Cutaneous melanoma is a heterogeneous entity in numerous respects, including location. Cutaneous head and neck melanoma (CHNM) accounts for between 12% and 30% of all melanomas and has a series of epidemiological, clinical, histologic, and prognostic features that distinguish it from melanomas in other parts of the body.1,2 It has, however, been studied less than other melanomas, both in Spain and internationally. Studies of this type can improve our understanding of how CHNM develops and who it affects, thus contributing to improved prevention and/or early diagnostic strategies.

The aim of this study was to describe the main clinical and histologic characteristics of CHNM, analyze the probability of metastasis and melanoma-specific death, and investigate changes over a period of 21 years in a series of Spanish patients.

We performed a retrospective study of all patients with a histologically confirmed diagnosis of CHNM at Hospital General Universitario Gregorio Marañón and associated healthcare centers in Madrid, Spain between January 1, 1995 and December 31, 2015. Both in situ (Tis) and invasive melanomas were included. We excluded mucosal and soft-tissue melanomas, melanoma of unknown primary, and melanomas in patients with missing information on tumor thickness or follow-up.

Clinical and histologic data were obtained from the patients’ medical records and pathology reports, anonymized, and entered into a database for analysis. The study period was divided into 3 periods of equal length—1995-2001, 2002-2008 and 2009-2015—to facilitate the segmented analysis of patients over time. The clinical characteristics were sex, tumor location (scalp, face, ear, and neck), and age, recorded as a continuous variable and categorized by groups (≤60, 61-74, and ≥75 years). Histologic data included histologic subtype (lentigo maligna, superficial spreading, nodular, other), presence of histologic ulceration, tumor thickness (recorded as a continuous and an ordinal variable [T1-T4]), presence of an invasive component, mitotic rate (mitoses/mm2), development of metastasis, and survival status. Information on mitotic rate was available for over 85% of patients with invasive melanoma. Mitotic figures were counted using the hot spot method.3 We created 3 categories—0, 1-5, and ≥6 mitoses/mm2—as other authors have reported that a rate of over 6 mitoses/mm2 has prognostic implications and may be associated with sentinel lymph node positivity.2,4,5 Biopsy samples for patients with missing data on mitotic rate in the pathology report were re-examined by 2 pathologists who were experts in melanocytic lesions.

Qualitative variables were analyzed using the χ2 or Fisher exact test. Benjamini-Hochberg correction was applied to adjust for multiple comparisons. Means of qualitative and quantitative variables were compared using the t test or analysis of variance, as appropriate. Statistical significance was set at a P value of less than .05 in all cases. The data were analyzed in SPSS version 25.

The head and neck area comprises just 9% of the body’s surface area but accounts for between 12% and 30% of all cutaneous melanomas.6–10 While cutaneous melanoma has an evident predilection for this area, we found no studies that have exclusively analyzed CHNM in Spain. We have therefore compared our findings to those of international studies and Spanish studies with subgroups of patients with CHNM.

According to studies of European-Caucasian populations, including Spanish series, patients with melanoma are more likely to be female.11–13 Some authors, however, have found that patients with CHNM are more likely to be male.14 Findings vary, however, as different studies from recent decades have described a predominance of both female6,15 and male patients.16,17 In our series, 53.9% of the patients with CHNM were women. This figure is very similar to that of 60% reported by Arranz-Sánchez et al.12 in a study of patients from Madrid diagnosed with cutaneous melanoma between 1990 and 2004.

Patients with CHNM are on average older than those with other types of melanoma, irrespective of histologic subtype.2,18 Mean age at diagnosis in CHNM, for example, ranges from 64 to 71 years compared with 50 to 58 years for other cutaneous melanomas.2,7,14 The patients in our series had a mean age of 71.3 years and almost half of them were 75 years or older. These data are consistent with findings from a population-based study carried out at an oncology center in Valencia, where patients older than 65 years accounted for more than half of all patients with CHNM.19 In our series, the proportion of patients with CHNM aged 75 years or older increased significantly over the study period. In 1995-2001, for example, they accounted for just 32.1% of all patients, but by 2009-2015, this figure had risen to 56.8%. This increase is consistent with the general increase observed for mean age at melanoma diagnosis in other countries.20 In the case of CHNM, it is probably due to the high proportion of lentigo maligna melanomas and the cumulative effect of chronic sun exposure.

Most of the CHNMs in our series were located on the face, coinciding with reports from other series that have described rates ranging from 53.2% to 78.5%.1,6,7,15 This is not surprising, as the face is a relatively large area exposed to high doses of UV radiation. Although the differences between the time periods were not statistically significant, we observed a progressive decrease in neck melanoma and a progressive increase in scalp melanoma.

Lentigo maligna accounted for 2 of every 3 CHNMs. The next most common histologic subtype was superficial spreading melanoma, followed by nodular melanoma. Lentigo maligna has classically been associated with the head and neck region, and most studies have described it as being the predominant subtype in this area.1,6,7,18,21 Nonetheless, some studies have reported a predominance of superficial spreading melanoma,16,17 which in our study ranked second but accounted for a significant proportion of cases (18.9%).

Over one-third of the CHNMs in our series (36.1%) were melanoma in situ. Other studies have reported varying rates ranging from 32% to 49.5%.2,7,22 Overall, these rates are higher than those reported for melanoma in situ located at other sites. According to the Spanish National Cutaneous Melanoma Registry, melanoma in situ accounts for approximately 16% of all melanomas.9 The proportion of melanomas in situ in our series almost doubled between 1995-2001 and 2009-2015, reflecting a progressive increase over the past 2 decades. A similar trend has been observed for CHNMs and melanomas in other locations12,13,23 and can probably be explained by earlier diagnosis thanks to the use of dermoscopy and follow-up protocols for at-risk patients.

Mean tumor thickness was 1.6mm, and decreased significantly over the study period (from 2.1mm in 1995-2001 to just 0.91mm in 2009-2015). This reflects a global trend towards increasingly thin melanomas in recent decades,12,23 although unlike us, most studies to date have not focused on CHNMs. A study of 98498 patients from several Surveillance, Epidemiology, and End Results (SEER) registries in the United States found that mean melanoma thickness decreased from 0.77mm to 0.65mm over a period of 20 years.24 T3 and T4 melanomas accounted for a considerable proportion (25.7%) of the CHNMs in our sample. This heterogeneity in tumor thickness—a high frequency of melanomas in situ and thick melanomas—has been described by other authors,7 and would appear to indicate the coexistence of numerous lentigo maligna melanomas and other fast-growing melanomas, such as nodular melanoma.

Mitotic rate is a quantifiable marker of cell proliferation in melanoma. It has been closely linked to survival and is considered by many authors to be an independent prognostic factor.3,25,26 Few studies, however, have analyzed this histologic marker in CHNM.27–29 Shen et al.28 found CHNMs in general to be associated with a high mitotic count,28 while others have found this to be the case for melanomas of the scalp.30 In our sample, 63.6% of patients had a mitotic rate of 1 or more mitoses/mm2. In a large series of patients with CHNM, Xie et al.27 reported that 65% of patients had at least 1 mitosis/mm2, which is very similar to our rate. The authors also found that 30.4% of patients had 5 mitoses/mm2 or more, again, similar to the rate observed for the highest mitotic activity category in our series (≥6 mitoses/mm2). These rates are higher than those described for cutaneous melanomas in other locations,31 although there was a decrease in the frequency of melanomas with high mitotic activity over the study period.

Finally, it is noteworthy that the prognosis of patients with CHNM improved over the 2 decades analyzed, with a significant reduction in metastasis and disease-specific deaths. The reduction in melanoma mortality is particularly striking. In 1995-2001, 19% of the patients in our series died of melanoma, but in 2009-2015, this figure was just 2.4%. This improvement in survival could be linked to the reductions observed in mean tumor thickness, mitotic activity, and frequency of ulceration.

We have presented a large series of patients diagnosed with CHNM in a tertiary hospital in Madrid.

The clinical and histologic characteristics of our sample largely coincide with those described in the literature and include a predominant facial location and a high proportion of lentigo maligna melanomas. We also observed that a significant percentage of invasive melanomas had high mitotic activity (≥6 mitoses/mm2).

Overall, the average age of patients with CHNM is rising, and considering that older persons are a growing demographic group in our society, more studies focusing on melanoma in this area are needed.

Our findings show that CHNM thickness is decreasing and that the risk of metastasis is lower than it was in the early years of our study.

The authors declare that they have no conflicts of interest.