NYU School of Medicine
David Polsky MD, PhD
Director, Cutaneous Oncology Section
Ronald O. Perelman Department of Dermatology
NYU Langone Medical Center
New York City
A key goal of oncology is to detect and remove cancers at an early stage, when they are localized to the primary site and malignant cells have not metastasized to other organs. Screening the entire population for early cancers, however, is costly and often of low yield. The yield could be increased if we focused primarily on those patients who are at increased risk of disease; however, other than age-related recommendations for tests such as mammography or colonoscopy, or recommendations specifically for patients in highly cancer-prone families, we do not yet have accurate enough tools to select high-risk patients for cancer screening.
Currently, several groups have been conducting genome-wide association studies to identify inherited genetic variations that might identify individuals at increased risk for cancer, including melanoma.1 These gene discovery efforts are not based on pre-existing hypotheses regarding the role of any particular gene or genetic pathway in the disease of interest, but they potentially can produce important breakthroughs in our understanding of various diseases, including cancers such as melanoma. These discoveries may lead to genetic tests to identify high-risk individuals.
An alternative approach to identify individuals at high risk for melanoma begins with a hypothesis to explain epidemiologic observations, such as the link between sunlight, fair skin, and melanoma, and investigates known genes that may play a role in the disease process. One epidemiologic observation that has been of great interest to our group is the difference in age-related melanoma incidence rates between men and women. Investigating this observation using the tools of genetic epidemiology led to our preliminary discovery of a genetic marker associated with increased risk of melanoma among women under 50 years of age.
Incidence Differs Markedly in Men and Women
|Figure 1: Melanoma of the Skin: SEER Incidence 2004-2006 White, by Age and Sex|
Melanoma incidence rates display a unique profile with respect to age and gender. For nearly all the other major forms of adult cancer, incidence rates increase exponentially with increasing age and in parallel fashion for men and women. Cancer rates are also generally higher among men than women across all age groups. In contrast, melanoma incidence is greater among women than men between age 20 and age 40, but greater among men than women at ages greater than 50. Although the age-related melanoma incidence rate among men displays the same exponential increase as is seen in other cancers, melanoma incidence in women never exhibits this exponential pattern of increase with increasing age.2 (See Figure 1.)
These differences raise many intriguing questions. Why is melanoma incidence greater in young women than in young men? Why doesn’t the incidence curve in women rise sharply with age as it does in men? Why doesn’t it resemble the curve for women with other common cancers such as lung and colon? Among the possible explanations for these trends is that women generally have more melanomas removed at the in-situ stage, which is not counted in the epidemiologic statistics, and this prevents the later development of invasive melanomas; or that age-related sun exposure patterns differ between men and women. It is possible that women become more aware of the risks of ultraviolet (UV) exposure and avoid the sun as they age, thereby reducing the incidence of sunlight/UV-driven melanomas among older women. Although these are interesting hypotheses, we are unaware of data to support or refute them.
An alternate explanation is that some melanomas in women are related to estrogen signaling. This hypothesis is based in part on the observation that the average age of menopause in the United States is 51,3 about the age at which melanoma incidence becomes greater in men than in women. The idea is that estrogen combines with excessive UV exposure and certain inherited genetic factors to transform benign melanoma precursor cells into malignant melanoma cells. When estrogen levels decline after menopause, the risk of melanoma also declines in women who have inherited the genetic factors that cooperate with estrogen to produce melanoma. Recently, one such genetic factor had been identified in other cancers, which prompted us to investigate its role in melanoma.
MDM2 SNP309 and Melanoma Risk in Women
A single nucleotide polymorphism (SNP) at position 309 of the MDM2 gene (SNP309) was found to be associated with the onset of several different cancers such as soft tissue sarcomas, diffuse large B-cell lymphoma, colorectal cancer, and non–small cell lung cancer in women under age 51.4 MDM2 is a key negative regulator of the tumor suppressor gene p53. In certain human tumors, overexpression of MDM2 is oncogenic, that is, associated with accelerated cancer progression due to excessive inactivation of p53.5 In addition, laboratory studies have demonstrated that the estrogen receptor complex activates MDM2 expression by binding to the SNP309 site, especially when the site is occupied by the G nucleotide. This leads to higher levels of expression of MDM2 and suppression of p53.5-7 These observations formed the basis for the hypothesis that estrogen signaling could be involved in cancer development related to the SNP309 G allele.
To investigate this hypothesis, we conducted a pilot study to examine the relationship between age at melanoma diagnosis and SNP309 genotypes for men and women.8 We examined a population of 227 newly diagnosed primary melanoma patients prospectively enrolled in the Interdisciplinary Melanoma Cooperative Group at the New York University Langone Medical Center from August 2002 to November 2006. The population was 41 percent female and 59 percent male, with a median age at diagnosis of 57 years. We found that the age of diagnosis among women with two copies of the G allele (GG genotype) was 13 years earlier (median 46 years) compared to women with either the TG or TT genotype (median 59 years); however, this difference was not statistically significant. Of note, by age 50, 11 of 21 women (52.4 percent) with the GG genotype were diagnosed with melanoma, compared to 15 of 68 women (22.1 percent) with either the TG or TT genotype (p=0.01).
|Table 1: Odds ratios for initial diagnosis of melanoma at ages less than those shown above, for the MDM2 SNP309 GG genotype compared to TG and TT genotypes|
We did observe a statistically significant increase in the odds ratios for a melanoma diagnosis in the GG genotype when we analyzed the patients by age groups (Table 1). Specifically, women with the GG genotype had an odds ratio of 3.89 for a melanoma diagnosis at ages under 50 years, and this increased to 4.62 for those under age 40. There was no significant age association found using a cutpoint of 60 years and older, and no significant association for any age cutpoint among men. These data demonstrate that women with the SNP309 GG genotype may be more likely to be diagnosed with melanoma at ages under 50 compared to women with either the TG or TT genotype. Taken together with the epidemiologic incidence data and our knowledge regarding the average age of menopause, these findings suggest that estrogen may play a role in the etiology of melanoma among women with a genetic predisposition such as MDM2 SNP309.
It is important to point out that this initial study has several limitations. We evaluated a relatively small number of patients, and all the individuals studied had been diagnosed with melanoma. As such, these findings suggest that MDM2 SNP309 must be cooperating with other genetic and environmental factors that contribute to melanoma. Indeed, analysis of the SNP309 polymorphism is complicated by a relatively high frequency of the GG genotype in the population. Definite population frequency data are not available through the National Center for Biotechnology Information, but frequency estimates in various published reports range from 15 to 30 percent of individuals. This is clearly much higher than the percentage of individuals who will develop melanoma, so SNP309 cannot be used as a sole marker of genetic risk for melanoma among young women. Based on this high frequency of the GG genotype, it is likely that analyzing melanoma risk using a study that compares melanoma patients with unaffected controls will yield lower odds ratios that may not be statistically significant. For example, no association between SNP309 and melanoma risk was found in a study using subjects enrolled in the Nurses Health Study, a large prospective cohort study that included research subjects with and without melanoma.9 It is difficult to compare the Nurses Health Study data with ours; however, because the average age of melanoma diagnosis in the Nurses Health Study was 63 compared to 57 in our group, their population had a much lower prevalence of the GG genotype (13 percent, vs. 24 percent in our cohort), and the investigators did not present an estimate of the risk of melanoma by age group and genotype. In addition, there may be other as yet unknown differences between our study population and theirs that may affect the association between SNP309 and melanoma risk.
Nevertheless, the unique pattern of age-related melanoma incidence among women remains intriguing and will be the subject of additional studies. As the current model of melanomagenesis is based on the cooperation of both genetic and environmental factors in the transformation of benign melanoma precursor cells into malignant melanoma, it is likely that inherited genetic variations such as SNP309 will play a role in explaining why some women develop melanoma before age 50 and others do not. Ultimately, a more sophisticated melanoma risk assessment model will need to include many genes, probably at least 10, in combination with clinical risk factors such as mole pattern and an assessment of sun and tanning bed exposure. It is possible that a woman’s SNP 309 genotype may be one of these important factors. With such a model in place, we would be able to recommend more intensive melanoma surveillance to those at high risk, hopefully intervening in the disease process at an early, curable stage.
1. Easton DF, Eeles RA. Genome-wide association studies in cancer. Hum Mol Genet Oct 15 2008; 17(R2):R109-115.
2. Ries L, Melbert D, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2005. http://seer.cancer.gov/csr/1975_2005/.
3. Nelson HD. Menopause. Lancet Mar 1 2008; 371(9614):760-770.
4. Bond GL, Levine AJ. A single nucleotide polymorphism in the p53 pathway interacts with gender, environmental stresses and tumor genetics to influence cancer in humans. Oncogene Feb 26 2007; 26(9):1317-1323.
5. Bond GL, Hu W, Bond EE, et al. A single nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans. Cell Nov 24 2004; 119(5):591-602.
6. Saville B, Wormke M, Wang F, et al. Ligand-, cell-, and estrogen receptor subtype (alpha/beta)-dependent activation at GC-rich (Sp1) promoter elements. J Biol Chem Feb 25 2000; 275(8):5379-5387.
7. Hu W, Feng Z, Ma L, et al. A single nucleotide polymorphism in the MDM2 gene disrupts the oscillation of p53 and MDM2 levels in cells. Cancer Res Mar 15 2007; 67(6):2757-2765.
8. Firoz EF, Warycha M, Zakrzewski J, et al. Association of MDM2 SNP309, age of onset, and gender in cutaneous melanoma. Clin Cancer Res Apr 1 2009; 15(7):2573-2580.
9. Nan H, Qureshi AA, Hunter DJ, Han J. A functional SNP in the MDM2 promoter, pigmentary phenotypes, and risk of skin cancer. Cancer Causes Control Mar 2009; 20(2):171-179.