[Skip to Content]
[Skip to Content Landing]
Figure.
Study Flowchart
Study Flowchart
Table 1.  
Participant Characteristics
Participant Characteristics
Table 2.  
Prevalence and Predictors of MC1R Testing Interest and Uptake
Prevalence and Predictors of MC1R Testing Interest and Uptake
Table 3.  
Predictors of Test Follow-through
Predictors of Test Follow-through
1.
Rouhani  P, Pinheiro  PS, Sherman  R,  et al.  Increasing rates of melanoma among nonwhites in Florida compared with the United States.  Arch Dermatol. 2010;146(7):741-746.PubMedGoogle ScholarCrossref
2.
Pollitt  RA, Clarke  CA, Swetter  SM, Peng  DH, Zadnick  J, Cockburn  M.  The expanding melanoma burden in California Hispanics: importance of socioeconomic distribution, histologic subtype, and anatomic location.  Cancer. 2011;117(1):152-161.PubMedGoogle ScholarCrossref
3.
Robinson  JK, Joshi  KM, Ortiz  S, Kundu  RV.  Melanoma knowledge, perception, and awareness in ethnic minorities in Chicago: recommendations regarding education.  Psychooncology. 2011;20(3):313-320.PubMedGoogle ScholarCrossref
4.
Roman  C, Lugo-Somolinos  A, Thomas  N.  Skin cancer knowledge and skin self-examinations in the Hispanic population of North Carolina: the patient’s perspective.  JAMA Dermatol. 2013;149(1):103-104.PubMedGoogle ScholarCrossref
5.
Kasparian  NA, McLoone  JK, Meiser  B.  Skin cancer-related prevention and screening behaviors: a review of the literature.  J Behav Med. 2009;32(5):406-428.PubMedGoogle ScholarCrossref
6.
Coups  EJ, Manne  SL, Heckman  CJ.  Multiple skin cancer risk behaviors in the US population.  Am J Prev Med. 2008;34(2):87-93.PubMedGoogle ScholarCrossref
7.
Coups  EJ, Stapleton  JL, Hudson  SV, Medina-Forrester  A, Natale-Pereira  A, Goydos  JS.  Sun protection and exposure behaviors among Hispanic adults in the United States: differences according to acculturation and among Hispanic subgroups.  BMC Public Health. 2012;12:985.PubMedGoogle ScholarCrossref
8.
Saraiya  M, Hall  HI, Uhler  RJ.  Sunburn prevalence among adults in the United States, 1999.  Am J Prev Med. 2002;23(2):91-97.PubMedGoogle ScholarCrossref
9.
Rodriguez  VM, Shuk  E, Arniella  G,  et al.  A qualitative exploration of Latinos’ perceptions about skin cancer: the role of gender and linguistic acculturation.  J Cancer Educ. 2017;32(3):438-446.PubMedGoogle ScholarCrossref
10.
Coups  EJ, Stapleton  JL, Manne  SL,  et al.  Psychosocial correlates of sun protection behaviors among US Hispanic adults.  J Behav Med. 2014;37(6):1082-1090.PubMedGoogle ScholarCrossref
11.
Friedman  LC, Bruce  S, Weinberg  AD, Cooper  HP, Yen  AH, Hill  M.  Early detection of skin cancer: racial/ethnic differences in behaviors and attitudes.  J Cancer Educ. 1994;9(2):105-110.PubMedGoogle Scholar
12.
Pipitone  M, Robinson  JK, Camara  C, Chittineni  B, Fisher  SG.  Skin cancer awareness in suburban employees: a Hispanic perspective.  J Am Acad Dermatol. 2002;47(1):118-123.PubMedGoogle ScholarCrossref
13.
Udayakumar  D, Tsao  H.  Moderate- to low-risk variant alleles of cutaneous malignancies and nevi: lessons from genome-wide association studies.  Genome Med. 2009;1(10):95.PubMedGoogle ScholarCrossref
14.
Goldstein  AM, Landi  MT, Tsang  S, Fraser  MC, Munroe  DJ, Tucker  MA.  Association of MC1R variants and risk of melanoma in melanoma-prone families with CDKN2A mutations.  Cancer Epidemiol Biomarkers Prev. 2005;14(9):2208-2212.PubMedGoogle Scholar
15.
Kanetsky  PA, Ge  F, Najarian  D,  et al.  Assessment of polymorphic variants in the melanocortin-1 receptor gene with cutaneous pigmentation using an evolutionary approach.  Cancer Epidemiol Biomarkers Prev. 2004;13(5):808-819.PubMedGoogle Scholar
16.
Landi  MT, Kanetsky  PA, Tsang  S,  et al.  MC1R, ASIP, and DNA repair in sporadic and familial melanoma in a Mediterranean population.  J Natl Cancer Inst. 2005;97(13):998-1007.PubMedGoogle ScholarCrossref
17.
Mitra  D, Luo  X, Morgan  A,  et al.  An ultraviolet-radiation-independent pathway to melanoma carcinogenesis in the red hair/fair skin background.  Nature. 2012;491(7424):449-453.PubMedGoogle ScholarCrossref
18.
Raimondi  S, Sera  F, Gandini  S,  et al.  MC1R variants, melanoma and red hair color phenotype: a meta-analysis.  Int J Cancer. 2008;122(12):2753-2760.PubMedGoogle ScholarCrossref
19.
Pasquali  E, García-Borrón  JC, Fargnoli  MC,  et al; M-SKIP Study Group.  MC1R variants increased the risk of sporadic cutaneous melanoma in darker-pigmented Caucasians: a pooled-analysis from the M-SKIP project.  Int J Cancer. 2015;136(3):618-631.PubMedGoogle Scholar
20.
Kanetsky  PA, Panossian  S, Elder  DE,  et al.  Does MC1R genotype convey information about melanoma risk beyond risk phenotypes?  Cancer. 2010;116(10):2416-2428.PubMedGoogle Scholar
21.
Kanetsky  PA, Rebbeck  TR, Hummer  AJ,  et al.  Population-based study of natural variation in the melanocortin-1 receptor gene and melanoma.  Cancer Res. 2006;66(18):9330-9337.PubMedGoogle ScholarCrossref
22.
Scherer  D, Nagore  E, Bermejo  JL,  et al.  Melanocortin receptor 1 variants and melanoma risk: a study of 2 European populations.  Int J Cancer. 2009;125(8):1868-1875.PubMedGoogle ScholarCrossref
23.
de Torre  C, Garcia-Casado  Z, Martínez-Escribano  JA,  et al.  Influence of loss of function MC1R variants in genetic susceptibility of familial melanoma in Spain.  Melanoma Res. 2010;20(4):342-348.PubMedGoogle ScholarCrossref
24.
Hacker  E, Nagore  E, Cerroni  L,  et al.  NRAS and BRAF mutations in cutaneous melanoma and the association with MC1R genotype: findings from Spanish and Austrian populations.  J Invest Dermatol. 2013;133(4):1027-1033.PubMedGoogle ScholarCrossref
25.
Puig-Butillé  JA, Carrera  C, Kumar  R,  et al.  Distribution of MC1R variants among melanoma subtypes: p.R163Q is associated with lentigo maligna melanoma in a Mediterranean population.  Br J Dermatol. 2013;169(4):804-811.PubMedGoogle ScholarCrossref
26.
Guida  S, Bartolomeo  N, Zanna  PT,  et al.  Sporadic melanoma in southeastern Italy: the impact of melanocortin 1 receptor (MC1R) polymorphism analysis in low-risk people and report of three novel variants.  Arch Dermatol Res. 2015;307(6):495-503.PubMedGoogle ScholarCrossref
27.
Ibarrola-Villava  M, Hu  HH, Guedj  M,  et al.  MC1R, SLC45A2 and TYR genetic variants involved in melanoma susceptibility in southern European populations: results from a meta-analysis.  Eur J Cancer. 2012;48(14):2183-2191.Google ScholarCrossref
28.
Demenais  F, Mohamdi  H, Chaudru  V,  et al; Melanoma Genetics Consortium.  Association of MC1R variants and host phenotypes with melanoma risk in CDKN2A mutation carriers: a GenoMEL study.  J Natl Cancer Inst. 2010;102(20):1568-1583.PubMedGoogle ScholarCrossref
29.
Kanetsky  PA, Hay  JL.  Marshaling the translational potential of MC1R for precision risk assessment of melanoma [published online December 15, 2017].  Cancer Prev Res (Phila). doi:10.1158/1940-6207.CAPR-17-0255Google Scholar
30.
Bonilla  C, Parra  EJ, Pfaff  CL,  et al.  Admixture in the Hispanics of the San Luis Valley, Colorado, and its implications for complex trait gene mapping.  Ann Hum Genet. 2004;68(pt 2):139-153.PubMedGoogle ScholarCrossref
31.
Klimentidis  YC, Miller  GF, Shriver  MD.  Genetic admixture, self-reported ethnicity, self-estimated admixture, and skin pigmentation among Hispanics and Native Americans.  Am J Phys Anthropol. 2009;138(4):375-383.PubMedGoogle ScholarCrossref
32.
Hay  JL, Berwick  M, Zielaskowski  K,  et al.  Implementing an internet-delivered skin cancer genetic testing intervention to improve sun protection behavior in a diverse population: protocol for a randomized controlled trial.  JMIR Res Protoc. 2017;6(4):e52.PubMedGoogle ScholarCrossref
33.
Bloss  CS, Ornowski  L, Silver  E,  et al.  Consumer perceptions of direct-to-consumer personalized genomic risk assessments.  Genet Med. 2010;12(9):556-566.PubMedGoogle ScholarCrossref
34.
Kinney  AY, Gammon  A, Coxworth  J, Simonsen  SE, Arce-Laretta  M.  Exploring attitudes, beliefs, and communication preferences of Latino community members regarding BRCA1/2 mutation testing and preventive strategies.  Genet Med. 2010;12(2):105-115.PubMedGoogle ScholarCrossref
35.
Vadaparampil  ST, McIntyre  J, Quinn  GP.  Awareness, perceptions, and provider recommendation related to genetic testing for hereditary breast cancer risk among at-risk Hispanic women: similarities and variations by sub-ethnicity.  J Genet Couns. 2010;19(6):618-629.PubMedGoogle ScholarCrossref
36.
Hamilton  JG, Shuk  E, Arniella  G,  et al.  Genetic testing awareness and attitudes among Latinos: exploring shared perceptions and gender-based differences.  Public Health Genomics. 2016;19(1):34-46.PubMedGoogle ScholarCrossref
37.
Ashida  S, Wilkinson  AV, Koehly  LM.  Social influence and motivation to change health behaviors among Mexican-origin adults: implications for diet and physical activity.  Am J Health Promot. 2012;26(3):176-179.PubMedGoogle ScholarCrossref
38.
Rodríguez  VM, Robers  E, Zielaskowski  K,  et al.  Translation and adaptation of skin cancer genomic risk education materials for implementation in primary care.  J Community Genet. 2017;8(1):53-63.PubMedGoogle ScholarCrossref
39.
National Cancer Institute.  Anyone Can Get Skin Cancer. Bethesda, MD: National Cancer Institute; 2011.
40.
Fitzpatrick  TB.  The validity and practicality of sun-reactive skin types I through VI.  Arch Dermatol. 1988;124(6):869-871.PubMedGoogle ScholarCrossref
41.
Snapinn  SM, Small  RD.  Tests of significance using regression models for ordered categorical data.  Biometrics. 1986;42(3):583-592.PubMedGoogle ScholarCrossref
42.
Peterson  B, Harrell  FE.  Partial proportional odds models for ordinal response variables.  J R Stat Soc Ser C Appl Stat. 1990;39(2):205-217.Google Scholar
43.
Stokes  ME, Davis  CS, Koch  GG.  Categorical Data Analysis Using SAS. 3rd ed. Cary, NC: SAS Institute Inc; 2012.
44.
Addie  S, Olson  S, Beachy  S,  et al.  Applying an Implementation Science Approach to Genomic Medicine Workshop Summary. Washington, DC: National Academies Press; 2016.Crossref
45.
Patenaude  AF, Guttmacher  AE, Collins  FS.  Genetic testing and psychology: new roles, new responsibilities.  Am Psychol. 2002;57(4):271-282.PubMedGoogle ScholarCrossref
46.
Aspinwall  LG, Stump  TK, Taber  JM,  et al.  Genetic test reporting of CDKN2A provides informational and motivational benefits for managing melanoma risk.  Transl Behav Med. 2018;8(1):29-43.PubMedGoogle ScholarCrossref
47.
McBride  CM, Alford  SH, Reid  RJ, Larson  EB, Baxevanis  AD, Brody  LC.  Putting science over supposition in the arena of personalized genomics.  Nat Genet. 2008;40(8):939-942.PubMedGoogle ScholarCrossref
48.
McBride  CM, Alford  SH, Reid  RJ, Larson  EB, Baxevanis  AD, Brody  LC.  Characteristics of users of online personalized genomic risk assessments: implications for physician-patient interactions.  Genet Med. 2009;11(8):582-587.PubMedGoogle ScholarCrossref
49.
Smit  AK, Espinoza  D, Newson  AJ,  et al.  A pilot randomized controlled trial of the feasibility, acceptability, and impact of giving information on personalized genomic risk of melanoma to the public.  Cancer Epidemiol Biomarkers Prev. 2017;26(2):212-221.PubMedGoogle ScholarCrossref
Original Investigation
June 2018

Interest and Uptake of MC1R Testing for Melanoma Risk in a Diverse Primary Care Population: A Randomized Clinical Trial

Author Affiliations
  • 1Department of Psychiatry and Behavioral Sciences, Memorial Sloan Kettering Cancer Center, New York, New York
  • 2Clinical Research Finance, Memorial Sloan Kettering Cancer Center, New York, New York
  • 3Division of Epidemiology, Department of Internal Medicine, University of New Mexico, Albuquerque
  • 4Department of Communication, University of Utah, Salt Lake City
  • 5Huntsman Cancer Institute, University of Utah, Salt Lake City
  • 6New Mexico Translation and Transcription, Albuquerque
  • 7CRTC Population Sciences Academic Unit, University of New Mexico, Albuquerque
  • 8Department of Family and Community Medicine, University of New Mexico, Albuquerque
  • 9Center for Mind and Body Health, PLLC, Charlottesville, Virginia
  • 10Department of General Internal Medicine, University of New Mexico, Albuquerque
  • 11Department of Anthropology, University of New Mexico, Albuquerque
  • 12Klein Buendel Inc, Golden, Colorado
  • 13Department of Internal Medicine, University of New Mexico, Albuquerque
  • 14Department of Dermatology, University of New Mexico, Albuquerque
JAMA Dermatol. 2018;154(6):684-693. doi:10.1001/jamadermatol.2018.0592
Key Points

Question  What are the prevalence and patterns of interest in MC1R testing in a diverse, primary care population?

Findings  In this randomized clinical trial that included 499 adults receiving an invitation to consider MC1R testing, nearly half logged on to the study website to consider testing; non-Hispanic whites and those with higher educational attainment were most likely to be interested in testing, compared with Hispanics and those with lower educational attainment.

Meaning  Genetic testing for common variation in skin cancer risk may be acceptable in the general population; addressing potential for reduced utilization in minority and less educated individuals may be warranted.

Abstract

Importance  Germline variants in the MC1R gene are common and confer moderate melanoma risk in those with varied skin types. Approaches to precision skin cancer prevention that include genetic information may promote risk awareness and risk reduction in the general population, including Hispanics.

Objective  To examine prevalence of interest in and uptake of MC1R testing in the general population and examine patterns across demographic and skin cancer risk factors.

Design, Setting, and Participants  A randomized clinical trial examined interest in and uptake of MC1R testing among patients at University of New Mexico General Internal Medicine clinics. Study participants were randomized to either a usual-care condition (National Cancer Institute skin cancer pamphlet for diverse skin types) or an MC1R test offer. Participants were registered clinic patients (≥6 months) and English or Spanish fluent. Of the 600 participants recruited to the overall trial, the present study included those 499 participants randomized to the MC1R test offer.

Interventions  Participants were presented with the option to log onto the study website to read 3 educational modules presenting the rationale, benefits, and drawbacks of MC1R testing.

Main Outcomes and Measures  Main outcomes include website log on (yes vs no), saliva test kit request (yes vs no), and saliva test kit return for MC1R testing (yes vs no). Demographic and skin cancer risk factors were examined as potential predictors of test interest and uptake.

Results  Of the 499 participants (220 [44%] non-Hispanic white, 242 [48%] Hispanic, 396 [79%] female; mean [SD] age, 54 [14.3] years), 232 (46%) elected to learn about MC1R testing by logging onto the website; 204 (88%) of those who logged on decided to request testing; and 167 (82%) of those who requested testing returned the kit. The strongest predictors of website log on were race/ethnicity and education (non-Hispanic whites were more likely to log on [odds ratio for Hispanics vs non-Hispanic whites, 0.5; 95% CI, 0.3-0.7], as were more highly educated individuals [odds ratio for more than high school vs high school or less, 2.7; 95% CI, 1.7-4.3]). The strongest predictor of ordering the test was sunburn history (odds ratio, 5.4; 95% CI, 2.3-12.9 vs no sunburn history).

Conclusions and Relevance  There were moderately high levels of MC1R test interest and uptake in this diverse sample. Addressing potential barriers to testing may be warranted as genomic information becomes integrated into general population approaches to the precision prevention of skin cancer.

Trial Registration  ClinicalTrials.gov identifier: NCT03130569

Introduction

Melanoma is a common malignant neoplasm, and disproportionate increases in melanoma, particularly thicker tumors with poorer prognoses, have been documented in Hispanics in states with high levels of year-round sun exposure.1,2 In ethnically and racially diverse populations, melanoma results in greater morbidity and mortality due to the disease often being identified at later stages, and because of low physician and patient awareness that melanomas occur in these populations.1,3,4 Most individuals do not use sunscreen, wear protective clothing, or seek shade on a regular basis,5 and in the United States, approximately 35% of the population uses sunscreen consistently.6 This behavior extends to Hispanics of varying skin types,7 and Hispanics in the United States have high sunburn rates.8 Lower skin cancer awareness and perceived skin cancer risk have been identified as particular barriers to adequate and consistent sun protection in Hispanics.3,9-12 Personalized genomic testing for melanoma risk may promote risk awareness and risk reduction in the general population, including Hispanics.

Variants of the melanocortin-1 receptor gene (MC1R) confer moderate melanoma and basal cell cancer risks in the general population.13 This gene is located on the long arm of chromosome 16 and is related to cutaneous pigmentation (eg, fair skin, red hair)14-16 and has effects unrelated to UV exposure.17 A great deal of accumulated evidence, including systematic analyses of candidate genes, genome-wide association studies, and a meta-analysis of 12 melanoma case-control studies involving 6000 individuals,18 has identified 9 risk variants for melanoma with odds ratios ranging from 1.42 (95% CI, 1.09-1.85) to 2.45 (95% CI, 1.32-4.55).19 Importantly, variation in MC1R is associated with melanoma risk after adjustment for hair color and skin type.15,16,20,21MC1R predicts melanoma risk in Spanish22-24 and Mediterranean populations,16,25-27 with several studies indicating that MC1R may confer greater risk in individuals with a darker compared with lighter phenotype.28 Across Hispanic and non-Hispanic populations, approximately 60% to 70% of individuals have at least 1 risk variant.20,22,28 As such, there is potential for MC1R feedback to motivate sun protection behavior in diverse population subgroups.29

The present study is drawn from a randomized clinical trial (RCT) examining reach and utility of MC1R testing in a diverse primary care setting in Albuquerque, New Mexico. Hispanics in Albuquerque, New Mexico, have substantial Spanish ancestry,30,31 so we expected to find a relatively high frequency of risk variants across Hispanic and non-Hispanic study participants.22 Over the past decade, the rapid pace of discovery of risk-influencing genes and the use of the internet as an important source of health information have evolved in parallel. Thus, for the RCT we developed internet delivery of information regarding MC1R testing, and participants could only order a saliva test kit online (see Hay et al32 or the Supplement for full trial methods). To date, uptake of internet direct-to-consumer personalized genomic testing has generally been concentrated among white, highly educated consumers.33 The literature examining responses to genetic information in Hispanics is limited, yet there are promising indications that Hispanics may have high interest in learning more about their genetic risks of developing cancers.34,35 Research has identified potential barriers to genetic risk communication in Hispanics, such as health literacy, language, and access,36 as well as potential facilitators of use such as a cultural orientation prioritizing family relationships and communication.37 All RCT study materials (website, risk feedback, surveys) were available in Spanish as well as English.38

There were 2 study aims: (1) to examine interest and uptake of MC1R testing and (2) to examine demographic and skin cancer risk factor covariates of interest and uptake of MC1R testing. These findings can be used to help shape the public health translation of personalized genomic testing for melanoma as it may become more widely available in coming years.

Methods
Participants

Bilingual project assistants approached primary care patients in the University of New Mexico (UNM) outpatient primary care clinics with invitation flyers (English and Spanish) and National Cancer Institute skin cancer information for diverse skin types (available in English and Spanish versions; Anyone Can Get Skin Cancer).39 Patients were eligible if they were registered in any UNM clinic for at least 6 months, assigned a UNM primary care physician, aged 18 years or older, and fluent in English or Spanish. Those who were eligible but refused study participation completed a 1-minute refuser survey that assessed reasons for refusal and demographic characteristics (ethnicity, race, sex, educational attainment, and age). All study procedures and materials were approved by the UNM institutional review board.

Procedure

Eligible patients completed written informed consent and a baseline assessment and were randomized to an invitation to consider personalized genomic testing (via MC1R) for skin cancer risk via logging onto the study website or to usual-care control (randomized 5:1; balanced across Hispanic vs non-Hispanic ethnicity). Usual-care controls did not receive an invitation to log on. Those randomized to the intervention arm could log onto the study website to read the 3 educational modules regarding MC1R testing and then register a test decision. Those without internet access were also offered the opportunity to view the website via paper form.

Measures

Outcome measures, including registration of a test decision (yes vs no), request of a saliva test kit (yes vs no), and return of a test kit (yes vs no), were assessed as primary assessments of study interest and uptake.

Predictor measures were as follows. All participants completed baseline assessments that included their ethnicity, race, sex, educational attainment, age, birth country, marital status, employment status, income, and internet access (ever; home). Skin cancer risk factors were assessed and included personal cancer history (cancer in general; skin cancer), family history of skin cancer, and skin type (burnability, tannability, and sunburn history [yes vs no; lifetime number]40).

Statistical Approach

Participant characteristics were reported overall and by ethnicity, and differences by ethnicity were tested using independent-samples t test for age and χ2 tests for other characteristics. Descriptive statistics were calculated for website log on (yes vs no), requesting a test kit (yes vs no), and providing a saliva sample for MC1R testing (yes vs no). We evaluated unadjusted and adjusted logistic regression models examining predictors of each of these outcomes. Next, an ordinal outcome41 representing extent of test follow-through was calculated and modeled using unadjusted and adjusted ordinal logistic regression models to examine each predictor (0 = no log on; 1 = log on only; 2 = log on and test kit request but failure to return it; 3 = log on, test request, and saliva kit return). Ordinal logit models are interpreted such that the odds ratio indicates the odds of more vs less extensive test follow-through; ordinal logit models assume the proportional odds assumption,42 which we tested via the score χ2 test.43 For all outcomes, the adjusted model was built starting with individually significant predictors from the unadjusted models, and using a backward technique with Akaike information criteria to reduce to the best-fit model. The α level was set to .05 and all tests were 2 sided. All statistics were conducted in SAS, version 9.4.

Results

The project assistants approached 1998 primary care patients, and 917 (46%) agreed to be screened for eligibility. Ineligibility (n = 191) was primarily due to being a registered UNM patient for less than 6 months (166 [87%]) or not having a primary care provider (36 [19%]). Of 726 eligible patients, 105 refused (predominantly due to lack of time or interest) and 621 consented to study participation (86% of eligible patients; 31% of all patients approached); 21 did not complete the baseline assessment. Study acceptance was significantly higher in non-Hispanic whites (93%) compared with Hispanics (89%), and those with higher (greater than high school, 94%) compared with lower educational attainment (high school or less, 84%), but did not differ on other demographic factors. Of the 600 participants who enrolled in the RCT and completed the baseline assessment, 499 were randomized to the MC1R testing website and make up the total sample for the current analyses (Figure).

Of these 499, most were non-Hispanic white (220 [44%]) or Hispanic (242 [48%]), female (396 [79%]), and 116 (23%) had a high school diploma or less. The mean age was 54 years (range, 19-85 years). Most participants were born in the United States (455 [92%]). Approximately half (263 [53%]) reported an annual income of $30 000 or more. Levels of internet access were high (ever, 416 [83%]; home, 409 [82%]). Few participants reported a personal history of skin cancer (31 [6%]) or other cancers (49 [10%]). A total of 118 participants (24%) reported a first-degree family history of skin cancer. In terms of skin phenotype, 195 (39%) reported that they burn easily when in the sun for 1 hour and 323 (66%) reported that they tan easily. Also, 289 (58%) had a history of sunburn and most of those (197 [68%]) reported 3 or more sunburns. Hispanics and non-Hispanic whites significantly differed across many of the demographic and skin cancer risk factors, indicating generally lower socioeconomic status and reduced skin cancer risk in Hispanics compared with non-Hispanic whites. Full descriptive statistics are reported in Table 1.

Prevalence and Predictors of MC1R Testing Interest and Uptake

As reported in Table 2, almost half of participants (232 [46%]) accepted the invitation and logged onto the study website (18 of the 232 viewed the study website via paper form, as per their preference). Website log on rate was higher in non-Hispanic whites compared with Hispanics, and higher in those with higher (greater than high school) compared with lower educational attainment (high school diploma or less). Those with Internet access (ever; home) were more likely to log on compared with those without access. Skin cancer risk factors were also related to website log on; participants with at least 1 first-degree relative who had a history of skin cancer and participants with a personal sunburn history were more likely to log on compared with those without these histories. In the adjusted analysis, only 2 variables—race/ethnicity and education—remained significant predictors of website log on.

Most participants (204 of 232 [88%]) who logged onto the website decided to request the saliva test kit. The remainder either refused testing (13 [6%]) or did not register a test decision (15 [6%]). Among participants who logged onto the website, the test kit request rate was higher in non-Hispanic whites, men, those with higher educational attainment, and those with internet access (ever or at home). Additionally, participants with a personal sunburn history were more likely to request a test kit. Adjusted analyses indicated that the most important predictor of test request was having a sunburn history.

Finally, most (167 of 204 [82%]) who requested testing completed and returned the saliva kit. The rate of returning the test kit was higher in non-Hispanic whites and older participants. In the adjusted analysis, neither Hispanic status nor age remained significant.

Predictors of Test Follow-through

As reported in Table 3, in unadjusted models of the ordinal outcome, race/ethnicity, education, internet access (ever or at home), family history of skin cancer, and personal sunburn history were significant predictors of increasing test follow-through. Specifically, non-Hispanic whites, those with higher educational attainment, those with internet access, those reporting a family history of skin cancer (first-degree relative or non–first-degree relative), and those with a sunburn history had higher test follow-through. In the adjusted model, race/ethnicity and education remained important predictors.

Discussion

A 2016 report from the National Academy of Sciences highlighted the need to address access issues in genomic medicine.44 Despite this need, for-profit companies are already marketing and offering genetic testing directly to consumers and although the direct-to-consumer model seeks to increase access, utilization has continued to be concentrated among non-Hispanic white, highly educated consumers.33

This model has largely bypassed behavioral research that could ensure broad utility and reach of this technology to diverse populations, arguing for the time-sensitive need to develop an empirical basis to maximize the benefits and minimize the harms of genomic feedback, even as evidence for specific gene variants and panels inevitably shifts over time. Psychosocial research has highlighted interest in, and outcomes of, genetic testing in high-risk families who present in specialized clinics and receive extensive genetic counseling,45 such as exceptionally high risk families offered testing for CDKN2A/p16.46 However, this research has been largely conducted in the context of familial disease, which does not shed light on how the general population will respond to precision skin cancer prevention approaches that may include genetic information.

In the present study, we found moderately high rates of interest and uptake of MC1R testing in a diverse primary care setting. These rates greatly exceed what was found in prior work offering multiplex genetic testing for risk for 8 conditions (including MC1R testing for melanoma risk)47 in primary care in Detroit, Michigan. For instance, nearly half (46.5%) of our sample logged on to the informational website compared with 30% in prior work; and of those who logged on, 87.9% requested a test kit compared with 50% in prior work.48 Finally, 82% followed through with providing a saliva sample compared with 30% who, in this prior study, came into the clinic to provide a blood sample. Possibly the use of a mailed saliva sample in this study rather than providing a blood sample in the earlier study facilitated testing. Additionally, most (62%) of those who ultimately followed through with testing had anticipated that they were “very likely” to choose testing (as assessed at initial website log on [J.L.H., D.B., K.Z., et al, unpublished data, 2017]) even before reading the website information; as such, the decision to be tested for some participants was made before website log on. Information that was presented in the website may have been instrumental for the 40% who were not certain about testing at enrollment. Given this demonstrated propensity toward testing for those who decided to log on to the website, the findings may not be entirely generalizable to the range of patients who might encounter messages recommending precision testing and counseling for skin cancer outside a medical clinic. Testing rates might well be higher in clinical practice if recommended by physicians.

There were some common predictors of interest in and uptake of MC1R testing, as well as test follow-through. Across our study outcome variables, the most important predictors were race/ethnicity and education, with those identifying as Hispanic, as well as those with lower educational attainment, showing lower rates of interest and uptake compared with non-Hispanic whites and more highly educated participants, consistent with demographic effects shown for interest in direct-to-consumer personalized genomic testing.33 It is important to point out that in the present study skin type variables were confounded with Hispanic status. As such, 181 (71%) non-Hispanic white participants reported a sunburn history compared with 108 (45%) Hispanics (P < .001), 112 (44%) non-Hispanic white participants reported that they sunburn easily, compared with 82 (34%) Hispanics, and 159 (62%) non-Hispanic white participants reported that they tan easily, compared with 163 (67%) Hispanics. Indeed, those with darker skin types often perceive lower risk for melanoma.3 More research is needed to explore barriers to genomic testing among racially and ethnically diverse and less educated patients, including lack of knowledge, lower genomic literacy, and lack of confidence in the medical system, to achieve maximum benefits of precision prevention for skin cancer and other chronic diseases in the broad population who stand to benefit from such technologies. In the present study, we obtained translations of all our study materials into New Mexican Spanish and conducted preliminary qualitative research to confirm the comprehensibility and acceptability of these materials38; yet further efforts to understand important barriers clearly remain. Comparatively lower interest in diverse populations might create or perpetuate health disparities in this population, and key factors that predict existing health disparities in this population might also be barriers. Importantly, even in those with a high school education or less, a sizable minority (33 [28%]) of participants went to the website to learn about MC1R testing, implying that less educated patients are reachable. In future analyses, we will examine comprehension of and satisfaction with the online information across educational and health literacy levels, which will help dictate future directions in adapting intervention material for diverse populations.

The importance of a history of sunburn emerged as an important predictor of participants’ decision to order a test kit once they considered the rationale, benefits, and drawbacks of MC1R testing. This indicates that having a history of sunburn is distinctly important in the decision to be tested rather than whether to seek information (log on) on skin cancer genetic testing. In future work, we will explore the psychosocial predictors of interest and uptake for MC1R testing, which will inform work exploring the potential for healthy behavior change after testing for other common genetic markers for melanoma risk.49

Limitations

There were notable study strengths and limitations. The sample was diverse and large. Also, the outcome measures were behavioral rather than self-reported. The trial was conducted in only 1 primary care health system in an academic setting and in a single location in the American Southwest, which may not fully generalize to other primary care systems. However, we did recruit and have all study materials (including our website) available in Spanish as well as English,38 setting the stage for dissemination of our MC1R educational website and risk feedback materials in other settings, and across languages. It is possible that we may have differentially recruited those with higher interest in testing; we will explore this possibility in subsequent psychosocial analyses. Although not all refusers (38 of 105) were willing to complete it, inclusion of a refuser survey was a study strength that allows us to clarify some factors related to study participation. Also in subsequent analyses we will examine psychosocial predictors of interest and uptake of MC1R testing.

Conclusions

We document relatively high rates of interest and follow-through for skin cancer genetic testing in primary care. While evidence will continue to accumulate concerning the reliability and utility of such markers in precision skin cancer risk assessments, this study has advanced the public health translation of skin cancer genetic testing in providing insight into how such information may be received in populations unselected for risk status drawn from the general population. While interest and test uptake was higher than in prior work, socioeconomic and demographic patterns in testing emerged. Identification of these disparities represents a valuable step providing guidance for future research to understand underlying mechanisms at play, directing us to identify future solutions to ensure ease of availability of genetic information seeking in the general population.

Back to top
Article Information

Accepted for Publication: February 17, 2018.

Corresponding Author: Jennifer L. Hay, PhD, Department of Psychiatry and Behavioral Sciences, Memorial Sloan Kettering Cancer Center, 641 Lexington Ave, 7th Floor, New York, NY 10022 (hayj@mskcc.org).

Published Online: May 9, 2018. doi:10.1001/jamadermatol.2018.0592

Author Contributions: Dr Hay had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Hay, Kaphingst, Sussman, Rodríguez, Li, Buller, Berwick.

Acquisition, analysis, or interpretation of data: Hay, Zielaskowski, Meyer White, Kaphingst, Robers, Guest, Talamantes, Schwartz, Rodríguez, Li, Schofield, Bigney, Hunley, Buller, Berwick.

Drafting of the manuscript: Hay, Zielaskowski, Sussman, Talamantes, Berwick.

Critical revision of the manuscript for important intellectual content: Hay, Zielaskowski, Meyer White, Kaphingst, Robers, Guest, Schwartz, Rodríguez, Li, Schofield, Bigney, Hunley, Buller, Berwick.

Statistical analysis: Hay, Zielaskowski, Li, Schofield.

Obtained funding: Hay, Berwick.

Administrative, technical, or material support: Zielaskowski, Meyer White, Robers, Guest, Schwartz, Rodríguez, Bigney, Hunley, Berwick.

Study supervision: Hay, Meyer White, Guest, Sussman, Berwick.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported in part by the National Cancer Institute (R01 CA181241), Support/Core Grant (P30 CA008748), and training grant (T32 CA009461).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We are indebted to Deborah Bowen, PhD, University of Washington, and Colleen McBride, PhD, Emory University, for their support in study inception, and to Stephanie Christian, MPH, and Jillisia James, BS, Department of Psychiatry and Behavioral Sciences, Memorial Sloan Kettering Cancer Center, for their administrative contributions in completing this manuscript. They did not receive compensation from a funding sponsor for their contributions. We also thank the study participants for their valued contributions to the research.

References
1.
Rouhani  P, Pinheiro  PS, Sherman  R,  et al.  Increasing rates of melanoma among nonwhites in Florida compared with the United States.  Arch Dermatol. 2010;146(7):741-746.PubMedGoogle ScholarCrossref
2.
Pollitt  RA, Clarke  CA, Swetter  SM, Peng  DH, Zadnick  J, Cockburn  M.  The expanding melanoma burden in California Hispanics: importance of socioeconomic distribution, histologic subtype, and anatomic location.  Cancer. 2011;117(1):152-161.PubMedGoogle ScholarCrossref
3.
Robinson  JK, Joshi  KM, Ortiz  S, Kundu  RV.  Melanoma knowledge, perception, and awareness in ethnic minorities in Chicago: recommendations regarding education.  Psychooncology. 2011;20(3):313-320.PubMedGoogle ScholarCrossref
4.
Roman  C, Lugo-Somolinos  A, Thomas  N.  Skin cancer knowledge and skin self-examinations in the Hispanic population of North Carolina: the patient’s perspective.  JAMA Dermatol. 2013;149(1):103-104.PubMedGoogle ScholarCrossref
5.
Kasparian  NA, McLoone  JK, Meiser  B.  Skin cancer-related prevention and screening behaviors: a review of the literature.  J Behav Med. 2009;32(5):406-428.PubMedGoogle ScholarCrossref
6.
Coups  EJ, Manne  SL, Heckman  CJ.  Multiple skin cancer risk behaviors in the US population.  Am J Prev Med. 2008;34(2):87-93.PubMedGoogle ScholarCrossref
7.
Coups  EJ, Stapleton  JL, Hudson  SV, Medina-Forrester  A, Natale-Pereira  A, Goydos  JS.  Sun protection and exposure behaviors among Hispanic adults in the United States: differences according to acculturation and among Hispanic subgroups.  BMC Public Health. 2012;12:985.PubMedGoogle ScholarCrossref
8.
Saraiya  M, Hall  HI, Uhler  RJ.  Sunburn prevalence among adults in the United States, 1999.  Am J Prev Med. 2002;23(2):91-97.PubMedGoogle ScholarCrossref
9.
Rodriguez  VM, Shuk  E, Arniella  G,  et al.  A qualitative exploration of Latinos’ perceptions about skin cancer: the role of gender and linguistic acculturation.  J Cancer Educ. 2017;32(3):438-446.PubMedGoogle ScholarCrossref
10.
Coups  EJ, Stapleton  JL, Manne  SL,  et al.  Psychosocial correlates of sun protection behaviors among US Hispanic adults.  J Behav Med. 2014;37(6):1082-1090.PubMedGoogle ScholarCrossref
11.
Friedman  LC, Bruce  S, Weinberg  AD, Cooper  HP, Yen  AH, Hill  M.  Early detection of skin cancer: racial/ethnic differences in behaviors and attitudes.  J Cancer Educ. 1994;9(2):105-110.PubMedGoogle Scholar
12.
Pipitone  M, Robinson  JK, Camara  C, Chittineni  B, Fisher  SG.  Skin cancer awareness in suburban employees: a Hispanic perspective.  J Am Acad Dermatol. 2002;47(1):118-123.PubMedGoogle ScholarCrossref
13.
Udayakumar  D, Tsao  H.  Moderate- to low-risk variant alleles of cutaneous malignancies and nevi: lessons from genome-wide association studies.  Genome Med. 2009;1(10):95.PubMedGoogle ScholarCrossref
14.
Goldstein  AM, Landi  MT, Tsang  S, Fraser  MC, Munroe  DJ, Tucker  MA.  Association of MC1R variants and risk of melanoma in melanoma-prone families with CDKN2A mutations.  Cancer Epidemiol Biomarkers Prev. 2005;14(9):2208-2212.PubMedGoogle Scholar
15.
Kanetsky  PA, Ge  F, Najarian  D,  et al.  Assessment of polymorphic variants in the melanocortin-1 receptor gene with cutaneous pigmentation using an evolutionary approach.  Cancer Epidemiol Biomarkers Prev. 2004;13(5):808-819.PubMedGoogle Scholar
16.
Landi  MT, Kanetsky  PA, Tsang  S,  et al.  MC1R, ASIP, and DNA repair in sporadic and familial melanoma in a Mediterranean population.  J Natl Cancer Inst. 2005;97(13):998-1007.PubMedGoogle ScholarCrossref
17.
Mitra  D, Luo  X, Morgan  A,  et al.  An ultraviolet-radiation-independent pathway to melanoma carcinogenesis in the red hair/fair skin background.  Nature. 2012;491(7424):449-453.PubMedGoogle ScholarCrossref
18.
Raimondi  S, Sera  F, Gandini  S,  et al.  MC1R variants, melanoma and red hair color phenotype: a meta-analysis.  Int J Cancer. 2008;122(12):2753-2760.PubMedGoogle ScholarCrossref
19.
Pasquali  E, García-Borrón  JC, Fargnoli  MC,  et al; M-SKIP Study Group.  MC1R variants increased the risk of sporadic cutaneous melanoma in darker-pigmented Caucasians: a pooled-analysis from the M-SKIP project.  Int J Cancer. 2015;136(3):618-631.PubMedGoogle Scholar
20.
Kanetsky  PA, Panossian  S, Elder  DE,  et al.  Does MC1R genotype convey information about melanoma risk beyond risk phenotypes?  Cancer. 2010;116(10):2416-2428.PubMedGoogle Scholar
21.
Kanetsky  PA, Rebbeck  TR, Hummer  AJ,  et al.  Population-based study of natural variation in the melanocortin-1 receptor gene and melanoma.  Cancer Res. 2006;66(18):9330-9337.PubMedGoogle ScholarCrossref
22.
Scherer  D, Nagore  E, Bermejo  JL,  et al.  Melanocortin receptor 1 variants and melanoma risk: a study of 2 European populations.  Int J Cancer. 2009;125(8):1868-1875.PubMedGoogle ScholarCrossref
23.
de Torre  C, Garcia-Casado  Z, Martínez-Escribano  JA,  et al.  Influence of loss of function MC1R variants in genetic susceptibility of familial melanoma in Spain.  Melanoma Res. 2010;20(4):342-348.PubMedGoogle ScholarCrossref
24.
Hacker  E, Nagore  E, Cerroni  L,  et al.  NRAS and BRAF mutations in cutaneous melanoma and the association with MC1R genotype: findings from Spanish and Austrian populations.  J Invest Dermatol. 2013;133(4):1027-1033.PubMedGoogle ScholarCrossref
25.
Puig-Butillé  JA, Carrera  C, Kumar  R,  et al.  Distribution of MC1R variants among melanoma subtypes: p.R163Q is associated with lentigo maligna melanoma in a Mediterranean population.  Br J Dermatol. 2013;169(4):804-811.PubMedGoogle ScholarCrossref
26.
Guida  S, Bartolomeo  N, Zanna  PT,  et al.  Sporadic melanoma in southeastern Italy: the impact of melanocortin 1 receptor (MC1R) polymorphism analysis in low-risk people and report of three novel variants.  Arch Dermatol Res. 2015;307(6):495-503.PubMedGoogle ScholarCrossref
27.
Ibarrola-Villava  M, Hu  HH, Guedj  M,  et al.  MC1R, SLC45A2 and TYR genetic variants involved in melanoma susceptibility in southern European populations: results from a meta-analysis.  Eur J Cancer. 2012;48(14):2183-2191.Google ScholarCrossref
28.
Demenais  F, Mohamdi  H, Chaudru  V,  et al; Melanoma Genetics Consortium.  Association of MC1R variants and host phenotypes with melanoma risk in CDKN2A mutation carriers: a GenoMEL study.  J Natl Cancer Inst. 2010;102(20):1568-1583.PubMedGoogle ScholarCrossref
29.
Kanetsky  PA, Hay  JL.  Marshaling the translational potential of MC1R for precision risk assessment of melanoma [published online December 15, 2017].  Cancer Prev Res (Phila). doi:10.1158/1940-6207.CAPR-17-0255Google Scholar
30.
Bonilla  C, Parra  EJ, Pfaff  CL,  et al.  Admixture in the Hispanics of the San Luis Valley, Colorado, and its implications for complex trait gene mapping.  Ann Hum Genet. 2004;68(pt 2):139-153.PubMedGoogle ScholarCrossref
31.
Klimentidis  YC, Miller  GF, Shriver  MD.  Genetic admixture, self-reported ethnicity, self-estimated admixture, and skin pigmentation among Hispanics and Native Americans.  Am J Phys Anthropol. 2009;138(4):375-383.PubMedGoogle ScholarCrossref
32.
Hay  JL, Berwick  M, Zielaskowski  K,  et al.  Implementing an internet-delivered skin cancer genetic testing intervention to improve sun protection behavior in a diverse population: protocol for a randomized controlled trial.  JMIR Res Protoc. 2017;6(4):e52.PubMedGoogle ScholarCrossref
33.
Bloss  CS, Ornowski  L, Silver  E,  et al.  Consumer perceptions of direct-to-consumer personalized genomic risk assessments.  Genet Med. 2010;12(9):556-566.PubMedGoogle ScholarCrossref
34.
Kinney  AY, Gammon  A, Coxworth  J, Simonsen  SE, Arce-Laretta  M.  Exploring attitudes, beliefs, and communication preferences of Latino community members regarding BRCA1/2 mutation testing and preventive strategies.  Genet Med. 2010;12(2):105-115.PubMedGoogle ScholarCrossref
35.
Vadaparampil  ST, McIntyre  J, Quinn  GP.  Awareness, perceptions, and provider recommendation related to genetic testing for hereditary breast cancer risk among at-risk Hispanic women: similarities and variations by sub-ethnicity.  J Genet Couns. 2010;19(6):618-629.PubMedGoogle ScholarCrossref
36.
Hamilton  JG, Shuk  E, Arniella  G,  et al.  Genetic testing awareness and attitudes among Latinos: exploring shared perceptions and gender-based differences.  Public Health Genomics. 2016;19(1):34-46.PubMedGoogle ScholarCrossref
37.
Ashida  S, Wilkinson  AV, Koehly  LM.  Social influence and motivation to change health behaviors among Mexican-origin adults: implications for diet and physical activity.  Am J Health Promot. 2012;26(3):176-179.PubMedGoogle ScholarCrossref
38.
Rodríguez  VM, Robers  E, Zielaskowski  K,  et al.  Translation and adaptation of skin cancer genomic risk education materials for implementation in primary care.  J Community Genet. 2017;8(1):53-63.PubMedGoogle ScholarCrossref
39.
National Cancer Institute.  Anyone Can Get Skin Cancer. Bethesda, MD: National Cancer Institute; 2011.
40.
Fitzpatrick  TB.  The validity and practicality of sun-reactive skin types I through VI.  Arch Dermatol. 1988;124(6):869-871.PubMedGoogle ScholarCrossref
41.
Snapinn  SM, Small  RD.  Tests of significance using regression models for ordered categorical data.  Biometrics. 1986;42(3):583-592.PubMedGoogle ScholarCrossref
42.
Peterson  B, Harrell  FE.  Partial proportional odds models for ordinal response variables.  J R Stat Soc Ser C Appl Stat. 1990;39(2):205-217.Google Scholar
43.
Stokes  ME, Davis  CS, Koch  GG.  Categorical Data Analysis Using SAS. 3rd ed. Cary, NC: SAS Institute Inc; 2012.
44.
Addie  S, Olson  S, Beachy  S,  et al.  Applying an Implementation Science Approach to Genomic Medicine Workshop Summary. Washington, DC: National Academies Press; 2016.Crossref
45.
Patenaude  AF, Guttmacher  AE, Collins  FS.  Genetic testing and psychology: new roles, new responsibilities.  Am Psychol. 2002;57(4):271-282.PubMedGoogle ScholarCrossref
46.
Aspinwall  LG, Stump  TK, Taber  JM,  et al.  Genetic test reporting of CDKN2A provides informational and motivational benefits for managing melanoma risk.  Transl Behav Med. 2018;8(1):29-43.PubMedGoogle ScholarCrossref
47.
McBride  CM, Alford  SH, Reid  RJ, Larson  EB, Baxevanis  AD, Brody  LC.  Putting science over supposition in the arena of personalized genomics.  Nat Genet. 2008;40(8):939-942.PubMedGoogle ScholarCrossref
48.
McBride  CM, Alford  SH, Reid  RJ, Larson  EB, Baxevanis  AD, Brody  LC.  Characteristics of users of online personalized genomic risk assessments: implications for physician-patient interactions.  Genet Med. 2009;11(8):582-587.PubMedGoogle ScholarCrossref
49.
Smit  AK, Espinoza  D, Newson  AJ,  et al.  A pilot randomized controlled trial of the feasibility, acceptability, and impact of giving information on personalized genomic risk of melanoma to the public.  Cancer Epidemiol Biomarkers Prev. 2017;26(2):212-221.PubMedGoogle ScholarCrossref
×