Humana Press Humana Press M E T H O D S I N M O L E C U L A R M E D I C I N E TM Melanoma Techniques and Protocols Edited by Brian J. Nickoloff, MD, PhD Molecular Diagnosis, Treatment, and Monitoring Melanoma Techniques and Protocols Molecular Diagnosis, Treatment, and Monitoring Edited by Brian J. Nickoloff, MD, PhD Molecular Medicine of Melanoma 3 3 From: Methods in Molecular Medicine, Vol. 61: Melanoma: Methods and Protocols Edited by: B. J. Nickoloff © Humana Press Inc., Totowa, NJ 1 The Many Molecular Mysteries of Melanoma Brian J. Nickoloff 1. Introduction Melanoma of the skin is one of the most rapidly increasing malignancies in both young and old patients (1,2). Not only is the incidence increasing, but the number of annual deaths from melanoma is also on the rise worldwide (3). In the United States, melanoma will be diagnosed in 43,000 new patients each year and be responsible for 7300 deaths (1 death every 72 min). The capacity of melanoma to develop in young patients is reflected by the rather alarming statistic that it has become one of the top causes of death in both men and women between the ages of 25 and 40 (3). Indeed, among Caucasian females, melanoma is the leading cause of death from malignancy between the ages of 25 and 29 (3). It is expected that by 2002, 1 in 70 Americans will develop melanoma during their lifetime (2). Also, melanoma is second only to adult leukemia as the leader in the number of potential years of life lost, which is significantly greater than for patients with cervical, breast, and colon malig- nancies (4). Despite the frequent presence of melanoma and major associated health problems around the globe, only recently have clinicians and labora- tory-based researchers begun to unravel some of the molecular mysteries of melanoma (5,6). The purpose of Melanoma: Methods and Protocols, published as part of the Methods in Molecular Medicine™ series, is to provide an up-to-date review of the many advances that have taken place during the past several years involving the pathophysiology, diagnosis, genetic analysis, and treatment approaches for patients with melanoma (7). Although the bad news is that the incidence as well as morbidity and mortal- ity rates for melanoma are on the rise, the good news is that our knowledge has tremendously increased across many clinical and scientific disciplines (5–7). The challenge for compiling a valuable multiauthored text containing contem- 4 Nickoloff porary viewpoints, scientific facts, clinical treatment protocols, and other dis- coveries is to select authors who can contribute their ideas and present the state-of-the art techniques from a rather broad-ranging set of perspectives. Thus, this book is written by a quite diverse group of individuals who share several unifying characteristics. First, the authors all are involved in the clini- cal practice of medicine either directly as surgeons, oncologists, tumor immu- nologists, or pathologists, or have decided to focus their investigative talents on working closely with these clinicians. Second, and perhaps most relevant for their selection to contribute a chapter in this book, is that they focus on the molecular basis of melanoma. Third, the authors have agreed to include in their respective chapters all relevant literature citations with an emphasis on the most recent available data. Fourth, the authors were encouraged to reduce their experimental procedures to a practical level so that others not familiar with specific techniques could use these important approaches in their own laboratories, hospitals, and cancer centers. Finally, despite the difficulty in translating scientific discoveries into clinical practice, each author was encouraged to select the most medically important advances in their respective areas and highlight the relevance of such findings for clinicians caring for patients with melanoma. This book provides a rich admixture of clinical perspectives, cutting-edge technological advances, including narrative overviews, as well as specific and detailed laboratory-based protocols. The emphasis on molecular biology throughout reflects the progress made in delineating the genetic basis for mela- noma, a forward-thinking approach to rendering molecular-based diagnostic reports, understanding the immunobiology of melanoma, initiating vaccine- based gene therapy to treat patients with melanomas, and using the tools of genomics (i.e., DNA sequencing, cDNA microarray analysis, and proteomics) to facilitate future progress in the field of melanoma. 2. From the Microscope to the Molecular Diagnosis of Melanoma During the past 15 yr as a practicing dermatopathologist, I have witnessed many changes in the field, particularly regarding pigmented skin lesions. Dur- ing my initial training in Boston, pathology reports of melanoma focused pri- marily on the Clark level and Breslow measurements of depths of invasion of the primary cutaneous lesion. In the early and mid-1980s, many academic der- matopathology units were struggling with delineation of accurate and repro- ducible criteria for potential precursor lesions of melanoma including dysplastic nevi and congenital nevi (8). By examining relatively large data- bases and using computer-generated multivariant analysis, numerous indepen- dent prognostic indicators were put forward to assist the clinician in the management of patients with melanoma (9). Thus, our current pathology Molecular Medicine of Melanoma 5 reports include the Clark level (defined as level I for in situ, Level II for mela- nomas partially infiltrating papillary dermis, Level III for lesions filling the papillary dermis, Level IV for melanomas extending into reticular dermis, and Level V for melanomas extending into sc fatty tissue), Breslow depth of inva- sion (expressed in millimeters of thickness from the granular-cell layer in the epidermis to the deepest portion in the dermis), presence or absence of regres- sion, surface ulceration, and microscopic satellitosis, to name a few (10). Although these rather objective measurements provide valuable prognostic information for the patient and physician, there is still a growing awareness and appreciation of the phenotypic complexity and capricious behavior of melanoma. Initially, it appeared that one of the most important determinants of the biologic behavior of melanoma was primary tumor thickness. The first sharp “break point” was set at 0.76 mm in thickness (Breslow measurement) and later changed to 0.85 mm. Thus, it was generally regarded that relatively “thin” melanomas had an extremely high cure rate, and such an anatomic consider- ation was frequently linked to the lack of vascularization of lesions in the upper dermis that did not grow beyond 1 or 2 mm in diameter before their removal. However, it has become clear that many other molecular determinants are important to the biologic behavior of melanoma, and the remainder of this chap- ter is devoted to a brief review of such molecules and the pathways they regu- late. A very real problem that remains for the dermatopathologist using only light microscopic criteria is the inability to predict metastatic behavior in rela- tively “thin” melanomas (11–14). Before delving into the next section, it is important to note that whereas many of the aforementioned “microstaging” criteria are relatively objective and reproducible among dermatopathologists, the classification of certain nevi that may be linked to melanoma such as “dysplastic nevi” has a higher degree of subjectivity (15). Indeed, despite a National Institutes of Health Consensus Panel meeting, and numerous attempts to define suitable histologic criteria, pathologists still are not able to agree con- sistently on these problematic pigmented legions (16). Given the limitations in rendering meaningful diagnosis when such an element of subjectivity is present, it becomes clear that moving from the microscope to a more mole- cular-based analysis of melanoma (Fig. 1) provides the opportunity to under- stand better the phenotypic complexity of nevi and melanoma. One of the most important new advances in this area has been the use of molecular staging of the sentinel lymph node in melanoma patients (17). 3. Importance of Sentinel Lymph Node Assessment As described in more detail in Chapter 17, surgical techniques have greatly advanced in the last decade and provide an opportunity to perform clinical staging of melanoma using the sentinel lymph node (SLN) biopsy (17). It is 6 Nickoloff based on the principle that the sentinel node is the first lymph node a metasta- sis encounters before entering into other lymph nodes (18). Because SLN biopsy can be performed under local anesthesia, and because it can detect sub- clinical metastatic disease when assessed using molecular-based techniques, it provides a new method to stage a patient without a period of clinical observa- tion previously requiring a certain period of time to elapse before the detection of palpable lymph nodes could be appreciated by the physician (19–21). A pathologist can generally detect 1 malignant melanoma cell in a back- ground of 10,000 lymphocytes in a lymph node using routine light microscopy (Fig. 2), but the addition of immunostaining can enhance this detection 10-fold (17). However, using reverse transcriptase polymerase chain reaction (RT-PCR) to detect a simple transcript—e.g., tyrosinase mRNA present in melanoma cells, but not B- or T-lymphocytes—can enhance the detection sen- sitivity by two to three orders of magnitude over immunostaining results (17). This is not just an academic exercise, because data clearly demonstrate the superior clinical correlation using molecular-based (i.e., RT-PCR) analysis of SLNs, compared to more routinely processed morphology-based visual assess- ments for patients with malignant melanoma. For example, if an SLN is upstaged (i.e., by routine light microscopic examination, it appears negative for presence of melanoma, but RT-PCR demonstrates the presence of over- looked or rare metastatic melanoma cells), then there is a significantly Fig. 1. Moving from a morphologic to a molecular-based diagnostic approach in melanoma. Molecular Medicine of Melanoma 7 increased chance of recurrence. The rate of recurrence and overall survival for 114 patients based on SLN analysis was as follows: histologically positive and RT-PCR positive (34% recurrence rate); histologically negative and RT-PCR negative (2% recurrence rate). But even when histology was negative, a posi- tive RT-PCR detection increased this 2% rate to a 13% rate (more than sixfold higher). It was determined that these differences in recurrence rates and sur- vival were statistically significant ( p = 0.02). Indeed, in both univariate and multivariate regression analysis, the histologic and RT-PCR status of the SLNs were the best predictors (Fig. 3) of disease-free survival (17). 4. Biologic Determinants of Melanoma Behavior This section provides an analysis of the critical biologic determinants that can supplement the light microscopic and molecular viewpoint, as previously mentioned, with an emphasis on those characteristics that are associated with metastasis. The focus on metastasis is important because despite improvement in clinical diagnosis, surgical techniques, and the use of novel treatments and adjuvant approaches, most melanoma deaths result from metastasis. There is less than a 5% chance of surviving for 5 yr in patients with metastatic mela- noma. Indeed, while considerable debate raged for years regarding the appro- Fig. 2. The relative sensitivity of detecting a metastatic melanoma cell in a lymph node comparing traditional routine hematoxylin and cosin (H&E) staining with light microscopy vs immunostaining vs a molecular analysis. (Adapted from ref. 17.) 8 Nickoloff priate surgical margin, such debate, in my view, focused too much attention on the local recurrence rates and not enough on the problem of metastasis. As already mentioned, significant advances have been made so that we can rou- tinely assess, by molecular techniques, the status of the SLN. After all, most patients do not succumb to local recurrence of their melanoma, but they do experience significant morbidity and mortality when their melanoma moves from the skin to extracutaneous sites. None of the randomized double-blind clinical studies of the width of surgical resection of melanoma ever pointed to a statistical significance on long-term survival—only rates of local recurrence. Having covered these histologic, surgical, and clinical perspectives, we now review some of the molecular determinants that can be useful in understanding and, it is hoped, predicting more reliably the progression of melanoma, includ- ing its metastasis beyond the confines of the epidermis and dermis. Before covering melanoma, it may be instructive first to review the biologic behavior of nevi, because many melanomas develop from such preexisting nevi. Whereas only 1% of individuals are born with nevi (i.e., congenital nevus), almost every individual will develop nevi beginning in adolescence and extending through adulthood. As documented by dermatologists, the num- Fig. 3. Molecular staging of melanoma. Molecular Medicine of Melanoma 9 ber of nevi or moles on each individual actually change over a lifetime, with many nevi coming and going with the passage of time. The molecular factors that prompt a single melanocyte in the basal cell layer of the epidermis in a teenager to change phenotypically into a nevus cell, and then initially prolifer- ate largely in a relatively tightly nested or clustered group to produce a junc- tional nevus, are not known. Neither is it clear as to the nature of the stimulus that triggers an exodus of the nevus cells from the epidermal compartment into the papillary dermis. However, a few recent molecular clues have emerged that point to the role of basic fibroblast growth factor (bFGF) and its receptor. It appears that nevus cells may use bFGF as a “lifesaver” by promoting the sur- vival of nevomelanocytes as they leave the confines of the epidermis where keratinocytes could supply this essential growth factor in an a paracrine fash- ion (22). Thus, when nevus cells are in the dermis, they acquire the capacity to produce their own bFGF in an autocrine fashion to ensure their independence of the epidermal-based constraints. As recently discussed, this autocrine switch may represent a double-edged sword, because the acquisition of the ability to produce a potent mitogen, coupled with the constitutive expression of the growth factor receptor, has been demonstrated in several oncologic models to represent an early event in the transformation process (23). Indeed, it has been documented that early stage melanoma cells cannot produce bFGF in abun- dance compared with late-stage melanoma cells (24). Another relevant molecular change controlling the migration of nevus cells from the epidermis to the dermis are the cadherin-mediated adhesive interactions (25). A large number of molecular markers have been documented to be corre- lated to the progression of melanoma. In general, it is possible to classify these changes as resulting from either an increase in the levels relative to normal keratinocytes or nevus cells, or a relative decrease in their expression. There are many examples of so-called gain-of-function molecular markers such as numerous growth factors, cytokines, and their receptors including keratinocyte growth factor, platelet-derived growth factor, stem cell factor, bFGF, and interleukin-1_ (IL-1_), IL-2`, IL-6, IL-7, IL-8, IL-10, and IL-12. In addition, melanoma cells express intercellular adhesion molecule-1, MUC-18, integrins (i.e., _V`3), and proteolytic enzymes (plasminogen activator) or CD95L (Fas ligand). To escape immunosurveillance, melanoma cells may also cease to express other molecules such as class I major histocompatibility complex anti- gens and CD95 antigen. Because monoclonal antibodies (MAbs) are available that can detect the presence or absence of many of these molecular markers, one wonders whether pathology reports that include semiqualitative assessments of such molecules could enhance the predictive value of our otherwise routine histologic analysis of primary cutaneous melanomas. After all, we have all had patients with a relatively thin melanoma (i.e., <0.85 mm) who have developed metastatic 10 Nickoloff lesions that we would not have accurately predicted using conventional microstaging criteria (12–14). Another important diagnostic dilemma for dermatopathologists is the iden- tification of a metastatic infiltrate in the lymph node or other extracutaneous sites when no primary cutaneous lesion is present. In approx 10% of loss involving metastatic melanoma, no primary site can be identified. If the meta- static cells are producing melanin, there is no difficulty in recognizing the malignancy as melanoma. However, in amelanotic malignant infiltrates, it is necessary to use immunohistochemical analysis to determine whether the tumor is related to melanoma. While ultrastructural studies using electron micros- copy can yield insight into the diagnosis by identifying melanomas or premelanomas, several MAbs have permitted assignment of metastatic lesions to the melanoma category (Fig. 4). These diagnostic reagents include use of detection of S-100 (highly sensitive, relatively nonspecific), gp100 (i.e., HMB-45), and newer MAbs to detect MART-1 (26–29). Fig. 4. Forward-looking depiction of sampling a pigmented lesion by needle biopsy followed by array analysis using microchip technology to assess thousands of mRNA transcripts. (Adapted from refs. 35–38.) Molecular Medicine of Melanoma 11 5. Future Directions Given the limitations in rendering precise and prognostically relevant pathology reports based solely on light microscopic criteria, it is likely that a more molecular-based approach will be forthcoming as the immunobiology and genetic basis of melanoma is better understood. From the practical per- spective, determining whether the melanoma cells express the `3 integrin appears to be the single best molecular determinant for distinguishing either benign nevomelanocytes or low-risk melanoma cells in the radial growth phase, from the high-risk melanoma cells in the vertical growth phase of primary melanomas. However, I believe we will rapidly shift our molecular analysis away from expression of single proteins such as `3 integrin, to a more compre- hensive analysis that will include examination of the presence and absence of dozens, if not hundreds or thousands, of different transcripts in small biopsy specimens of pigmented skin lesions. Indeed, the Human Genome Project is revolutionizing the practice of biology and medicine in several respects (30). Cancers such as melanoma can be viewed as a systems problem, and using global tools of genomics, the information pathway responsible for conversion of a benign melanocyte to a melanoma cell can be understood (30). As has been shown elegantly by Duggan et al., (31) as well as by many others (32–38), assays of genes on various chips can permit the simultaneous analysis of numerous transcripts. The goal of this next generation of diagnostic tests will be to assign specific “signatures” or to fingerprint a distinctive constellation of both positive and negative transcripts that will have better prognostic value. Not only can this technology assist the pathologist in better cataloguing of various phenotypes of melanoma, but with more experience this approach will also facilitate more customized treatment protocols. For example, there may be considerably greater heterogeneity in the behavior of melanomas besides the current distinc- tion of radial vs vertical growth phases of melanoma. A more prognostically sophisticated classification scheme based on differential transcription profiles may yield several distinctive phenotypes. Within each tumor classification, further distinctions may be made with clinical experience based on therapeutic responsiveness, so that not only will new diagnostic categories be created but also therapeutic decisions based on such molecular analysis will be forthcom- ing. By examining hundreds, if not thousands, of target molecules, the full range of biologically relevant pathways can be analyzed including molecules that regulate cell-cycle progression, transcription factors, signal transduction, adhesion molecules, cytokine production profiles, growth factors, apoptotic resistance/sensitivity proteins, immunoregulatory cell surface molecules, and chemotactic polypeptides. [...]... a large percentage of melanoma tumor samples (MAGE-A1 in 40%, MAGE-A2 in 70% of metastatic melanomas, MAGE-A3 in 65% of melanoma samples, and MAGE-A10 in 21% of primary melanoma lesions and in 47% of metastatic melanoma tissues) and are therefore promising targets for immunotherapy At present it is not clear why these testis-specific genes are activated in certain malignancies and what their normal... identification and generation of different CTL lines, it was postulated that melanomas express multiple antigenic peptides, which can be recognized by CD8+ T-cells (13) Since then, a series of antigens and antigenic peptides has been identified for melanoma and also for other tumor types, using different techniques, discussed later in this chapter 1.2 Melanoma Antigens Almost all the antigens identified in melanoma, ... Characterization of Melanoma- Derived Antigens Marten Visser, Markwin P Velders, Michael P Rudolf, and W Martin Kast 1 Introduction In the last decade, many antigens expressed by tumors and recognized by the immune system have been identified Melanoma was among the first tumors found to express such tumor-associated antigens, and, therefore, melanoma is currently one of the best and extensively studied... diagnostic profiles of melanoma References 1 Hall, H I., Miller, D R., Rogers, J D., and Bewerse, B B (1999) Update on the incidence and mortality from melanoma in the United States J Am Acad Dermatol 40, 35–42 2 Parker, S L., Tong, T., Bolden, S., and Wingo, P A (1997) Cancer statistics Cancer Clin 47, 5–27 3 Dennis, L K (1999) Analysis of the melanoma epidemic, both apparent and real Arch Dermatol... growth factor genes by melanoma cells but not normal melanocytes J Invest Dermatol 97, 20–26 25 Seline, P C., Norris, D A., Horikawa, T., Fujita, M., Middleton, M H., and Morelli, J G (1996) Expression of E- and P-cadherin by melanoma cells decreases in progressive melanomas and following ultraviolet irradiation J Invest Dermatol 106, 1320–1324 26 Wick, M R., Swanson, P E., and Rocamura, A (1988) Recognition... 6 in malignant melanoma cell lines suppressing either their tumorigenicity (2) or metastasis (3,4) However, the suppressor genes involved have yet to be identified Human melanoma cell lines UACC903, UACC903(+6), and SRS3 were derived from two steps of genetic manipulation (2,5) Specifically, the parental malignant melanoma cell line UACC903 was derived from a primary melanoma specimen and displays anchorage-independent... (e.g., small KimWipe) and place in a 50-mL conical tube 5 Place the microarrayed glass slide over the tissue and seal the cap tightly 6 Ensure that the slide is level and stable in a 65°C incubator Allow the hybridization to proceed overnight 7 Wash the slide in 0.5X SSC and 0.01% SDS until the cover slip falls off 8 Wash in 0.5X SSC and 0.01% SDS for 15 min 9 Wash in 0.06X SSC and 0.01% SDS for 15... melanoma by Clark Level and Breslow thickness Cancer 71, 3737–3743 10 Balch, C M., Murad, T M., Soong, S.-J., et al (1978) A multifactorial analysis of melanoma: prognostic histopathological factors comparing Clark’s and Breslow’s staging methods Am Surg 188, 732–742 Molecular Medicine of Melanoma 13 11 Gromet, M A., Epstein, W L., and Blois, M S (1978) The regressing thin melanoma: a distinctive... application of Clark classification and the Breslow thickness measurements in cutaneous malignant melanoma Int J Cancer 26, 159–163 16 Goldsmith, L A., Askin, F B., Chang, A E., et al (1992) Diagnosis and treatment of early melanoma NIH consensus development panel on early melanoma JAMA 268, 1314–1319 17 Shivers, S C., Wang, X., Weigua, L., et al (1994) Molecular staging of malignant melanoma: correlation with... H K., Geller, A C., Miller, D R., Prout, M N., and Lew, R A (1990) Years of potential life cost: another indicator of the impact of cutaneous malignant melanoma on society J Am Acad Dermatol 23, 308–310 5 Kamb, A (1996) Human melanoma genetics J Invest Dermatol 1, 177–182 6 Sauter, E R and Herlyn, M (1998) Molecular biology of human melanoma development and progression Mol Carcinogenesis 23, 132–143 . E TM Melanoma Techniques and Protocols Edited by Brian J. Nickoloff, MD, PhD Molecular Diagnosis, Treatment, and Monitoring Melanoma Techniques and Protocols Molecular Diagnosis, Treatment, and. of Melanoma 3 3 From: Methods in Molecular Medicine, Vol. 61: Melanoma: Methods and Protocols Edited by: B. J. Nickoloff © Humana Press Inc., Totowa, NJ 1 The Many Molecular Mysteries of Melanoma Brian. in understanding and, it is hoped, predicting more reliably the progression of melanoma, includ- ing its metastasis beyond the confines of the epidermis and dermis. Before covering melanoma,