Primary extramammary invasive Paget’s vulvar disease: what is the standard, what are the challenges and what is the future for radiotherapy?


Primary Extra-mammary Paget’s disease (EMPVD) is a rare form of skin cancer. Vulva is one of the most com- mon sites of involvement accounting for approximately 1 % of all cases [1, 2]. It usually occurs in postmeno- pausal women between 50 and 80 years of age [3]. EMPVD originates from basal epidermal cells and ex- tends beyond the apparent edges of the lesion. Surgery is the standard treatment [4], but even with the use of more radical procedures, local recurrence remains com- mon (15–61 % of cases) due to microscopic extension, positive margins, and multicentric disease [3, 5].

Despite the high recurrence rate, there is little consen- sus regarding the value of radiotherapy (RT) in EMPVD. Most prior studies have been derived from limited cases of single institution experience. The use of RT, alone or in combination with surgery in the management of EMPVD, warrants further investigation. The aim of this review was to evaluate the potential benefit of RT for EMPVD. 

Materials and methods

The key words used for the search were: «Extramam- mary Invasive Paget’s disease», «Paget’s vulvar disease», «Radiotherapy» and synonyms). A literature review was performed based on database search in: Cochrane Gynecological Cancer Group Trials Register, Cochrane Register of Controlled Trials (CENTRAL), MEDLINE and EMBASE up to September 2015.

The exclusion criteria were: 1) pre-clinical studies, 2) not English language and 3) studies with no or inappro- priate intervention (Fig. 1).

The search of the literature identified 34 papers. Twenty seven publications were excluded after the study of their summaries, as they are not related to «Primary Extramammary Invasive Paget’s Vulvar Disease: what is the standard, what are the challenges and what is the fu- ture for radiotherapy?».

The Quality rating of included studies was based on the COCHRANE RISK OF BIAS TOOL. The quality of the studies was medium (Figs. 2 and 3). All of the stud- ies were small in sample size. In general, study design was attributed to the rarity of the condition. 


From a review of all published studies up to September 2015, a total of 57 invasive EMPVD patients underwent RT (Tables 1 and 2). From the total of the read papers, we created an infographic that contains: 1) Geographical distribution of cases 2) The patient’s characteristics re- lated to the age at the time of diagnosis. 3) The charac- teristics related to the diameter of the tumor before radiotherapy (Fig. 4).

Besa et al. [6] reported the clinical course of 2 EMPVD patients that underwent definitive RT. They were treated to a total dose of 44–64 Gy in 23–28 fractions. The 

investigators concluded that dose greater than 50 Gy to those who are medically unfit for surgery and for organ preservation could be indicated. For the cases of EMPVD mixed with adenocarcinoma, the use of adju- vant postoperative RT in doses greater than 55 Gy should be considered because of the high risk of local re- currence after surgery alone.

Luk et al. [7] delivered adjuvant RT (60 Gy) in one EMPVD patient with stage III vulvar carcinoma. RT techniques included anteroposterior opposed photon fields to pelvis and vulva, then electron boost to right vulva with 10 MeV electron field. The patient had a very good local control but she died at 15 months after RT of distant metastasis.

Son et al. [8] treated 3 patients with EMPVD. Two of them received definitive RT. The first case received 55.8 Gy to the primary and the second received 81.6 Gy to the primary and 45.6 Gy to the inguinal nodal areas. The last case received adjuvant RT to a total dose of 55.8 Gy. The RT fields encompassed 2–3 cm radially the clinically visible disease. In all cases 6 MV photons and 9 MeV electrons were used.

Tanaka et al. [9] treated 2 EMPVD patients with de- finitive RT. Both received 60 Gy with electrons. The pa- tients were also treated with CO2 laser therapy and Mohs micrographic surgery respectively because of local recurrence. They had a complete response and a disease free survival of 3 and 7 years respectively.

Hata et al. [10] underwent RT to 12 EMPVD patients to the primary, pelvic and inguinal lymph nodes. The au- thors used anteroposterior opposed ports with 4–6-MV X-rays, followed by a local RT boost to the gross tumor site using 7–13 or 6–15 MeV electrons. The RT fields included the gross tumor or the tumor bed (including the positive surgical margin), with a 2–5-cm margin. A 

bolus with a 5–10-mm water-equivalent thickness was used to compensate for the skin surface dose. Total doses of 45–70.2 Gy (median 60) were delivered with a fraction size of 1.8–2.2 Gy. A total dose of 59.4–70.2 Gy (median 60.6) was given to the gross primary and meta- static inguinal lymph nodes. In case of positive surgical margin the patients received a dose of 45–64.8 Gy (median 50) to the tumor bed. The 2- and 5-year local progression- free rates were 91 and 84 %, respectively. The overall and cause-specific survival rates were 100 % for both at 2 years and 53 % and 73 % at 5 years, respectively. None of the patients who received RT to the local lymph node regions developed a recurrence and this suggests that inclusion of the draining groin lymph nodes into the RT field might be important in optimizing treatment outcome. No therapy- related Grade 3 or greater toxicity was observed.

Hata et al. [11] treated seven EMPVD patients with definitive RT therapy. Two cases were irradiated to the local tumor site and to the regional (pelvic and inguinal) lymph node area through antero-posterior opposed ports with 4–6 MV X-rays. This was followed by a local RT boost to the gross tumor site using 6–13 MeV elec- trons. The remaining five patients underwent local RT only to the tumor site, using 4–14 MV X-rays or 6– 15 MeV electrons. RT fields were set up to include the primary lesion with a 2–5 cm margin. The RT field was reduced after doses of 44–45 Gy, and total doses of 

59.4–70.2 Gy were subsequently delivered to the gross tumors, including enlarged metastatic inguinal lymph nodes. The overall local control rate was 71 % at both 3 and 5 years. The disease-free, and overall survival rates in all patients were 58, and 92 % at 3 years, and 46, and 79 % at 5 years, respectively. No therapy-related toxic- ities of grade 3 or greater were observed.

Cai et al. [2] delivered RT in five patients (1: preopera- tive, 4: postoperative). The radiation fields were set up according to the tumor size with a 2–3 cm margin. The total dose ranged between 57–60 Gy. During a follow-up of 7–169 months, the median overall survival was 70.8 months in invasive cases and 21.3 months in cases with adnexal adenocarcinomas. A higher local recur- rence was associated with the presence of a positive margin, with associated adnexal adenocarcinomas hist- ology, and presence of dermal invasion.

Hata et al. [12] delivered RT in a total of 14 EMPVD patients. The patients with regional lymph node metas- tases underwent RT with 4–15 MV X-rays to the local tumor site and the lymph nodal areas. A local RT boost to the gross disease or involved single node (45–80.2 Gy, median 60 Gy) was delivered using 6–13 MeV electrons. The patients without nodal involvement received RT to the tumor, using 4–15 MV X-rays or 6–15 MeV elec- trons. The prophylactic dose delivered to the regional lymph node area was 41.4–50.4 Gy. RT fields included the tumor, or the tumor bed with a 2–5 cm margin. A bolus with a 5–10 mm water equivalent thickness was used to compensate for the surface dose in the area of tumor. At a median follow-up period of 41 months, the local control was 88 and 46 % in 3 and 5 years respect- ively. The DFS rate was 54 % in 3 years and the overall survival was 62 % in 5 years. The authors [13, 14] found that tumor dermal invasion and presence of regional lymph node metastasis were the most significant prog- nostic factors for both distant metastasis and survival. Due to the fact that it is usually difficult to keep the bolus into close contact with the vulva, the tumor surface may have possibly been covered with lower RT doses than those prescribed, due to this technical difficulty. The au- thors [12–14] also concluded that there was no significant 

difference in the local recurrence rate between EMPVD patients treated with margins of >2 and ≤2 cm.

Itonaga et al. [15] studied the outcome of 7 EMPVD patients. All patients were treated with curative intent and they received a median dose of 50 Gy. The patients achieved complete response within the irradiated volume during a median follow-up period of 71.4 months. The 5-year locoregional progression-free survival and overall survival were 91.7 and 84.3 %, respectively.

Hata et al. [16] delivered adjuvant post-operative RT in 4 EMPVD patients. A median total dose of 59.4 Gy (range, 45–64.8 Gy) was delivered to the tumour bed with a margin of at least 2 cm. A bolus with a 5–10-mm water equivalent thickness was used to compensate for the surface dose. At a median follow-up period of 38 months all patients had local control. The OS and CSSrateswereboth92%in3years,and62and71%in 5 years, respectively. No therapy-related acute or late toxicities of grade ≥ 3 were observed. 

Karam et al. [17] used the data derived from the Sur- veillance, Epidemiology and End Results (SEER) pro- gram, and undertook a retrospective analysis of the treatments approaches and outcome of 92 EMPVD pa- tients. The authors found a significantly worse disease specific survival (DSS) associated with the use of radi- ation therapy even when combined with site directed surgery. The authors found that patients who underwent surgery alone had a most favorable prognosis with a mean DSS compared to patients who did not undergo surgery or RT (346.8 vs. 255.1 months, 95 % CI 221.1–289.2, p = 0.002), to patients who received RT alone (mean DSS 143.4 months, 95 % CI 119.2–167.5, p = 0.004) and with patients who underwent surgery and RT (mean DSS 120.6 months, 95 % CI 93.6– 147.6, p < 0.001). These outcomes may be explained by the fact that the use of RT was associated with lo- cally advanced stage or recurrent disease. Interpret- ation of this analysis is difficult because radiation 

fields, doses and schedules were not standardized be- tween patients.

Several studies [13, 14, 18–20] have reported various prognostic factors that predict the poorer outcome of EMPVD, including dermal invasion, lymph node metas- tasis, stage [13, 14], tumor coexisting histology of adeno- carcinoma [2] or vulvar carcinoma, positive surgical margins [2], HER-2/neu overexpression [19] and high expression of Ki-67 [20].


Wide surgical excision remains the standard therapeutic approach of EMPVD although this may not always dem- onstrate acceptable rates of local control [2]. Since the lesions are often multifocal and the margins are irregu- lar, involvement of microscopic margins occurs in ap- proximately 40 –75 % of patients following surgical excision [2]. Postoperative RT could be considered in presence of cases with positive surgical margin, lymph node metastasis, multifocal disease, associated adnexal adenocarcinomas.

However, surgery is sometimes not possible, because many patients are elderly, and complete excision can be difficult owing to the tumor location. Definitive RT as a first-line treatment used only in a small number of pa- tients could be an alternative choice in case of medical contraindication or surgical difficulties because of exten- sive disease, tumor location and perineal involvement [5].

The optimal radiation dose for EMPD has not been established [21]. With the limited data available in the studies, it is not possible to form any conclusions with regards to local control as a treatment and disease-free survival. Moreover, in practice RT doses, fields, tech- niques and fractionation may vary widely [22].

A radiation field that encompasses the gross tumor vol- ume with a 2–5 cm margin radially has been used in most patients [2]. RT of the inguinal, pelvic or para-aortic lymph nodal chains can be considered only for the cases with lymph node metastases within these areas [2, 10].