ORIGINAL STUDY Melanoma Hyperpigmentation Is Strongly Associated With KIT Alterations Julie M. Wu, MD,* Hector Alvarez, MD, PhD,* Patricia Garcı ´a, MSc,† Pamela L. Rojas, PhD,† Grace Wong, BA,‡ Anirban Maitra, MD, PhD,* Cristina Antonescu, MD,‡ and Elizabeth A. Montgomery, MD* Abstract: KIT alterations have been identified in melanoma and treatment with imatinib has met with some success. However, the relationship between KIT and melanoma histology remains un- characterized, and its role in melanoma pathogenesis unknown. We evaluated 70 melanomas from 70 patients seen at a single institution from 1997 to 2008. Cases were analyzed for KIT protein expression relative to histologic variables: subtype, sun damage, tumor in- filtrating lymphocytes, melanoma in situ, vertical growth phase (VGP), location, and hyperpigmentation. Twenty-eight cases dem- onstrated 3+ membranous staining. Univariate analysis revealed 5 significant variables: sun damage (inverse, P = 0.015), tumor location (trunk.extremities.head and neck, P = 0.005), subtype (epi- thelioid.spindle, mixed.desmoplastic, P , 0.001), VGP (inverse, P = 0.024), and hyperpigmentation [22/26 (85% hyperpigmented cases) and 6/44 (14% nonhyperpigmented cases), P , 0.001]. Upon multivariate analysis, only hyperpigmentation and VGP remained statistically significant (P = 0.002, P = 0.019). Mutational analyses for KIT exons 9 and 11, and BRAF were performed on cases with 3+ labeling. Two of 27 of cases contained mutations in KIT exon 11, whereas only 1 case contained a V600E BRAF mutation, suggesting that KIT and BRAF mutations may be redundant events. Although KIT mutations were uncommon overall, pigmentation in conjunction with immunohistochemistry and nodular growth phase raised their frequency to 2 (40%) of 5 cases. We expand the context of KIT aberrations to involve areas other than acral and mucosal sites and demonstrate an inverse relationship between KIT abnormalities and sun damage. There is a strong correlation to hyperpigmentation that overrides factors including sun damage, tumor location, and histologic subtype, which may be used to identify cases with KIT aberrations. Key Words: melanoma, pigment, KIT, BRAF, CD117 (Am J Dermatopathol 2009;31:619–625) INTRODUCTION Metastatic melanoma has a dismal prognosis that is unchanged by chemotherapy and radiation. Intense scrutiny of molecular pathways is underway for the identification of new therapeutic targets, and recent developments implicate KIT (CD117) in at least a subset of melanomas. 1–12 More encouragingly, patients with metastatic melanoma have been shown to respond, either partially or dramatically, to the treatment with imatinib. 1,13,14 The KIT protein is a transmembrane tyrosine kinase receptor involved in melanogenesis, gametogenesis, and hematopoiesis. An activating mutation in the gene encoding the KIT protein is detected in the majority of gastrointestinal stromal tumors, and targeting the activated tyrosine kinase has resulted in efficacious treatment outcomes. The distribution of KIT mutations in gastrointestinal stromal tumors has been well characterized with respect to gender prevalence, tumor location, patient age, and response to imatinib therapy. 15–18 In contrast, the role of KIT aberrations in melanoma is not as well delineated. Recent studies have highlighted a prevalence of KIT mutations in mucosal melanomas on the order of 20%–36%. 2,13,19 Furthermore, KIT aberrations appear to be restricted to acral and mucosal sites and to skin with chronic sun damage, such that the distribution is inversely related to that of BRAF mutations. 19 Although the nature of KIT mutations has been explored within the paradigm of chronic sun damage, the relationship between KIT aberrations and histologic parameters such as morphologic subtype, tumor location, vertical growth phase (VGP), and pigmentation remains to be determined. In this report, we studied 70 melanomas taken from 70 patients to discern the relationship between KIT protein labeling by immunohistochemistry and KIT gene mutations and multiple histologic parameters and BRAF alterations. MATERIALS AND METHODS Melanoma biopsies and excisions obtained between 1997 and 2008 were retrieved from The Johns Hopkins Surgical Pathology archives. The study was conducted with the approval of Institutional Review Board. Cases were selected by histologic type to include representative samples of each subtype: epithelioid (38), spindle (19), mixed (10), and desmoplastic (3) (Table 1). Epithelioid subtypes were defined as those composed of tumor cells that displayed voluminous cytoplasm with rounded, ‘‘plump’’ contours (Fig. 1). Spindle From the *Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD; †Department of Pathology, Universidad de La Frontera, Temuco, Chile; and ‡Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, NY. Reprints: Elizabeth A. Montgomery, MD, Department of Pathology, Johns Hopkins Medical Institutions, Weinberg 2242 Pathology, 401 North Broadway, Baltimore, MD 21231–2410 (e-mail: emontgom@jhmi.edu). Copyright Ó 2009 by Lippincott Williams & Wilkins Am J Dermatopathol  Volume 31, Number 7, October 2009 www.amjdermatopathology.com | 619