According to data published by the Spanish national epidemiology institute (Carlos III Health Institute), the annual incidence of melanoma in Spain is close to the adjusted annual incidence for the European population of 6.14 per 100000 population in men and 7.26 per 100000 population in women.1 In the past 20 years, a substantial increase in the incidence of this tumor has been registered in countries such as Australia, New Zealand, and the United States,2,3 but in Spain we lack the data that would allow us to analyze the true scale of the problem in this country. However, the findings of various regional studies indicate that the incidence in Spain may differ from that of other European countries.4–9 For example, up to 15 new cases per year per 100000 inhabitants have been reported in northern Europe, that is, double the incidence observed in Spain.1,10–12

The increase in incidence rates ranges from 3% to 7%, making melanoma one of the cancers with the fastest growing incidences in recent years. In Europe, approximately 60000 new cases are diagnosed every year.

The data on the characteristics of melanoma in Spain are derived essentially from cancer registries, which collect data on tumors diagnosed in different provinces or autonomous communities. These registries currently cover more than 25% of the Spanish population, and their geographic distribution throughout the country means that they provide valuable information on the variability of the incidence of the tumor in Spain. There are 14 fully operational, population-based general cancer registries in Spain: Albacete, Asturias, Canary Islands, Cantabria, Cuenca, Girona, Granada, Mallorca, Murcia, Navarre, Basque Country, La Rioja, Tarragona, and Zaragoza.4,13,14 However, the data on melanoma are incomplete given that these are general cancer registries. They do not collect information on risk factors, disease characteristics, or the treatments received by patients because their primary objective is epidemiologic surveillance, that is, to record incidence, prevalence, and overall disease-free survival.

The Spanish Cutaneous Melanoma Registry (Registro Nacional de Melanoma Cutáneo [RNMC]) was created in 1997 to record the incidence, clinicopathologic characteristics, and prognosis of melanoma at diagnosis in Spain. The data provided by this registry are useful for epidemiologic surveillance and can provide guidance to the preventive campaigns organized regularly by the Spanish Academy of Dermatology and Venereology (AEDV). It is a simple and flexible registry that is well accepted by the participating clinicians. The number of cases reported annually has remained stable at approximately 900 from 1997 through 2011. The registry has been kept running largely thanks to the efforts of the various coordinators who have participated in the dissemination of the aims and findings of the RMNC over these 14 years. The impact of the findings has led to an extension of the project with the aim of gathering information on patient outcomes and studying the effectiveness of the treatments currently available. This second phase began in June 2012, after it was presented to the members of the AEDV at the national Dermatology Congress in Oviedo, Spain, in 2012.

The aim of this article is to analyze the characteristics of the tumors reported during the period the registry has been operational and the changes in the main prognostic factors.

The RNMC is a national registry run by the AEDV Foundation. The organizational structure consists of a national coordinator, a scientific committee of 5 dermatologists responsible for coordinating the registry (each one in a given region), a scientific coordinator, and external consultants. The operating procedures are laid down in 6 documents: seasonal scheduling, a financial plan, a coordination plan, a standard operating procedures manual, a statistical analysis manual, and a document defining the procedures for reporting and dissemination. All these documents are based on the tenets of good registry practice published by the Agency for Healthcare Research and Quality (AHRQ) and the ethical guidelines for the creation and use of registries for biomedical research published by the Carlos III Health Institute.15,16

The patients included on the registry had to have a first diagnosis (incident cases) or a previous diagnosis (prevalent cases) of melanoma. The result from the first diagnostic biopsy, which confirmed melanoma diagnosis, had to be available. In addition, the case history had to include a minimum basic dataset comprising the date of the first diagnosis of melanoma, date of birth, sex, date of assessment, province where the diagnosis was made, and disease status at diagnosis (localized disease, regional metastasis, or distant metastasis).

Initially, the registry record included the patient identifier, date of diagnosis, tumor-related epidemiologic data, site, histopathology, and tumor size. As the project progressed, other variables were added because they were considered necessary for complete assessment of the patients. No data are available on these additional variables for the cases registered prior to the date on which these changes came into effect. The variables introduced in 2005 were presence of ulceration in the tumor, lactate-dehydrogenase (LDH) level, and classification according to the American Joint Committee on Cancer (AJCC) Staging System.

Given that the RNMC is a population-based registry, no formal sample size calculation was performed.

The analyses of tumor depth and extension only include data from patients with invasive melanoma and excludes cases in which the sex of the patient was unknown.

Descriptive statistics (frequencies and percentages) were calculated for qualitative variables. Comparisons between groups for such variables were performed using the Fisher exact test or the χ2 test, as appropriate.

Means with their corresponding 95% CIs and medians were calculated for quantitative variables. For such variables, group comparisons were performed using the t test or analysis of variance (ANOVA), with Bonferroni or Games-Howell corrections to control for errors due to multiple comparisons. Multiple linear regression analyses were used to assess changes in the age of the patients at the time of diagnosis, tumor thickness (Breslow index [BI]) according to the year of diagnosis of the tumor, age, and sex of the patient. Statistical significance was set at .05. The statistical analysis was performed with the SPSS program, version 14.0 (SPSS Inc).

Of the patients included in the registry, 56.5% were women (n=7584) and 43.5% (n=5833) were men. The sex of the patient was not recorded in 211 cases (1.5%). Table 2 shows the data for patients with invasive melanoma.

The mean age for the whole group was 57 years (95% CI, 56.4-57 years), the median age was 58 years, and the range was from 3 to 100 years. A statistically significant difference between the sexes was observed in the age of the patients at the time of diagnosis, with men being diagnosed on average 2.2 years later (P<.0001). For women, the mean age was 55.8 years (95% CI, 55.4-56.2 years), the median age was 56 years, and the range was 3 to 100 years. For men, the mean age was 57.9 years (95% CI, 57.5-58.4 years), the median age was 60 years, and the range was 4 to 96 years.

The distribution by ages was as follows: 1.3% (n=175) less than 20 years; 19.5% (n = 2585) between 21 and 40 years; 34.7% (n=4601) between 41 and 60 years; 36.3% (n=4811) between 61 and 80 years, and 8.2% (n=1086) over 80 years. The proportion of girls and women was significantly higher in the group aged under 20 years and the group aged between 21 and 40 years (P<.0001). In the group aged 61 to 80 years, the proportion of men significantly exceeded that of women (P<.0001). There were no significant differences in the proportions of men and women in the groups aged 41 to 60 years and those aged over 80 years. Table 2 shows the age distribution for the group of patients with invasive melanoma.

In the group of patients as a whole, the most frequent clinicopathologic variant was superficial spreading melanoma (SSM) (n=7481; 62.6%), followed by nodular melanoma (NM) (n=2014; 16.8%), and lentigo maligna melanoma (LMM) (n=1705; 14.3% [1705 cases]). Acral lentiginous melanoma (ALM) and mucosal melanoma accounted for 5.8% (n=698) and 0.5% (n=61), respectively. Significant differences between the sexes were observed in the clinicopathologic type of tumor, with a greater proportion of SSM in women (n=4351, 64.2% vs n=3130, 60.4% in men) and NM in men (n=1010, 19.5% vs n=1004, 14.8% in women) (P<.0001). Table 2 shows the data corresponding to the patients with invasive melanoma. NMs occur more frequently in men from 21 years of age onwards (15.1% vs 13.3%) but the difference only becomes significant from 41 years onwards, with a frequency of 21% in men and 16.9% in women (P=.021) in patients aged 41 to 60 years, 25% in men and 18.7% in women (P<.0001) in patients aged 61 to 80 years, and 36.6% in men and 27.2% in women (P=.055) in patients aged over 80 years. In those under 20 years, the proportion of NM was identical in both sexes (23%).

Depth of Invasion and Tumor Thickness

A total of 2044 patients (16%) had in situ melanoma. The proportions of patients with in situ melanoma by 5-year periods in the registry were as follows: 1997-2001, 16.1%; 2002-2006, 16.1%; and 2007-2011, 19.2%. The proportion was significantly greater in the most recent 5-year period with respect to the previous 2 periods (P<.0001).

The following analyses were only performed for invasive melanomas. The mean BI was 1.97mm (95% CI, 1.91-2.03mm), with a median of 1mm. Statistically significant differences in BI were observed between men and women (P<.0001) with a mean BI 0.387mm (95% CI, 0.264-0.511mm) greater in men. The mean BI was 2.19 mm (95% CI, 2.09-2.28mm) in men and 1.80mm (95% CI, 1.72-1.88mm) in women.

In the multivariate analysis, BI was studied as a function of patient age and sex. Regardless of age, BI was greater in men than in women (P<.0001) by, on average, 0.346mm (95% CI, 0.223-0.469mm).

Regardless of sex, the BI increased significantly with increasing age (P<.0001), such that for each increase in age of a year, the BI on diagnosis increased 0.025mm (95% CI, 0.022-0.029mm), that is, 0.25mm every 10 years. Figure 2 depicts this trend.

Figure 2.

Depth of invasion (Breslow index in mm) at diagnosis of invasive melanoma, by patient age and sex.

(0,16MB).

Patients were classified according to T stage (tumor size) as per the AJCC guidelines for melanoma.17 Significant differences were observed in the proportion of patients of each sex who belonged to each BI category (Table 2), with a BI less than 1mm being more common in women than men (P<.0001) and a BI between 2.01 and 4mm or greater than 4mm being more common in men (P<.0001). Figure 3 shows the proportion of patients in each risk group by age.

Figure 3.

Proportion of patients in each Breslow group, by age group, for invasive melanoma.

(0,11MB).

The most frequently observed Clark level was Clark level III (n=3719; 37.4%). Statistically significant differences between men and women were observed in the proportion of patients in all categories (P<.0001), with Clark level II being more frequent in women and levels IV and V being more frequent in men (Table 2).

It is important to highlight that 342 clinicians have participated in reporting cases to the RNMC since it became operational in 1997. There was little change in the notification rate, reflecting the efforts of all participants. Although the number of cases recorded in some autonomous communities was below that expected (Table 1), on a nationwide level, the rate of notification was high for a specific tumor registry. However, the ideal situation would be to make melanoma a disease of mandatory notification, as is the case in the United States since 2010 (http://wwwn.cdc.gov/nndss/document/NNDSS_history_spreadsheet_2013_for_web_v1.pdf) and Australia (http://www.health.nsw.gov.au/phact/Documents/is6-disease-notification.pdf). A minimum dataset can be filled out quickly and would contribute to the generation of very important data on disease control (www.registromelanoma.org). The highest number of cases were notified in 1997 when the project first started because of the participation of pathologists, although their interest waned over time. It is therefore very important to involve all clinicians responsible for the care of patients with melanoma in the notification and follow-up of patients to achieve the important goals of the project.

The frequency of tumors was 13% higher in women than men during the study period (Table 1). The greater number of tumors observed in women is in line with other international and Spanish registries.1–9

The mean age on initial diagnosis of the tumor was 57 years, a figure similar to that reported in other Spanish studies.4,7,8 A statistically significant increase was seen in the age of the patient at diagnosis by year of diagnosis (P<.0001), in both men and women (Fig. 4). Some authors have reported a similar observation,18,19 but others have not reported any statistically significant differences.20 The progressive increase in age could be explained by the failure of prevention campaigns to reach older individuals, especially men, so that these patients first consult with advanced primary tumors. Such an observation has in fact been reported in studies of melanoma prevention.21

The most frequent site was the trunk, followed by the legs, in line with Spanish and international studies. The legs were more frequently affected in women and the trunk was more frequently affected in men (Table 2).7,8,22 The cause of these differences should be investigated. They could conceivably be due to differences in exposure to sunlight between men and women.

The clinicopathologic type most frequently reported was SSM, in both men and women, followed by NM. These findings are consistent with the data published in Spanish and international studies.7,8,22

Assessment of tumor thickness is very important given that this variable has been shown to affect prognosis.23–29 The results from the RNMC between 1997 and 2011 show a nonsignificant trend towards increasing melanoma thickness (BI) with year of diagnosis (Figure 5). The progressive increase observed can be considered an indication of how early the tumors are being diagnosed. Similar findings have been reported in other Spanish studies.18 Analysis of tumor size distribution by 5-year period shows that the proportion of cases with tumors having a thickness greater than 4mm has not increased over the 14-year study period (Fig. 6). However, the proportion of thin tumors has increased over time, as has the proportion of in situ melanomas. These findings coincide with those observed in other geographic areas.30,31

Women were found more often in the group with a BI less than 1mm than men (P<.0001). In men, tumors of greater thickness are more frequent, that is, those with worse prognosis (Table 2). The BI is greater in men than women (P<.0001), regardless of age (Fig. 2), indicating that the tumors are more aggressive on diagnosis in men. In addition, NMs were more frequent in men aged over 41 years. This clinicopathologic type of melanoma has been associated with higher mortality and a more aggressive clinical behavior than other types of melanoma given its faster rate of growth than other tumors.32,33

In addition to targeting the general population, prevention campaigns should focus particularly on men over 50 years, a population in whom thicker tumors are observed. This finding may possibly be explained by a longer delay in diagnosis given that a correlation was observed in men between older age and thicker tumors on diagnosis.

In conclusion, we would like to highlight the value of a specific population-based registry of patients with melanoma and the importance of keeping the registry running. The AEDV Foundation, convinced of this need, has decided to broaden the objectives of the project to enable prospective follow-up of patients and the study of outcomes according to treatments received and known prognostic factors. A webpage has been created (www.registromelanoma.org) to facilitate access to the registry data for the participating specialists, and participation in the project has been opened to all specialists who treat melanoma patients.

The authors declare that they have followed their hospital's protocol on the publication of data concerning patients and that all patients included in the study have received sufficient information and have given their written informed consent to participate in the study.

The authors declare that no private patient data are disclosed in this article.

The study was sponsored by the Foundation of the Spanish Academy for Dermatology and Venereology (AEDV), and received funding from Schering-Plough SA and Bristol Myers Squibb SA. E-C-BIO SL were responsible for designing the study, managing consent, quality control, and the statistical analysis under contract from the AEDV Foundation.

The authors declare that they have no conflicts of interest.