Incidence of neurocutaneous melanosis in Japanese pediatric patients with congenital melanocytic nevi

Neurocutaneous melanosis (NCM) is a rare, non-hereditary neurocutaneous disorder characterized by excessive melanocytic proliferation in the skin and central nervous system. As no major studies have covered the incidence of NCM among Japanese patients with congenital melanocytic nevi (CMN), we prospectively investigated the incidence of NCM among Japanese patients who underwent initial treatment for CMN. The relationship of CMN and NCM was also investigated. Japanese pediatric patients with CMN under 1 year of age were included between January 2020 and November 2022, and all patients underwent brain MRI to check for NCM in this study. NCM lesions were most frequently seen in the amygdala, followed by the cerebellum, brainstem, and cerebral hemispheres. NCM was diagnosed on brain MRI in 31.6% of the 38 patients with CMN and in 25.0% of patients with no prior examination or treatment. Distribution and size of CMN, number of satellite nevi, rugosity and nodules were strongly associated with the existence of NCM, and these findings may guide a future registry study with a large cohort of CMN patients.


Patients
This prospective study was performed in accordance with the Declaration of Helsinki and was approved by Kyoto University Graduate School and Faculty of Medicine, Ethics Committee.A total of 40 pediatric patients under 1 year old [20 males, 20 females; average age, 5.3 months (median age, 4 months; range, 3-12 months)] with CMN, who were referred to our institute from all over Japan between January 2020 and November 2022, were prospectively included in this study (Fig. 1).Written informed consent was obtained from the parents of all patients.The upper age limit was defined because the development of myelination tends to obscure hyperintensity lesions on T1WI.All patients underwent brain MRI as screening for NCM.We defined patients with NCM in this study as those patients who showed a hyperintense lesion in the brain parenchyma and/or abnormal leptomeningeal enhancement on T1WI.
The exclusion criteria were as follows: motion artifacts during MRI; and patients with pathologies other than melanocytic nevi; and treatment for CMN at another institute.

Analysis 1
This analysis was conducted to determine the relationship between characteristics of CMN and NCM, and included all patients with untreated CMN.The incidence of NCM was also calculated.

Analysis 2
Since no registry study of CMN had been performed in Japan, the incidence of NCM among patients with CMN was unknown.Prior to any future registry study, we investigated the incidence of NCM among treatment-naive patients for CMN.We focused on patients who had never undergone brain MRI and had not been treated for CMN before referral to our institute (no previous MRI examination and treatment group).

Evaluation of neurological manifestations
All patients were evaluated for neurological symptoms by a board-certified pediatric neurologist (T.Y., with 16 years of experience in pediatric neurology).

Evaluation of CMN
The size, anatomical localization and characteristic features of CMN were recorded and evaluated by a board-certified plastic surgeon (M.S., 23 years of experience in plastic surgery) for all patients, based on the categorization proposed by Krengel et al. 19 .
Anatomical distributions of CMNs on the skin surface of patients were categorized as follows: bonce (head and facial region), bolero (mainly on the upper back, including the neck), back (usually in a round shape and not involving the buttocks or shoulders), bathing trunk (mainly involving the genital region and buttocks, excluding the shoulders and neck), breast/belly (only distributed on the chest or abdomen and showing no overlap with bolero or bathing trunk), body extremity (only located on the extremities and excluding the shoulders and genital region) and body (a pattern combining bolero and bathing trunk patterns, and affecting almost the entire body) 20,21 .

Statistical analysis
Relationships between NCM and characteristics of CMN were evaluated using Fisher's exact test.Values of p < 0.05 were considered statistically significant.JMP Pro software (version 16.2.0;SAS Institute, Cary, NC, USA) was used for all statistical analyses.

Patients
Forty patients with CMN under one year old were included in this study.No patients were excluded due to motion artifacts during MRI.No family history of CMN or melanosis was identified.Two female patients treated for CMN before being referred to our institution were excluded, so a final total of 38 pediatric patients (20 males, 18 females) were included for Analysis 1.In the next step, 32 patients (17 males, 15 females) with CMN who had not undergone any MRI of the brain before referral were included for Analysis 2.
Only one female patient with bilateral amygdala lesions (right-side predominance) showed repetitive epileptic seizures and mild electroencephalographic examination abnormality during the course of her disease.
Pathological diagnosis for CMN was performed for all patients in our institution, and all specimens were confirmed to be negative for malignancy.

Diagnosis of NCM
Brain MRI was performed for 38 patients.Other than hyperintensity on T1WI, no patients showed brain abnormalities such as hydrocephalus, Chiari malformation, or syringomyelia.Representative cases of NCM are shown in Figs. 2 and 3.
The relationships between characteristics of CMN and NCM are shown in Table 2. Significant association with the presence of NCM were shown by Fisher's exact test for distribution of CMN (p = 0.002), size of CMN (p = 0.01), number of satellite nevi (p = 0.002), rugosity (p = 0.009) and nodules (p = 0.006).

CMN
The relationship between the distribution of CMN and the presence of NCM is shown in Table 3.In terms of anatomical localization, the most frequent sites of CMN were the scalp and back, followed by the flank.These localizations of CMN overlapped in most patients.No patients were categorized into "multiple CMN" or showed CMN on the hands, feet or lower legs.

Discussion
In this study, NCM was found in 31.6%(Analysis 1, 12/38) and 25.0% (Analysis 2, 8/32, no previous MRI examination and treatment group) of patients with CMN.The incidence of large CMN has traditionally been estimated as around one in 20,000-50,000 newborns 3,5,22,23 .The incidence of NCM among patients with CMN has varied between studies because of differences in study designs and the rarity of the pathology 3,5,22,23 .The accurate incidence of NCM among patients with CMN outside western countries remains unclear, since only a few studies have been published to date 24,25 .The evidence accumulated in the present study may contribute, at least among Japanese populations, to alleviating some of the anxiety of patients, families, and physicians before patients undergo brain MRI.The distribution of NCM lesions in this study, particularly the predominance of the amygdala, was consistent with findings from previous studies 3,7,11,[14][15][16][26][27][28] . Distriution, size and characteristic features (rugosity and nodules) of CMN and the number of satellite nevi were strongly associated with the presence of NCM.Among these, a tendency was seen for NCM to show a "bathing trunk"-type distribution, large-sized CMN and larger number of satellite nevi.CMN on the posterior axis (paraspinal, head, and neck regions) and the presence of a greater number of satellite nevi have appeared to represent the strongest risk factors for NCM in past studies [12][13][14]29,30 .The relationship between characteristics of CMN and the incidence of NCM should be clarified in further studies with larger cohorts.Whether all patients with CMN should undergo brain MRI remains an open question.A positive result on MRI for NCM may impact decisions regarding timing and extent of surgical management for CMN, because brain MRI can suggest whether NCM is stable or progressive and indications for aggressive treatment of CMN will be considered based on the prognosis of NCM 31 .Patients with NCM may be asymptomatic, but occasionally present with headache associated with hydrocephalus, seizure, mental retardation, and various abnormalities of the CNS, including Dandy-Walker syndrome, Chiari malformation, syringomyelia, and lissencephaly 7,15,32 .Brain MRI is currently the most appropriate modality to screen for coexistence of these abnormalities.However, some studies have cast doubt on the effectiveness of routine MRI itself as a screening test for NCM, particularly in patients who are neurologically asymptomatic 22,24 .
CMN are present at birth and have developed in utero in the absence of ultraviolet exposure.CMN is caused by somatic mosaicism in melanocytic precursors, and NRAS mutations can be found in 80% to 95% of CMNs 33 , and no BRAF mutation was found 34 .The decision for resection of CMN is complicating due to many factors, including family wishes, size and location of CMN, patient age, general health, and prognosis 17 .Intralesional injection of proinflammatory squaric acid dibutylester (SADBE) reportedly achieved major regressions of CMN in mice 35 .Although still in preclinical state, such topical approach may be an alternative of surgical resection.
The ectoderm cells destined to generate neural crest precursor cells undergo an epithelial-to-mesenchymal conversion and then migrate to diverse sites in the embryo where they differentiate into multiple derivatives such as neurons and melanocytes of the skin 36 .The epidermis represents the outermost layer of the skin, which is predominantly formed by proliferating and differentiating keratinocytes, and the epidermis also contains melanocytes, Merkel cells, and Langerhans cells.Of note, the skin constituents communicate cross-functionally through local nervous, immune, and endocrine systems to maintain essential epidermal functions 37,38 .
Several limitations should be noted in this study.First, the number of patients with CMN was limited.Although our institute is a high-volume center for patients with CMN in Japan, the relatively short enrollment period and inherent rarity of this pathology would have contributed to limiting the number of patients enrolled in this study.Further registry studies are expected in this area.Second, the number of patients with NCM itself was also quite limited.However, the evidence related to the incidence of NCM confirmed from brain MRI in this prospective cohort is valuable despite the small cohort.Third, particularly among infants, accurate evaluation of neurological and clinical manifestations is often difficult.Some neurologically symptomatic patients may have been overlooked in this study.Fourth, the prognosis of NCM was not evaluated in this study.
In conclusion, NCM was found in 31.6% of patients with CMN in this Japanese cohort, and in 25.0% of patients with no prior examination or treatment.The distribution, size and particular characteristic features (rugosity and nodules) of CMN and number of satellite nevi appear strongly associated with the existence of NCM.The results accumulated in the present study may guide a future registry study with a large cohort of CMN patients.

Figure 1 .
Figure 1.Flowchart of enrollment is shown.

Figure 2 .
Figure 2. A 4-month-old male with NCM.Hyperintensity lesions on T1WI are present in bilateral amygdalas (a) and right temporal cortex (b).T2WI shows slightly low signal intensities in the corresponding lesions (c,d).

Table 1 .
Characteristics of all patients with NCM.

Table 2 .
Relationship between characteristics of NCM and CMN.The size of the CMN was categorized by

Table 3 .
Distribution of CMN and presence of NCM.