Extramammary Paget disease (EMPD) is a rare cutaneous neoplasm that arises in apocrine gland-rich areas other than the mammary glands (1-3). It originates from the apocrine gland cells and extends and proliferates within the epithelium. EMPD can be classified into primary and secondary EMPD manifestations (1). While primary EMPD arises as an intraepithelial neoplasm of the epidermis, secondary EMPD presents as an epidermotropic spread of malignant cells or a direct extension from an underlying internal malignancy. As the treatment strategies and prognoses for primary and secondary EMPDs differ, accurate diagnosis based on detailed histological evaluation is required (4). Most patients with primary EMPD have a favorable prognosis because the tumor grows slowly and is usually limited to the epidermis. However, approximately 10% of the cases exhibit aggressive behavior and progress to invasive disease, metastasizing to local lymph nodes and distant organs (5).

Primary vulvar EMPD (VPD) is the most common type of EMPD, accounting for approximately 1-2% of all vulvar neoplasms. VPD is usually diagnosed between 50 and 80 years of age, and the disease is more common in Caucasian women (6, 7). VPD is typically multifocal and occurs mostly in the labia majora. The neoplasms appear as ill-defined, patchy, erythematous, or eczema-like skin lesions with varying symptoms, most commonly pruritus. Approximately 16% of patients have underlying carcinomas or invasive VPD (8). Wide local excision (WLE) or simple vulvectomy is the standard treatment for VPD (1, 6). However, the frequent microscopic extension, indistinct borders, and multifocality associated with the procedure make it difficult to achieve negative resection margins (RMs), raising the need for more careful RM evaluation through microscopic examination (9, 10).

The recurrence rate of VPD after surgery is high, with a reported range of 20-70% and average rates of 35% and 33% for the intraepithelial and invasive types, respectively (11-13). This high recurrence rate necessitates repeated operations and mutilation of the vulva, leading to significant long-term complications and a high morbidity risk (9, 14). Several clinicopathological factors, including tumor location, size, invasion depth, RM status, associated malignancy, and treatment modality, have been found to be associated with recurrence (11, 14, 15). However, most previous studies were limited by small sample sizes, mixed patient populations, inconsistent treatment modalities, and short follow-up periods, making it difficult to reach a consensus on the predictive factors for recurrence and prognosis in patients with VPD.

To date, some studies have investigated the influence of the RM status on the recurrence risk in patients with EMPD. Some authors have reported lower VPD recurrence rates when a negative RM was achieved (16, 17). A recent case series regarding scrotal EMPD revealed that after achieving negative RM, no patient in the cohort experienced recurrence (18). In contrast, another study showed that a positive RM was strongly associated with a higher risk of recurrence in patients with EMPD, and the risk of a positive RM was significantly higher after wild local excision compared to after other surgical procedures (11). Similarly, the recurrence rate of VPD is high despite aggressive surgery. Although alternative therapeutic modalities, such as radiation therapy (RT) and photodynamic therapy, have emerged as treatment options for EMPD, their results are often variable (15).

In this study, to determine the factors predicting recurrence of VPD, we analyzed the clinical features, pathological characteristics, RM status, recurrence, and survival rates of patients with VPD. Our data highlight the clinical significance and prognostic implications of RM status in predicting recurrence and outcome, suggesting the necessity for adjuvant therapies in patients with VPD and a positive RM.

Patients and Methods

Case selection. This study (protocol number: 2023-07-155) was approved by the Institutional Review Board of the Samsung Medical Center (Seoul, Republic of Korea). Due to the retrospective nature of this study, the Institutional Review Board waived the requirement for investigators to obtain signed informed consent from the patients or their legal guardians. A pathology database was thoroughly searched for primary noninvasive or invasive EMPD involving the vulva or perineum. We identified 45 patients with primary VPD between March 2003 and February 2022. We analyzed a consecutive cohort of 45 patients with VPD who were treated at the Department of Obstetrics and Gynecology at the Samsung Medical Center (Seoul, Republic of Korea). None of the patients had a history of other malignancies. Moreover, none of the patients received preoperative neoadjuvant chemotherapy, RT, or concurrent chemoradiotherapy.

Clinicopathological data collection. The following clinical details were obtained from the electronic medical records of 45 patients with VPD: age at initial diagnosis, presenting symptoms, surgical treatment, post-operative adjuvant treatment, post-operative recurrence, recurrence-free survival (RFS) period, survival status at the time of the last follow-up, and overall survival (OS) period. Microscopic examination was performed on formalin-fixed, paraffin-embedded VPD specimens retrieved from the archives of the Department of Pathology and Translational Genomics at the Samsung Medical Center (Seoul, Republic of Korea). The histological diagnosis of VPD was established according to the World Health Organization Classification of Female Genital Tumors (19). Pathologically, VPD was defined as an in-situadenocarcinoma of the vulvar skin characterized by large, round epithelial cells scattered singly or arranged in small clusters along the epidermis. The tumor cells are predominantly concentrated in the lower layers of the epidermis but can extend upwards through the superficial layers as single cells or small groups. They possess abundant mucin-containing or vacuolated cytoplasm and large, pleomorphic nuclei with conspicuous nucleoli (19). Immunohistochemically, the tumor cells of primary VPD consistently express cytokeratin 7 (CK7), low-molecular-weight CK (LMW-CK), and carcinoembryonic antigen (CEA). All available hematoxylin and eosin-stained and immunostained slides obtained from these patients were examined by two board-certified pathologists specializing in gynecological oncology. The following pathological details were collected: tumor location, tumor size, dermal invasion, invasion depth, RM involvement, and safety distance.

Statistical analysis. Pearson’s chi-squared test, Fisher’s exact test, or the linear-by-linear association test was used to determine the association between RM involvement and the clinicopathological characteristics of patients with VPD. A univariate survival analysis was conducted to identify the significant factors predicting worse RFS and OS. Curves for the RFS and OS were generated using the Kaplan-Meier method, and disparities in the curves were analyzed using the log-rank test. The Cox proportional hazards model (95% confidence interval) with the backward elimination method (stepwise withdrawal of non-significant variables with p≥0.05 from the model) was used for the multivariate analysis of the RFS and OS. All statistical analyses were performed using IBM SPSS Statistics for Windows (version 23.0; IBM Corp., Armonk, NY, USA). The level of statistical significance was set at p<0.05.

Results

Clinicopathological characteristics. Table I summarizes the clinicopathological characteristics of the 45 patients who underwent surgery for VPD. The mean age at diagnosis was 63.6 years (range=37-78 years). Twenty-six patients (57.8%) were 62 years or older, and 19 (42.2%) were aged below 62 years. Regarding symptoms, 19 patients (42.2%) presented with pruritus. Erythematous erosions were observed over the vulva in seven (15.6%) patients. The remaining 19 patients (42.2%) did not report any symptoms. All patients underwent surgery, including WLE (31/45; 68.9%), simple vulvectomy (8/45; 17.8%), and radical vulvectomy (6/45; 13.3%). A pathological diagnosis was established based on the characteristic histological and immunophenotypical features of EMPD and the presence of dermal invasion. Twenty-seven cases (60.0%) were diagnosed as VPD confined to the epidermis, while 18 (40.0%) showed dermal invasion. Specifically, 12 cases (26.7%) had dermal invasion with a depth of <1 mm (microinvasive VPD), while in six cases (13.3%), the invasion depth measured ≥1 mm (invasive VPD). Regarding the location, the tumors were located in the unilateral vulval region in 12 patients (26.7%), the bilateral vulval region in 23 patients (51.1%), and the vulva and clitoris in 10 patients (22.2%). Eighteen patients (40.0%) had RM involvement, and 15 (33.3%) had a safety distance of more than 2 cm. Twelve patients (26.7%) received adjuvant RT. One patient (2.2%) with invasive VPD received adjuvant chemotherapy. Ten patients (22.2%) experienced recurrence. The mean RFS was 35.8 months (range=4.3-89.1 months). Four patients (8.9%) died of the disease. Forty-one patients (91.1%) survived to the end of the follow-up period, with a mean OS of 40.0 months (range=10.1-97.4 months).

Factors predicting RFS and OS. Table II summarizes the results of univariate and multivariate analyses for the RFS in patients with VPD. The univariate analysis revealed that RM involvement was the only significant factor predicting worse RFS (p=0.013; Figure 3A). At the multivariate level, RM involvement was retained in the model, indicating that this parameter was an independent predictor of the RFS (p=0.024; hazard ratio=4.859; 95% confidence interval=1.275-18.519). In contrast, age, type of surgical treatment, tumor laterality and size, dermal invasion, and the closest distance between the tumor and RM (safety distance) were not significantly associated with the RFS. Table III summarizes the results of univariate and multivariate analyses for the OS. In the univariate analysis, RM involvement (p=0.001; Figure 3B), adjuvant RT (p=0.038; Figure 3C), and post-operative recurrence (p=0.027; Figure 3D) were significantly associated with worse OS. However, these parameters did not independently predict the RFS at the multivariate level (p=0.277, p=0.253, and p=0.216, respectively). There were no significant differences in OS according to age, type of surgical treatment, tumor laterality and size, and dermal invasion.

Factors associated with RM involvement. Table IV summarizes the relationship between RM status and the clinicopathological characteristics. The type of surgical treatment was significantly associated with RM involvement. None of the six patients who underwent radical vulvectomy showed a positive RM, whereas approximately 37.5% (3/8) and 48.4% (15/31) of the patients who underwent simple hysterectomy and WLE, respectively, showed positive RM. The differences in the frequency of RM involvement among the surgery types were significant (p=0.036). In contrast, age, tumor laterality and size, and dermal invasion were not significantly associated with RM involvement.

Discussion

In this study, we found a significant association between the RM status and clinical outcomes regarding patients with VPD. A positive RM was significantly associated with worse survival in patients with VPD. The multivariate survival analysis revealed that a positive RM was the only independent predictor of poor RFS. We also observed a significant correlation between the type of surgical procedure and RM involvement. The post-operative recurrence rate was significantly higher in patients who underwent WLE and simple vulvectomy than in those who underwent radical vulvectomy. In addition, we demonstrated that post-operative RT could improve the prognosis of patients with VPD.

The borders of EMPD are extremely indistinct, making it difficult to determine the extent of the lesion (20). To ensure complete excision, a wide safety distance is recommended. However, the incidence of incomplete excisions with positive RMs was estimated to be relatively high (4.8-57.6%), which is likely attributable to clinically ill-defined tumor borders, unexpected tumor spread beyond the clinical tumor border, and multifocal distribution (20-22). Shaco-Levy et al. (23) observed that 40% of patients with VPD and positive RMs experienced post-operative recurrence compared to 18% of patients with negative RMs. Pierie et al. (4) reported a recurrence rate of 72% among 18 patients with EMPD and positive RMs and no recurrence among 14 patients with negative RMs. These results are consistent with our data, demonstrating significantly worse RFS in patients with positive RMs, with a hazard ratio of 4.859 following the multivariate analysis. Long et al. (11) analyzed the prognostic implications of RM involvement in 154 patients with EMPD and found that a positive RM was associated with a 3.5 times higher risk of recurrence. They also reported a strong correlation between the surgical procedure and the RM status. The patients who underwent WLE had a significantly higher rate of RM involvement (33.3%) than those who underwent other surgeries (3.4%). This finding is consistent with our observation that patients who underwent WLE (48.4%) and simple vulvectomy (37.5%) had significantly higher rates of positive RMs than those who underwent radical vulvectomy (0.0%). Altogether, these results indicate that obtaining a negative RM may be important in reducing the risk of recurrence in women who undergo surgery for EMPD. However, the choice of surgical procedure and treatment plan is highly individualized and depends on various factors, including the tumor extent, the patient’s overall health, and the surgeon’s expertise. WLE for VPD is often associated with positive RMs, which can result in recurrence. In contrast, radical excision with wide margins may deteriorate organ function and the quality of life of patients. Therefore, collaboration between qualified healthcare professionals is crucial to help determine the most appropriate multidisciplinary approach for each case.

RT appears to be a promising treatment option for EMPD. It may be used as first-line treatment in patients with unresectable EMPD or as post-operative therapy in patients with positive RMs. A previous study found that all patients with EMPD who received post-operative adjuvant RT achieved local control after a median follow-up of 38 months, although a small subset of patients developed distant metastases at 6-43 months after RT (4). In another study regarding 41 patients with genital EMPD who underwent RT with curative intent, the local progression-free and disease-free survival rates were 88% and 55%, respectively, at 3 years. The overall and cause-specific survival rates were 93% and 96% at 3 years and 68% and 84% at 5 years, respectively. RT is safe and effective for patients with EMPD. It can contribute to prolonged survival owing to good tumor control and is a promising curative treatment option.

Although the prognosis of noninvasive EMPD is favorable (24), once metastasis occurs, the prognosis worsens (25). Some chemotherapeutic regimens have been proposed for metastatic EMPD; however, no standardized treatments have been established. Although docetaxel monotherapy and low-dose 5-fluorouracil/cisplatin therapy are frequently used, their efficacies are limited (1). Therefore, prospective studies are warranted to identify predictive biomarkers and novel therapies for advanced EMPD. Given the biological resemblance between EMPD and breast carcinoma (1, 24), human epidermal growth factor receptor 2 (HER2) is recognized as a therapeutic target in patients with HER2-positive EMPD. HER2 over-expression has been reported in 15-60% of EMPD cases and correlates with EMPD progression (26, 27). Concerning anti-HER2 treatment for EMPD, lapatinib has been reported to be effective against HER2-positive EMPD (28, 29). Furthermore, trastuzumab alone or with cytotoxic agents has been reported to be effective against metastatic or advanced EMPD cases harboring pathogenic v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2 (ERBB2) mutation (30-35). Two antibody-drug conjugates (ADCs), trastuzumab deruxtecan (T-DXd) and trastuzumab emtansine (T-DM1), have been approved for the treatment of HER2-positive breast carcinoma (36). Concerning anti-HER2 ADCs for EMPD, there has been one case report in which a 68-year-old Asian man with ERBB2-amplified scrotal EMPD was treated with T-DM1 and achieved complete remission (30). A recent in vivostudy demonstrated that HER2 over-expression was more frequent in patients with invasive EMPD than in those with in situ or microinvasive EMPD. Treatment with T-DXd or T-DM1 was found to eradicate an EMPD patient-derived xenograft model with no recurrence observed after 10 weeks (24). HER2-targeted ADCs may be promising treatment options for patients with ERBB2-mutant or HER2-over-expressing EMPD. In addition, genomic and immunohistochemical profiling has revealed that the phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) is frequently mutated in EMPD, and the androgen receptor (AR) is over-expressed in the majority of the cases (37, 38). Successful treatment with an androgen deprivation regimen containing bicalutamide and leuprolide acetate has been reported in a single case of EMPD involving the pubis and inguinal region (39). Mutations in the phosphatidylinositol-3 kinase/Akt pathway indicate the potential for targeted therapeutic trials with newer generations of PIK3CA inhibitors, either alone or in combination with other agents (40). Patients with PIK3CA-mutant EMPD showing AR-positive or estrogen receptor (ER)-positive/HER2-negative immunophenotypes may benefit from combined treatment with PIK3CA inhibitors and anti-AR or anti-ER therapies (40). Recent studies have provided new insights into the molecular mechanisms underlying the development and progression of EMPD. Since the efficacy of conventional chemotherapies for advanced or metastatic EMPD is limited, novel therapies based on recent basic and clinical studies that shed light on the understanding of the pathogenesis of EMPD should be developed.

In conclusion, we investigated the clinicopathological and prognostic characteristics of patients with VPD to identify the predictive factors for recurrence and survival. We found that a positive RM was an independent predictive factor for worse RFS. Patients with VPD who underwent WLE and simple vulvectomy experienced post-operative recurrence more frequently than those who underwent radical vulvectomy. Additionally, we observed that adjuvant RT could improve the prognosis of patients with VPDs. We believe that our findings will aid in formulating treatment strategies and effective clinical decision-making for patients with VPD with regard to this highly understudied topic of concern within the fields of dermatologic oncology.

Acknowledgements

This work was supported by the Samsung Medical Center Grant (SMO1230291) and the National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIT) (2023R1A2C2006223).