Primary invasive carcinoma associated with penoscrotal extramammary Paget's disease: a clinicopathological analysis of 56 cases

Transcription

1 Sexual Medicine Primary invasive carcinoma associated with penoscrotal extramammary Paget's disease: a clinicopathological analysis of 56 cases Bo Dai*, Yun-Yi Kong, Kun Chang*, Yuan-Yuan Qu*, Ding-Wei Ye*, Shi-Lin Zhang* and Hai-Liang Zhang* Departments of *Urology and Pathology, Fudan University Shanghai Cancer Center, and Department of Oncology, Fudan University Shanghai Medical College, Shanghai, China B.D and Y.Y.K. contributed equally to this work. Objectives To investigate the clinicopathological features, therapeutic strategies, and prognostic factors of patients with penoscrotal invasive extramammary Paget s disease (EMPD). Patients and Methods We retrospectively collected clinical, pathological, and follow-up data of 56 men with invasive penoscrotal EMPD. Histopathological features of the primary skin lesion including tumour size, surgical margin status, depth of invasion and lymphovascular invasion were examined. Results The median age was 67 years and median longest diameter of lesion was 5 cm. All patients were treated with wide surgical excision and 22 patients with clinically positive regional lymph nodes underwent therapeutic regional lymph node dissection. At the end of the study, 44.6% of patients developed distant metastasis and 39.3% of patients had died from disease. Univariate analysis showed that patients with one of the following poor prognostic factors: depth of invasion of lower dermis or deeper, presence of lymphovascular invasion and regional lymph node metastasis at diagnosis, had significantly shorter cancer-specific survival time. Multivariate analysis found that depth of invasion was the only independent prognostic factor. Conclusion The prognosis of invasive EMPD is significantly associated with depth of invasion, lymphovascular invasion and regional lymph node status. More aggressive therapy and more rigorous follow-up should be recommended for patients with these poor prognostic factors. Keywords extramammary Paget s disease, neoplasm invasiveness, neoplasm metastasis, prognosis, treatment outcome Introduction Extramammary Paget s disease (EMPD) is a rare cutaneous malignancy that mostly arises in areas rich in apocrine sweat glands, e.g. the vulva, scrotum, perineum, perianal area and axilla [1 3]. Histologically, it is characterised by intra-epidermal proliferation of neoplastic cells identical to those seen in classical mammary Paget s disease [1 3]. Most patients with EMPD have a good prognosis, because the tumour cells grow slowly and spread only in the epidermis [1 3]. However, in some cases the tumour cells can present aggressive behaviour and invade the dermis and subcutaneous tissue [4 13]. Once invasion occurs, this disease should be defined as invasive EMPD or invasive carcinoma associated with EMPD [1,2]. Although radical excision may obtain comparatively good control of primary skin lesions of invasive EMPD, systemic metastases may arise in some cases and lead to poor outcomes [4,6,7,14 17]. Previous studies investigating the clinicopathological characteristics, management and outcome of EMPD suggested that invasive EMPD had a significantly higher metastatic rate and worse prognosis than non-invasive EMPD [4,6,7]. However, in those studies, patients with invasive EMPD constituted only a small part of the whole study population and factors that specifically influence the prognosis of invasive EMPD were not clarified [4,6,7]. Therefore, in the present study, we investigated the clinical and pathological features, therapeutic strategies, and prognostic factors of patients with invasive penoscrotal EMPD to improve knowledge of this rare disease. BJU International 2014 BJU International doi: /bju BJU Int 2015; 115: Published by John Wiley & Sons Ltd. wileyonlinelibrary.com

2 Dai et al. Patients and Methods From January 1980 to December 2012, 168 male patients with penoscrotal EMPD were diagnosed and treated at our institution. The diagnosis of EMPD was confirmed by biopsy in all patients. At diagnosis, a thorough physical examination and diagnostic tests, including cystoscopy, colonoscopy, chest X-ray, abdominal and pelvic CT were performed. Serum tumour markers including PSA, α-fetoprotein (AFP) and carcinoembryonic antigen (CEA) were measured in all patients who were diagnosed after year After reviewing the clinical records, pathological slides and follow-up data of these patients, we found 66 male patients with invasive penoscrotal EMPD at diagnosis. For the specific purpose of this study, we excluded eight patients who had synchronous or metachronous internal noncutaneous malignancies (two prostate cancers, four rectal cancers, one colon cancer and one lung cancer) and two patients who developed distant metastasis at diagnosis and did not undergo surgical excision of primary skin lesions. Thus, 56 patients were included in the present study. Among them, 16 cases were previously reported in two published articles focusing on the frozen section-guided wide local excision technique or the management of metastatic EMPD [14,18]. Informed consent was obtained from each patient and the study was approved by our Institutional Ethics Committee. Pathological slides of surgical specimens were reviewed by two experienced dermatopathologists. Histopathological features of the primary skin lesion including tumour size, surgical margin status, depth of invasion and lymphovascular invasion were examined. The depth of invasion was stratified into two groups: upper dermis (papillary layer) and lower dermis (reticular layer) or deeper. The pathological stage was determined according to the 7th edition of the American Joint Committee on Cancer (AJCC) TNM classification of skin appendageal carcinoma [19]. Because of the specific location of the lesions, we defined both inguinal and pelvic lymph nodes as regional nodes (as for penile cancer). Patients were routinely followed by their treating physicians. Routine follow-up consisted of physical examination, pelvic CT, abdominal ultrasonography and chest X-ray every 3 months during the first 2 years, every 6 months during the next 3 years, and annually after 5 years. Pearson s chi-squared test and Fisher s exact test were used to compare categorical variables. Survival was calculated by the Kaplan Meier method with the log-rank test to assess differences between groups. Multivariate analysis of prognostic factors was done using the Cox proportional hazard model. Due to the limited number of patients, only the variables that appeared to have a significant impact on survival by univariate analysis were included into the multivariate analysis. All statistical tests were done using SPSS 16.0 with a P < 0.05 considered to indicate statistical significance. Results The clinicopathological characteristics of the patients are summarised in Table 1. All the patients were male Chinese. Manifestations of invasive EMPD have a wide clinical spectrum. Patients may initially present with erythematous or eczematous-appearing patches in the scrotum and surrounding skin. The lesions then progress to thick plaques and ultimately to nodules or masses, sometimes with erosion and ulceration (Fig. 1). Pruritus, pain, burning or other local discomfort were reported by some patients. Dermatologists or GPs clinically misdiagnosed 40 (71.4%) patients with chronic eczema or dermatitis and treated them with different types of ointments for years. All patients received histopathological diagnoses after incisional biopsy of skin lesions. The median (range) time from the awareness of symptoms to diagnosis was 30 (3 120) months. Tumour markers were measured in 42 patients. In all, 38.1% (16/42) of the patients had elevated CEA Table 1 Clinicopathological characteristics of the 56 male patients with invasive penoscrotal EMPD. Characteristics Value Median (range): Age, years 67 (44 86) Delay in diagnosis, months 30 (3 120) Longest diameter of lesion, cm 5 (2 12) N (%): Lesion location: Scrotum and penile shaft 40 (71.4) Scrotum only 16 (28.6) Primary treatment: Lesion excision only 34 (60.7) Lesion excision + unilateral ILND 10 (18.9) Lesion excision + bilateral ILND 6 (10.7) Lesion excision + bilateral ILND + bilateral PLND 6 (10.7) Surgical margin status: Positive 16 (28.6) Intraepithelial tumour positive 15 (26.8) Invasive tumour positive 0 Intraepithelial and invasive tumour positive 1 (1.8) Negative 40 (71.4) Depth of invasion: Upper dermis (papillary layer) 21 (37.5) Lower dermis (reticular layer) or deeper Lower dermis (reticular layer) 23 (41.1) Subcutaneous tissue 9 (16.1) Deep extradermal structures (i.e., penis, testis, spermatic cord) 3 (5.3) Lymphovascular invasion: Yes 20 (35.7) No 36 (64.3) Pathological stage*: T1-2N0M0 33 (58.9) T3-4N0M0 1 (1.8) TanyN1-3 M0 Unilateral/bilateral inguinal lymph node metastasis 16 (28.6) Unilateral/bilateral pelvic lymph node metastasis 6 (10.7) Adjuvant therapy: Chemotherapy 13 (23.2) Radiation therapy 5 (8.9) *7th edition of AJCC TNM classification of skin carcinoma. 154 BJU International 2014 BJU International

3 Invasive penoscrotal extramammary Paget's disease Fig. 1 Clinical morphology of invasive EMPD. (A) Erythematous and eczematous change in penoscrotal and pubic area, with multiple reddish or brownish coalescing nodules. (B) Ill-defined, elevated plaque with cerebriform appearance in scrotal area. A the primary skin lesions, 22 patients with clinical evidence (enlarged or fine-needle aspiration biopsy positive) of regional lymph node metastasis underwent therapeutic regional LND, including 10 cases of unilateral inguinal LND (ILND), six cases of bilateral ILND, and six cases of bilateral ILND plus bilateral pelvic LND (PLND). Pathological examination showed that 16 (28.6%) patients had positive surgical margins. Among them, seven patients were treated with re-excision 2 weeks after the first surgery, five were treated with adjuvant radiation therapy, and the remaining four refused any further therapy. B Lymphovascular invasion was found in 20 patients and depth of invasion is listed in detail in Table 1 (Fig 2). In 22 patients who underwent therapeutic regional LND, 16 had confined inguinal lymph node metastasis and six had both inguinal and pelvic lymph node metastasis. Among them, 59.1% (13/22) of the patients received adjuvant chemotherapy and the remaining nine refused or were unfit for chemotherapy. The adjuvant chemotherapy consisted of four cycles of combined chemotherapy. The exact chemotherapy regimens were as follows: (1) 5-fluorouracil 500 mg/m 2 daily days 1 5, cisplatin 25 mg/m 2 daily days 1 3, every 3 weeks, in five patients; (2) docetaxel 75 mg/m 2 day 1, 5-fluorouracil 500 mg/m 2 daily days 2 5, cisplatin 25 mg/m 2 daily days 2 4, every 3 weeks, in six patients; (3) docetaxel 60 mg/m 2 day 1, cisplatin 25 mg/m 2 daily days 1 3, every 3 weeks, in two patients. After a median (range) follow-up of 39 (6 130) months, 44.6% (25/56) of the patients developed distant metastasis and 39.3% (22/56) of the patients had died from disease by the end of the study. The three most common sites of distant metastasis were the non-regional lymph nodes, bone and lung (Table 2). The treatment and outcome data of patients with recurrence and/or metastatic disease are summarised in Figure 3. levels ( 5 μg/l) at diagnosis. Prostate biopsies were taken in two patients with elevated PSA levels at diagnosis, showing no evidence of malignancy. In all patients, the primary skin lesions were treated with wide surgical excision of an area at least 2 cm distal from the tumour margin, deep to subcutaneous fat and up to the fascia. Frozen-section analysis was performed in lesions with ill-defined margins to guide the excision as described in our previous study [18]. According to the size and location, the skin defect was reconstructed by primary closure (44 cases), local scrotal flap (eight) or split-thickness graft (four). The split-thickness skin grafts were harvested from each anterior thigh. No prophylactic regional lymph node dissection (LND) or dynamic sentinel node biopsy was performed in our patients with cn0 disease. Within 1 month of the excision of Univariate analysis revealed that patients with one of the following poor prognostic factors: depth of invasion of lower dermis or deeper, presence of lymphovascular invasion and regional lymph node metastasis at diagnosis had significantly shorter cancer-specific survival (CSS) (Table 3, Figs 4 6). We further performed a multivariate analysis using the above three factors and found that depth of invasion was the only independent prognostic factor of CSS in patients with invasive EMPD (Table 4). Discussion To our knowledge, the prognosis and response to treatment for invasive EMPD have not been completely clarified and few studies with large number of patients have been reported. In the present study, we report the clinicopathological and survival data of 56 male patients with invasive penoscrotal EMPD initially treated by surgery. All these patients had no distant metastasis at diagnosis and underwent radical BJU International 2014 BJU International 155

4 Dai et al. A B Fig. 2 Histopathology of invasive EMPD. (A) Involving the upper dermis. (B) Infiltrating the subcutaneous fat tissue. (C) Extending to the dartos muscle of the scrotum. (D) With lymphovascular invasion. (haematoxylin and eosin, original magnification 200). C D Table 2 Summary of site of metastasis. Site of distant metastasis* N (%) Non-regional lymph node metastasis 15 (26.8) Bone 8 (14.3) Lung 5 (8.9) Liver 3 (5.4) Brain 3 (5.4) Muscle 2 (3.6) *Eight patients developed multi-organ metastasis. resection of primary skin lesions. Moreover, within 1 month after the primary skin lesion resection, therapeutic regional LND was performed in 22 patients with clinically positive regional lymph nodes. However, the prognosis is extremely poor, 44.6% (25/56) of the patients developed distant metastasis and 39.3% (22/56) of the patients had died from their diseases by the end of the present study. More importantly, our present data indicate that the prognosis of our patients was influenced by several clinicopathological factors, because patients with one of the following poor prognostic factors: depth of invasion of lower dermis or deeper, presence of lymphovascular invasion and regional lymph node metastasis at diagnosis had significantly shorter CSS than those without these adverse factors. In further multivariate analysis, we found that depth of invasion was the only independent prognostic factor of CSS in our present patients. As the penoscrotal area belongs to anogenital region, which is the most common site of EMPD, we think that the present results are representative of invasive EMPD. Previously Karam et al. [8] investigated the prognosis of 1439 patients with invasive EMPD using the Surveillance, Epidemiology and End Results (SEER) programme database and reported that a poor outcome was associated with older age and advanced stage. Due to the inherent limitation of the SEER database, the stage was only classified into three groups: localised, loco-regional spread and distant spread and no information about the depth of invasion or lymphovascular invasion of primary lesion was studied [8]. Therefore, the results from the SEER database study had very limited value for guiding clinical practice. Another large study of invasive EMPD was from a Japanese cohort and included 51 patients (41 scrotal and 10 vulvar cases) [5]. The authors focused on the clinicopathological evaluation and found that lymphatic or venous invasion, mitosis, and nodal metastasis were all significantly more frequent in patients with dermal invasion >1 mm, as compared to those with dermal invasion <1mm [5]. A severe limitation of their study is that they did not provide follow-up data. Therefore, the deeper depth of invasion could not be considered as a poor prognostic predictor directly. The first strength of our present study is that we collected detailed clinical, pathological and follow-up data of our patients and defined CSS as the primary endpoint. After statistical analysis, we found that depth of invasion, lymphovascular invasion and regional lymph node metastasis were three prognostic factors for our patients and depth of invasion was the only independent prognostic factor. The clinical significance of our findings include: i) a routine pathological report of EMPD should describe depth of invasion in detail and provide information about 156 BJU International 2014 BJU International

5 Invasive penoscrotal extramammary Paget's disease Fig. 3 Patient treatment, recurrence and outcome data. WE, wide excision; RLND, regional LND; LNM, lymph node metastases; TCM traditional Chinese medicine; DOD; die of disease. Invasive EMPD (56 cases) Initial treatment (56 WE, 22 RLND) Stage (pn1/pn0, 22/34) First recurrence, 31 cases (pn1/pn0, 20/11) No recurrence, 25 cases (pn1/pn0, 2/23) Systemic metastases, 20 cases (pn1/pn0, 15/5) Local recurrence and/or regional LNM, 11 cases (pn1/pn0, 5/6) Die from other disease, 6 cases (pn1/pn0, 0/6) Alive with no disease, 19 cases (pn1/pn0, 2/17) Chemotherapy, radiation, TCM treatment DOD (17 cases), alive with disease (3 cases) Only local recurrence (5 cases) WE of recurrence (5 cases) Only regional LNM (3 cases) RLND (3 cases) Local recurrence + regional LNM (3 cases) WE of recurrence + RLND (3 cases) Alive with no disease (5 cases) DOD due to systemic metastases (3 cases) Alive with no disease (1 case) DOD due to systemic metastases (2 cases) lymphovascular invasion; ii) for patients with poor prognostic features, more rigorous follow-up should be recommended and more aggressive adjuvant therapy should be considered. The second strength of our present study is that we excluded patients with synchronous or metachronous internal noncutaneous malignancies. According to the criteria proposed by Kanitakis [2], EMPD should be classified into two different forms due to distinct pathogenetic mechanisms. One form is the primary or cutaneous EMPD form, which seems to originate in the skin (epidermis or the underlying apocrine sweat glands) [2]. The other form is secondary form of EMPD, which is associated with an underlying (distant) adenocarcinoma and this form would represent an epidermotrophic metastasis of the underlying tumour [2]. It has been reported that the secondary form of EMPD has a worse prognosis than the primary/cutaneous EMPD [1,2,20]. By excluding patients with synchronous or metachronous internal noncutaneous malignancies, we only included the primary/cutaneous EMPD in the present study. Therefore, our present data reconfirmed that the primary/cutaneous EMPD, which is initially limited to the epithelium, can progress to an invasive tumour and eventually produce lethal metastases. EMPD is characterised by ill-defined margins and sometimes concomitant bacterial or fungal infection makes the clinical margins more indistinct. To obtain clear margins, we performed wide surgical excision of at least 2 cm beyond the boundary of all lesions and in difficult cases frozen-section analysis was used to guide the excision. However, positive surgical margins were still found in 28.6% of the patients. Recently, in a pooled analysis of surgery for EMPD, Bae et al. [21] reported that Mohs micrographic surgery was superior to wide local excision in obtaining a safety surgical margin and decreasing the rate of local recurrence. Therefore, we have begun to learn the Mohs micrographic surgery technique (beginning of this year). Due to the rarity of this disease, there is no published guideline about the adjuvant therapy for EMPD after primary lesion resection. Previous studies suggested that the local recurrence rate was higher in patients with positive surgical margins [10,22]. Therefore, we performed re-excision or adjuvant radiation therapy for our present patients with positive surgical margins and obtained comparatively good local control rate. Another retrospective study on vulvar EMPD also recommended that adjuvant radiation therapy BJU International 2014 BJU International 157

6 Dai et al. Table 3 Univariate analysis of variables affecting CSS. Variables Number of patients Died from disease, n (%) All patients (39.3) Age, years < (31.3) (50) Delay in diagnosis, months < (31.8) (44.1) Longest diameter of lesion, cm < (46.2) (33.3) Surgical margin status Positive 16 9 (56.3) Negative (32.5) Depth of invasion Upper dermis (papillary layer) 21 2 (9.5) <0.001 Lower dermis (reticular layer) (57.1) or deeper Lymphovascular invasion Yes (60) <0.001 No (27.8) Regional lymph node metastasis at diagnosis Yes 34 9 (26.5) <0.001 No (59.1) *Log-rank test. P* Fig. 5 CSS curves of patient subgroups by the lymphovascular invasion status. LI, without lymphovascular invasion; LI+, with lymphovascular invasion LI (n = 36) CSS, % LI+ (n = 20) Time, months Fig. 6 CSS curves of patient subgroups by the regional lymph node status. pn, without regional lymph node metastasis at diagnosis; pn+, with regional lymph node metastasis at diagnosis. 100 Fig. 4 CSS curves of patient subgroups by the depth of invasion. UD, depth of invasion of upper dermis; LD or deeper, depth of invasion of lower dermis or deeper. 100 UD (n = 21) 80 CSS, % pn (n = 34) CSS, % pn+ (n = 22) Time, months Time, months LD or deeper (n = 35) should be considered for patients with a positive margin after surgical excision for better local control [9]. It is widely accepted that adjuvant chemotherapy plays no role in non-invasive EMPD, which has a high curative rate. For invasive EMPD, to prevent distant metastasis and to improve survival, adjuvant chemotherapy was given to some patients in previous studies [4,5,7]. The rationale of using chemotherapy in the adjuvant setting may be due to the fact that systemic chemotherapy has been confirmed to produce complete or partial response in some patients with advanced EMPD [16,23,24]. In the present study, adjuvant chemotherapy was recommended for our patients with pn+ disease and altogether 59.1% (13/22) of the patients with pn+ disease received adjuvant chemotherapy. However, due to the limited number of patients, the value of adjuvant chemotherapy could not be evaluated at the end of the present study. Because there is no long-term effective treatment for disseminated metastatic EMPD, we think that adjuvant chemotherapy should be recommended for patients with invasive EMPD with poor prognostic features, young age and good performance status. 158 BJU International 2014 BJU International

7 Invasive penoscrotal extramammary Paget's disease Table 4 Multivariate analysis of variables affecting CSS. Variables HR (95% CI) P Depth of invasion (lower dermis or deeper vs 7.54 ( ) upper dermis) Lymphovascular invasion (yes vs no) 0.73 ( ) Regional lymph node metastasis (yes vs no) 3.59 ( ) HR, hazard ratio. For patients with invasive EMPD with clinically positive regional lymph nodes and no evidence of distant metastasis, therapeutic regional LND should be performed to take advantage of the curative opportunity [1,25]. However, the value of prophylactic LND for patients with cn0 invasive EMPD is still controversial [25,26]. Some researchers have suggested that elective LND should be recommended for patients with depth of invasion of reticular dermis or deeper [25]. Others suggest that LND should not be performed on a prophylactic basis in patients with clinically negative nodes [26]. Because ILND and PLND are associated with significant morbidities and there is no guideline support, we did not perform prophylactic LND in our patients with clinically negative nodes. During follow-up, six patients (Fig. 3) developed regional lymph node metastasis without distant metastasis and were treated by therapeutic LND. Five (83.3%) of them had died from disease by the end of the study. Because histopathological existence of the tumour is often recognised far before the clinically visible lesion, these patients would have had pathologically positive lymph nodes if they had undergone prophylactic LND. Experience from penile cancer also suggests that risk factors including histological grade and lymphovascular invasion are independent predictors for occult metastasis in clinically node-negative (cn0) patients [27]. However, in patients with invasive EMPD, there is no prospective study investigating this issue. In order to answer the question of whether patients will benefit from the prophylactic LND or not, we plan to perform prophylactic bilateral ILND in our patients with invasive EMPD with clinical N0 disease and poor prognostic factors in the future. The limitations of the present study are its relatively small cohort size, retrospective nature, and data from a single centre. Because it is almost impossible to conduct a large prospective study for such a rare disease, we think that our present findings can be used to discriminate patients with high-risk invasive EMPD from those with a comparatively good prognosis. In conclusion, our present data indicated that the prognosis of invasive EMPD is not significantly associated with the size of tumour but with the depth of invasion, lymphovascular invasion and regional lymph node status. More aggressive therapy and more rigorous follow-up should be recommended for patients with these poor prognostic features. Funding Sources This study was supported in part by grants from Guide Project of Science and Technology Commission of Shanghai Municipality (No a7300) and grants from Shanghai Natural Science (No. 12ZR ). Conflict of Interest None declared. References 1 Lam C, Funaro D. Extramammary Paget s disease: summary of current knowledge. Dermatol Clin 2010; 28: Kanitakis J. Mammary and extramammary Paget s disease. JEurAcad Dermatol Venereol 2007; 21: Shepherd V, Davidson EJ, Davies-Humphreys J. Extramammary Paget s disease. BJOG 2005; 112: Hatta N, Yamada M, Hirano T, Fujimoto A, Morita R. Extramammary Paget s disease: treatment, prognostic factors and outcome in 76 patients. Br J Dermatol 2008; 158: ShiomiT,NoguchiT,NakayamaHetal.Clinicopathological study of invasive extramammary Paget s disease: subgroup comparison according to invasion depth. J Eur Acad Dermatol Venereol 2013; 27: Lai YL, Yang WG, Tsay PK, Swei H, Chuang SS, Wen CJ. Penoscrotal extramammary Paget s disease: a review of 33 cases in a 20-year experience. Plast Reconstr Surg 2003; 112: Hegarty PK, Suh J, Fisher MB et al. Penoscrotal extramammary Paget s disease: the University of Texas M. D. Anderson Cancer Center contemporary experience. J Urol 2011; 186: Karam A, Dorigo O. Treatment outcomes in a large cohort of patients with invasive extramammary Paget s disease. Gynecol Oncol 2012; 125: Cai Y, Sheng W, Xiang L, Wu X, Yang H. Primary extramammary Paget s disease of the vulva: the clinicopathological features and treatment outcomes in a series of 43 patients. Gynecol Oncol 2013; 129: Choi YD, Cho NH, Park YS, Cho SH, Lee G, Park K. Lymphovascular and marginal invasion as useful prognostic indicators and the role of c-erbb-2 in patients with male extramammary Paget s disease: a study of 31 patients. J Urol 2005; 174: Siesling S, Elferink MA, van Dijck JA, Pierie JP, Blokx WA. Epidemiology and treatment of extramammary Paget s disease in the Netherlands. Eur J Surg Oncol 2007; 33: Tanaka R, Sasajima Y, Tsuda H et al. HER2 protein overexpression and gene amplification in extramammary Paget s disease. Br J Dermatol 2013; 168: Hata H, Abe R, Hoshina D et al. MUC5AC expression correlates with invasiveness and progression of extramammary Paget s disease. JEur Acad Dermatol Venereol 2014; 28: ZhuY,YeDW,YaoXDetal.Clinicopathological characteristics, management and outcome of metastatic penoscrotal extramammary Paget s disease. Br J Dermatol 2009; 161: Hendi A, Brodland DG, Zitelli JA. Extramammary Paget s disease: surgical treatment with Mohs micrographic surgery. J Am Acad Dermatol 2004; 51: Mochitomi Y, Sakamoto R, Gushi A et al. Extramammary Paget s disease/carcinoma successfully treated with a combination chemotherapy: report of two cases. J Dermatol 2005; 32: Matsushita S, Yonekura K, Mera K, Kawai K, Kanekura T. Successful treatment of metastatic extramammary Paget s disease with BJU International 2014 BJU International 159

8 Dai et al. S-1 and docetaxel combination chemotherapy. J Dermatol 2011; 38: Zhu Y, Ye DW, Chen ZW, Zhang SL, Qin XJ. Frozen section-guided wide local excision in the treatment of penoscrotal extramammary Paget s disease. BJU Int 2007; 100: Edge SB, Byrd DR, Compton CC et al. AJCC Cancer Staging Manual, 7th edn. New York, NY: Springer, Chapter Yoon SN, Park IJ, Kim HC et al. Extramammary Paget s disease in Korea: its association with gastrointestinal neoplasms. Int J Colorectal Dis 2008; 23: Bae JM, Choi YY, Kim H et al. Mohs micrographic surgery for extramammary Paget s disease: a pooled analysis of individual patient data. J Am Acad Dermatol 2013; 68: Wang Z, Lu M, Dong GQ et al. Penile and scrotal Paget s disease: 130 Chinese patients with long-term follow-up. BJU Int 2008; 102: Oguchi S, Kaneko M, Uhara H, Saida T. Docetaxel induced durable response in advanced extramammary Paget s disease: a case report. J Dermatol 2002; 29: Fujisawa Y, Umebayashi Y, Otsuka F. Metastatic extramammary Paget s disease successfully controlled with tumour dormancy therapy using docetaxel. Br J Dermatol 2006; 154: Tsutsumida A, Yamamoto Y, Minakawa H, Yoshida T, Kokubu I, Sugihara T. Indications for lymph node dissection in the treatment of extramammary Paget s disease. Dermatol Surg 2003; 29: Hatta N, Morita R, Yamada M et al. Sentinel lymph node biopsy in patients with extramammary Paget s disease. Dermatol Surg 2004; 30: Graafland NM, Lam W, Leijte JA et al. Prognostic factors for occult inguinal lymph node involvement in penile carcinoma and assessment of the high-risk EAU subgroup: a two-institution analysis of 342 clinically node-negative patients. Eur Urol 2010; 58: Correspondence: Ding-Wei Ye, Department of Urology, Fudan University Shanghai Cancer Center, DongAn Road 270, Shanghai , China. dwye@shca.org.cn Abbreviations: CEA, carcinoembryonic antigen; CSS, cancer-specific survival; EMPD, extramammary Paget s disease; (I)(P)LND, (inguinal) (pelvic) lymph node dissection; SEER, Surveillance, Epidemiology and End Results. 160 BJU International 2014 BJU International