JAAD ONLINE RESEARCH LETTERS Determinants of unfavorable presentation of primary cutaneous melanoma To the Editor: ‘‘Which patients are at higher risk of developing melanoma?’’ and ‘‘Which patients are at higher risk of being diagnosed at later stage and thus likely to have poor survival?’’ are crucial questions to be answered to be able to specify target groups for more purposeful clinical examination of suspicious lesions, secondary prevention measures, and aware- ness campaigns for skin melanoma. In general, high TNM stage is associated with poor prognosis, and features that portend a potentially poorer prognosis are older age, male sex, site (acral, head, or neck), increased Breslow tumor thickness, higher Clark level, ulceration, higher number of mitoses, vertical growth phase, regression, absence of a host inflam- matory response, increased tumor vascularity, an- giotropism, vascular invasion, neurotropism, marked atypia, and satellite metastasis. 1 This study aimed to further elucidate the association between TNM staging at the time of diagnosis and patient characteristics (such as education, living in large cities or not, family history of melanoma, and family size) and histology and location of tumor in a nationwide family cancer database, where mela- noma in all individuals and their family members were recorded since 1958 with close to 100% com- pleteness and pathologic verification. 2 In addition, we aimed to provide determinants of primary tumor thickness (T), regional lymph node metastasis (N), and distant metastasis (M) separately. The population-based Swedish Cancer Registry has collected TNM data since 2002. We analyzed 5102 melanoma patients until 2006. Ordinal logistic regression analysis was performed to model stages according to age at diagnosis, sex, education, resi- dential region, histology and location of tumor, and family size and familial melanoma. Of all patients (excluding cases with unspecified characteristics), those living in large cities accounted for 37.0%. The histology was superficial spreading, 68.6%; nodular, 22.7%; lentigo maligna, 6.1%; acral lentiginous, 1.2%; and five other known histology types (n ¼ 52), 1.4%. The location of the tumor was on the trunk, 40.6%; limbs, 46.5%; and head and neck, 12.9%. There was a positive family history of melanoma in 4.7%. Higher TNM stages were significantly associated with older age, male sex, nodular or acral lentiginous histology, location of tumor in lower limb or face, and living in smaller cities or rural areas (Table I). When a patient’s parent, especially mother, had history of melanoma, they were diagnosed system- atically at lower stage (odds ratio of the parents, 0.49 [95% CI, 0.24-1.01]; odds ratio of the mother, 0.21 [95% CI, 0.05-0.87]) and had younger age at diagnosis (mean age, 49 vs 53 in those without parental melanoma) or higher incidence of superficial spreading melanoma (60.5% vs 54.3%) and lentigo maligna (6.2% vs 2.6%). When adjusted for other factors, education and family size did not show significant association with TNM. This study identified risk factors associated with an unfavorable presentation of melanoma, includ- ing older age, male sex, nodular or acral lentigi- nous histology, and location of tumor in lower limb or face, which are in line with other studies. 2-4 Aside from known risk factors and the aforemen- tioned negative prognostic factors, living in smaller cities or rural areas should be considered as negative prognostic factors in the clinical scrutiny of patients and in screening and awareness campaigns. Mahdi Fallah, MD, PhD, a Elham Kharazmi, MD, PhD, b Hauke Thomsen, PhD, a Jan Sundquist, MD, PhD, c,d and Kari Hemminki, MD, PhD a Divisions of Molecular Genetic Epidemiology a and Cancer Epidemiology, b German Cancer Re- search Center, Heidelberg, Germany; Center for Primary Health Care Research, c Lund University, Malmo ¨, Sweden; and Stanford Prevention Re- search Center, d Stanford University School of Medicine, Stanford, California Funding sources: None. Conflicts of interest: None declared. Correspondence to: Mahdi Fallah, MD, PhD, Department of Molecular Genetic Epidemiology, German Cancer Research Center, Im Neuen- heimer Feld 580, 69120, Heidelberg, Germany E-mail: M.Fallah@dkfz.de REFERENCES 1. Payette MJ, Katz M 3rd, Grant-Kels JM. Melanoma prognostic factors found in the dermatopathology report. Clin Dermatol 2009;27:53-74. 2. Hemminki K, Ji J, Brandt A, Mousavi SM, Sundquist J. The Swedish Family-Cancer Database 2009: prospects for histology- specific and immigrant studies. Int J Cancer 2010;126: 2259-67. JAM ACAD DERMATOL JULY 2011 e5