Spitz melanoma is a very rare disease in the pediatric population, and differs from adult melanoma [1],[2], with variation even between infants (0–10 years of age) and adolescents (11–19 years of age) [3],[4],[5]. In the prepubescent population, the majority of melanomas are of the pediatric Spitz melanoma/atypical Spitz tumor type, as this patient was, with a second biologically separate category of melanomas arising in association with giant congenital nevi or de novo. Prognostically, the former tends to be more low-grade or borderline malignant in behavior, despite metastasis to lymph nodes [6],[7], while the latter is aggressive and invariably fatal. A note is made of recent changes in categorization by the WHO in 2018, which distinguishes spitzoid melanoma, a purely morphological diagnosis, from Spitz melanoma, a subset of spitzoid melanoma with both the morphological features and the specific genetic signatures of mitogen-activated protein kinase (MAPK) activating alterations, which is present in this case [8]. This case report highlights age as a misleading characteristic in the diagnosis of Spitz melanoma, and a literature review is performed to examine trends, presentation, and prognosis in previous diagnoses of pediatric Spitz melanoma.

A 4-years-8-months-old Caucasian girl presented with an 18-month history of a 2.5 × 2.0 cm polypoid lesion with a 5 mm base on the right lateral lower chest wall, which clinically resembled a pyogenic granuloma (Figure 1). It had been enlarging since it was first discovered, frequently bled and wept, and was painful to the touch. There were no abnormal nodes palpable in the groin, axilla, or neck. The patient had no known past medical history or relevant family history. Specifically, there was no family history of melanoma or other genetic syndromes that predisposed this patient to developing malignancies.

The lesion was removed by a pediatric surgeon (BHC) approximately three weeks later with 3 mm clinical margins. She had a postoperative cellulitis wound infection that resolved with antibiotics. The pathology was initially reported as an atypical Spitz nevus with positive deep margins with concerns for melanoma. However, there was hesitation to diagnose the lesion as melanoma due to the patient’s age. Interdepartmental dermatopathology review agreed with the diagnosis of atypical Spitz nevus, with concerns for melanoma. An external dermatopathology review again diagnosed the lesion as an atypical Spitz tumor with high grade melanocytic atypia and features concerning for melanoma. A further quaternary care dermatopathology review made the diagnosis of a pediatric-type Spitz melanoma with ALK rearrangement.

Subsequently a local expert oncologic surgeon was consulted (BH), and there was discussion of management with other North American children’s hospitals. The patient also underwent further workup, including a complete staging with computed tomography (CT) of the chest, abdomen, and pelvis as well as a head magnetic resonance imaging (MRI) which showed no evidence of metastatic disease. A sentinel lymph node study showed focal activity in the right axilla in keeping with activity in sentinel nodes. This led to the decision to proceed with a scar re-excision with 2 cm clinical margins, flap reconstruction, and sentinel lymph node biopsy. Six nodes were excised from the right axilla, five of which were positive for metastatic melanoma.

The patient recovered well and has subsequently undergone close clinical follow-up with clinical exams and ultrasound imaging of the chest for recurrence. As of 16 months following the initial resection, the patient remains free of disease.

Investigations

Routine histology of the original lesion

Basic histology revealed a 2.5 cm polypoid tumor with neutrophilic crusts (see Figure 2 and Figure 3). The dermal tumor contained both epithelioid and spindle cells arranged in solid nests and sheets with abundant eosinophilic cytoplasm with no maturation, and demonstrated significant cellular atypia throughout, with bizarre nuclei and prominent nucleoli, nuclear pseudoinclusions, and multi-nucleated giant cells (see Figure 4 and Figure 5). Mitotic rate was brisk at 5/mm². Tumor depth was at least 9.5 mm, with positive deep margins.

Immunohistochemistry

The tumor cells were strongly positive for melan-A, WT-1, S100, and CyclinD1. It was focally positive for HMB45 in the upper dermis, a feature found more commonly in benign nevi. It was also patchy positive for p16 (CDKN2A), demonstrating partial loss of expression. This was later confirmed with fluorescence in situ hybridization (FISH) which showed homozygous deletion of this tumor suppressor gene. Anaplastic lymphoma kinase was diffusely positive in the tumor cells. Ki67 was diffusely positive, demonstrating a high proliferation index. Figure 6 and Figure 7 show micrographs of selected stains.

The tumor cells were negative for AE1/AE3 (pankeratin), SMA, CD34, myogenin, and desmin.

Molecular pathology

Fluorescence in situ hybridization testing using a melanoma kit for RREB1, MYB, CEP6, and CyclinD1 (11q13) probes was positive. While this did not equate to a diagnosis of malignant melanoma, this kit was found to be positive in 80–90% of diagnosed melanomas, and negative in 95–100% of benign nevi. However, benign Spitz nevi do have a higher incidence of positivity (10–35%), possibly due to polyploidy [9],[10].

Fluorescence in situ hybridization demonstrated homozygous loss of p16 (CDKN2A) tumor suppressor gene in the tumor. Break-apart FISH showed an ALK rearrangement.

Next-generation sequencing was performed on the tumor which detected no clinically relevant variants (i.e., class 1–3A) [11] in BAP1, BRAF, CDK4, CDK6, CDKN2A, GNA11, GNAQ, KIT, NRAS, EIF1AX, SF3B1, or HRAS. Sanger sequencing did not reveal a telomerase reverse transcriptase (TERT) promoter mutation in this tumor, which was a favorable prognostic factor.

Diagnosis of melanoma in the pediatric population is challenging, as there are numerous benign histologic mimics, especially in the Spitz nevus/Spitz melanoma spectrum. Other possible differential diagnoses can include Reed nevi, cellular blue nevi, or pigmented epithelioid melanocytoma. An extensive literature search for cases of cutaneous pediatric melanoma (0–10 years of age) was performed and showed that since 1987, only 19 individual cases of cutaneous pediatric melanomas have been published, highlighting the rarity of the diagnosis in this age group. See Table 1 for details on these case reports. Pediatric-type Spitz melanomas rarely develop distant metastases despite nodal metastases [6],[12]. This is supported by our review; a large proportion (11/19) of the published cases appear to be clinically well on follow-up at around 12 months, 6 of these 11 cases reported a Breslow depth greater than 4 mm and had positive sentinel lymph nodes.

Another study [13] reported clinical findings similar to our own case, in which the presented lesion appeared to mimic a pyogenic granuloma. The lesion evolved into a rubbery, erythematous papule over one month, and began to ulcerate and bleed, similar to our presented case. While a pyogenic granuloma would be overwhelmingly more likely than melanoma in a 10-month-old child, our case demonstrates that melanoma should be considered in the differential, and that changes in the lesion’s appearance should be closely monitored.

Management of Spitz melanoma and atypical Spitz lesions in the pediatric population generally derives from guidelines for adult melanoma [12]. A sentinel lymph node biopsy (SLNB) is often performed to assess for the need for a complete lymph node dissection and adjuvant therapy. In children, while this is prognostically significant, it does not improve melanoma-specific survival [14].

Histologically our patient’s lesion met criteria for a malignant lesion. However, given the history and the patient’s young age, three separate groups of pathologists from two quaternary academic centers were hesitant to label the tumor as malignant. Despite the use of immunohistochemistry, it was in fact only after exhaustive molecular analysis of the tumor at a third, subspecialized academic center in conjunction with the investigations performed prior that allowed the final diagnosis of pediatric-type Spitz melanoma to be made.

This case demonstrates two major lessons. The first is that age can be a misleading factor in reporting atypical melanocytic lesions, and may deter pathologists from diagnosing a tumor as malignant. While clinical history is indispensable in pathology practice, the patient’s age should not distract us from the histology. Children can get melanoma, although the clinical behavior generally differs from melanoma in adults. The second lesson highlights the usefulness of molecular pathology as a future adjunct in the practice of pathology. Entities that overlap morphologically or that present with conflicting histological features may be more easily categorized based on the presence of genetic variants or deletions. In our case, p16 deletion and ALK translocation were supporting features that in conjunction with histology clinched the diagnosis of melanoma.