Anatomic Subtype Differences in Extramammary Paget Disease

Key Points

Question

How do disease presentation and treatment in extramammary Paget disease (EMPD) differ by anatomic subtype?

Findings

This meta-analysis of 135 studies found that, despite typically being intraepidermal, vulvar disease was more likely treated with radical surgery. Nearly one-third of vulvar cases recurred, predominantly locally in the skin and mucosa and one-third of perianal EMPD cases recurred; however, one-third of those recurrences presented as regional or distant metastasis and penoscrotal EMPD was least likely to recur.

Meaning

Based on these anatomic subtype differences, updated practice may include less morbid treatments for vulvar EMPD and closer surveillance for local recurrence in vulvar EMPD and metastasis in perianal EMPD.

This meta-analysis examines demographic and tumor characteristics and treatment approaches for different extramammary Paget disease subtypes and presents recommendations for diagnosis and treatment.

Abstract

Importance

Extramammary Paget disease (EMPD) is a rare, highly recurrent cutaneous malignant neoplasm of unclear origin. EMPD arises most commonly on the vulvar and penoscrotal skin. It is not presently known how anatomic subtype of EMPD affects disease presentation and management.

Objective

To compare demographic and tumor characteristics and treatment approaches for different EMPD subtypes. Recommendations for diagnosis and treatment are presented.

Data Sources

MEDLINE, Embase, Web of Science Core Collection, and Cochrane Reviews CENTRAL from December 1, 1990, to October 24, 2022.

Study Selection

Articles were excluded if they were not in English, reported fewer than 3 patients, did not specify information by anatomic subtype, or contained no case-level data. Metastatic cases on presentation were also excluded.

Data Extraction and Synthesis

Abstracts of 1295 eligible articles were independently reviewed by 5 coauthors, and 135 articles retained. Reporting was in accordance with Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guidelines. The analysis was cunducted in August 2019 and updated in November 2022.

Findings

Most vulvar EMPD cases were asymptomatic, and diagnosis was relatively delayed (mean, 25.1 months). Although most vulvar EMPD cases were intraepidermal (1247/1773 [70.3%]), radical surgeries were still performed in almost one-third of cases. Despite this aggressive surgical approach, 481 of 1423 (34%) recurred, commonly confined to the skin and mucosa (177/198 [89.4%]). By contrast, 152 of 1101 penoscrotal EMPD cases (14%) recurred, but more than one-third of these recurrences were regional or associated with distant metastases (54 of 152 [35.5%]). Perianal EMPD cases recurred in one-third of cases (74/218 [33.9%]), with one-third of these recurrences being regional or associated with distant metastasis (20 of 74 [27.0%]). Perianal EMPD also had the highest rate of invasive disease (50% of cases).

Conclusions and Relevance

The diagnosis and treatment of EMPD should differ based on anatomic subtypes. Considerations for updated practice may include less morbid treatments for vulvar EMPD, which is primarily epidermal, and close surveillance for local recurrence in vulvar EMPD and metastatic recurrence in perianal EMPD. Recurrences in penoscrotal subtype were less common, and selective surveillance in this subtype may be considered. Limitations of this study include the lack of replication cohorts and the exclusion of studies that did not stratify outcomes by anatomic subtype.

Introduction

Extramammary Paget disease (EMPD) is a rare, highly recurrent cutaneous malignant disease of unclear origin. EMPD typically arises on apocrine-rich skin, most commonly the vulvar and penoscrotal skin, and less frequently in the perianal region. Primary EMPD refers to disease confined to cutaneous surfaces, whereas secondary EMPD is associated with an underlying visceral adenocarcinoma. Recent clinical practice guidelines recommend different treatment approaches for intraepidermal vs invasive EMPD.¹ One key difference is the appropriateness of topical treatment of primary intraepidermal disease when surgery is highly morbid and unnecessary. National Cancer Registry studies have revealed several factors associated with overall survival in EMPD, including male sex, vaginal extension, and primary radiotherapy.² However, the effect of anatomic subtype on disease presentation, treatment, and prognosis has not been fully elucidated.

The purpose of the current study was to (1) compare demographic and tumor characteristics, as well as treatment approaches, for different EMPD subtypes; and (2) present recommendations for diagnosis and treatment.

Methods

A systematic review of the diagnosis and treatment of EMPD subtypes was performed. A clinical librarian searched MEDLINE, Embase, Web of Science Core Collection, and Cochrane Reviews CENTRAL from December 1, 1990, to October 24, 2022, using variations of the phrase extramammary Paget’s disease and controlled vocabulary (MeSH and Emtree) when applicable. Reference lists were evaluated for eligible studies. Studies related to EMPD diagnosis, workup, treatment, or follow-up were included. Articles were excluded if they were not in English, reported fewer than 3 patients, related to the management of the underlying malignant disease alone, did not specify all anatomic locations, or contained no case-level data. The abstracts of 1295 eligible articles were reviewed by 3 coauthors (N.K, J.L.O., B.W., J.X.W., V.H.; eFigure in Supplement 1), and 135 articles were retained when (1) cases were limited to a single anatomic subtype (ie, perianal, vulvar, or penoscrotal) or (2) case-level data allowed distinction among these 3 anatomic subtypes (eTable in Supplement 1).^{3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137} Among the included articles, 113 (83%) included 10 or more cases of EMPD, suggesting reasonable cohort sizes considering the rarity of the disease. Cases that were metastatic on presentation were excluded. The review was not registered.

Statistical analysis was conducted using Microsoft Excel (version 16.6; Microsoft). Weighted average was calculated by multiplying the reported means by the number of cases and dividing the sum from the reporting studies by the total number of cases. Studies reporting values other than the mean (eg, the median) were excluded. A weighted standard deviation was calculated similarly. Analysis of variance analysis or paired t test was performed for grouped or paired mean values, respectively. χ² test was calculated for variables reported as frequencies. Reporting of this review was in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guidelines.

Results

Demographics and tumor characteristics for each EMPD subtype are summarized in Table 1. Though penoscrotal EMPD cases presented in considerably older patients (weighted mean age, 70.5 years; 95% CI, 70.4-70.6 years; P < .001), the difference was small compared with vulvar (weighted mean age, 67.8 years; 95% CI, 67.6-68.0 years) and perianal cases (weighted mean age, 68.years ; 95% CI, 68.1-68.9 years). Vulvar EMPD cases were larger in diameter (mean, 7.6 cm; 95% CI, 7.56-7.64 cm) than cases of penoscrotal and perianal EMPD (mean 5.8 cm; 95% CI, 5.7-5.9 and mean 5.3 cm; 95% CI, 4.9-5.7 cm, respectively; P < .001). Diagnostic delay, defined as the mean lesion duration prior to diagnosis, was nearly 2 years for vulvar and perianal EMPD (25.1 months; 95% CI, 24.1-26.1 months and 21.7 months; 95% CI, 20.9-22.5 months, respectively), and these were significantly more delayed compared with penoscrotal EMPD (mean 12.0; 95% CI, 12.0-12.0 months; P < .001). Pruritus was the most common presenting symptom for all anatomic regions, most frequent in penoscrotal EMPD (281 of 306 cases [91.8%]) and least frequent in vulvar EMPD (366 of 3573 cases [10.2%]; P < .001). Whereas 115 of 228 perianal EMPD (50.4%) were invasive, most vulvar and penoscrotal cases were confined to the epidermis (70% and 63%, respectively; P < .001). Microinvasive disease, defined as EMPD that is limited to the epidermis and papillary dermis, was reported in a minority of cases across subtypes (5.2%, 2.2%, 0 in vulvar, penoscrotal, and perianal cases, respectively). Nearly one-third of perianal cases (22/65 [33.8%]) were recurrent on presentation, compared with 150 of 627 vulvar cases (23.9%) and only 75 of 1073 penoscrotal cases (7.0%) (P < .001).

Table 1. Demographic and Tumor Characteristics of Extramammary Paget Disease by Anatomic Subtype.

Characteristic	No./total No. (%)			P value
	Vulvar	Penoscrotal	Perianal
Age, mean (95% CI), y	67.8 (67.6-68)	70.5 (70.4-70.6)	68.5 (68.1-68.9)	<.001
No. of cases	2361	1435	302
Sex
Female	100	0	132/302 (43.7)	<.001
Male	0	100	170/302 (56.3)
Lesion size, mean (95% CI), cm	7.6 (7.56-7.64)	5.8 (5.7-5.9)	5.3 (4.9-5.7)	<.001
No. of cases	410	217	17
Duration of lesion, mean (95% CI), mo	25.1 (24.1-26.1)	12.0 (12-12)	21.7 (20.9-22.5)	<.001
No. of cases	664	908	80
Pruritus as most reported symptom	366/3573 (10.2)	281/306 (91.8)	20/47 (42.6)	<.001
Intraepidermal	1247/1773 (70.3)	351/557 (63.0)	113/228 (49.6)	<.001
Microinvasive	92/1773 (5.2)	12/557 (2.2)	0/228
Dermal or beyond	434/1773 (24.5)	194/557 (34.8)	115/228 (50.4)
Recurrent on presentation	150/627 (23.9)	75/1073 (7.0)	22/65 (33.8)	<.001

Across all subtypes, the most common treatment was wide local excision: 730 of 1268 (57.6%), 1067 of 1213 (88.0%), and 162 of 195 (83.1%) for vulvar, penoscrotal, and perianal EMPD, respectively (Table 2). However, significant differences in treatment modalities were also noted. Radical excision, defined as surgery removing a large amount of normal tissue and potentially sacrificing normal function of tissues, was performed most commonly in vulvar EMPD, where 398 of 1268 cases (31.4%) were treated by radical vulvectomy. By comparison, only 17 of 195 perianal cases (8.7%) and no penoscrotal cases in the cohort underwent radical excision (P = .005). Mohs micrographic surgery (MMS) was seldom the primary surgical modality in the included cases, most frequently reported in penoscrotal disease (131/1213 [10.8%]). Topical therapy (imiquimod, 5-fluorouracil, and/or photodynamic therapy) was rare as monotherapy, with the highest frequency reported in vulvar disease (97 of 1268 vulvar EMPD [7.6%]). Radiotherapy was an uncommon primary treatment approach.

Table 2. Treatment Approaches by Anatomic Subtype^a.

Variable	No./total No. (%)
	Radical surgeryb	Wide local excision	Mohs micrographic surgery	Radiotherapy	Topical therapy
Vulvar	398/1268 (31.4)	730/1268 (58.0)c	22/1268 (1.7)	21/1268 (1.7)	97/1268 (7.6)
Penoscrotal	0	1067/1213 (88.0)	131/1213 (10.8)	6/1213 (0.50)	9/1213 (0.80)
Perianal	17/195 (8.7)	162/195 (83.1)	8/195 (4.1)	8/195 (4.1)	0

^a P = .005.

^b Radical surgery was defined as surgery that was nonsparing to critical functions of the genitalia or anus. Topical therapies included imiquimod, 5-fluorouracil, and photodynamic therapy.

^c Wide local excision: in the case of vulvar extramammary Paget disease, 245 of these cases were specified as skinning vulvectomy.

Overall rates of recurrence were lowest in penoscrotal EMPD (152/1101 [13.8%]), and similar among vulvar (481/1423 [33.8%]) and perianal disease cases (74/218 [33.9%], P < .001) (Table 3). For vulvar EMPD, most recurrences were local (177/198 [89.4%]). Distant metastasis in vulvar EMPD was rare (21/198 [10.6%]), and the liver was the most commonly affected organ. There were no reported cases in the sample of nodal recurrences in vulvar EMPD. Penoscrotal EMPD recurred as either local (98/186 [64.4%]), nodal (15/152 [9.9%]), or distant disease (39/152 [25.6%], typically in the liver, lung, or bone). Similarly, perianal EMPD recurred as either local (54/74 [73.0%]), nodal (3/74 [4.1%]) or distant (17/74 [23.0%]) disease.

Table 3. Recurrences by Anatomic Subtype^a.

Variable	No./total No. (%)
	Overall	Local	Nodal	Distant
Vulvar	481/1423 (33.8)	177/198 (89.4)	0	21/198 (10.6)
Penoscrotal	152/1101 (13.8)	98/152 (64.4)	15/152 (9.9)	39/152 (25.6)
Perianal	74/218 (33.9)	54/74 (73.0)	3/74 (4.1)	17/74 (23.0)

^a P < .001.

Discussion

The present systematic review and meta-analysis of case-level data on EMPD stratified by anatomic site reveals several clinically relevant differences in diagnostic delay, reported surgical interventions, and patterns of recurrence. These differences help generate subtype-specific recommendations that may improve disease recognition, surgical interventions, and surveillance by anatomic subtype.

Regarding diagnosis, this study found that vulvar EMPD, the most common EMPD subtype, remains underdiagnosed and may be overtreated with unnecessary, highly morbid radical surgeries for what is largely an epidermal disease.⁵¹ Regarding diagnosis, vulvar EMPD presented with significantly larger lesions and a longer delay between presentation and diagnosis. This may be due to several factors, such as the need for enhanced patient and clinician awareness of this tumor on the female genitalia. From the patient standpoint, vulvar EMPD may often be asymptomatic; thus patients may not present to a physician until disease is more advanced. Genital examination and referral to gynecology for a vulvovaginal examination are needed to improve detection of this subtype. Regarding treatment, consistent with prior reports,¹³⁸ we found that margin-controlled surgery was offered less often in this subtype, despite data showing lower recurrence rates with MMS compared with wide local excision or radical vulvectomy.¹ This may be due to referral patterns to gynecologic oncology first, despite evidence showing a higher rate of highly morbid short- and long-term complications with radical surgeries, including deep surgical site infections and permanent urinary and sexual dysfunction,^139,140 where the same is not reported in other surgical or nonsurgical modalities.^13,141,142 Alternatively, MMS may be offered less frequently than nonsurgical therapies, such as imiquimod cream. More recent data show that recurrence rates with MMS can be further reduced with the use of CK7 immunohistochemistry on frozen section tissue during MMS.³⁰ Because vulvar EMPD cases can be large, scouting biopsies or optical confocal microscopy evaluating for subclinical extension may help to reduce the morbidity of treatment while providing a chance of surgical cure. For epidermal-only cases, clinical expert guidelines have suggested the use of topical therapy as a primary treatment where surgery is overly aggressive,^1,72,142,143 or in the adjuvant setting where additional surgical resection to achieve clear margins would result in excess morbidity.^26,144,145 Multicenter, prospective studies are needed to further address treatment approaches in this disease. Consistent with the predominantly intraepidermal nature of vulvar EMPD, almost all recurrences were local, with the rest presenting as distant metastasis of the liver or bone. Close surveillance of the primary tumor site (external and internal genitalia) should be considered given that most recurrences in this subtype are in the surrounding skin and mucosa.

Penoscrotal EMPD, the second most common subtype, was more readily diagnosed. It presented with a pruritic rash in nearly all cases. A pruritic patch or plaque on the penoscrotal skin that is persistent or poorly responsive to treatment should raise suspicion for EMPD as a possibility. Recurrence rates were lowest in this anatomic subtype, which may be due to the nature of the disease or the fact that margin control surgery was more commonly offered. Because penoscrotal EMPD was 2 and 10 times more likely to develop nodal metastasis compared with perianal and vulvar EMPD subtypes, sentinel lymph node biopsy (SLNB) might be particularly useful in this subtype for detecting early microscopic nodal disease. However, current data on SLNB in penoscrotal EMPD show mixed results,^{4,52,146,147,148,149,150} and additional prospective studies are needed before wider adoption of SLNB procedure can be recommended. Surveillance in this tumor subtype could focus on both skin examination and imaging for early detection of recurrence in select cases.

Although perianal EMPD was the least common subtype, diagnostic delay was similar to that of vulvar subtype. This may be attributable to perianal EMPD also being asymptomatic in a high proportion of cases in addition to limitations on self examination of this body region. In contrast to vulvar and penoscrotal EMPD, where invasive disease was uncommon, perianal EMPD was invasive in half of the included cases, suggesting it may be the most aggressive subtype, and this is supported by the high nodal and distant metastasis recurrence rates. Additional contributing factors for poor outcomes in this subtype may also be the increased risk of associated colorectal malignant diseases.^151,152 Given the high nodal and distant metastasis rates, surveillance in this subtype could focus on frequent skin examinations and imaging, as with penoscrotal disease.

Limitations

Limitations of this study include that data were confined to studies reporting outcomes by anatomic subtype, and thus excluded those that reported mixed data by multiple anatomic subtypes and where case-level data could not be sorted. Due to the retrospective nature of the studies, follow-up data were inconsistent, and this may influence outcomes. Future cooperative oncologic studies may address this in a prospective fashion. Moreover, based on the data presented, delineation of outcomes by depth of invasion (including microinvasive disease) or treatment modality could not be performed. The effect of secondary EMPD could not be elucidated. Lastly, this systematic review was not registered.

Conclusion

This systematic review and meta-analysis found significant differences in the diagnosis and treatment of EMPD by anatomic subtype. Although vulvar disease is more commonly intraepidermal and less likely to metastasize than penoscrotal or perianal EMPD, it is more frequently treated with radical resection, which may constitute overly aggressive treatment. Margin-controlled techniques may have utility in reducing the likelihood of local regional recurrence and metastasis while limiting surgical morbidity, but more data on the relative benefits of different surgical treatments for the 3 EMPD subtypes are needed. Close surveillance for local recurrence in vulvar EMPD and metastatic recurrence in perianal EMPD and potentially selectively in penoscrotal EMPD is warranted.

Supplement 1.

eTable. Table summarizing included studies and risk of bias according to ROBINS-I

eFigure. Study Flow diagram

Supplement 2.

Data Sharing Statement