SEARCH

SEARCH BY CITATION

Keywords:

  • melanoma;
  • time trends;
  • epidemic;
  • Hispanic

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

BACKGROUND

Hispanics comprise almost one-third of the population of California, are the most rapidly increasing racial/ethnic group in the state, and represent almost one-third of all Hispanics in the U.S. California has among the highest rates of melanoma in the world, yet little is known about trends in melanoma in its Hispanic population.

METHODS

Trends in invasive and in situ melanoma incidence data and melanoma mortality data, between 1988 and 2001, from the California Cancer Registry were analyzed. Trends in the Hispanic population were compared with those in the non-Hispanic white population. Time trends in tumors of differing thicknesses and histology were assessed.

RESULTS

There was a statistically significant 1.8% per year increase in incidence of invasive melanomas among Hispanic males and a similar but nonstatistically significant increase in invasive melanoma among Hispanic females between 1988 and 2001. Among Hispanic males and females tumors thicker than 1.5 mm at presentation increased at 11.6% per year (95% confidence interval [CI], 8.1, 15.2) and 8.9% per year (95% CI, 4.7, 13.3), respectively.

CONCLUSION

Rates of invasive melanoma have increased markedly among Hispanics in California since 1988. In contrast to trends in the non-Hispanic white population, increases in melanoma in Hispanics have been confined to thicker tumors, whose prognosis is poor. We recommend that efforts be undertaken immediately to target both primary and secondary melanoma prevention messages to Hispanic communities. Cancer 2006. © 2006 American Cancer Society.

Little is known about the occurrence of melanoma in the Hispanic population of the U.S. This is partly because the disease is sufficiently rare that sample size precludes the calculation of rates in many population-based registries throughout the U.S. However, the Hispanic population in California is large, comprising 11.97 million people, and is increasing rapidly: at the 2000 Census 32.4% of the California population was Hispanic, up from 25.8% in 1990.1

The growing Hispanic population in California is occurring in one of the areas of highest melanoma incidence in the world.2 Both the immigrant and indigent Hispanic populations in California share one of the most important melanoma risk factors: birth and childhood in an area of high sun exposure potential,3 with 94.3% of the non-California-born Hispanic population coming from Mexico or Central America.

Melanoma is less common in nonwhite than in white populations worldwide.2 In California, Hispanic males had a melanoma rate of 2.8 per 100,000 compared with 17.2 per 100,000 in non-Hispanic males between 1988 and 1993.4 However, melanoma is an important problem in the Hispanic population, accounting for 1.2% of all male cancers and 1.6% of all female cancers in California's Hispanic population, comparable in magnitude to esophageal carcinoma and Hodgkin disease.5 We are aware of no specific melanoma prevention interventions in California (or elsewhere) targeted at the Hispanic population.

We assessed recent trends in melanoma incidence and mortality in the Hispanic population of California and contrasted these with trends among the non-Hispanic white population.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Data were obtained from the California Cancer Registry (CCR: www.ccrcal.org). Since 1988, statewide cancer data have been reported in a uniform way. This population-based cancer surveillance system represents a cooperative relationship between hospitals and other cancer diagnostic or treatment facilities, regional registries, and the California Department of Health Services (CDHS). It comprises 10 regional registries that report cancer incidence data to the Cancer Surveillance Section of CDHS: About 140,000 new cancer cases and about 50,000 cancer deaths are reported to the CCR each year.

Incidence Data

Cancer incidence data are based on new cases of cancer that were first diagnosed among California residents from January 1, 1988, to December 31, 2001, and were reported to the CCR as of November 2003. Cancer reporting by hospitals, physicians, and laboratories is mandated by state law. Records from pathology laboratories throughout the state are reviewed routinely in an effort to identify malignancies that have not been previously reported. Tumors that are identified only through pathology records and death certificates are ‘followed back’ to the physician of record to complete patient records. Data on histologic type and tumor thickness are abstracted from the patient's medical records and pathology reports. Coding of histologic type and tumor thickness are completed according to the International Classification of Diseases for Oncology (ICDO).6

Incident melanomas were classified as in situ or invasive. Invasive tumors were further classified by their thickness (in mm) and histologic type. Thickness was categorized in the same groups typically found in survival analyses7 and representing the levels most commonly used to describe the changing incidence of melanoma (< 0.75 mm, 0.75–1.49 mm, 1.50 mm and greater, and unknown). Typically, patients with lesions < 0.75 mm experience excellent survival rates (greater than 90% at 10 years), whereas prognosis for melanomas 1.5 mm or greater is significantly poorer (around 60% at 10 years).7 We grouped melanomas using ICDO-36 by histology into superficial spreading melanoma (SMM), 8743; nodular melanoma (NM), 8721; acral lentiginous melanoma (ALM), 8744; Hutchinson melanotic freckle (HMF), 8742; and malignant melanoma (MM), 8720. All other melanomas (8722–8741, 8745–8790) were omitted from histology-specific analyses.

Mortality Data

The CCR obtains computerized files containing information on cancer-related deaths from the California Department of Health Services' Center for Health Statistics. Death certificate master files were used for all years. Cause of death is coded by the International Classification of Diseases, Ninth Edition (ICD-9) for deaths occurring from 1988–1998.8 Beginning in 1999 and thereafter, cause of death is coded by the International Classification of Diseases, Tenth Edition (ICD-10).9

Defining Hispanic and Non-Hispanic White Populations

Race/ethnicity is grouped into the mutually exclusive categories of non-Hispanic white, non-Hispanic black, Hispanic, and six other groups according to the race/ethnicity reported in medical records. Persons in any category with a last name on the 1980 U.S. Census list of 12,497 Hispanic surnames are also categorized as Hispanic. Maiden name, when present, is used instead of last name to identify Hispanic women by surname. The use of surname to identify persons of Hispanic ethnicity was adopted by the CCR because of the recognized underreporting of Hispanic ethnicity on the medical record and death certificate.10 Overall statewide cancer incidence and mortality rates for Hispanics, based on this definition, are about 14% higher than those based on medical record and death certificate alone, and rates for non-Hispanic whites are about 1.4% lower.

Denominator Data

The CCR annual population estimates from 1988–2001 by age, sex, and race/ethnicity were used for the calculation of rates. They are based on data from the 1990 and 2000 U.S. population censuses, with linear interpolation for the intercensal years and extrapolation for 1988–1989 and 2001. The 1990 Census counts are classified in mutually exclusive groups of non-Hispanic white and Hispanic.

Statistical Methods

By using the year 2000 U.S. Census population as the standard for direct age adjustment, we calculated annual gender- and race-specific age-adjusted incidence rates (AAIRs) between 1988 and 2001. Estimated annual percent changes (EAPC) in incidence rates between 1988 and 2001 were calculated using the JoinPoint regression program.11 For analyses of trends by tumor thickness level and histologic type we aggregated data into one 2-year period (1988–1989) and four 3-year periods (1990–2001 inclusive) because data were too sparse to calculate meaningful sex-specific, thickness-specific, or histology-specific rates in Hispanics using annual data.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Overall Trends in Melanoma in California's Hispanic and Non-Hispanic White Populations

Between 1988 and 2001 the incidence of invasive melanoma increased at a statistically significant rate of 3.9% per year among non-Hispanic white males, 3.3% per year among non-Hispanic white females, and 1.8% per year among Hispanic males (Table 1). This overall 1.8% annual increase consisted of a 7.3% annual increase (95% confidence interval [CI], 1.2, 13.7) in the period between 1996 and 2001 (Fig. 1). Rates of invasive melanoma among Hispanic females increased at 0.6% per year, but this increase was not statistically significant.

Table 1. Average Annual Age-Adjusted Incidence Rates (AAIR) per 100,000 Population (Standardized to the Year 2000 U.S. Population) and Estimated Annual Percentage Change (EAPC) in Invasive and In Situ Melanomas, and for Melanoma Deaths, by Gender in Non-Hispanic Whites and Hispanics in California, 1988 to 2001
 GenderEthnic groupNo. of casesAge-adjusted incidence rate (per 100,000)Overall trend, 1988–2001 (EAPC)
Invasive melanoma incidenceMaleNon-Hispanic white31,33526.7 (26.4–27.0)3.9 (2.9–4.2)
  Hispanic11784.1 (3.8–4.4)1.8 (0.1–3.2)
 FemaleNon-Hispanic white22,30216.9 (16.6–17.1)3.3 (2.4–4.2)
  Hispanic16244.1 (3.9–4.3)0.6 (−1.2–2.5)
In situ melanoma incidenceMaleNon-Hispanic white15,22413.1 (12.9–13.3)9.0 (7.5–10.6)
  Hispanic3421.4 (1.3–1.6)4.7 (−0.1–9.7)
 FemaleNon-Hispanic white10,7217.9 (7.8–8.1)9.6 (8.5–10.8)
  Hispanic6101.6 (1.5–1.8)5.7 (2.5–9.0)
Melanoma mortalityMaleNon-Hispanic white64075.6 (5.4–5.7)−0.4 (−1.3–0.5)
  Hispanic2881.1 (1.0–1.3)1.1 (−2.8–5.2)
 FemaleNon-Hispanic white36112.6 (2.5–2.7)−1.7 (−2.5–−0.9)
  Hispanic2330.7 (0.6–0.8)0.3 (−2.6–3.2)
thumbnail image

Figure 1. (A) Male annual age-adjusted incidence rates by race in California (1988-2001) for invasive melanoma of the skin. (B) Female annual age-adjusted incidence rates by race in California (1988-2001) for invasive melanoma of the skin. Trends in age-adjusted incidence (rates per 100,000 population standardized to the year 2000 U.S. population) of invasive melanoma in the Hispanic and non-Hispanic white population of California, 1988–2001, with estimates of linear trends (line segments are results of JoinPoint regression analysis).

Download figure to PowerPoint

The rate of in situ melanomas increased rapidly among non-Hispanic white males and females between 1988 and 2001 (9.0% and 9.6% per year, respectively) (Fig. 2). AAIRs for in situ melanomas among Hispanic males and females increased at rates similar to each other (4.7% and 5.7% per year, respectively), but there was substantial uncertainty in this estimate among Hispanic males (Table 1).

thumbnail image

Figure 2. (A) Male annual age-adjusted incidence rates in California (1988-2001) for melanoma of the skin. (B) Female annual age-adjusted incidence rates in California (1988-2001) for melanoma of the skin. Trends in age-adjusted incidence (rates per 100,000 population standardized to the year 2000 U.S. population) of in situ melanoma and melanoma mortality in the Hispanic and non-Hispanic white population of California, 1988–2001, with estimates of linear trends (line segments are results of JoinPoint regression analysis).

Download figure to PowerPoint

Only among non-Hispanic white females did melanoma mortality decline substantially and statistically significantly between 1988 and 2001 (at 1.7% per year) (Fig. 2). There appeared to be slight declines in melanoma mortality among non-Hispanics white males (0.4% per year) and increases in mortality for Hispanic males and females (1.1% and 0.3% per year, respectively), but none of these findings were statistically significant.

Trends by Tumor Thickness

For both males and females, Hispanics had a larger proportion of thick lesions than thin lesions. Although 53.7% of invasive melanomas among non-Hispanic white males were less than 0.75 mm at diagnosis, only 43.9% were among Hispanic males. Tumors greater than 1.5 mm at diagnosis accounted for 24.4% of tumors among non-Hispanic white males, but 35.4% among Hispanic males.

Increases over time in invasive melanomas occurred approximately uniformly by tumor thickness among non-Hispanic white males (Fig. 3) and females (not shown). However, among Hispanic males the increase in thick tumors (> 1.5 mm) was far more pronounced than increases in thin (< 0.75 mm) or moderate (0.75–1.49 mm) tumors (Fig. 3) and the same was true for Hispanic females (not shown). The incidence of thin tumors (< 0.75 mm) increased between 1988 and 2001 among non-Hispanic males, with an annual increase of 5.4% (95% CI, 4.4, 6.5), and there was a smaller and nonstatistically significant increase in thin tumors among Hispanic males (EAPC = 1.4; 95% CI, −2.1, 4.9). For Hispanic males, thick tumors (> 1.5 mm) had an annual increase of 15.4% (95% CI, 11.5, 19.4), whereas among non-Hispanic white males the annual increase was 11.6% (95% CI, 8.1, 15.2). Thick tumors among Hispanic females increased at 8.9% per year (95% CI, 4.7, 13.3), whereas thin tumors among Hispanic females only increased at 0.7% (95% CI, −2.3, 3.8).

thumbnail image

Figure 3. (A) Non-Hispanic white male age-adjusted incidence rates by thickness in California (1988-2001) for invasive melanoma of the skin. (B) Hispanic male age-adjusted incidence rates by thickness in California (1988-2001) for invasive melanoma of the skin. Trends in age-adjusted incidence of invasive melanoma in the Hispanic and non-Hispanic white population of California, 1988–2001, by tumor thickness.

Download figure to PowerPoint

Trends by Histologic Type

The proportion of melanomas classified as malignant melanoma was similar in non-Hispanic and Hispanic males (47.3% and 49.8%, respectively), but the remainder were distributed differently by race: Hispanic males had a different distribution of nodular and superficial spreading melanomas, and acral lentiginous melanomas compared with non-Hispanic white males (12.0% were NM, 23.6% were SSM, and 5.1% were ALM in Hispanic males, whereas 9.0% were NM, 30.7% were SSM, and 0.6% were ALM in non-Hispanic white males). Among Hispanic males, SSM, NM, and ALM rate increases within each tumor thickness group were similar to those observed for all invasive melanomas combined (not shown).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Rates of invasive melanoma have increased markedly among Hispanic males in California since 1988. In contrast to trends in the non-Hispanic white population, the increase has been confined to thicker tumors rather than thinner ones. Despite the finding that the increase in melanoma among Hispanic females was not statistically significant, the increase in thick tumors (> 1.5 mm) was substantial and statistically significant. These are problematic trends because thicker melanomas have a substantially poorer prognosis than thin melanomas.12–14 Although melanoma mortality appears to be declining in the non-Hispanic white population (almost certainly in females), mortality is rare in the Hispanic population, but is more likely to be either unchanged or increasing. These trends also have important ramifications for melanoma prevention, because primary and secondary melanoma prevention efforts are focused on white (i.e., non-Hispanic) populations.

We are aware of only one other report detailing melanoma rates in a Hispanic population in the U.S.,4 and that study did not investigate time trends. All other time trend analyses of melanoma incidence or mortality data we are aware of either specifically include only non-Hispanic whites, or make no distinction between Hispanic and non-Hispanic whites.

The observed increases in melanoma rates in the Hispanic population are unlikely to be attributable to increased efforts at screening for the disease. If this was the case we would expect to see increases among thinner lesions (and possibly among in situ lesions, both of which are more likely than thick invasive cancers to be found by screening). Whereas there appeared to be substantial increases in in situ cancers in non-Hispanic white males and females, there were none noted in Hispanic males, and only very small increases in in situ cancers among Hispanic females. Among invasive cancers, there was little or no change over time in the thinnest lesions (< 0.75 mm) in the Hispanic population.

The observed increases in melanoma rates in the Hispanic population are unlikely to be due to changing classifications of race/ethnicity (i.e., who is considered “Hispanic” rather than “non-Hispanic white”). The denominators we used for the calculation of rates included populations from the 1990 and 2000 Censuses, which had differing definitions of “Hispanic.” The 2000 Census allowed respondents to check more than one race/ethnicity, and although only 2.4% did so, this undoubtedly increased the estimate of the Hispanic population. Increasing the Hispanic population denominator would result in a decrease in observed cancer rates. Although the definition of Hispanic race/ethnicity for melanoma incident cases and deaths relies on subjective reports in medical records and on the death certificate, there is no reason to believe these would systematically change over time, and only among melanomas (many other cancers are not increasing in the Hispanic population in California5).

Similarly, the observed increases in thick tumors among the Hispanic population are unlikely to be due to changing classification of tumor thickness over time. The rates of ‘unknown’ thickness declined over time, requiring that one or more thickness categories must have experienced an increase in counts as a direct result. However, the rate of decline in ‘unknown’ lesion thickness was similar in Hispanics and non-Hispanic whites (who did not experience increasing incidence only among thicker lesions). Moreover, the actual number of tumors with an ‘unknown’ thickness is insufficient to account for all of the increases observed in thick lesions in Hispanic males.

Histologic type and thickness of tumors are not independent of each other: nodular melanomas are more often thicker, and superficial spreading melanomas are more often thinner. We observed that the overall proportion of both acral lentiginous and nodular melanomas among Hispanics were higher than among non-Hispanic whites. The increases in thick melanomas among Hispanics are unlikely to be attributed to an increase in either acral lentiginous or nodular melanomas over time, because we observed no changes over time in either histologic type within each thickness level.

Although there is a vast literature on prospects for primary prevention of melanoma (through sun avoidance education) and secondary prevention (through various screening methods) in white populations, little is known about prospects for melanoma prevention in Hispanic populations. The few studies that have focused on the Hispanic population suggest that little has been achieved in addressing melanoma prevention behaviors in this population.

For example, Hispanic high school students have a significantly lower prevalence of sunscreen use compared with non-Hispanic whites.15 Two studies have compared perception of melanoma risk between non-Hispanic whites and Hispanics, and both studies showed that Hispanics had a poorer awareness of skin carcinoma risk factors, and a lower perception of their own risk than non-Hispanic whites.16, 17 One study17 was stratified for skin type and showed that among respondents with the same melanoma risk according to their skin color, perception of risk and awareness of melanoma risk factors was still lower among Hispanics than among non-Hispanic whites. Hispanics have a reported similar prevalence of sunburn as non-Hispanics.18 This is a particularly important observation because sunburn prevalence in effect represents an estimate of sun exposure dose-adjusted for skin type, because with sunburn the skin has received sufficient erythmal dose to produce a response.

Data on secondary prevention efforts (i.e., screening) that might help address the observed increases in thick lesions in this population are likewise scant and, where they exist, do not bode well for prevention efforts. In a recent comparison of skin carcinoma prevention techniques, more than twice as many non-Hispanic whites had performed skin self-examination in the past year compared with Hispanics (32% vs. 15%17). Among 384 individuals participating in a work site skin carcinoma screening program, Hispanics were less likely to report that they would seek immediate followup care for suspicious skin lesions identified through screening.16 These two studies are the only ones we are aware of that have addressed screening practices among Hispanics, and both indicate that screening for melanoma among Hispanics is not as prevalent as among non-Hispanic whites, and where it does occur, is likely to be less effective.

We recommend that efforts are undertaken immediately to target both primary and secondary prevention messages to Hispanic communities. This effort should include information on sun avoidance and protection from the sun (i.e., seeking shade and sunscreen use), as well as information on self-screening and recommendations on regular skin checks by a qualified professional—all these efforts should be specifically targeted to members of the Hispanic community and to physicians who treat Hispanic patients. We also recommend that future analyses of melanoma rates include, wherever possible, the distinction between Hispanic and non-Hispanic trends, so that the success of prevention activities can be monitored.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors thank Bradford Scott for literature review assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    U.S. Census Bureau. Census 2000 Summary File 1. Available from http://www.census.gov/
  • 2
    Parkin DM, Whelan SL, Ferlay J, Teppo L, Thomas DB. Cancer incidence in five continents. IARC Scientific Publications No. 155, Volume VIII. Lyon, France: IARC, 2002.
  • 3
    Mack TM, Floderus B. Malignant melanoma risk by nativity, place of residence at diagnosis, and age at migration. Cancer Causes Control. 1991; 2: 401411.
  • 4
    Cress RD, Holly EA. Incidence of cutaneous melanoma among non-Hispanic whites, Hispanics, Asians, and blacks: an analysis of California cancer registry data, 1988–93. Cancer Causes Control. 1997; 8: 246252.
  • 5
    Cockburn M, Deapen D. Cancer incidence and mortality in California: trend by race/ethnicity 1988–2001. Sacramento, CA: Department of Health Services, 2004.
  • 6
    Percy C, Fritz A, Jack A, et al. International Classification of Diseases for Oncology (ICD-O), 3rd ed. Geneva: World Health Organization, 2000.
  • 7
    Balch CM, Soong SJ, Gershenwald JE, et al. Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint Committee on Cancer Melanoma Staging System. J Clin Oncol. 2001; 19: 36223634.
  • 8
    World Health Organization. International Classification of Diseases, Ninth Revision. Geneva: World Health Organization, 1977.
  • 9
    World Health Organization. International Statistical Classification of Diseases and Related Health Problems, Tenth Revision. Geneva: World Health Organization, 1992.
  • 10
    Stewart S, Glaser S, Horn-Ross P, West D. SEER study of methods to classify Hispanic cancer patients (final report: contract N01-CN-05224). Union City, CA: Northern California Cancer Center, 1993.
  • 11
    Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for JoinPoint regression with applications to cancer rates. Stat Med. 2000; 19: 335351.
  • 12
    Levi F, Randimbison L, La Vecchia C, Te VC, Franceschi S. Prognostic factors for cutaneous malignant melanoma in Vaud, Switzerland. Int J Cancer. 1998; 78: 315319.
  • 13
    MacKie RM, Hole D, Hunter JA, et al. Cutaneous malignant melanoma in Scotland: incidence, survival, and mortality, 1979–94. The Scottish Melanoma Group. Br Med J. 1997; 315: 11171121.
  • 14
    MacKie RM, Hole DJ. Incidence and thickness of primary tumours and survival of patients with cutaneous malignant melanoma in relation to socioeconomic status. Br Med J. 1996; 312: 11251128.
  • 15
    Hall HI, Jones SE, Saraiya M. Prevalence and correlates of sunscreen use among US high school students. J School Health 2001; 71: 453457.
  • 16
    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: 105110.
  • 17
    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: 118123.
  • 18
    Saraiya M, Hall HI, Uhler RJ. Sunburn prevalence among adults in the United States, 1999. Am J Prevent Med. 2002; 23: 9197.