A Scoping Review and Population Study Regarding Prevalence and Histopathology of Juvenile Vulvar Melanocytic Lesions. A Recommendation

Cases of vulvar melanocytic lesions in juveniles are rarely reported. We analyze the evidence regarding vulvar melanocytic lesions in juveniles with or without vulvar lichen sclerosus to help decision making by clinicians and pathologists. A scoping review on vulvar melanocytic lesions with or without vulvar lichen sclerosus, including malignant vulvar melanomas, in females up to age 18 years was performed. In addition, the histopathology records of the cohort of all such lesions in The Netherlands from 1991 through 2020 were investigated, and a structured analysis of tissue samples of the subset of cases with lichen sclerosus was performed. The literature study performed confirms that vulvar melanomas in juveniles are extremely rare and that published case reports are often disputed. In The Netherlands, there are no cases of malignant vulvar melanomas up to age 18 years recorded from 1991 through 2020. Atypical histopathological features are often found in biopsies of vulvar nevi in juveniles, especially with concomitant lichen sclerosus, confirming earlier case studies in the literature. We conclude that even with atypical findings, vulvar melanocytic lesions in juveniles have a benign course. To avoid unnecessary and possibly mutilating procedures, we advise referral to an expert center and adaption of existing guidelines for vulvar melanocytic lesions in juveniles.


INTRODUCTION
Vulvar melanocytic lesions in the background of lichen sclerosus (LS) are an enigma for both clinicians and pathologists, especially when found in a juvenile (a child or adolescent up to age 18 years). The nevus may first be noted when the child is seen because of complaints due to vulvar LS (VLS), and the lesion may be damaged by scratching. In certain locations such as genitalia and acra, melanocytic lesions are considered special-site nevi and may show histopathological signs of atypia, although having a benign course (Clark et al., 1998).
Of the case reports of malignant vulvar melanomas (MVMs) in juveniles, most were found in the background of LS (La Spina et al., 2016). A number of these diagnoses were challenged (Carlson et al., 1997;Schaffer and Orlow, 2005). Clinicians are encouraged to be liberal with biopsies of pigmented vulvar lesions in adult women to rule out melanoma and expedite early detection (Heller, 2013). Regarding children, there is no such directive (Trager, 2004). Pathologists generally advise a diagnostic excision if there are atypical features in a biopsy of a vulvar melanocytic tumor. Such a directive for juveniles might lead to unnecessary and possibly mutilating excisions.
Which specific aspects of melanocytic lesions of the vulva in juveniles should clinicians and pathologists be aware of? What is the role, if any, of concomitant LS? What is the prevalence of MVM in juveniles?
We examine the risk of vulvar melanocytic lesions in a juvenile being or becoming malignant by performing a literature study together with a nationwide population study. The histopathological characteristics of vulvar melanocytic lesions found in the literature are then applied to a case series of the national cohort with concomitant LS. We discuss the need for specific guidelines for the care of juveniles with vulvar melanocytic lesions.

Scoping review of the literature
The literature search yielded 3,400 publications, with 140 eligible after screening. On the basis of full texts, 66 publications were included, as shown in Figure 1 and summarized in Table 1.

Case reports of vulvar melanoma in juveniles.
We discerned eight publications with case reports of vulvar melanomas in girls aged 9-18 years encompassing, in all, nine patients Filippetti and Pitocco, 2015;Friedman et al., 1984;Hassanein et al., 2004;Hulagu and Erez, 1973;La Spina et al., 2016;Rosamilia et al., 2006;Webber et al., 2008), six with concomitant VLS and three without. No recurrence or metastasis was reported. Three of the diagnoses of MVM in a juvenile with VLS Hassanein et al., 2004) were questioned (Carlson et al., 1997;Hassanein and Wilkinson, 2005;Schaffer and Orlow, 2005). Schaffer and Orlow (2005) also noted the resemblance to the clinicopathological features of LS nevi in the case of an individual aged 14 years published by Friedman et al. (1984), with two vulvar lesions diagnosed as superficial melanomas with concomitant VLS. This child (Friedman et al., 1984) underwent wide excision of both labia minora and superficial inguinal lymph node dissection in which there was no residual MVM found, only VLS. We found only one case (Rosamilia et al., 2006) reporting a positive lymph node in a juvenile. Currently, the consensus is that such a single metastasis of a melanocyte to an adjacent node is not necessarily indicative of malignancy (Mooi, 2014). No publications report recurrence or mortality during follow-up.
Case series of MVM including one or more juveniles.
A total of 13 case series were identified including at least one juvenile with an MVM (Ariel, 1981;Chung et al., 1975;De Simone et al., 2008;Eleno Beierbach et al., 2020;Jaramillo et al., 1985;Morgan et al., 1988;Nagarajan et al., 2017;Räber et al., 1996;Rao et al., 1990;Rouzbahman et al., 2015;Sinasac et al., 2019;Trimble et al., 1992;Wechter et al., 2004), of which nine included just one juvenile. These series often stemmed from a historical archive. It is not always clear whether these were unique cases because the same databases were used in several publications. An age gap between the one juvenile in a series and the rest of the subjects was often seen (Table 1). Unique records (n = 3,400)    Childhood vulvar melanoma Case report 2 cases, ages 9 and 11 Yes Show the clinical and histological findings in two cases with VLS; shave biopsies and subsequent excisions showed AGN, with no residual melanoma. For discussion on whether these two cases are truly malignant melanomas, see Carlson and Mihm (1997) and  Filippetti   Schaffer and Orlow, 2005 Reaction to article by Hassanein et al. (2004) Case report, letter 1 case, age 8 Yes Reference both Carlson et al. (2002Carlson et al. ( , 1997 and Clark et al. (1988) that 1 of 3 genital nevi is misdiagnosed and that diagnosis is even more difficult in the setting of LS; state that activated melanocytic phenotype related to cytokine milieu and altered extracellular matrix in LS; bridge a number of cases. , , Friedman et al. (1984), Hassanein et al. (2004), and Hassanein and Wilkinson (2005), concurring with Clark et al. (1998); refute diagnosis in article by Hassanein et al. (2004) and discuss a case. Spatz et al. (1998)     Case reports of vulvar nevi in juveniles other than melanomas.
Case series of vulvar nevi including juveniles illustrating diagnostic pitfalls.
There were seven publications describing a series of vulvar nevi (Carlson et al., 2002(Carlson et al., , 1997Gleason et al., 2008;Hunt et al., 2014;Michalova et al., 2017;Ribé, 2008;Shabrawi-Caelen et al., 2004). Most of these articles address the diagnostic pitfalls in cases with VLS. In a case series of 11 subjects with four juveniles (Carlson et al., 2002), the clinicopathological findings for melanocytic nevi occurring in LS, which can mimic malignant melanoma, are described. In one series of 58 pigmented genital lesions in women (Ribé, 2008), there were six Seven publications were included (Blessing et al., 1991;Cohen Goldemberg et al., 2020;Hieta et al., 2019;Ragnarsson-Olding et al., 1993;Rock et al., 1990;Sanchez et al., 2016;Woolcott et al., 1988). In a prospective study of 301 women seen in general gynecologic practice (Rock et al., 1990), 12% had a pigmented lesion of some kind or hyperpigmentation, whereas 2.3% of the subjects had a vulvar nevus. The one patient aged 19 years in this cohort had the only dysplastic nevus. In a population study with regional data from the United States in 1973-2010, of 1,463 cases of vulva or vaginal melanoma, 13 cases were aged 10-19 years. A 10-year survival in the juveniles was 100% (Sanchez et al., 2016). A brief communication based on population data from Finland (Hieta et al., 2019) studied the possible relation between LS and MVM and found a relative risk for MVM of 341 for women with VLS compared with that for women without. The age of the subjects was not given. A national Scottish study (Blessing et al., 1991) found 41 cases of MVM in the period 1979-1989, with one subject being aged 11 years and all the other cases being aged 40 years. Recent data from Brazil (Cohen Goldemberg et al., 2020) found three adolescents with MVM of 801 mucosal melanomas documented in females. Data from Sweden (Ragnarsson-Olding et al., 1993) documenting MVM found 245 cases before 1984 with an average age of 67.7 years, a range of 18-91 years, and one juvenile. A total of 13 publications on the histopathology and immunohistochemistry of melanomas and melanocytic lesions met the inclusion criteria (Ahn et al., 2016;Blessing, 1999;Brenn, 2011Brenn, , 2018Christensen et al., 1987;Clark et al., 1998;Cook, 2010;Elder, 2006;Fu et al., 2021;Haupt and Stern, 1995;Mason et al., 2011;Skelton et al., 1991;Wick, 2015). Clark et al. (1998) introduced the term atypical melanocytic tumor of the genital type, subsequently abbreviated to AGN. An AGN has a relatively specific morphology and may be regarded as belonging to the class of nevi with special-site features. The specific features of AGN comprise characteristics such as symmetry and the presence of borders, which ultimately can only be investigated on a diagnostic excision. Therefore, Brenn (2011) advocates diagnostic excision to adequately differentiate these lesions from melanoma. He also states that in the background of lichen sclerosis, the differentiation of vulvar nevi from melanoma is even more challenging, likely because inflammation can lead to cytological atypia in melanocytes and that epidermal atrophy and clefting can mimic the epidermal consumption sometimes seen in melanoma.
Systematic and comprehensive reviews.
Three systematic reviews were included. Gadducci et al. (2018) showed that vulvar melanoma is a rare disease in women, with a mean age of onset of 54-76 years, a very poor prognosis, and a median overall survival of 41 months. This publication includes but does not differentiate for juveniles. From 1970 to 2020, there were 100 case reports of vulvar malignancies up to age 21 years, of which six were melanomas, all with VLS (Strickland and Fadare, 2021). A review on vulvar melanocytic lesions (Murzaku et al., 2014) found the median age of patients with common vulvar nevi to be 28-33 years, whereas the median age of those with atypical melanocytic tumor of the genital type was 17-26 years.
In all the literature studied in this scoping review, 42 possible MVM in juveniles were reported with no mortalities ( Table 2). were found. In the histology reports in the Pathologische-Anatomisch Landelijk Geautomatiseerd Archief, Dutch Pathology Registry, atypia or dysplasia was recorded in 29 cases (4.6%), and LS was recorded in 16 cases (2.6%). In six cases with VLS, the pathologist reported atypia or dysplasia (36% of the cases with VLS). Follow-up data of the 627 subjects through October 2021 revealed no melanomas but two nongenital premelanoma lesions. A patient aged 8 years with a vulvar melanocytic nevus in 1994 developed a lentigo maligna of the eyelid at age 32 years. In another case, a patient aged 18 years with a vulvar compound nevus in 1995 was diagnosed with melanoma in situ on the lower leg at age 43 years. The average follow-up was 13.43 years, range of 10 months to 30 years.

Histopathology of a cohort of juvenile cases of a vulvar nevus with concomitant JVLS
Of the total 627 juvenile vulvar nevi, the records of 16 cases reported concomitant JVLS, tissue samples were available from 12 of these 16 cases (75%) of vulvar melanocytic lesions with concomitant JVLS in the Netherlands in 1991-2020. Five biopsies of vulvar nevi in juveniles without JVLS were retrieved from our own archive for comparison. Cases with and without VLS were compared (Table 3). Moderate to severe cytological atypia was seen in 83% of the cases with VLS and 40% of the cases without. In cases with VLS, architectural atypia was seen in 83% versus that seen in 40% without VLS. Focal pagetoid spread was found in about 40% regardless of the presence of VLS. A slightly increased proliferation fraction was only seen in the Ki-67 staining when VLS was present. Furthermore, the dyshesive pattern was Features of LS such as inflammation and homogenized stroma were logically more often seen in VLS cases. PRAME was always negative. These characteristics are illustrated in Figures 3 and 4.

DISCUSSION
This scoping review shows that juvenile vulvar melanoma is extremely rare, and this is substantiated by the population data from the Netherlands. We confirm with a large case series that in a background of juvenile VLS, histopathological analysis can show atypical features that could raise suspicion of melanoma despite having a benign course. Thus, we conclude that (diagnostic) excision biopsies in this age group are not indicated even when atypical histological features are observed. It may be questioned whether an MVM in a juvenile exists at all. The literature reveals a minimal number of cases, and even these cases have led to polemics regarding the correct diagnosis. Only one lymph node metastasis was ever reported (Wechter et al., 2004), whereas only a distant metastasis is regarded as the definitive proof of malignancy in melanocytic tumors (Mooi, 2014). One case of the extremely rare malignant blue nevus in an adult, probably stemming from a vulvar blue nevus observed in adolescence, is reported to have led to an ovarian metastasis 31 years after the nevus was first seen (Spatz et al., 1998). It would be instructive to review the histology of a number of these cases with a panel of experts.
Cohort studies and epidemiologic studies of MVM that include juveniles are nearly all studies from before the year 2000 and are always with a favorable outcome for juveniles (Table 2) compared with the vast majority of MVM that show a poor prognosis. The question arises regarding the possibility that these earlier documented lesions were misdiagnosed or that these melanomas in juveniles have such a favorable prognosis compared with adult vulvar melanoma and conventional cutaneous melanoma that the term melanocytoma (low-grade melanocytic neoplasm) (Elder et al., 2020;Schaffer and Orlow, 2005) would be more appropriate. Vulvar melanoma in a juvenile has not been seen in 30 years in the Netherlands, nor did any melanomas develop after a vulvar nevus in a juvenile. It is highly unlikely that the two premalignant lesions reported 24 and 25 years after a vulvar melanocytic lesion were related to the original nevus, considering their locations (eyelid and lower leg). Thus, our scoping review and population study together give credibility to questioning the existence of MVM in juveniles.
Genital nevi in a background of LS are regarded as more challenging to differentiate from MVM (Brenn, 2011). AGN lesions are more likely found in younger subjects (Brenn, 2018;Ribé, 2008). Considering the unlikelihood that a lesion is an MVM in a juvenile, we question the advice (Brenn, 2011(Brenn, , 2018 to excise the lesion in juvenile cases. Nevertheless, our study shows a doubling of biopsies in the last 30 years in the Netherlands, likely reflecting defensive medicine (Welch et al., 2021).
Despite a worrisome histological pattern that is seen in many cases of vulvar melanocytic lesions in juveniles, subsequent melanomas did not develop. In the tissue samples we examined, comprising three quarters of all cases of vulvar nevi with JVLS in the Netherlands over the past 30 years, we mainly found cytological and architectural atypia, in combination with a slightly higher proliferation fraction. A dyshesive pattern and inflammation were often seen, which in the absence of VLS are atypical features raising suspicion of melanoma (inflammation mostly in mucosal melanoma) (Busam et al., 2019). This dyshesive pattern is described (Clark et al., 1998) in a subset of AGN. In contrast to the description of AGN, vulvar melanocytic lesions in the background of VLS are usually without distinct borders and have lentiginous growth (Brenn, 2011). Kurman et al. (2019) elucidates the differences between atypical vulvar nevi and dysplastic vulvar nevi. Dysplasia implies architectural and cytological atypia but with some additional features. "Characteristics that help to distinguish AGN from melanoma include the presence of dermal maturation, the sparsity of mitotic activity, and the absence of necrosis or ulceration" (Kurman et al. [2019] citing Murzaku et al. [2014]). The use of the term dysplastic nevus has varied over the years. Furthermore, some interpret dysplasia as a premalignant condition, which is not the case in the series of biopsies we examined. In our opinion, these vulvar nevi in the context of VLS are a distinct subset of nevi that show atypical features mainly owing to the inflammatory reaction and are not indicative of premalignancy. Thus, using the term dysplasia in this context has no additional value above the combination architectural and cytological atypia, and the term dysplastic should be avoided in this context. In our series, at a median follow-up of 13 years, no melanoma or metastasis occurred. This supports a policy of utmost restraint, confirming that there is little or no necessity for a diagnostic biopsy or excision of a vulvar melanocytic lesion in a juvenile. If biopsied, a diagnosis of melanoma should be avoided because these lesions do not show distant metastasis when found in juveniles (Mooi, 2014).
The strength of this study lies in the systematic and comprehensive evaluation of the literature combined with a large population study in the Netherlands and verifying the population data with a second independent national database.
Limitations include the fact that owing to anonymity, there is no clinical information. In addition, a limitation is that the literature is, for the most part, descriptive. Neither were we able to elucidate the relationship of LS with the atypical histopathological features observed, one of the questions motivating this study. Future guidelines should differentiate for age and anatomical location, with a high threshold of suspicion before biopsy of vulvar lesions in juveniles and a low threshold for referral, with advice for having a consultation in a center of expertise. Previous studies on vulvar nevi or vulvar melanoma in juveniles have generally been anecdotal. This scoping review shows just how unlikely a vulvar nevus in a juvenile is a melanoma. The population data show that most vulvar nevi in a background of VLS in juveniles had atypical features with no subsequent metastasis. Clinicians should be extremely hesitant to excise a vulvar nevus in a juvenile, and pathologists should beware of overdiagnosis when interpreting findings in a biopsy, especially in a setting of LS. Clinicians and pathologists must work together so that clinical decision making is well-founded when a vulvar melanocytic lesion with or without VLS in a juvenile shows signs of atypia. Our findings reflect the need for specific guidelines for care of juveniles with vulvar melanocytic lesions, and considering the rarity of such lesions and the consequences of potentially unnecessary diagnostics or treatment, we recommend referral to a center where knowledge can be maximized.

Scoping review of the literature
A scoping review of the literature was performed systematically according to the preferred reporting items for systematic reviews and meta-analysis guidelines (Tricco et al., 2018).
All published literature on the clinics and histopathology of vulvar nevi with or without VLS or vulvar melanoma in juveniles up to September 2021 were studied. The search strategy was developed encompassing five databases: Embase, Medline ALL, Web of Science Core Collection, Cochrane Central Register of Controlled Trials, and Google Scholar 100 top ranked. Screening for relevance based on title and abstract was done by one reviewer (BM), uncertainties were discussed with other authors (ALM, IAMvdA), and if necessary, the material and methods and results sections of the publication were perused to ensure that no case of juvenile vulvar melanoma was overlooked. Final eligibility was based on full-text reading (BM) and consultation when needed (ALM, IAMvdA). References in the included publications were checked for relevant publications. Included articles were grouped and summarized according to their focus: case report and series, epidemiology, histology, immunohistochemistry, and systematic review.
Cohort of nevi with and without a background of JVLS demographics from the Netherlands A search was performed through the national cytohistopathology database of the Netherlands, Pathologische-Anatomisch Landelijk Geautomatiseerd Archief, Dutch Pathology Registry, for biopsy reports on females aged 18 years in the period 1991-2020, including anatomical locations of vulva, labium, clitoris, or perineum in which a melanocytic lesion of any kind was recorded. In addition, all followup histology reports of these subjects were retrieved. Descriptive statistics regarding demographics and histological findings were calculated using the Statistical Package for the Social Sciences 25.
An independent query was submitted to the National Cancer Registry of the Netherlands (Netherlands Comprehensive Cancer Organization) for any melanoma of the genital tract in females aged up to age 18 years.
Histopathology and immunohistochemical analysis of a series of melanocytic lesions with and without a background of JVLS The vast majority of publications on the histopathology of nevi of special sites, including vulvar nevi, and MVM refer to a few seminal publications (Brenn, 2011;Clark et al., 1998;Elder, 2006) when discussing special-site nevi of the vulva. Using the characteristics described in these publications as well as the World Health Organization publication (Elder et al., 2020) and textbooks (Busam et al., 2019;Kurman et al., 2019), the data were analyzed in a standardized and semiquantitative manner. The stains H&E as well as SOX10 (clone SP267, Cell Marque, Rocklin, CA), HMB-45 (Ventana Medical System, Oro Valley, AZ) (Skelton et al., 1991), and Ki-67 (clone 30-9, Ventana Medical System) (Carlson et al., 2002;La Spina et al., 2016;Pinto et al., 2012) (markers for melanocytic lesions and proliferation) and PRAME (clone EPR20330, Abcam, Cambridge, United Kingdom) (Lezcano et al., 2018) (a marker that is preferentially expressed in melanoma) were applied. From a previous study, (Morrel et al., 2020), all cases of nevi in a background of JVLS in the Netherlands were identified, and material, if available, was obtained. In addition, we identified and retrieved material from all cases of vulva nevi in juveniles from a single center. Scoring was done by two dermatopathologists (ALM, JD) using the list of predefined histopathological features based on the literature.

Data availability statement
Raw data were generated through the Pathologische-Anatomisch Landelijk Geautomatiseerd Archief, Pathology Registry of The Netherlands, queries numbers LZV2020_195A1 and LZV2020_195A2, and at repository EMCD18041 of the Department of Dermatology, Erasmus MC University Medical Center (project identification number 3608-Oracle 7402). The data that support the findings of this study are available from the corresponding author (ALM) on reasonable request.