DOI: 10.1111/jdv.18190 JEADV ORIGINAL ARTICLE Skin lesions in neurofibromatosis type 2: diagnostic and prognostic significance of cutaneous (plexiform) schwannomas A. Plana-Pla,1,2 B. Garcıa,3,4 M. Munera-Campos,1,2 N. Catasus,4 E. Serra Arenas,5 I. Blanco,3,4 E. Castellanos Perez,4,6 I. Bielsa,1,2,* NF2 Multidisciplinary Clinics HUGTiP-ICO-IGTP 1 Dermatology Department, Hospital Universitari Germans Trias I Pujol (HUGTiP), Badalona, Spain 2 Departament de Medicina, Universitat Auto noma de Barcelona, Barcelona, Spain 3 Clinical Genetics and Genetic Counseling Unit, Clinical Genetics Service, Northern Metropolitan Clinical Laboratory, Germans Trias i Pujol University Hospital (HUGTiP), Barcelona, Spain 4 Clinical Genomics Research Unit, Germans Trias i Pujol Research Institute (IGTP-PMPPC), Barcelona, Spain 5 Hereditary Cancer Group, Germans Trias i Pujol Research Institute (IGTP-PMPPC), Barcelona, Spain 6 Clinical Genomics Unit, Clinical Genetics Service, Northern Metropolitan Clinical Laboratory, Germans Trias i Pujol University Hospital (HUGTiP), Barcelona, Spain *Correspondence: I. Bielsa. E-mail: ibielsa.germanstrias@gencat.cat Abstract Background Neurofibromatosis type 2 (NF2) is a genetic disease characterized by the appearance of multiple tumours in the nervous system. Cutaneous lesions are common and may provide useful diagnostic and prognostic information, but they have not been widely studied. Objectives To characterize cutaneous lesions in a Spanish cohort of patients with NF2 and investigate associations with clinical and genetic severity. Methods We studied the clinical and histologic characteristics of cutaneous lesions in 49 patients with NF2 and ana- lysed correlations with phenotype- and genotype-based severity scores. We collected information on the presence/ absence of cutaneous lesions, location, age at onset, type of lesion, and histologic features. We also studied level of sys- temic involvement and genetic mutations involved. Results Forty-nine patients (31 women [63.3%] and 18 men [36.7%]) were analysed, and 33 (67.3%) had cutaneous lesions presumed to be schwannomas. According to their clinical form, they were distributed as follows: 24 patients (48%) had deep tumours, 21 (42%) had plaque-like lesions, and 3 (6%) had superficial tumours. Histologic examination from 27 lesions analysed out 23 patients showed classic schwannoma or hybrid schwannoma-neurofibroma features in the 8 deep tumours biopsied and plexiform schwannoma features in the 17 plaque-like lesions and the 2 superficial tumours analysed. Early onset (first 2 decades of life) was reported by all patients with plaques and superficial tumours. In our cohort, 100% of the patients with plaque-like lesions and superficial tumours with microscopic features of plexi- form schwannoma were in the 2 groups with the most severe clinical phenotypes, and 82.6% of them were in the 3 most severe genotype-based classes. Conclusions and Relevance Cutaneous lesions, specially plexiform schwannomas, are common in NF2, and they usually appear at an early age providing useful diagnostic and prognostic information. These tumours are part of the spectrum of cutaneous manifestations in this disease. Although its diagnostic and prognostic value has been pointed out, there are few studies focussed on their analysis. Received: 17 November 2021; Accepted: 7 April 2022 Conflict of interest The authors have no conflict of interest. Funding information None. JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. 2 Plana-Pla et al. Introduction young patients (4–28 years) or patients with a severe Neurofibromatosis type 2 (NF2) (MIM101000) is an autosomal phenotype.14–21 Moreover, in 2018, our group described a series dominant disorder caused by mutations in the NF2 gene, which of 7 patients with cutaneous plexiform schwannomas who all codes for a tumour-suppressor protein known as merlin. It is later developed a severe NF2 phenotype.22 characterized by the development of schwannomas, meningio- The aim of this study was to describe the clinical and histo- mas, ependymomas, mononeuropathy, eye disorders, and cuta- logic characteristics of cutaneous lesions in a cohort of patients neous lesions.1,2 with NF2 and to explore associations with phenotype and geno- Patients have traditionally been classified into 2 main groups type based on the FGSS. according to clinical severity: those with Gardner-type NF2, a less severe form of disease with onset in adulthood, low tumour Material and methods burden and low mortality, and those with Wishart-type NF2, a All the procedures were performed in accordance with the prin- more severe form of disease with onset in childhood or adoles- ciples of the Declaration of Helsinki and met the ethical stan- cence, high tumour burden and high mortality.1,2 It is known dards of the Institutional Review Board at Germans Trias i Pujol that these differences can be largely due to the type of constitu- Research Center, which approved this study. tional mutation carried, which constitute a fairly good predictor From 2010 to 2020, we prospectively studied all patients with of clinical severity.3–5 The UK NF2 Genetic Severity Score (GSS) a clinical and genetic diagnosis of NF2 evaluated by the NF2 was designed to predict clinical phenotype based on type of Multidisciplinary Clinics in the Spanish National Reference Cen- mutation detected and considering several clinical severity fac- tre (CSUR) of Phacomatoses HGTP-ICO-IGTP at Hospital Uni- tors.6 GSS classifies patients into 3 severity groups—tissue versitari Germans Trias i Pujol in Badalona, Spain. Informed mosaic, classic, and severe—based on the pathogenic germline consent was obtained from all patients. We collected informa- NF2 variant present. Our research group recently validated the tion on age at the time of this evaluation, sex, age at diagnosis, GSS in NF2 patients seen at the Spanish National Reference Cen- and presence of schwannomas and other NF2-related tumours ter (CSUR) for phakomatoses and proposed a revised version and cutaneous lesions. Data regarding the degree of systemic incorporating data from functional assays and the predicted involvement of 45 of our patients were previously collected in mutational effect on merlin.7 This new version, called the Func- the article published recently by Catasus et al.7 tional Genetic Severity Score (FGSS), may have greater prognos- The following information was noted for cutaneous lesions: tic accuracy as it was found to reduce intragroup variability and location (head and neck, upper or lower extremities, trunk), offer enhanced stratification of mosaic patients.7 number and type of lesions and age at onset. The lesions were Different cutaneous lesions have been described in association classified as CAL macules or cutaneous schwannomas. Cutane- with NF2, although little has been published on their clinical ous schwannomas were further divided into 3 groups based on utility for early diagnosis or on their association with disease their clinical morphology: (i) deep tumours, defined as firm, severity. The largest series in the literature describe cafe-au-lait nodular lesions attached to the deep planes with a smooth sur- (CAL) macules and cutaneous schwannomas, but none of them face and no skin changes; (ii) superficial tumours, defined as soft provide detailed information on clinical or histologic findings or superficial papules or nodules, sometimes with a wrinkled sur- possible associations with NF2 severity.1,2,8 In 2019, Halliday face; and (iii) plaques, defined as soft, slightly raised, roughened et al. did publish a large series of 86 patients in which they strati- lesions with hypertrichosis and a darker colour than the sur- fied different clinical variables, including skin lesions, according rounding skin. to genetic severity by GSS. Nevertheless, this study only included We also analysed the histologic features of cutaneous lesions paediatric patients and was not focussed on cutaneous findings.9 that had been excised over the 10-year study period for diagnos- The largest study to focus exclusively on cutaneous abnormali- tic purposes or at the patient’s request. This study was based ties in NF2 was published in 1997 by Mautner et al.,10 who pro- solely on light microscopy of haematoxylin and eosin–stained vided a detailed clinical picture of cutaneous lesions and their slides. For the immunohistochemical analysis, fixed cells were association with disease phenotypical severity. To our knowl- stained with primary polyclonal rabbit anti-S100 (Dako) and edge, no studies to date have analysed associations between cuta- mouse monoclonal anti-Melan A (Ventana Medical Systems). neous lesions and genetic severity in adult patients with NF2. Neurofilament antibody staining was performed in some cases. We believe that a greater understanding of the cutaneous Genetic testing of blood, or tissue when available, was per- manifestations of NF2 might facilitate early diagnosis and pro- formed using a custom NGS Panel (I2HCP), which comprises vide useful prognostic information, as on reviewing the litera- 135 genes associated with cancer predisposition syndromes and ture we observed that skin lesions are a common presenting RASopathies,23 as described by Catasus et al.7 symptom in children with NF211–13 and that all cases of cutane- Patients were grouped into 6 classes using the FGSS, which ous plexiform schwannomas described to date have occurred in predicts clinical phenotype based on the type of mutation JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. Importance of cutaneous schwannomas in NF2 3 Table 1 Clinical characteristics and patient distribution according Table 1 Continued to severity classifications Patient characteristics Patient characteristics Meningiomas (total N [%]) 28 (57.1%) Demographic Intracranial (N [%]) 27 (55.1%) Sex (N [%]) Spinal (N [%]) 15 (30.6%) Male 18 (36.7%) Ependymomas (total N [%]) 11 (22.4%) Female 31 (63.3%) Extravestibular schwannomas (total N [%]) 33 (67.3%) Current age (years), (mean [SD]) 45.84 (18.62) Intracranial schwannomas (N [%]) 15 (30.6%) Age at diagnosis (years), (mean [SD]) 29.49 (19.08) Spinal schwannomas (N [%]) 27 (55.1%) Diagnosis before age 25 (N [%]) 25 (51%) Peripheral schwannomas (N [%]) 16 (32.7%) Familiar history of NF2 (N [%]) Familiar NF2 4 (8.2%) Sporadic/de novo NF2 45 (91.8%) Classification of NF2 patients by phenotype severity (N [%]) detected, ranging from class 1 (less severe) to class 6 (most Group 1 21 (42.9%) severe) (Table S1). Patients were also grouped according to phe- Group 2 13 (26.5%) notype severity scores calculated using well-known prognostic Group 3 4 (8.2%) markers24,25 at the time of evaluation, as described by Catasus Group 4 11 (22.4%) et al.7 Phenotype severity was classified into 4 categories, where Classification of NF2 patients by FGSS categories (N [%]) group 1 represented the most severe form of disease and group 4 Class 1 17 (34.7%) the less severe (Table S2). Class 2 0 (0) All data were recorded on a spreadsheet (Microsoft Excel Class 3 3 (6.1%) 2019) and expressed as numbers (n) and percentages (%) or Class 4 8 (16.3%) means  standard deviation (SD). Categorical phenotype var- Class 5 9 (18.4%) iables (e.g. sex and cutaneous features) were compared using Class 6 12 (24.5%) the v2 test. Trends with severity were investigated using the Cutaneous manifestations Mantel–Haenszel linear-by-linear v2 test of association. Spear- Cutaneous lesions (N [%]) man’s rank-order correlation analysis was used to assess cor- Present 33 (67.3%) Absent 16 (32.7%) relations between disease severity and patient age and -au-lait macules (CAL macules) Cafe number of skin lesions. The preliminary analysis (visual Presence of CAL macules (N [%]) 19 (38.8%) inspection of scatterplots) showed the relationships to be Proportion of patients by number of CAL macules monotonic. Statistical analyses were performed in SPSS V.25 0 30 (61.2%) (IBM Statistics). Standard summary statistics are reported 1 6 (12.2%) throughout with statistical significance of inferences set at 2 2 (4.1%) P < 0.05. 3 2 (4.1%) 4 6 (12.2%) Results 5 1 (2%) 6 1 (2%) Description of NF2 cohort 7 1 (2%) The study population included 49 patients diagnosed of NF2 Deep tumours according to Manchester NF2 diagnostic criteria,1,2 45 of them Presence of cutaneous schwannomas (N [%]) 24 (49%) previously described in Catasus 2021.7 There were 31 women Plaque-like lesions (63.3%) and 18 men (36.7%) (female-to-male ratio, 1.7 : 1). Presence of plaques (N [%]) 21 (42.9%) Mean (SD) age was 45.48 years at the time of analysis and Superficial tumours Presence of tumours (N [%]) 3 (6.1%) 29.49 years at the time of diagnosis; 51% of patients had been Plaque-like lesions + Superficial tumours diagnosed with NF2 before the age of 25 years. Bilateral schwan- Presence of plaques + tumours (N [%]) 23 (46.9%) nomas were found in 89.8% of the patients, 55.1% presented Non-cutaneous manifestations spinal schwannomas, and 32.7% had peripheral schwannomas Vestibular schwannomas (total N [%]) 48 (98%) (Table 1). Bilateral (N [%]) 44 (89.8%) Unilateral (N [%]) 4 (8.2%) Cutaneous lesions Age at diagnosis (years) (mean [SD]) 32 (17.89) Thirty-three out of 49 NF2 patients (67.3%) had cutaneous manifestations (Table 2). Almost half of them (38.8%) had at JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. 4 Plana-Pla et al. Table 2 Cutaneous lesions Cutaneous manifestations Type of lesion No. of Location Appearance Average per patient Histopathology patients -au-lait Cafe 19 (38.8%) Trunk 58.6%; Extremities 39.6%; 31.5% congenital o during 1st 1.2 CLM per patient (3.05 per None of the were macules Head and neck 1.7% year; 68.4% from 1 year to patient counting only biopsied (CAL macules) 12 years old patients with macules) Deep tumours 24 (49%) Trunk 15.3%; 20.5% Head and 0% Congenital o during 1st 1.59 deep tumour per patient 6 cases with classic Neck; 64% Extremities year; 25% from 1 year old to (3.2 per patient counting schwannoma features; 12 years old; 29% only patients with 2 cases with a hybrid Adolescence; 37.5% from 18 deep tumours) schwannoma- years old to 29 years old; neurofibroma pattern 8.3% after 30 years old Plaque-like 21 (42.9%) Trunk 65%; Head and neck 8.9%; 4.7% Congenital o during 1st 1.36 per patient (3.1 per 17 cases with plexiform lesions Extremities 25.3% year; 57.1% from 1 year old to patient counting only schwannoma features 12 years old; 28.5% patients with plaque-like Adolescence; 4.7% After lesions) adolescence but before 20 years old Superficial 3 (6.1%) Extremities 71.4%; Trunk 28.5% 100% from 1 year to 0.14 per patient (2.3 per 2 cases with plexiform tumours 12 years old patient counting only schwannoma features patients with superficial tumours) least one CAL macule (mean [SD], 1.2 [1.8]; range, 0–7). The Twenty-four patients (49%) had deep tumours presumed to macules were all light brown, and 58.6% were located on the be schwannomas (Fig. 1). Most of them (64%) were located on trunk. They had been present at birth or had appeared in the the limbs and followed the course of the nerves. They had first years of life. appeared in the first 2 decades of life in 57% of patients. Figure 1 (a) Deep schwannoma, located on the wrist of a patient with NF2. (b and c) Plaque-like plexiform schwannoma, defined as soft, slightly raised, roughened lesions with hypertrichosis and a darker colour than the surrounding skin, in a patient with NF2. (d) Plexiform schwannoma characterized by bundles of spindle cells surrounded by a thin fibrous capsule in the superficial dermis. The cells are posi- tive for s-100. (e) Superficial schwannoma in a patient with NF2. JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. Importance of cutaneous schwannomas in NF2 5 Histology of the 8 deep tumours that had been biopsied or associated with severest NF2 phenotype (Tables 1,3). Patients excised over the study period confirmed a diagnosis of schwan- in the higher severity genotype classes were more likely to noma in all cases. The tumours exhibited 2 distinct histologic have any cutaneous lesion than those in the lower severity patterns: the classic schwannoma pattern (6 cases) and a hybrid classes (Table 3). Twenty-five of the 33 patients with cutane- schwannoma-neurofibroma pattern (2 cases). The classic pattern ous lesions were in the top 3 classes of severity. The oppo- showed a round, very well-circumscribed, encapsulated, dermal site trend was observed for patients without these lesions. tumour formed by spindle cells in cellular areas (some of which The association between cutaneous lesions and predicted contained nuclear palisading and Verocay bodies) together with severity based on the causative germline mutation was more evi- a number of hypocellular areas with a loose myxoid stroma and dent for tumours than for CAL macules, although it was statisti- blood vessels. The hybrid pattern was characterized by an area cally significant in both cases (Table 3). Of the 19 patients with with classic schwannoma features and an adjacent area with neu- CAL macules, 15 (78.9%) were in classes 4 to 6, 2 (10.5%) were rofibroma features comprising a mixed proliferation of haphaz- in class 3, and another 2 were in class 1. The majority of patients ardly arranged nonencapsulated Schwann cells, fibroblasts and with deep tumours (n = 19, 79.2%) were also in classes 4 to 6, collagen with dispersed axons. Immunohistochemical staining while those without these tumours were more evenly distributed was positive for s-100 and negative for neurofilaments in the across the classes. Finally, 19 (82.6%) of the 23 patients with schwannoma area and positive for s-100 and neurofilaments in superficial tumours or plaques were in classes 4 to 6, while 16 the neurofibroma area. (61.5%) of those without these lesions were in classes 1 to 3. The Similarly, nearly half of the patients included in this study differences were statistically significant (Table 3). (42.9%) had at least one plaque-like lesion defined as soft, The above results indicate that the presence of cutaneous slightly raised, roughened lesions with hypertrichosis and a dar- lesions in patients with NF2 is associated with severity predicted ker colour than the surrounding skin (Fig. 1b,c). Furthermore, by the type of NF2 variant present. 65% of them were located on the trunk. All the patients said their plaques had always been present, suggesting they were con- NF2 cutaneous manifestations and association with genital or had appeared in childhood or adolescence. Deep phenotype severity tumours, by contrast, were much less likely to develop in early To study the correlation between cutaneous lesions and pheno- life (just 57% of cases). One plaque from seventeen patients, for type at the time of this study, we quantified phenotype severity a total of seventeen plaques, had been excised and biopsied over using a numeric nine-scale based on the sum of transformed the 10-year study period, and they all showed the same plexi- quantitative variables based on well-known NF2 prognosis form schwannoma pattern characterized by bundles of spindle markers and grouped into four different groups, ranging from cells surrounded by a thin fibrous capsule and Verocay bodies severe phenotype (group 1) to very mild phenotype (group 4) and Antoni A areas scattered throughout the superficial and (Table S2). deep dermis. There were no Antoni B areas. Immunohisto- There were 21 patients (42.9%) in group 1 (early onset/high chemically, the cells were positive for s-100 and negative for neu- tumour burden/high mortality) and 11 (22.4%) in group 4 (late rofilaments (Fig. 1d). onset/low tumour burden/low mortality). Of the remaining Finally, very few patients (6%) had superficial tumours patients, 13 (26.5%) were in group 2 and 4 (8.2%) in group 3 (Fig. 1e). These were mainly located on the extremities and had (Table 1). all appeared during childhood. From them, two of these Cutaneous lesions were more common in patients with severe tumours had been biopsied and showed the typical plexiform phenotypes (groups 1/2) than in those with milder disease schwannoma pattern observed in the plaque-like lesions. (groups 3/4) (P < 0.001) (Table 4). Twenty of the 33 patients (60.6%) were in group 1, 10 (30.3%) were in group 2, and 3 NF2 cutaneous manifestations and association with (9.1%) were in group 3. The distribution of patients without predicted NF2 severity cutaneous lesions followed the opposite trend. Similar distribu- To study the association between cutaneous manifestations tion patterns were observed on analysing the cutaneous lesions and NF2 severity predicted by the type of germline mutation by category, particularly for lesions assumed to be cutaneous and determine the potential prognostic value of these mani- schwannomas, regardless of their clinical form. festations, all the patients in the cohort were grouped into Most of the patients with CAL macules had a severe clinical the 6 FGSS classes. From them, 17 patients (34.7%) were phenotype, with 11 patients (57.9%) in group 1 and 6 (31.6%) assigned to class 1, the lowest severity class, corresponding in group 2. The distribution of patients without CAL macules to probable or confirmed mosaicism. Of the remaining was more uniform. The presence of CAL macules was signifi- patients all harbouring a pathogenic germline NF2 variant, cantly associated with a more severe phenotype (P < 0.05). 11 were assigned to mild and moderate NF2 phenotype, Tumour lesions were even more strongly associated with a severe while seventeen were classified as 5 or 6, the classes phenotype. All the patients with deep tumours were either in JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. 6 Plana-Pla et al. Table 3 Demographic and clinical data according to FGSS classification FGSS severity 1 2 3 4 5 6 Correlation and significance “Less Severe” “More Severe” Demographic data N (% total) Distribution of patients 17 (34.7%) 0 3 (6.1%) 8 (16.3%) 9 (18.4%) 12 (24.5%) N (% gender) Gender χ2 =1.95; P = 0.589 Female 13 (41.9%) 0 2 (6.5%) 5 (16.1%) 4 (12.9%) 7 (22.6%) Male 4 (22.2%) 0 1 (5.6%) 3 (16.7%) 5 (27.8%) 5 (27.8%) Clinical data: cutaneous manifestations Presence of Skin lesions* χ2 =12.73; P <0.001 skin lesions Present 5 (15.2%) 0 (0%) 3 (9.1%) 8 (24.2%) 6 (18.2%) 11 (33.3%) N (% total) Absent 12 (75%) 0 (0%) 0 (0%) 0 (0%) 3 (18.8%) 1 (6.3%) CAL macules* χ2 =6.49; P =0.04 Present 2 (10.5%) 0 (0%) 2 (10.5%) 7 (36.8%) 3 (15.8%) 5 (26.3%) Absent 15 (50%) 0 (0%) 1 (3.3%) 1 (3.3%) 6 (20.0%) 7 (23.3%) Deep tumors* χ2 =11.49; P =0.02 Present 3 (12.5%) 0 (0%) 2 (8.3%) 6 (25.0%) 6 (25.0%) 7 (29.2%) Absent 14 (56.0%) 0 (0%) 1 (4.0%) 2 (8.0%) 3 (12.0%) 5 (20.0%) Superficial tumors χ2 =9.08; P =0.05 Present 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 3 (100%) Absent 17 (37.0%) 0 (0%) 3 (6.5%) 8 (17.4%) 9 (19.6%) 9 (19.6%) Plaque-like lesions* χ2 =8.65; P =0.03 Present 3 (14.3%) 0 (0%) 1 (4.8%) 4 (19.0%) 4 (19.0%) 9 (42.9%) Absent 14 (50%) 0 (0%) 2 (7.1%) 4 (14.3%) 5(17.9%) 3 (10.7%) Superficial tumors + plaque-like χ2 =12.89; P <0.001 lesions* Present 3 (13%) 0 (0%) 1 (4.3%) 4 (17.4%) 4 (17.4%) 11 (47.8%) Absent 14 (53.8%) 0 (0%) 2 (7.7%) 4 (15.4%) 5 (19.2%) 1 (3.8%) Asterisk (*) indicates statistical significance (P < 0.05) in trends (X2) and correlations (rs) of measures with phenotypical severity classification. CAL macules, Cafe -au-lait macules; N, Number; NF2, neurofibromatosis type 2; SD, standard deviation. group 1 (17 patients, 70.8%) or group 2 (5 patients, 20.8%). The Discussion presence of cutaneous tumours was significantly associated with Dermatologists are not usually involved in the diagnosis, treat- more severe disease (P < 0.001). ment, or follow-up of patients with NF2, meaning that Superficial lesions were markedly distributed in the most skin changes are often underdiagnosed, poorly defined and severe phenotype groups. All 3 patients with superficial tumours characterized, and their importance underestimated. Our find- were in group 1, while the 21 patients with plaque-like lesions ings confirm that accurate diagnosis of cutaneous lesions by der- were either in group 1 (16 patients, 76.2%) or 2 (5 patients, matologists provides valuable diagnostic and prognostic 23.8%). Patients without plaques were more likely to have information in the setting of NF2. milder disease. The 23 patients with superficial lesions (tumours About 70% of patients in our cohort had cutaneous lesions, or plaques) all had a severe phenotype (group 1 or 2). Those which is similar to rates observed in other series.1,2,10 without superficial lesions were more uniformly distributed. CAL macules are less common in NF2 than in NF1. This is true Overall, superficial lesions were significantly associated with a in terms of the percentage of patients affected—close to 40% in more severe phenotype (P < 0.001). our cohort and others1,2,10 (vs. 95% for NF126)—and the number Number of cutaneous lesions (CAL macules, superficial of macules per patient—1.2 in NF2 vs. 6 or more in NF1. CAL lesions, and deep tumours) was also associated with greater phe- macules are paler and have less distinct borders than those of NF1. notype severity. Their diagnostic value is thus low, due to their relative infrequency The above results indicate that cutaneous lesions, and superfi- and the fact that they are observed in many other diseases, includ- cial tumours in particular, are strongly associated with a severe ing RASopathies, Fanconi anaemia and tuberous sclerosis.27 They NF2 phenotype. The association in this case was stronger than are also found in healthy individuals.28 Nevertheless, NF2 should that observed using the FGSS. always be suspected in a patient with CAL macules and other JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. Importance of cutaneous schwannomas in NF2 7 Table 4 Demographic data and presence of skin lesions according to phenotypical severity classification PHENOTYPICAL severity 1 2 3 4 Statistics “More severe” “Less severe” Demographic data N (% total) Distribution of patients 21 (42.9%) 13 (26.5%) 4 (8.2%) 11 (22.4%) N (% gender) Gender χ2=0.04; P = 0.843 Female 14 (45.2%) 7 (22.6%) 2 (6.5%) 8 (25.8%) Male 7 (38.9%) 6 (33.3%) 2 (11.1%) 3 (16.7%) Mean (SD) Age at diagnosis* 16.95 (± 7.52) 26 (± 8.94) 43.75 (± 5.91) 52.36 (± 12.90) rs=-0,81; P <0.001 Current age* 37.19 (± 10.86) 42.38 (± 15.87) 61.75 (± 8.73) 60.64 (± 13.34) rs=0.59; P <0.001 Diagnosis before 25yo* 19 (76%) 6 (24.0%) 0 (0%) 0 (0%) χ2=26.58; P <0.001 N (% total) Familiar history of NF2 1 (25%) 2 (50%) 0 (0%) 1 (25%) χ2=0.67; P =0.796 Clinical data: cutaneous manifestations Presence of Skin lesions* χ2=30.765; P <0.001 skin lesions Present 20 (60.6%) 10 (30.3%) 3 (9.1%) 0 (0%) N (% total) Absent 1 (6.3%) 3 (18.8%) 1 (6.3%) 11 (68.8%) CAL Macules* χ2=7.21; P = 0.028 Present 11 (57.9%) 6 (31.6%) 2 (10.5%) 0 (0%) Absent 10 (33.3%) 7 (23.3%) 2 (6.7%) 11 (36.7%) Deep Tumors* χ2=17.43; P <0.001 Present 17 (70.8%) 5 (20.8%) 2 (8.3%) 0 (0%) Absent 4 (16%) 8 (32%) 2 (8%) 11 (44%) Superficial tumors χ2=2.724; P = 0.099 Present 3 (100%) 0 (0%) 0 (0%) 0 (0%) Absent 18 (39.1%) 13 (28.3%) 4 (8.7%) 11 (23.9%) Plaques* χ2=19.22; P <0.001 Present 16 (76.2%) 5 (23.8%) 0 (0%) 0 (0%) Absent 5 (17.9%) 8 (28.6%) 4 (14.3%) 11 (39.3%) Superficial tumors + χ2= 23.77; P <0.001 plaques* Present 18 (78.3%) 5 (21.7%) 0 (0%) 0 (0%) Absent 3 (11.5%) 8(30.8%) 4 (15.4%) 11 (42.3%) Number of skin Number of CAL Macules* 2.10 (± 2.41) 0,85 (± 1.14) 0.75 (± 0.96) 0 rs=0.418; P = 0,014 lesions Number of superficial 0.33 (± 0.86) 0 0 0 rs=0.268; P = 0.061 Mean (SD) tumors Number of plaques* 2.86 (± 3.07) 0,54 (± 0.78) 0 0 rs= 0.666; P = 0.001 Number of superficial 3.19 (± 2.96) 0,54 (± 0.78) 0 0 rs=0.740; P <0.001 tumors + plaques* Deep tumors* 2.62 (± 2.25) 1,54 (± 3.43) 0.75 (± 0.96) 0 rs=0.617; P <0.001 Asterisk (*) indicates statistical significance (P < 0.05) in trends (X2) and correlations (rs) of measures with phenotypical severity classification. CLM, Cafe -au-lait macules; N, Number; NF2, neurofibromatosis type 2; SD, standard deviation. manifestations such as mononeuropathy (in particular facial paral- childhood and show typical features of a plexiform schwannoma ysis), eye disorders or meningiomas. on histology. Most studies of NF2 to date have used different systems Deep schwannomas were observed in 49% of patients in our (based on lesion depth) to classify cutaneous schwannomas, with cohort. While they are clinically non-specific (they are similar to some authors using 2 categories (superficial lesions/plaques and other subcutaneous lesions, such as lipomas, cysts or fibrohistio- deep lesions)1,2 and others using 3 (superficial lesions/plaques, cytic tumours), they should always raise suspicion of NF2, espe- intracutaneous lesions and subcutaneous lesions).10 Based on cially in patients with multiple lesions. Deep schwannomas, our findings, we propose 2 main categories: (i) deep schwanno- however, can also occur in isolation or in patients with schwan- mas with histologic features of classic schwannoma or hybrid nomatosis. Schwannomatosis is caused by mutations in the schwannoma-neurofibroma that can appear at different times of SMARCB1 or LZTR1 genes (located on chromosome 22 and life and (ii) superficial schwannomas in the form of plaques, near the NF2 gene) and is characterized by schwannomas involv- nodules or small soft tumours that appear at birth or in ing the peripheral or cranial nerves or the spine. Unlike NF2, it JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. 8 Plana-Pla et al. does not manifest with bilateral vestibular involvement, menin- prevalence and incidence of NF2, a sample size of 49 is not insig- giomas, ependymomas or eye disorders.29 nificant. Furthermore, ours is the first study to prospectively col- Superficial schwannomas would appear to be the most useful lect information on cutaneous lesions in patients with NF2 and diagnostic lesions because of their highly characteristic clinical to perform a detailed case-by-case analysis. and histological findings. Clinically, they present as nodules, In summary, cutaneous lesions are common in NF2 and are small soft tumours or plaques that are slightly darker than the useful from both a diagnostic and prognostic perspective, partic- surrounding skin and have a wrinkled surface, hypertrichosis ularly in the case of superficial plexiform schwannomas, which and a texture similar to that of an atrophic neurofibroma. Histo- have highly specific clinical and histologic features. logically, they exhibit the highly characteristic features of a plexi- form schwannoma. Superficial schwannomas are relatively Informed consent common. In our cohort, they were present in 46% of patients, a The patients in this manuscript have given written informed rate similar to that seen in other series,1,2 and unlike deep consent to publication of their case details. schwannomas, most of them had developed in childhood or adolescence. Because they can be easily identified by an experi- Data availability statement enced dermatologist and subsequently confirmed by histology, The data that support the findings of this study are available on they are practically pathognomonic for NF2. Isolated plexiform request from the corresponding author. The data are not pub- schwannomas are very rare and have mostly been described in licly available due to privacy or ethical restrictions. NF2. When they occur in a healthy individual or in a patient with schwannomatosis, they tend to be solitary, late-onset References lesions with an extracutaneous location.30 NF2, however, must 1 Evans DG. Neurofibromatosis type 2 (NF2): a clinical and molecular always be suspected on detection of a cutaneous plexiform review. Orphanet J Rare Dis 2009; 4: 16. schwannoma, even if the patient has no other symptoms or fam- 2 Asthagiri AR, Parry DM, Butman JA et al. Neurofibromatosis type 2. ily history. The index of suspicion must be particularly high in Lancet 2009; 373: 1974. 3 Selvanathan SK, Shenton A, Ferner R et al. Further genotype-phenotype cases of multiple lesions or paediatric onset. correlations in neurofibromatosis 2. Clin Genet 2010; 77: 163. Cutaneous schwannomas are useful for guiding diagnosis in 4 Smith MJ, Higgs JE, Bowers NL et al. Cranial meningiomas in 411 neuro- NF2, but in some cases, they may be key. Patients with mosai- fibromatosis type 2 (NF2) patients with proven gene mutations: clear positional effect of mutations, but absence of female severity effect on age cism, for example, do not always have detectable mutations in at onset. J Med Genet 2011; 48: 261–265. the blood, but if they have a skin lesion, tissue can be used for 5 Hexter A, Jones A, Joe H et al. Clinical and molecular predictors of mor- genetic testing, thus facilitating diagnosis. In an earlier study by tality in neurofibromatosis 2: a UKnational analysis of 1192 patients. J our group, 7 patients were diagnosed with NF2 following the Med Genet 2015; 52: 699–705. 6 Halliday D, Emmanouil B, Pretorius P et al. Genetic severity score pre- demonstration of 2-hit mutations in the tissue of cutaneous dicts clinical phenotype in NF2. J Med Genet 2017; 54: 657–664. tumours.22 7 Catasus N, Garcia B, Galvan-Femenıa I et al. Revisiting the UKgenetic Cutaneous lesions can also provide prognostic information. severity score for NF2: a proposal for the addition of a functional genetic In our cohort, all the patients with superficial schwannomas and component. J Med Genet 2021; jmedgenet-2020-107548. 8 Ferner RE. The neurofibromatoses. Pract Neurol 2010; 10: 82–93. microscopic features of plexiform schwannoma, and 91.6% of 9 Halliday D, Emmanouil B, Vassallo G et al. Trends in phenotype in the those with deep schwannomas were in the top 2 phenotype English paediatric neurofibromatosis type 2 cohort stratified by genetic severity groups, while 82.6% of those with superficial schwanno- severity. Clin Genet 2019; 96: 151–162. mas and 79.2% of those with deep schwannomas were in the top 10 Mautner VF, Lindenau M, Baser ME, Kluwe L, Gottshaclk J. Skin abnor- malities in neurofibromatosis 2. J Arch Dermatol 1997; 133: 1539–1543. 3 genotype severity classes. The presence of cutaneous plexiform 11 Ruggieri M, Lanetti P, Polizzi A et al. Earliest clinical manifestatios and schwannomas, therefore, is correlated with both clinical and natural history of neurofibromatosis type 2 (NF2) in childhood: a study genetic severity. The link with clinical severity, however, is par- of 24 patients. Neuropediatrics 2005; 36: 21–34. ticularly strong, probably because genotype-based classification 12 Tibussek D, H€ ubsch S, Berger K, Schaper J, Rosenbaum T, Mavatepek E. Infantil onset neurofibromatosis type 2 presenting with peripheral facial systems cannot always accurately predict phenotype, particularly palsy, skin patches, retinal hamartoma and foot drop. Klin pediatr 2009; in patients with mosaicism.7 It is reasonable to assume that a 221: 247–250. patient with severe NF2—and consequently higher tumour bur- 13 Gugel I, Grimm F, Zipfel J et al. Age at onset and presenting symptoms of neurofibromatosis type 2as prognostic factors for clinical course of ves- den—will have more cutaneous lesions. Knowing, however, that tibular schwannomas. Cancers 2020; 12: 2355. these lesions appear at a young age facilitates diagnosis and may 14 Val-Bernal JF, Figols J, Vazquez-Barquero A. Cutaneous plexiform have prognostic relevance. Early identification of severe disease schwannoma associated with neurofibromatosis type 2. Cancer 1995; 76: is crucial in NF2, as early treatment facilitates the management 1181–1186. 15 Reith JD, Goldblum JR. Multiple cutaneous plexiform schwannomas. and significantly improves prognosis. Reporto f a case and review of the literatura with particular reference to Our study is limited by its low statistical power due to the rel- the association with types 1 and 2 neurofibromatosis and schwannomato- atively small study population. Nonetheless, considering the low sis. Arch Pathol Lab Med 1996; 120: 399–401. JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology. Importance of cutaneous schwannomas in NF2 9 16 Ishida T, Kuroda M, Motoi T, Oka T, Imamura T, Machinami R. Pheno- 25 Baser ME, Friedman JM, Aeschliman S et al. Predictors of the risk of mor- typic diversity of neurofibromatosis 2: association with plexiform tality in neurofibromatosis type 2. Am J Hum Genet 2002; 71: 715–723. schwannoma. Histopathology 1998; 32: 264–270. 26 Nunley KS, Gao F, Albers AC, Bayliss SJ, Gutmann DH. Predictive value 17 Lim HS, Jung J, Chung KY. Neurofibromatosis type 2 with multiple plex- of cafeau lait macules at initial consultation in the diagnosis of neurofi- iform schwannomas. Int J Dermatol 2004; 43: 336–340. bromatosis type 1. Arch Dermatol 2009; 145: 883. 18 Miyakawa T, Kamada N, Kobayashi T et al. Neurofibromatosis type 2 in 27 Hernandez-Martin A, Duat-Rodriguez A. An update on neurofibromato- an infant with multiple plexiform schwannomas as first symptom. J Der- sis type 1: not just cafe-au-lait spots, freckling and neurofibromas. An matol 2007; 34: 60–64. update. Part 1. Dermatological clinical criteria diagnostic of the disease. 19 Yeon Ko J, Eun Kim J, Hoon Kim Y, Suck RY. Cutaneous plexiform Actas Dermosifiliogr 2016; 107: 454–464. schwannomas in a patient with neurofibromatosis type 2. Ann Dermatol 28 Landau M, Krafchik BR. The diagnostic value of cafeau-lait macules. J 2009; 21: 402–405. Am Acad Dermatol 1999; 40: 877. 20 Jo~ao Cruz M, Baudrier T, Gil-Da-Costa MJ, Azevedo F, Mota A. Multiple 29 Hadfield KD, Newman WG, Bowers NL et al. Molecular characterisation cutaneous plexiform schwannomas revealing neurofibromatosis type 2 in of SMARCB1 and NF2 in familial and sporadic schwannomatosis. J Med a child – report of a novel mutation in this rare association. Eur J Derma- Genet 2008; 45: 332. tol 2011; 21: 1010–1011. 30 Berg JC, Scheithauer BW, Spinner RJ, Allen CM, Koutlas IG. Plexiform 21 Nguyen TV, Matthews MR, Barrera FF, Browning JC. Multiple cutaneous schwannoma: a clinicopathologic overview with emphasis on the head plexiform schwannomas as initial presentation of neurofibromatosis 2 in and neck regi on. Hum Pathol 2008; 39: 633–640. a 9-year-old. Pediatr Dermatol 2012; 29: 536–538. 22 Castellanos E, Plana A, Carrato C et al. Early genetic diagnosis of Neuro- Supporting information fibromatosis type 2 from skin plaque plexiform schwannomas in child- hood. JAMA Dermatol 2018; 154: 341. Additional Supporting Information may be found in the online 23 Castellanos E, Gel B, Rosas I et al. A comprehensive custom panel design version of this article: for routine hereditary cancer testing: preserving control, improving diag- nostics and revealing a complex variation landscape. Sci Rep 2017; 7: 39348. Table S1. Functional genetic severity score. 24 Iwatate K, Yokoo T, Iwatate E et al. Population characteristics and pro- Table S2 Phenotype severity classification. gressive disability in neurofibromatosis type 2. World Neurosurg 2017; 106: 653–660. JEADV 2022 Ó 2022 The Authors. Journal of the European Academy of Dermatology and Venereology published by John Wiley & Sons Ltd on behalf of European Academy of Dermatology and Venereology.