The assessment of melanoma by light microscopy with hematoxylin-eosin staining remains an often subjective process. However, there are additional diagnostic measures that may be of utility, such as immunohistochemical staining and genetic evaluation. Adjunctive genetic assessment to augment the diagnosis of melanoma includes comparative genomic hybridization or fluorescent in situ hybridization to assess for gains or losses in genetic material, or gene expression profiling in some form, to ascertain the expression of genes associated with malignancy. Although these techniques may bolster the dermatopathologic assessment of melanoma, none of them, at the present time, are singularly diagnostic. Additional developments in the genetic assessment and prognostication of melanoma are anticipated.
Key points
- •
Melanoma is an often subjective diagnosis made using light microscopy and H&E staining, and the dermatopathologic impression may be augmented by other means.
- •
Additional means for augmenting a histologic impression of melanoma include immunohistochemical (IHC) staining and/or genetic testing.
- •
In problematic lesions, IHC stains used to refine or refute the diagnosis of melanoma may include Melan-A/MART-1, S100, Ki67, p16, and HMB-45.
- •
In problematic lesions, adjunctive genetic assessment of melanoma used to refine or refute the diagnosis of melanoma include comparative genetic hybridization (CGH), fluorescent in situ hybridization (FISH), and gene expression profiling (GEP).
- •
There exist new genetic means to stratify risk and gauge prognosis in melanoma, and as additional adjunctive therapy expands, and surveillance protocols are further refined, growth in this area is anticipated.
Introduction
At present, the histologic diagnosis of skin cancer remains a critical step in the evaluation and management of skin disease. For the foreseeable future, and from a medicolegal standpoint in particular, a histologic report of cancer is requisite for additional intervention. For example, melanoma requires histologic staging information (eg, Breslow depth, presence/absence of ulceration, dermal mitotic activity) to select appropriate management. Even entrance into oncologic trials for melanoma, using new agents and new protocols, is affected by this histologic staging information.
In dermatopathology, the underlying practical means for rendering a cancer diagnosis has changed little in the last 140 years. Each work day, dermatopathologists in the United States rely heavily on hematoxylin and eosin (H&E) staining and standard light microscopy to render most diagnoses. Moreover, even when a diagnosis is augmented by an adjunctive technique of some sort, H&E evaluation still remains a cornerstone or “back-bone” of the diagnostic schema.
Yet, reliance on simple histomorphologic assessment of tumors (ie, the physical appearance and arrangement of cancerous cells under the microscope) to predict genetic potential (ie, possession of genetic characteristics to become fatal metastatic disease) is not without consequence. Parenthetically, in the social sciences, reliance on outward appearances to predict behavior is considered an inefficient means of analysis (ie, so-called profiling) ; yet in dermatopathology, variations on this same type of pattern recognition are ubiquitously used.
In truth, particularly with regard to melanocytic neoplasia, there remain fundamental disagreements, even among dermatopathology experts, as to what represents melanoma. Additionally, it has been demonstrated, repeatedly, that with regard to the histomorphologic assessment of melanoma, discrepancies among key histologic features exist. It has been estimated by some experts that about 10% of pigmented lesions examined at expert tertiary melanoma centers may defy a singular confident diagnosis.
Therefore, if specialty care providers in dermatology and dermatopathology desire to know what a skin neoplasm is capable of doing, at least in terms of yielding metastatic disease and patient demise, it is imperative that they develop and adopt means of genetic assessment, for it is precisely such genetic potential that is of concern in caring for patients.
This article focuses on available techniques used for the evaluation of melanocytic neoplasms. The reason for this focus is practical in nature: this is where the bulk of available adjunctive genetic techniques currently lie. At present, there are not significant and useful genetic techniques, in widespread use, for the evaluation of basal cell carcinoma or squamous cell carcinoma, although there is some interest in the latter. The evaluation of dermatofibrosarcoma protuberans, by means of a break-apart fluorescent in situ hybridization (FISH) probe that identifies a specific translocation, permeates clinical practice, yet this tumor is relatively rare.
Although cutaneous lymphoma may be the subject of genetic studies to demonstrate clonality, again as an adjunctive measure to light microscopy and immunohistochemistry (IHC) studies, clonality is not the equivalent of malignancy, nor is the absence of a detected clone exclusive of malignancy. Hence, given its lesser utility, and the fact this is a largely subspecialty concern, such a discussion is better reserved for an exclusive discourse on lymphoproliferative disorders.
Introduction
At present, the histologic diagnosis of skin cancer remains a critical step in the evaluation and management of skin disease. For the foreseeable future, and from a medicolegal standpoint in particular, a histologic report of cancer is requisite for additional intervention. For example, melanoma requires histologic staging information (eg, Breslow depth, presence/absence of ulceration, dermal mitotic activity) to select appropriate management. Even entrance into oncologic trials for melanoma, using new agents and new protocols, is affected by this histologic staging information.
In dermatopathology, the underlying practical means for rendering a cancer diagnosis has changed little in the last 140 years. Each work day, dermatopathologists in the United States rely heavily on hematoxylin and eosin (H&E) staining and standard light microscopy to render most diagnoses. Moreover, even when a diagnosis is augmented by an adjunctive technique of some sort, H&E evaluation still remains a cornerstone or “back-bone” of the diagnostic schema.
Yet, reliance on simple histomorphologic assessment of tumors (ie, the physical appearance and arrangement of cancerous cells under the microscope) to predict genetic potential (ie, possession of genetic characteristics to become fatal metastatic disease) is not without consequence. Parenthetically, in the social sciences, reliance on outward appearances to predict behavior is considered an inefficient means of analysis (ie, so-called profiling) ; yet in dermatopathology, variations on this same type of pattern recognition are ubiquitously used.
In truth, particularly with regard to melanocytic neoplasia, there remain fundamental disagreements, even among dermatopathology experts, as to what represents melanoma. Additionally, it has been demonstrated, repeatedly, that with regard to the histomorphologic assessment of melanoma, discrepancies among key histologic features exist. It has been estimated by some experts that about 10% of pigmented lesions examined at expert tertiary melanoma centers may defy a singular confident diagnosis.
Therefore, if specialty care providers in dermatology and dermatopathology desire to know what a skin neoplasm is capable of doing, at least in terms of yielding metastatic disease and patient demise, it is imperative that they develop and adopt means of genetic assessment, for it is precisely such genetic potential that is of concern in caring for patients.
This article focuses on available techniques used for the evaluation of melanocytic neoplasms. The reason for this focus is practical in nature: this is where the bulk of available adjunctive genetic techniques currently lie. At present, there are not significant and useful genetic techniques, in widespread use, for the evaluation of basal cell carcinoma or squamous cell carcinoma, although there is some interest in the latter. The evaluation of dermatofibrosarcoma protuberans, by means of a break-apart fluorescent in situ hybridization (FISH) probe that identifies a specific translocation, permeates clinical practice, yet this tumor is relatively rare.
Although cutaneous lymphoma may be the subject of genetic studies to demonstrate clonality, again as an adjunctive measure to light microscopy and immunohistochemistry (IHC) studies, clonality is not the equivalent of malignancy, nor is the absence of a detected clone exclusive of malignancy. Hence, given its lesser utility, and the fact this is a largely subspecialty concern, such a discussion is better reserved for an exclusive discourse on lymphoproliferative disorders.
Diagnostic melanoma assessment
Lack of a Gold Standard
The assessment of melanoma is sometimes highly subjective. There are disagreements, even among experts, as to which challenging melanocytic neoplasms represent atypical/dysplastic/Clark nevi or melanoma, or which challenging spitzoid lesions represent classic Spitz nevi or atypical Spitz nevi/tumors or spitzoid melanoma. Complicating matters is that there is no singular gold standard, short of actual biologic behavior over time, which can prove something is, or is not, melanoma.
Moreover, where a tertiary care facility has overturned an assessment of melanoma rendered by an outside facility, or in the alternative, where a tertiary facility has rendered an assessment of melanoma, where an outside facility had not, there is bias to assume this second opinion is correct. Although perhaps a reasonable first assumption, given the expertise and depth of experience at a tertiary facility, in the absence of an adverse outcome experienced by the patient, there is no way to guarantee a second opinion is any more, or any less, correct than the first opinion rendered.
What if a person had melanoma, but it was treated as a severely atypical nevus, and the person simply survived? What if a person had a severely atypical nevus, but received treatment of melanoma that was unnecessary, but difficult to separate from melanoma survival?
In sum, the lack of a definitive gold standard, except in cases where an adverse outcome is experienced by the patient, poses a challenge to investigations of genetic diagnostic techniques ; this fact must be kept in mind in all of the discourse to follow.
Immunohistochemical Stains
Although the realm of H&E staining and light microscopy has changed little in the last century, the development and addition of IHC stains is a more recent development. The groundwork for modern IHC techniques was laid in the 1940s through 1960s, and widespread use of IHC began in earnest in the late 1980s.
In brief, IHC represents a means to detect specific antigens in or on cells based on an antigen-antibody reaction that is recognized at the light microscopic level because of final application of a material that produces a visible color. Antigen retrieval is performed, and then a primary monoclonal antibody is applied. This antibody is directed against a specific tissue antigen. A secondary antibody is then applied that localizes to the first antibody. Conjugated to this secondary antibody are molecules of either horseradish peroxidase enzyme or alkaline phosphatase enzyme. Finally, a chromagen is applied that reacts with the conjugated enzyme to yield brown or red pigment deposition that is visualized under the microscope. The existence of both brown and red chromagen systems mean that under certain conditions two antibodies may be tested at the same time, yielding two different colors (discussed later).
No Single Melanoma Marker Exists
There are several IHC stains are useful for studying melanocytic lesions, but it is import to stress bluntly that no single melanoma marker exists by IHC. Although some IHC markers are useful in suggesting a melanocytic process is benign or malignant, this is merely adjunctive information that must be interpreted in the context of all available data, including assessment by H&E staining alone. Certain patterns of IHC staining, although sometimes suggestive of melanoma, may at other times simply be spurious and inconsequential.
IHC markers are only adjunctive tools that must be placed in context, and do not necessarily supersede traditional morphologic clues provided by light microscopy, to include asymmetry, pagetoid scatter, nuclear pleomorphism, hyperchromasia, irregular nuclear contour, irregular nesting of melanocytes, poor maturation in the dermis, and dermal mitotic activity.
Immunohistiochemical Stains Useful for Diagnosis of Melanoma
The following IHC markers are often used in the evaluation of pigmented lesions and the characteristics of each is discussed next.
S100
A sensitive marker for melanocytic neoplasms, with about 95% of primary cutaneous melanomas expressing this antibody. Desmoplastic/spindle cell melanoma often expresses S100, even though these types of melanoma are typically negative for melan-A/MART-1 and HMB-45. However, S100 expression is not specific for only melanoma, and S100 is also expressed by normal melanocytes, nervous tissue, myoepithelial cells, adipocytes, chondrocytes, and Langerhans cells.
Melan-A/MART-1
Benign and malignant melanocytes express the Melan A/MART-1 antigen, which is detected by the A103 and M2-7C10 clonal antibodies, respectively. In practical clinical parlance the terms are interchangeable. Although a sensitive and specific marker for melanocytes, it is important to recognize that desmoplastic/spindle cell melanoma is consistently negative for Melan A/MART-1 antibodies. Lastly, Melan A/MART-1 can yield a confusing pattern of staining called “pseudonesting,” where the immunostain marks a disputed mixture of inflammatory cells, melanocytes, and keratinocytes at the dermoepidermal junction that can lead to erroneous impressions of melanoma in situ.
Ki-67/MIB-1
Ki-67/MIB-1 is a proliferation marker used to identify cells that are progressing through the cell cycle and preparing for mitosis. In one study, common and dysplastic nevi showed about 1% of melanocyte nuclei to mark (most often near the dermoepidermal junction or in the superficial dermis), whereas in melanoma, about 16.4% of melanocyte nuclei marked, particularly in the deeper extents. This and other investigations led to a postulated “rule of thumb” for difficult pigmented lesions where benignity is favored by a low Ki67/MIB-1 proliferative index of less than 2%, whereas malignancy is favored by a Ki67/MIB-1 higher proliferative index of greater than 10%, with indeterminate lesions occupying a middle ground.
Moreover, Melan A/MART-1 IHC staining, using a red chromagen, may be combined with Ki67 IHC staining, using a brown chromagen, to yield a “double stain” (KiMart) that may be used to assess the proliferative index of melanocytes, with lesser confounding by other cell types (see case study 1).
HMB-45
HMB-45 is an antibody directed at the gp100 protein of melanocytes. In benign acquired nevi HMB-45 expression is often limited to melanocytes at dermoepidermal junctional and in the papillary dermis, with lesser marking in the deeper dermis. This pattern reinforces the normal maturation of nevi, and it is referred to as “zonation.” Lack of zonation with HMB-45 is a concerning finding in some situations; however, melanocytic neoplasms with dusty cytoplasm, such as clonal nevi or deep penetrating nevi, represent important exceptions and do not “zone” with HMB-45, but this is not evidence of malignancy per se.
p16
p16 is a tumor-suppressor protein encoded on chromosome 9. Most benign nevi manifest a mosaic pattern of p16 expression, with more than 25% of cells marking with either nuclear or cytoplasmic expression. In one study examining p16 expression among 46 atypical Spitz tumors and 42 melanoma specimens, decreased nuclear expression of p16 was three-fold more likely in melanoma than in Spitz tumors ( P = .004), and loss of both nuclear and cytoplasmic dermal p16 expression was eight-fold more likely in melanoma ( P = .01). In another study of p16 staining, using 15 desmoplastic Spitz nevi and 11 desmoplastic melanomas, all cases of Spitz nevi showed p16 expression, whereas 81.2% of desmoplastic melanomas were p16 negative, and 18.8% of melanomas were only weakly p16 positive. It is important to stress that loss of p16 expression is not simply an immediate marker of melanoma, but loss of p16 expression could indicate biallelic loss of chromosome 9p21, which in turn, is an adverse prognostic marker in spitzoid neoplasia. Particularly in the realm of spitzoid neoplasia, lesions with lost p16 expression by IHC should be treated with greater suspicion than those that express p16 in a normal fashion.
Case study 1
A 14-year-old boy presented with a dark-brown, pedunculated lesion on the flank that had been growing over the last 4 months. The lesion was sampled, and despite his relative youth, the results of H&E/light microscopy and IHC staining were interpreted as melanoma ( Fig. 1 ). This diagnosis was subsequently reviewed at a second academic institution and there was full concurrence in the diagnosis of melanoma.
Use of genetic testing in melanoma
Current widely available genetic testing modalities to augment a diagnostic impression formed by H&E/light microscopy are divided into three classes: (1) comparative genomic hybridization (CGH), (2) FISH, and (3) gene expression profiling (GEP).
Comparative Genomic Hybridization
CGH is a technique that detects genome-wide changes in DNA copy number, but it does not detect actual mutations. Current microarray-based methods of CGH allow for maximum resolution, and can even detect single copy number changes. Use of chromosomal analysis in the diagnosis of melanocytic lesions is premised on a concept that genomic aberrations are common in melanoma, but generally absent in benign nevi, with most melanomas possessing an average of greater than five genomic aberrations. Common genetic anomalies found in melanoma include gains in chromosomes 6p, 1q, 7p, 7q, 8q, 17q, and 20q, and/or losses of 9p, 9q, 10q, 10p, 6q, and 11q. Spitz nevi often demonstrate isolated gains of 11p or 7q, but these changes are not often observed in melanoma. Some benign proliferations arising in congenital nevi may also manifest chromosomal aberrations by CGH, but these changes are generally different than those of melanoma.
Array-based CGH has the advantage of examining the entire genome for alterations in copy number, but it requires a homogenous area of tumor cells, and there is a higher cost and longer turnaround time for results (often 3–4 weeks), at least in comparison with other technologies. Recently, Ali and colleagues used CGH to evaluate 10 spitzoid melanocytic lesions and reported that 9 of 10 cases had long-term clinical outcomes that correlated with the test result, whereas in 1 of 10 cases the result was uncertain. In general, sensitivity of about 92% to 96% and specificity of about 87% to 100% has been reported for the use of CGH in the diagnosis of melanoma.
Fluorescent In Situ Hybridization
FISH is a technique that allows direct visualization of cytogenic abnormalities in cancer, such as chromosomal deletions, amplifications, and translocations. Loss of heterozygosity is demonstrated by examining DNA copy number using specific loci, and centromere-specific probes can identify gains and losses of entire chromosomes in melanomas. FISH probes are applied directly to formalin-fixed, paraffin-embedded tissue sections, allowing assessment of interphase cells. Early cytogenic studies in melanoma identified reproducible chromosomal aberrations, including losses on chromosomes 6, 8, 9, and 10, and gains on chromosome 1.
Although not assessing the entirety of the genome is a limitation, FISH has some advantages over CGH in some situations. FISH probes can be used on less homogenous populations of cells, and the technique is generally more expedient and of lesser cost. FISH-based detection of chromosomal abnormalities is governed by the sensitivity and specificity of the probes and technique used. Multiple probes are required for sufficient sensitivity in the diagnostic evaluation of melanocytic lesions. Often several probes, each with a distinct fluorophore, are combined into a single multicolor FISH analysis. Also, the mathematical algorithms that control detection thresholds are adjusted to impact sensitivity and specificity.
Initially, a set of four probes targeting 6p25 (RREB1), 6q23 (MYB), 11q13 (CCND1), and centromere 6 (CEP6) was offered for commercial use. Positive results for melanoma were defined by the presence of one or more of the following four criteria: (1) greater than 55% nuclei with more 6p25 than CEP6 signals, (2) greater than 29% nuclei with greater than two 6p25 signals, (3) greater than 40% nuclei with fewer 6q23 than CEP6 signals, or (4) greater than 38% nuclei with greater than two 11q13 signals. These four probes have undergone evaluation in studies, with sensitivities and specificities ranging from 72% to 100% and 94% to 100%, respectively.
However, not all parties found this original probe set (that examined only portions of chromosomes 6 and 11) to be of diagnostic utility in difficult and ambiguous melanocytic lesions. In fact, a recent re-examination of the original probe set found three major problems in the technique: (1) low sensitivity and specificity in morophologically ambiguous melanocytic neoplasms, (2) relatively low sensitivity and specificity in spitzoid neoplasms, and (3) occurrence of false-positives caused by tetraploidy in Spitz nevi and nevus with atypical epitheloid cell component (so-called “clonal nevi”). For this reason, subsequent work sought to alter the original probe set, and now the probes used typically include 9p21 (CDKN2a), 6p25 (RREB1), 11q13 (CCND1), and 8q24 (MYC).
Because of improved functionality with the addition of 11q13 and 8q24 probes, the commercial probe set from the largest single national provider (NeoGenomics, Inc, Hatfield, PA) was altered to a five-probe set consisting of 9p21 (CDKN2a), 6p25 (RREB1), 11q13 (CCND1), 8q24 (MYC), and Cen 9 (centromere of chromosome 9). Applying internally validated cutoffs, the NeoSITE Melanoma test (NeoGenomics, Inc) is reported to provide 86% sensitivity and 90% specificity, but also cases with borderline positive results can transpire, and should be interpreted with caution.
Despite the improvements made in probe constituents, exclusion of tetraploidy, and algorithm advancement, there remain critics of the use of FISH in ambiguous lesions, particularly in those neoplasms with spitzoid features. Massi and colleagues recently questioned the information gleaned from FISH testing of spitzoid neoplasm in those less than 18 years (where spitzoid lesions are more common), and no significant correlation between FISH status and clinical outcome was observed. A recent review has denounced FISH testing as a magic bullet for melanocytic processes, and instead, it has proffered that FISH represents an effective compromise “between cost, technical complexity, and diagnostic accuracy.”
Case study 2
A 25-year-old man presented with a pigmented lesion of 8 × 8 mm on the left arm that had undergone growth in the last 3 months. IHC staining revealed loss of p16 among many of the cells, but equivocal results with regard to assessment of the proliferative index via KiMart, and equivocal results with regard to zonation with HMB-45 ( Fig. 2 ).
Related posts:
Clinical Diagnosis of Skin Cancer
Enhancing Skin Cancer Diagnosis with Dermoscopy
Electrical Impedance Spectroscopy in Skin Cancer Diagnosis
Using Raman Spectroscopy to Detect and Diagnose Skin Cancer In Vivo
Assessing Genetic Expression Profiles in Melanoma Prognosis
Smartphone-Based Applications for Skin Monitoring and Melanoma Detection
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
