1 INTRODUCTION

Agminated melanocytic nevi are characterized by the presence of multiple cutaneous clustered nevi.^1-3 These lesions can arise on an underlying nevus spilus, normal-appearing background skin, after radiation or chemotherapy, and in immunosuppressed states.^4-10 A commonly recognized pattern of agminated nevi is that of agminated Spitz nevi (ASN).^11-19 The World Health Organization defines Spitz nevi as melanocytic nevi with characteristic spitzoid cytomorphologic features harboring an HRAS mutation/gain or a kinase fusion involving ALK, ROS1, NTRK1-3, RET, MET, BRAF, or MAP3K8.^20-23 ASN have been shown to occur in the setting of nevus spilus as a result of a mosaic HRAS mutation, involving the entire area of the nevus spilus, along with copy number gains of 11p in the focal spitzoid proliferations.^{14, 16, 20, 24, 25}

However, there is limited knowledge regarding the pathophysiology of ASN outside of the setting of nevus spilus, and the potential for genomic fusions to play a role in agminated nevi.^{11, 15, 26} In this study, we present the clinical, histologic, and molecular findings of three cases of agminated melanocytic neoplasms harboring genomic fusions and therefore propose that genomic fusions can play a role in the development of agminated Spitz or other fusion-mediated melanocytic neoplasms.

2 MATERIALS AND METHODS

3 RESULTS

3.1 ROS1 fusion: Case 1

A 6-year-old male presented with a rapidly enlarging 15 mm well-defined red nodule on the right superior ear with a clinical impression of pyogenic granuloma (Table 1). Morphologically (Figure 1A–D), the lesion showed a compound polypoid and exophytic architecture with a junctional component surmounted by epithelial hyperplasia. Larger nests of melanocytes were present superficially and there was good maturation with descent. Cytology was epithelioid and spindle shaped with convincingly spitzoid features. The mitotic index was 3/mm². PRAME immunohistochemistry (IHC) was negative. Whole exome RNA-sequencing identified a ZCCHC8::ROS1 gene fusion. FISH studies for melanoma were negative. TERT-promoter mutation and CDKN2A mutations assessed by NGS studies were negative. The tumor was excised with negative margins and a final diagnosis of ROS1 fusion atypical Spitz tumor was given. About 5 months later, the patient presented with four new red papules inferior and outside of the surgical scar (Figure 2), showing highly similar histopathologic (Figure 1E, F) features to the original biopsy. ROS1 break-apart FISH studies performed on the second lesion also identified a break-apart in the ROS1 gene. ddPCR showed no evidence of a TERT-promoter mutation. The lesion was re-excised with clear margins. Fourteen months after the initial recurrence, the patient presented with another red papule outside the original re-excision site. A wide excision was performed with clear margins and no evidence of distant disease was found in CT scans. An additional 4 months of follow-up showed no evidence of clinical recurrence thus far.

TABLE 1. Summary of clinical and genomic data in three cases of agminated presentation of fusion-driven melanocytic neoplasms.

Case	Age	Gender	Location	Clinical impression	Diagnosis	Follow-up (months)	Recurrence	Fusion
1	6-year old	M	Right ear	15 mm well-defined red nodule. R/o pyogenic granuloma	AST	23	Yes, four new red papules at 5 months after excision and one new red papule at 19 months after excision	ZCCHC8::ROS1
2	4-month old	F	Left foot	Multiple blue–black nodules. Thrombosed vascular malformation vs. atypical melanocytic neoplasm	PFMT-PEM	18	Yes, additional 2 mm solitary grayish blue papule at 13 months after excision	ATP2B4::PRKCA
3	42-year old	F	Left shin	Several pigmented brown papules. Malignant melanoma with in-transit metastasis	PFMT-PEM	2	No	MPZL1::PRKCA
	Mean follow-up time: 14.3 months

• Abbreviations: AST, atypical Spitz tumor; PFMT-PEM, PRKCA fusion melanocytic tumor-pigmented epithelioid melanocytoma.

3.2 PRKCA fusion: Cases 2 and 3

In Case 2, a 4-month-old female presented with multiple blue–black nodules on the left plantar surface (Figure 3) present since birth, while in Case 3, a 42-year-old female presented with several pigmented papules located on the left distal shin (Figure 4). The larger brown papule for Case 3 had been present for years with additional clustered pigmented lesions arising in the last year. The clinical impression for Cases 2 and 3 was thrombosed vascular malformation versus atypical melanocytic neoplasm and malignant melanoma with in-transit metastasis, respectively. The histology for Cases 2 (Figure 5) and 3 (Figure 6) were similar. Both cases showed predominately dermal melanocytic proliferations of pigmented epithelioid melanocytes of intermediate size forming a solid cellular nodule centrally. The cells were highly monotonous with vesicular nuclei, prominent nucleoli, and pigmented cytoplasm. In the more peripheral regions, there were smaller aggregates of similar epithelioid melanocytes as well as some dendritic melanocytes with melanophages separated by dense bundles of collagen. The agminated lesions in both cases showed identical histology with no signs of tumor progression. Whole exome RNA-sequencing identified an ATP2B4::PRKCA fusion for Case 2 and an MPZL1::PRKCA gene fusion on two of the lesions that underwent sequencing for Case 3. Pathogenic variants in TERT-promoter were not identified in NGS studies for Case 2 and in ddPCR for Case 3. In both cases, the tumor was completely resected with clear margins, with a final diagnosis of a PRKCA fusion melanocytic tumor that currently falls under the rubric of pigmented epithelioid melanocytoma. Approximately 13 months later, the patient from Case 2 presented with an additional 2 mm solitary grayish blue papule outside the scar from the original re-excision. This lesion showed similar histopathologic features to the primary lesion. Mutational profiling by NGS studies on the new lesion also identified an ATP2B4::PRKCA fusion. The lesion was re-excised with clear margins and there has been no evidence of clinical recurrence for an additional 5 months. For Case 3, the patient has had no evidence of recurrence during a brief follow-up period of 2 months.

4 DISCUSSION

For many years, mutations in HRAS were the only known pathogenic driver for Spitz neoplasms.³² Since then, HRAS has also been recognized as a driver for nevus spilus and investigators have shown that while the entire nevus spilus carries an HRAS mutation, agminated Spitz arising in a nevus spilus also have copy number gains in 11p, where the HRAS gene resides.^{14, 16, 20, 24, 25} Since these discoveries, much has been learned about the pathogenesis of Spitz including the fact that a significant majority of Spitz, as well as some other classes of melanocytic neoplasms such as PRKCA fusion melanocytic nevi/tumors, are the result of genomic fusions.^{20, 21, 33-35} However, there have been very few studies investigating whether fusions can play a role in agminated presentations of melanocytic neoplasms.

Understanding whether genomic fusions can play a role in the development of agminated Spitz has significant clinical consequences. Spitz tumors can cause great anxiety for clinicians and pathologists. Pathologists often have trouble definitively excluding the possibility of a malignancy. While the overwhelming majority of these cases have an excellent prognosis, they are often reported as tumors of uncertain malignant potential. The presence of a multiplicity of lesions or recurrence of a lesion outside the scar in the setting of a lesion of uncertain malignant potential may be interpreted as an indicator of malignancy with the recurrence being a localized metastasis. This raises a highly disparate potential differential diagnosis of benign agminated Spitz nevus/tumor versus clinical Stage 3 melanoma with local regional metastasis, which has a 5-year survival rate of 69% (i.e., AJCC eighth edition N2c).³⁶ Considering the great disparity in prognosis and optimal therapy, being able to optimally distinguish the agminated disease from Spitz melanoma with local recurrence is critical.

In this study, we identified three cases in which the differential diagnosis included agminated Spitz, agminated PRKCA fusion melanocytic nevus/tumor, or melanoma with local recurrence. In all three cases, we favor the diagnosis of a fusion-based agminated melanocytic neoplasm. The concept of the presence of mosaicism of a gene fusion giving rise to multiple identical tumors has been demonstrated by other investigators both in cutaneous and non-cutaneous tumors. There have been several reports of gene fusion mosaicism related to renal cell carcinoma (RCC) development involving a chromosome 3 translocation.^37-41 Bodmer et al. reported a familial case of RCC involving a t(2;3)(q35;q21) chromosomal translocation with mosaicism.^{37, 38} One patient developed three different anatomically distinct tumors on two different occasions, all of which were found to have a translocation-derived der(3) chromosome loss. Subsequent sequencing of the VHL tumor suppressor gene revealed distinct mutations in two of the three tumors, suggesting the tumors developed independently and not as metastases of a primary tumor. Non-cutaneous gene fusion mosaicism has also been demonstrated in acute lymphoblastic leukemia (ALL) including TEL::AML1, KMT2A::AFF1 (previously known as MLL::AF4) and BCR::ABL1 rearrangements.^42-45 Interestingly, the TEL::AML1 (now referred to as ETV6::RUNX1) gene aberration has been shown to be prenatal in origin and is detectable in blood spots of 1% of randomly selected newborns.^46-51 However, only approximately 1% of newborns harboring the translocation develop ALL,^{44, 52} suggesting that embryonic somatic mosaicism allows for the proliferation of mutations that can lead to neoplasia.

There are various other studies supporting the concept of gene fusions in ASN. Gene fusion mosaicism in ASN has been discussed in two reported GOPC::ROS1 positive cases^{11, 15} and in one case with a complex three-way rearrangement involving TRPM1, PUM1, and LCK.²⁶ Goto et al. described two patients with clustered Spitz neoplasms harboring a GOPC::ROS1 fusion detected in the tumoral cells as well as in the normal-appearing basal melanocytes located in the interlesional epidermis.¹¹ Similarly, the presence of a translocation in the melanocytes of the tumor and surrounding normal epidermis was also observed in a novel TRPM1::PUM1::LCK ASN case.²⁶ Robertson et al. reported a case of facial ASN involving a GOPC::ROS1 kinase fusion.¹⁵ Sequencing results revealed a GOPC::ROS1 genetic rearrangement affecting chromosome 6 that was not present in DNA extracted from blood, suggesting that the Spitz nevi emerged from cutaneous genetic mosaicism.

In all three of our cases, identical gene fusions involving protein kinases were identified in both the primary and secondary lesions. In Case 1, the secondary lesions occurred following the excision of the initial lesion, while in Cases 2 and 3 the patient presented with multiple lesions. In all cases, there were no signs of tumor progression between the primary and secondary lesions. None of the primary lesions met histologic criteria for malignancy as the cytology was relatively bland, there was no evidence of ulceration, epidermal consumption or subcutaneous involvement. In the secondary lesions, the histologic grade was comparable with no evidence of greater cytologic atypia, increased mitotic count, necrosis, or other features typical of metastatic lesions. At the genomic level, the secondary lesions also lacked features suggestive of tumor progression. In all the tested cases, the primary and secondary lesions showed no evidence of TERT-promoter mutations and were negative for chromosomal copy number aberrations by the melanoma FISH assay.

All the patients continue to be followed clinically and none of them have had evidence of lymph node or systemic involvement and they all remain disease-free following re-excision. Overall, we believe the studies from Goto et al.^{11, 26} in conjunction with our current study, support the concept that some agminated melanocytic neoplasms including agminated Spitz or agminated PRKCA fusion melanocytic nevi/tumors are the result of mosaic gene fusions. Longer follow-ups and a larger study with additional cases will further help support this hypothesis. However, we believe that since the consequence of overdiagnosis of an agminated Spitz or PRKCA fusion melanocytic nevi/tumors can result in considerable overtreatment of patients with benign lesions, it is important to report these findings at this time. We believe that this is another example of how understanding molecular genomics and the pathogenesis of various subtypes of melanocytic neoplasms can result in considerable advances in our diagnostic acumen and improve patient quality of care.

CONFLICT OF INTEREST STATEMENT

Dr. Pedram Gerami has served as a consultant for Castle Biosciences and has received honorarium for this. Other authors declare no conflict of interest.

ETHICS STATEMENT

This study was approved by of our Institutional Review Board (IRB) (STU00001127). Given the retrospective chart review design, a waiver of HIPPA authorization was granted for research purposes and informed consent was not required for this study.