign dermatoses: vulvar eczema, contact dermatitis, inverse psoriasis, tinea cruris, candidiasis, lichen simplex chronicus, and erythroplasia of Queyrat. The clinical rule that should prompt biopsy is any persistent, treatment-resistant eczematous plaque in the genital, perianal, or intertriginous region of an older patient, particularly when topical corticosteroids and antifungals have failed to produce sustained improvement. Biopsy technique matters: a 4-6 mm punch biopsy from the most indurated or erythematous area, extending to at least mid-dermis, is recommended. Multiple biopsies from different areas of a large lesion may be necessary to identify areas of dermal invasion. Vulvar EMPD may be multifocal, and careful examination under adequate lighting with a Wood lamp (which can sometimes enhance contrast of the involved skin) is advisable to map the clinical extent before biopsy.

The histopathologic hallmark of EMPD is the presence of Paget cells within the epidermis. Large, round to oval cells with abundant pale-staining (clear or lightly eosinophilic) cytoplasm, vesicular nuclei, and prominent nucleoli. Paget cells may be found as single cells scattered throughout the epidermis, in small clusters or nests, or in a pagetoid pattern spreading along the basilar epidermis and ascending through all epidermal layers. The background epidermis often shows acanthosis, and the underlying dermis may demonstrate a chronic inflammatory infiltrate. In primary EMPD, the disease is typically confined to the epidermis (in situ), though 10-20% of primary EMPD cases show dermal invasion at the time of diagnosis. When invasion occurs, it ranges from microinvasion (single cells or small nests in the papillary dermis) to frank invasive adenocarcinoma extending into the reticular dermis or subcutis. The depth of dermal invasion is a critical prognostic factor. Immunohistochemistry (IHC) is essential for confirming the diagnosis and, most importantly, for distinguishing primary from secondary EMPD. The standard IHC panel includes CK7, CK20, GCDFP-15 (gross cystic disease fluid protein-15), and additional markers as clinically indicated.

The distinction between primary and secondary EMPD is the most consequential diagnostic determination in this disease, as it fundamentally alters the malignancy workup, surgical planning, and prognosis. Primary EMPD (75-85% of cases) represents an in situ adenocarcinoma of apocrine gland origin arising de novo within the epidermis. The Paget cells in primary EMPD are believed to originate from either intraepidermal apocrine duct cells or pluripotent epidermal stem cells with glandular differentiation. Primary EMPD is predominantly an in situ disease with slow growth, and when invasive disease develops, it tends to progress through predictable stages of microinvasion before frank dermal involvement. Secondary EMPD (10-25% of cases) represents cutaneous extension (epidermotropic spread) of an underlying visceral malignancy. Most commonly colorectal adenocarcinoma (especially for perianal EMPD), urothelial carcinoma of the bladder or urethra, or cervical/endometrial carcinoma. In secondary EMPD, the Paget cells originate from the visceral tumor and migrate into the overlying epidermis, producing an identical clinical appearance to primary EMPD. The prognosis of secondary EMPD is determined by the stage and biology of the underlying visceral malignancy, which is often advanced at the time of EMPD diagnosis. Because the clinical and histopathologic distinction between primary and secondary EMPD on H&E alone can be difficult, IHC profiling (CK7, CK20, GCDFP-15, CDX2) is essential for every case. Once the IHC profile is established, the workup proceeds accordingly.

EMPD is one of the best indications for Mohs micrographic surgery among all skin tumors, and the rationale is compelling. The hallmark feature of EMPD that makes it ideally suited for Mohs is its extensive subclinical extension with irregular, asymmetric margins. Multiple mapping studies have demonstrated that the subclinical spread of EMPD routinely extends 2-5 cm beyond the clinically visible border, with some cases showing subclinical extension exceeding 7 cm. This subclinical extension is unpredictable. It does not follow a uniform radial pattern but instead spreads asymmetrically along the epidermis, making clinical margin estimation extremely unreliable. Wide local excision (WLE) with predetermined 1-2 cm margins, which is the traditional approach, has reported recurrence rates of 30-60% in the literature, reflecting the inadequacy of clinically estimated margins for this tumor. Mohs surgery reduces recurrence rates to approximately 8-16%, a substantial improvement attributable to the complete margin assessment that accounts for the irregular subclinical spread. However, standard Mohs with H&E frozen sections has important limitations for EMPD. Scattered individual Paget cells can be extremely difficult to identify on frozen H&E sections, particularly at low cell density at the tumor periphery. This has led to the development of immunostain-guided Mohs techniques that significantly improve tumor detection at the margin.

The majority of EMPD (80-90%) is diagnosed as in situ disease confined to the epidermis. However, 10-20% of cases show dermal invasion at the time of initial diagnosis, and invasive EMPD has a fundamentally different prognosis and management approach compared to in situ disease. Dermal invasion transforms EMPD from a locally persistent but indolent neoplasm into a potentially lethal disease with metastatic capability. There is no universally accepted AJCC staging system specific to EMPD. Several classification systems for invasive EMPD have been proposed based on depth of invasion, with the Ohara classification being one of the most commonly referenced. The depth of dermal invasion is the single most important prognostic factor for invasive EMPD. Microinvasive disease (invasion limited to the papillary dermis, depth less than 1 mm) carries an excellent prognosis with very low metastatic risk, while deep invasion into the reticular dermis or beyond carries a substantial risk of lymph node and distant metastasis.

While Mohs micrographic surgery or staged excision with CK7 margin control remains the definitive treatment for EMPD, several non-surgical and adjuvant therapies play important roles for patients who are not surgical candidates, have unresectable disease, recurrent in situ disease, or metastatic EMPD. These modalities should be viewed as complementary to. Not substitutes for. Surgical excision when surgery is feasible.

The prognosis of EMPD is strongly determined by the in situ versus invasive classification, depth of invasion for invasive disease, and whether the EMPD is primary or secondary. In situ primary EMPD has an excellent overall prognosis. It is a slow-growing neoplasm that rarely metastasizes when confined to the epidermis, and disease-specific mortality is under 1%. However, in situ EMPD has a frustratingly high recurrence rate regardless of treatment modality: 30-60% after wide local excision and 8-16% after Mohs/staged excision with immunostain-guided margins. Recurrences may appear years or even decades after initial treatment, underscoring the need for lifelong surveillance. Invasive primary EMPD has a substantially worse prognosis that is directly proportional to the depth of invasion. Microinvasive disease (less than 1 mm depth) retains a favorable prognosis with less than 5% metastatic risk. However, deeply invasive EMPD (greater than 3 mm) is associated with 5-year disease-specific survival rates of 30-50%, comparable to intermediate-thickness melanoma. Nodal metastasis further worsens prognosis, with 5-year survival dropping below 20% in patients with regional nodal involvement. Secondary EMPD prognosis is determined primarily by the stage and histology of the underlying visceral malignancy, not by the cutaneous disease itself.