Spitzoid Melanoma


There is currently no consensus definition on what constitutes a “spitzoid” melanoma. The term is most often used in a practical sense referring to a diagnostic pitfall: a melanocytic neoplasm that at first glance simulates the appearance of a Spitz nevus but is classified as melanoma, either based on clinical follow-up (distant metastasis or death), compelling scientific evidence (e.g., detection of genomic aberrations associated with melanoma), or consensus review by experienced pathologists. Hence one traditionally used definition of spitzoid melanoma is a melanocytic neoplasm that could be or has been diagnostically confused with a Spitz nevus or atypical Spitz tumor ( Figs. 18.1–18.3 ). This definition allows for significant latitude in the use of the term “spitzoid melanoma,” because pathologists vary not only in what they regard as “spitzoid” but also in their threshold for classifying a lesion as melanoma (versus nevus or tumor of indeterminate biologic potential).

Fig. 18.1
(A) Melanoma from the head of a 10-year-old boy. Several experts classified this lesion as a Spitz nevus or Spitz tumor, but brain metastases developed and the child died of metastatic melanoma. The lesion displays some spitzoid features (epidermal hyperplasia, clefts between epidermis, and junctional nests), but its silhouette is asymmetric. (B) Nuclear atypia and mitotic figures are seen. (C) Fluorescence in situ hybridization (FISH) targeting RREB1 in red showing many copy number gains at 6p25 in the vast majority of the cells. (D) FISH targeting CCND1 showing copy number gains at 11q13 in the vast majority of cells. (E) FISH targeting CDKN2A showing homozygous deletions at 9p21 in nearly all cells.

Fig. 18.2
Melanoma From the Trunk of a 27-Year-Old Man.
Several experts favored a diagnosis of atypical Spitz tumor. (A) The lesion lacks maturation and displays an expansile base. Fluorescence in situ hybridization (FISH) studies were performed but were negative for copy number aberrations at the targeted loci. (B) Fusiform tumor cells display severe nuclear atypia. (C) A sentinel lymph node biopsy showed rare subcapsular melanoma cells. This patient subsequently developed several large bulky lymph node metastases.

Fig. 18.3
Melanoma Misdiagnosed as Spitz Nevus.
(A) Symmetric silhouette of a compound pigmented spindle and epithelioid cell neoplasm. (B) Cleft between junctional nest and epidermis. (C) The tumor shows evidence of maturation and breaks up into single cells at its base. (D) A few mitotic figures are seen at the base. Metastatic melanoma was found in the regional lymph node draining the site of this tumor 6 years after its removal. The patient died of metastatic melanoma. The primary was positive by the melanoma FISH test.

(Courtesy Dr. K.J. Busam.)

The recent discovery that the initiating oncogenic driver in the majority of spitzoid neoplasms is a translocation often involving ROS, RET, BRAF, NTRK , or ALK , as opposed to driver mutations in BRAF or NRAS , provides an opportunity for a molecular pathway-based tumor classification. A malignant Spitz tumor (or Spitzian melanoma) could be defined as a melanocytic neoplasm with an initiating genomic driver typical of Spitz nevi/tumors. The problem with this approach is that genomics are often not available or attainable.

In this chapter, we use the following three criteria for designating a melanocytic neoplasm as spitzoid melanoma: (1) the lesion displays microscopic features that could lead to confusion with a Spitz nevus/tumor, but an adverse event such as distant metastasis has occurred in a way that make the tumor a plausible primary melanoma (see Fig. 18.3 ); (2) the lesion looks spitzoid, but histopathologically there are sufficient atypical features, such as nuclear pleomorphism or deep and atypical mitoses, that a melanoma is much more plausible than a nevus or borderline tumor ( Fig. 18.4 ), and (3) a spitzoid neoplasm is found to harbor genomic aberrations, such as homozygous deletions of 9p21, which have been strongly associated with adverse prognosis ( Figs. 18.1, 18.5 and 18.6 ). There is anecdotal-level evidence that spitzoid melanoma with documented Spitz nevus/tumor–related genomic aberrations, such as kinase fusions ( Fig. 18.7 ), may have a better prognosis than conventional melanoma of similar thickness.

Fig. 18.4
(A) Melanoma with expansile nodular growth surrounded by sclerotic stroma. (B) Large epithelioid tumor cells with pink cytoplasm display nuclear atypia. Mitoses are present. The majority of tumor cells had homozygous deletion of 9p21 by fluorescence in situ hybridization (FISH).

Fig. 18.5
Melanoma on the Arm of an 8-Year-Old Boy.

(A) Erythematous plaque with satellites at the superior pole of the lesion. The tumor was associated with multiple in-transit metastases and bulky lymph node disease. After complete excision the boy has been disease free for 10 years now.

(B) Histopathology showing an ulcerated tumor with sheetlike growth.

(C) The tumor cells display spitzoid features, nuclear pleomorphism, and mitoses.

(D) There is homozygous deletion of 9p21 in nearly 100% of the tumor cells. The Cep9 control showed a completely normal diploid pattern in all cells. Array CGH studies documented deletion of 9p21 as the only chromosomal copy number change.

Fig. 18.6
Malignant Spitz Tumor of Childhood.
(A) Shave biopsy of a spitzoid lesion from a 2-year-old girl. (B) The tumor cells display atypia. (C) Lesional melanocytes are immunohistochemically negative for p16. (D) Array CGH reveals loss of 9p. (E) FISH documents homozygous deletions of 9p in all lesional melanocytes (only the centromeric probe signal is seen). (F) Metastatic melanoma in the regional lymph node.

(Courtesy Dr. K.J. Busam.)

Fig. 18.7
Spitzoid Melanoma From a 12-Year-Old Girl.
(A) Compound melanocytic neoplasm of spindle and epithelioid cells with complex growth pattern and high cellularity. (B) Large aggregates of tumor cells underneath the ulcer. (C) The tumor cells display nuclear pleomorphism. They lacked expression of p16 and nearly all lesional melanocytes displayed homozygous deletions of p16. The regional lymph node was positive for metastatic melanoma.

Lesions reported as spitzoid melanoma in the literature are likely a heterogeneous group of neoplasms. The term spitzoid melanoma has been applied by pathologists to a range of lesions, including conventional melanomas with a subpopulation of tumor cells with spitzoid cytology or nevoid melanomas with secondary epidermal changes often associated with Spitz nevi, such as epidermal hyperplasia or clefting between the hyperplastic epidermis and the melanocytic nests. We do not recommend using the term “spitzoid” melanoma for tumors with only minor spitzoid features. We propose restricting the term to melanomas with shared microscopic and genomic features of Spitz nevi/tumors (i.e., a Spitz nevus or tumor is a serious consideration in the differential diagnosis).

Clinical Findings

From a clinical perspective, spitzoid melanomas may be indistinguishable from Spitz nevi or tumors. Lesions can be macular, present as a small papule, or as a large nodule. They may be pigmented or amelanotic. Some of the lesions may have a verrucous epidermal hyperplasia resembling verruca. Most lesions will have at least some degree of symmetry. Dermoscopically, as in ordinary nevi, there may be either homogeneous pink or brown appearance or reticular or globular appearance; alternatively, there may be one of the more specific patterns for a spitzoid neoplasm such as either (1) starburst pattern, (2) tiered globular pattern, or (3) negative pigment network with or without shiny white streaks. Although clinical and dermoscopic features can help designate a lesion as spitzoid, they cannot distinguish between benign and malignant lesions. Spitz nevi can have clinically disturbing features such as history of rapid growth, spontaneous bleeding, erosions, or ulceration. On the other hand, a spitzoid melanoma may clinically present as a discrete circumscribed symmetric nodule.

Histopathologic Findings

There is a spectrum of microscopic features that can be seen in association with a spitzoid melanoma, paralleling that of Spitz nevi and atypical Spitz tumors. Some lesions are pigmented, others are amelanotic. Some involve the epidermis and dermis, others only the dermis. Some tumors are composed of predominantly large epithelioid melanocytes, others of fusiform cells or a mixture of spindle and epithelioid cells. The degree of cytologic atypia is variable, but often nuclear pleomorphism, a lack of maturation, and mitotic figures, especially deeply located mitoses, are present (see Figs. 18.3 and 18.4 ). Like Spitz nevi, a spitzoid melanoma may be associated with epidermal hyperplasia, clefts, Kamino bodies, and inflammation.

Several attempts have been made to distinguish spitzoid melanoma from atypical Spitz nevi/tumors by microscopic criteria alone. This can be very difficult and may at times be impossible without additional evidence from ancillary studies. Instead of rendering a specific diagnosis, Spatz and Barnhill have proposed to grade spitzoid neoplasms according to features associated with risk for recurrence. Their risk stratification system includes five parameters: (1) age older than 10 years, (2) lesional diameter greater than 10 mm, (3) involvement of subcutaneous fat, (4) ulceration, and (5) mitotic activity. The more of these parameters that are present and the higher the mitotic count, the higher the risk for recurrence. However, this system has not been independently validated by others. A more recent study, which included 75 atypical spitzoid neoplasms with long-term follow-up, found the following factors most related to aggressive behavior: (1) high mitotic count, (2) deep mitotic figures, (3) high-grade cytologic atypia, and (4) ulceration. Thus, not surprisingly, larger tumors with high-grade nuclear atypia, ulceration, and deep mitoses are at higher risk for recurrence ( Table 18.1 ). In other words, they are more likely malignant and may be designated as spitzoid melanoma (see Fig. 18.4 ). However, the problem is that there are many tumors, particularly in children, that may meet multiple of the above criteria but are associated with an indolent clinical course, attesting to the difficulty of establishing a diagnosis of spitzoid melanoma by light microscopic parameters alone.

TABLE 18.1
Histopathology Features Associated With Risk for Recurrence
Morphologic Feature Spatz 1999 (Conditional Probability Ratio >1.5) Gerami 2014 ( P -Value <.05)
Age 1.72 (11–17 years) Not significant
Ulceration 2.40 0.005
Mitotic count 1.72 (6–8/mm 2 ) 0.001
5.16 (>8/mm 2 )
Deep marginal mitoses Not Assessed 0.005
Diameter >10.0 mm 1.72 Not Assessed
Clark level 2.15 (Level 5) Not significant
Nuclear atypia Not Assessed 0.013
Asymmetry Not Assessed 0.02

Hence for a diagnosis of spitzoid melanoma, unless the microscopic features for malignancy are compelling and there is a consensus among experts on the diagnosis, we recommend attempting to obtain additional evidence from ancillary studies, such as genomic testing, bearing in mind that this method has its own limitations and pitfalls. Not all spitzoid melanomas display cytogenetic aberrations readily detectable by conventional tests (fluorescence in situ hybridization [FISH], comparative genomic hybridization [CGH]). On the other hand, some benign Spitz nevi or indolent Spitz tumors may have cytogenetic aberrations which have not been associated with adverse outcome and therefore do not warrant a diagnosis of melanoma. Hence interpretation of CGH and FISH requires a working knowledge of the significance of a range of potential chromosomal aberrations that may be seen in Spitz lesions. Based on the experience that fully malignant spitzoid melanomas are very rare in children, pathologists tend to have a high threshold for rendering a diagnosis of spitzoid melanoma in this age group and tend to default to reporting lesions as atypical Spitz tumor when the histopathologic and/or cytogenetic findings of a Spitz lesion are too atypical for a “nevus.” There is a greater willingness to accept a diagnosis of spitzoid melanoma in an adult, especially if associated with chronic sun damage (located in head and neck area, presence of marked solar elastosis).

Immunohistochemical Findings

Immunohistochemical studies are of limited value for the diagnosis of spitzoid melanomas. For amelanotic intradermal nodular tumors, using antibodies for melanocyte differentiation antigens can help to support the melanocytic nature of the tumor. On occasion, using HMB-45 can help because diffuse labeling for gp100 would be unusual for a Spitz nevus and therefore adds to concern for a possible melanoma.

Some spitzoid melanomas express p16; others do not. However, complete loss of labeling of a spitzoid melanocytic neoplasm with atypical features may prompt further analysis by fluorescence in situ hybridization.

Molecular Findings

Genomic aberrations associated with spitzoid melanoma likely include many of those that have been found in Spitz nevi and Spitz tumors, such as translocations involving BRAF, NTRK1, ALK, ROS, RET , or MET . HRAS aberrations have to date not yet been reported in a bona fide spitzoid melanoma. It seems that Spitz lesions with an isolated gain of 11p and HRAS amplifications or mutations represent a stable variant of Spitz nevus, which is unlikely to progress to melanoma. It is not known how many spitzoid melanomas contain BRAF, NTRK1, ALK, ROS, and RET translocations. Based on preliminary experience, BRAF translocations are not uncommonly associated with melanoma ( Fig. 18.8 ). In a publication involving a set of 17 ALK -translocated, 17 NTRK1 -translocated, and 14 BRAF -translocated spitzoid neoplasms that included 4 spitzoid melanomas, 3 were BRAF-translocated and 1 was ALK-translocated.

Fig. 18.8, Malignant Spitz Tumor of Childhood.

Although the detection of such a translocation in a melanocytic tumor supports its genomic classification of a spitzoid melanocytic neoplasm, the translocations per se do not inform about whether the lesions are benign or malignant.

For diagnostic or prognostic purposes, the detection of segmental genomic gains or losses has been found to be more relevant. Several studies have identified homozygous deletions of 9p21 as a characteristic feature of many spitzoid melanomas and as having a statistically significant correlation to distant metastatic behavior. Although most conventional melanomas have on average approximately seven segmental chromosomal copy number aberrations, some spitzoid melanomas display only a single aberration, such as homozygous deletion of 9p21, which implies risk for local, regional, and/or distant metastatic disease. In a study of 75 atypical spitzoid neoplasms with long-term follow-up, among 11 cases with disease progression beyond a sentinel lymph node, 9 had homozygous deletions of 9p21, and in several of the cases this was the only identified chromosomal aberration in cases assessed by both CGH and FISH. However, it should be kept in mind that this was a case-control study enriched for aggressive cases. In a pediatric cohort study of 246 children with atypical Spitz tumors, 19 had homozygous deletions of 9p21. Follow-up was available in 85 patients, and 4 patients had tumor spread beyond the sentinel lymph node. Two patients had recurrence following complete excision, and one had distant metastatic disease. Three of the four patients with tumor spread beyond the sentinel lymph node and both patients with recurrent disease had homozygous deletions of 9p21. Hence, assuming no adverse events in the patients lost to follow-up, 4 of 19 patients with homozygous deletions of 9p21 had tumor spread beyond the sentinel lymph node, and 2 of 19 had tumor recurrence following wide excision which consisted of one case with in-transit metastasis and a second case with distant metastasis. The cohort study findings suggest the majority of patients with homozygous 9p21 deletions will not develop recurrent or metastatic disease; however, identification of this subgroup is of significant importance because it greatly distills down the group of atypical spitzoid neoplasms to a select group with significantly higher risk. In our opinion, cases that have the combination of high-risk morphologic features in conjunction with high-risk molecular aberrations, such as, but not limited to homozygous 9p21 deletion, may be best classified as spitzoid melanoma ( Table 18.2 ). We do not consider homozygous deletions of 9p21 per se as proof of malignancy. They need to involve most tumor cells and be associated with atypical histopathologic findings to warrant a tumor classification of melanoma.

TABLE 18.2
Various Molecular Aberrations Commonly Detected in Spitzoid Neoplasms
Genomic Events in Spitz Nevi Genomic Changes That Can Be Seen in Spitz Nevi and Indolent Spitz Tumors Aberrations Favoring Melanoma
  • HRAS mutations

  • Copy number gains in HRAS (11p)

  • Fusions involving RET, ROS, BRAF, ALK, NTRK1 a

  • Tetraploidy

  • 6q23 deletion

  • Heterozygous 9p21 deletion

  • 3p21 deletions

  • 7q34 gains

  • Copy number gains involving a kinase fusion

  • Tetraploidy

  • Whole chromosome changes

  • Homozygous 9p21 deletion b

  • Gains in 11q13

  • Gains in 6p25

  • Gains in 8q24

  • Multiple segmental gains and/or losses of different chromosomes

  • TERT promoter mutations/methylation

a Fusions can be seen across the full spectrum of Spitz lesions (nevus, atypical tumor, and melanoma).

b Homozygous deletions in the vast majority of tumor cells.

It has also become evident that some clonal segmental copy number aberrations are acceptable for a benign melanocytic nevus. For example, spitzoid neoplasms in which the only detectable chromosomal copy number aberration is a 6q23 (MYB) deletion usually have an indolent course. Other examples of segmental copy number changes in benign melanocytic neoplasms include 3p21 deletions in BAP1-related spitzoid neoplasms and segmental copy number gains involving the kinase domain of the tyrosine kinase involved in a translocation such as BRAF, NTRK1, ALK, ROS, RET, or MET . Although the presence of these chromosomal aberrations is often associated with an indolent clinical course, it does not guarantee a benign diagnosis. Such aberrations may be seen in Spitz tumors with lymph node deposits or on occasion (usually in context with other cytogenetic aberrations) even in a spitzoid melanoma. However, it is important not to equate the presence of any chromosomal aberrations with malignancy. There is a range of chromosomal copy number changes that may be seen in benign melanocytic nevi (spitzoid or otherwise). Currently, due to the lack of long-term follow-up studies, we do not know yet the significance of many chromosomal aberrations, including those seen in conventional melanoma in the setting of atypical spitzoid neoplasms.

Some studies have suggested that, in addition to homozygous 9p21 deletions, clonal segmental gains of 6p25 with or without accompanying 6q23 deletions, as well as 11q13S are also indicative of higher risk for recurrence. Both of these chromosomal aberrations had a statistically significant correlation with disease progression beyond the sentinel lymph node in one study. However, in the setting of atypical spitzoid neoplasm, both of these changes are less of a marker for aggressive behavior than homozygous deletion of 9p21 (see Figs. 18.5–18.7 ). Considering how few cases of atypical spitzoid neoplasms ever lead to distant metastases, it is difficult to truly quantitate the risk level.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here