JEADV ISSN 1468-3083
ORIGINAL ARTICLE Blackwell Publishing Ltd
Comparison between familial and sporadic cutaneous melanoma in Valencia, Spain E Nagore,*† R Botella-Estrada,† Z Garcia-Casado,‡ C Requena,† C Serra-Guillen,† B Llombart,† O Sanmartin,† C Guillen† † Dermatology, Instituto Valenciano de Oncologia, Valencia, Spain ‡ Molecular Biology, Instituto Valenciano de Oncologia, Valencia, Spain
Keywords clinical, familial, melanoma, risk factors *Correspoding author, tel. +3496 111 4016; fax +3496 1114346; E-mail:
[email protected] Received: 28 September 2007, accepted 11 December 2007 DOI: 10.1111/j.1468-3083.2008.02682.x
Abstract Some clinical, pathological and genetic features have been associated to familial melanoma, particularly multiple melanoma and earlier age at diagnosis. To compare the clinical, epidemiological and pathological differences between familial and sporadic melanoma patients in Valencia, Spain, a series of 959 patients with cutaneous melanoma were selected at a single institution. For this study the following variables were selected: age, sex, melanoma site and presence of solar lentigines on the melanoma surrounding skin, histological subtype, tumor thickness, stage, family and personal history of cutaneous melanoma and of other neoplasias, personal history of non-melanoma skin cancer, past personal history of severe sunburns, cutaneous phenotype (phototype, hair and eyes colors number of common nevus, number of atypical nevi, presence of solar lentigines). Forty-one (4.28%) familial and 918 sporadic melanoma were identified. Among the multiple variables studied, a younger age at diagnosis (median age of 42 vs 53 years), higher frequency of the presence of at least one clinically atypical nevus (36.1% vs 17.7%), multiple melanomas (12.2% vs 3.4%) and red/blonde hair (33.3% vs 18.9%), and a lower rate of cases with solar lentigines in melanoma site (33.3% vs 56.3%) were found for familial cases. Except for hair color and age, the other variables remained statistically significant after the multivariate study. Interestingly, no acral melanomas were found among the familial cases. In summary, phenotypic risk factors for familial melanoma are a tendency to develop multiple melanomas, to have clinically atypical nevi and to present less actinic damage at the melanoma site. All these results enhance the relevancy of genetic susceptibility associated to the ability to produce atypical nevi and partly to a higher sensitivity to the sun.
Population studies suggest that approximately 6% to 14% of melanoma cases occur in a familial setting.1 Similarly, a family history of melanoma represent an important risk factor for melanoma.2,3 Previous studies suggested that some characteristics may be associated with the familial occurrence,1,4–10 although only a few of them have evaluated clinical phenotypic variations.1,4,8,10 Besides, all these studies have been performed on populations from Northern European ancestry. Most of these studies
have reported younger ages at diagnosis of melanoma,6,7 relatively thinner melanomas,4,6,11 and greater frequencies of multiple primary melanomas in individuals with familial melanoma vs. persons with sporadic melanomas.6,8,10,11 Few studies have been performed about what phenotypic features are associated to familial melanoma. In a study carried out in France, red hair colour and the presence of atypical moles were associated to familial melanoma.10
© 2008 The Authors JEADV 2008, 22, 931– 936 Journal compilation © 2008 European Academy of Dermatology and Venereology
931
Clinical characterization of familial melanoma
Nagore et al.
It seems interesting to know clinical features that are associated to the family aggregation in the context of a better knowledge about biologic and genetic characteristics of melanoma.12–14 These clinical features might be helpful to search new genes of predisposition to melanoma development. To better understand the clinical profile of familial melanoma cases, we have undertaken an epidemiological study from a series of patients followed up at Instituto Valenciano de Oncolog’a, Valencia, Spain. The goal of this article is to compare the clinical, histologic and epidemiological characteristics of 41 patients with a familial cutaneous melanoma with 918 patients with sporadic cutaneous melanoma.
Patients and methods The study was approved by the institutional review board. All cutaneous melanoma patients who were examined in the Department of Dermatology of the Instituto Valenciano de Dermatologia between January 1, 2000 and February 28, 2007 were eligible for the study. Criteria for enrolment included both newly diagnosed cutaneous melanoma and followed up subjects. Data were obtained from the database of the Department, which includes both clinical and pathological data from each patient. All the data were prospectively obtained from interview and physical examination by a senior dermatologist specifically trained for melanoma care at the visit and immediately recorded in the database. Melanoma diagnosis was confirmed for each patient or relative by pathologic material and/or pathology reports. The following variables were selected: 1 Epidemiological features included age at diagnosis, sex, melanoma site (head/neck, upper limbs, trunk, lower limbs, and hand/foot), histological subtype (lentigo malignant melanoma, superficial spreading melanoma, nodular melanoma, acral lentiginous melanoma, other/ no specified), tumour thickness and stage. In addition, personal and family history of melanoma, personal history of non-melanoma skin cancer, personal history of other non-cutaneous malignancies, family history of other neoplasias (only first-degree relatives, retrieved by interview) and lifetime history of severe sunburns (defined by the presence of blisters or pain for at least two days and coded as presence or absence) were included. 2 Phenotype characteristics included phototype, eye colour (fair vs. dark), hair colour (red/blonde vs. brown/ black), a count of total body common naevi (coded as: < 50, 50–100, > 100), presence of at least one clinically atypical naevus, presence of solar lentigines and presence of solar lentigines at melanoma site. We did 932
not consider the past history of freckles, and thus, we included as solar lentigines only those typical persistent macular areas of brown pigmentation with a striking irregular border that are commonly seen either after acute or chronic sun exposure and thus were included as a clinical marker of a sun-damaged skin. We considered International Agency for Research on Cancer (IARC) definitions for common and atypical naevi.15 The IARC protocol for identifying and recording naevi in epidemiological studies uses the following criteria to identify atypical naevi: there must be a macular component in at least one area; in addition, at least three of the following features must be present: (i) border not well defined, (ii) size 5 mm or more, (iii) colour variegated, (iv) contour uneven, (v) presence of erythema. A patient was considered to belong to the familial melanoma category when at least one first-degree relative had pathologically demonstrated cutaneous melanoma. Only the proband, defined as the patient who had been attended in the first place to our Department, was included in the analysis. The distribution of the clinical, histologic and epidemiological categorical variables was compared by the usual chi-squared test. Fisher’s exact test was used when categories contained numbers of five or less. Analysis of variance test was performed for continuous variables. Univariate and multivariate logistic regression analyses were performed to obtain an estimate of risk of being a familial melanoma. P < 0.05 was considered significant.
Results Of 959 patients who entered the study, 41 (4.3%) had a positive family history of cutaneous melanoma. Demographic data are given in Table 1. The distribution of sexes was similar in both groups. The patients in the familial melanoma group were significantly younger than those in the sporadic group (median: 42 vs. 53). Thus, 56% of the patients in the familial group vs. 36.7% in the sporadic group were younger than or equal to 45 years [odds ratio (OR), 2.20; 95% confidence interval (95% CI), 1.17–4.13]. No difference was observed between the groups for the location of the tumour, clinical stage at diagnosis, histologic subtypes and Breslow thickness. However, it was remarkable that no melanoma was found on acral sites in the familial melanoma group and, accordingly, no acral lentiginous melanoma was present in this group. We observed a similar distribution in the familial and sporadic groups for the familial history of other malignancies, although a trend to present relatively more cases in the sporadic group was found. No difference
© 2008 The Authors JEADV 2008, 22, 931– 936 Journal compilation © 2008 European Academy of Dermatology and Venereology
Nagore et al.
Clinical characterization of familial melanoma
Table 1 Demographic features distribution and melanoma characteristics of 959 study participants
Features Age Mean (SD) Median (range) ≤ 45 years > 45 years Sex Men Women Melanoma site Head/neck Upper limbs Trunk Lower limbs Hands/feet Histological subtype LMM SSM NM ALM Other/non-specified Breslow thickness (m.v. = 136) Mean (SD) Median (range) Stage (m.v. = 6) In situ Localized Locorregional Distant Multiple melanomas (m.v. = 0) No Yes Personal history of NMSC (m.v. = 377) No Yes Second malignancies (m.v. = 114) No Yes Family history of cancer† (m.v. = 141) No Yes Past personal history of severe sunburns (m.v. = 213) No Yes
Familial melanoma (N = 41) n (%)
Non-familial melanoma (N = 918) n (%)
45.2 (17.0) 42 (21–87) 23 (56.1) 18 (43.9)
52.1 (16.5) 53 (6–96) 337 (36.7) 581 (63.3)
0.008
15 (36.6) 26 (63.4)
410 (44.7) 508 (55.3)
0.308
8 (19.5) 6 (14.6) 14 (34.1) 13 (31.7) 0 (0)
162 (17.6) 128 (13.9) 346 (37.7) 203 (22.1) 79 (8.6)
0.253
2 (4.9) 31 (75.6) 6 (14.6) 0 (0) 2 (4.9)
67 (7.3) 575 (62.6) 174 (18.9) 42 (4.6) 61 (6.6)
0.431
2.14 (2.73) 1.25 (0.10–25.00) 93 (10.2)
0.305
1.65 (1.36) 1.25 (0.10–5.00) 6 (14.6)
P*
0.012
0.749
31 (75.6) 4 (9.8) 0 (0)
707 (77.5) 104 (11.4) 8 (0.9)
36 (87.8) 5 (12.2)
887 (96.6) 31 (3.4)
0.016
17 (81.0) 4 (19.0)
506 (90.0) 53 (10.0)
0.258
34 (91.9) 3 (8.1)
839 (92.6) 67 (7.4)
0.871
29 (74.4) 10 (25.6)
544 (62.0) 333 (38.0)
17 (74.2) 19 (52.8)
374 (46.3) 434 (53.7)
0.120 0.912
SD, standard deviation; m.v., missing values *From Chi-squared or analysis of variance tests when appropriate. †Only first-degree relatives (melanoma and non-melanoma skin cancer excluded).
was observed between both group cases for the occurrence of non-melanoma skin cancer (a slightly but not significantly higher percentage of cases in the familial group was observed: 4 of 21 [19%] vs. 53 of 506 [10%]) and non-cutaneous malignancies. A higher proportion of patients in the familial melanoma group had multiple primary melanomas
(5 of 41, 12.2%) than in the sporadic group (31 of 918, 3.4%). This difference was significant (OR, 3.97; 95% CI, 1.46–10.82). No difference was observed for the presence of past history of severe sunburns. Details about phenotype differences are given in Table 2. A significantly higher proportion of patients in the familial
© 2008 The Authors JEADV 2008, 22, 931– 936 Journal compilation © 2008 European Academy of Dermatology and Venereology
933
Clinical characterization of familial melanoma
Nagore et al.
Table 2 Phenotype features distributions Familial melanoma (N = 41) Non-familial melanoma (N = 918) n (%) n (%) P
Features Phototype (m.v. = 83) I–II III–V Eye colour (m.v. = 133) Fair Dark Hair colour (m.v. = 162) Blond/red Brown/black Common melanocytic naevi count (m.v. = 160) < 50 51–100 > 100 Presence of at least one clinically atypical melanocytic naevus (m.v. = 132) No Yes Presence of solar lentigines (m.v. = 273) No Yes Presence of solar lentigines at melanoma site (m.v. = 273) No Yes
16 (43.2) 21 (56.8)
316 (37.8) 520 (62.2)
0.504
16 (44.4) 20 (55.6)
292 (37.0) 498 (63.0)
12 (33.3) 24 (66.7)
144 (18.9) 617 (81.1)
0.033
26 (74.3) 5 (14.3) 4 (11.4)
616 (80.6) 96 (12.6) 52 (6.8)
0.531
23 (63.9) 13 (36.1)
651 (82.3) 140 (17.7)
0.005
8 (22.2) 28 (77.8)
109 (14.6) 639 (85.4)
0.208
22 (66.7) 11 (33.3)
333 (46.4) 384 (53.6)
0.023
0.364
Table 3 Univariate and multivariate regression analysis Univariate*
Multivariate†
Variables
OR
95% CI
OR
95% CI
Multiple melanomas Absence of solar lentigines at melanoma site Presence of at least one clinically atypical naevus Age at diagnosis ≤ 45 years Fair hair colour
3.97 2.31 2.64 2.20 2.14
1.46–10.82 1.10–4.83 1.30–5.33 1.17–4.13 1.05–4.39
5.62 3.17 2.26 – –
1.87–16.86 1.42–7.05 1.04–4.93 – –
*Only statistically significantly variables by contingency tables are included. †Forward stepwise multivariate logistic regression. OR are only referred for variables finally included in the model.
group had blond/red hair (33.3% vs. 18.9% in the sporadic group; OR, 2.14; 95% CI, 1.05– 4.39), had at least one clinically atypical naevus (36.1% vs. 17.7% in the sporadic group; OR, 2.64; 95% CI, 1.30–5.33) and did not have solar lentigines at melanoma site (66.7% vs. 46.4% in the sporadic group; OR, 2.31; 95% CI, 1.10 – 4.83). In addition, a lower proportion of solar lentigines and a higher proportion of phototypes I/II, fair eye colour and of high number of common melanocytic naevi was found in the familial group, but these differences did not reach significance. Multivariate regression analysis (Table 3) showed that the most representative variables associated to familial 934
melanoma were the presence of multiple melanomas (OR, 5.62; 95% CI, 1.87–16.86), the absence of solar lentigines at melanoma site (OR, 3.17; 95% CI, 1.42–7.05) and the presence of at least one clinically atypical naevus (OR, 2.26; 95% CI, 1.04–4.93).
Discussion The frequency of familial cutaneous melanoma varies according to the geographical area. Differences are due to the relationship between the phenotypic characteristics of each population and the climate as well
© 2008 The Authors JEADV 2008, 22, 931– 936 Journal compilation © 2008 European Academy of Dermatology and Venereology
Nagore et al.
as to the source of data about family history. 16,17 In a hospital-based population, we found a frequency of 4.3%. This figure does not substantially differ of those that are shown in Caucasian populations in the literature, except for white population in Australia or New Zealand.4,6–8,10,17–19 In the present study, we have studied in detail the characteristics of a series of patients with familial melanoma in a Mediterranean area. The characterization of familial cases has been shown to result in early detection of new melanomas by elaborating specific education and clinical surveillance plans for all the members.11,20 Among the multiple factors included in the statistical analysis, few showed significant differences between sporadic and familial cases. In univariate analysis, patients in the familial group had a younger age at diagnosis, more cases with multiple melanomas, fair (red/blond) hair or at least one clinically atypical naevus and less cases with signs of sun damage at the melanoma site, as evaluated by the presence of solar lentigines. When these variables were included in a multivariate analysis, only the presence of multiple melanomas and clinically atypical naevus and the absence of solar lentigines at the melanoma site maintained their significance in the model. Although it did not appear as a relevant factor in the statistical analysis, it was noticeable that no familial melanoma presented a melanoma on hand or foot locations. In addition, familial cases showed a tendency to present with lower tumour thickness and still manifested at earlier clinical stages at diagnosis. Concerning statistically significant risk factors, a younger age at diagnosis for familial cases has previously been reported in the literature.4,5,7,8,10,11 In our study, 56.1% of patients in the familial group vs. 36.7% in the sporadic group were younger than or equal to 45 years. This difference lost its significance after adjusting for hair colour, presence of clinically atypical naevi, presence of multiple melanomas and absence of lentigines at melanoma size, possibly because of an insufficient number of patients. It is believed to be due to an earlier occurrence because of a genetic predisposition, although a more rapid detection could also influence.10 Consequently, we found thinner tumours in familial cases albeit below significance. However, a study found that patients with familial melanomas still manifested at earlier ages at diagnosis even when only the first cases in each family were considered.7,11 The higher frequency of multiple melanomas in individuals with a familial form of cutaneous melanoma is an unquestionable fact, broadly shown in the literature and intimately related with the presence of mutations in CDKN2A gene.21,22 In our study, the presence of multiple primary melanomas was the strongest predictive factor of
Clinical characterization of familial melanoma
being a familial melanoma case, even after multivariate analysis. On the other hand, patients in the familial group had at least one clinically atypical naevus in a significantly higher proportion, as it has previously been reported10 particularly in families with the atypical naevus syndrome.19,23 This is particularly remarkable because we found a similar distribution for past personal history of severe sunburns between familial and sporadic melanoma patients, an environmental factor that has a strong association with the number of clinically atypical naevi.24 This finding supports that at least a part of the familial aggregation relies on a trend to develop atypical naevi. The absence of solar lentigines at melanoma site was the second most strongly associated factor with a familial melanoma after multivariate analysis. This finding could indicate that genetic alterations favour a smaller requirement of sun damage to produce melanoma. Polymorphisms determining red hair colour in the MC1R gene, which have been shown to increase the risk for cutaneous melanoma, could have a role here. A previous study showed that red hair was the most strongly associated factor with familial melanoma.10 In our series, we gathered blond and red colour in the same category for statistical purposes because of sample size, and we also found that fair hair colour was a risk factor for being a familial melanoma, although it lost its significance after multivariate analysis. Taken together, our findings concur with the hypothesis of at least two pathways to cutaneous melanoma pathogenesis: one associated with a trend to melanocyte proliferation (‘naevi-prone’ people) and the other with a personal sun sensitivity.12,13 In addition, all these features suggest that familial melanoma patients are in fact a very heterogeneous population most likely from both genotypic and phenotypic point of view. In addition, one more feature should be stressed. It can be observed that an important percentage of family cases did not have the most important risk factors for familial melanoma (87.8% of patients did not have multiple primary melanomas, 33.3% had solar lentigines at melanoma site and 63.9% did not present any clinically atypical naevus). Furthermore, 12% of familial cases did not have any of those risk factors (data not shown). Thus, possible explanations for these cases require additional studies, although coincidental familial clustering, low penetrance susceptibility genes (not related to nevogenic ability) and phenotype definition problem are some of the most plausible explanations. In summary, our study shows that familial occurrence of cutaneous melanoma was significantly associated with the development of multiple melanomas, the presence of clinically atypical naevi and the absence of clinically detected sun damage at melanoma site. The presence of
© 2008 The Authors JEADV 2008, 22, 931– 936 Journal compilation © 2008 European Academy of Dermatology and Venereology
935
Clinical characterization of familial melanoma
Nagore et al.
these three criteria in a patient with cutaneous melanoma could be of help to identify familial cases and warrants to collect thorough pedigree information, although it remains important to look for additional familial cases for every melanoma patient.
References 1 Ang CG, Kelly JW, Fritschi L, Dowling JP. Characteristics of familial and non-familial melanoma in Australia. Melanoma Res 1998; 8: 459–464. 2 Hemminki K, Lonnstedt I, Vaittinen P. A population-based study of familial cutaneous melanoma. Melanoma Res 2001; 11: 133–140. 3 Gandini S, Sera F, Cattaruzza MS et al. Meta-analysis of risk factors for cutaneous melanoma. III. Family history, actinic damage and phenotypic factors. Eur J Cancer 2005; 41: 2040–2059. 4 Barnhill RL, Roush GC, Titus-Ernstoff L, Ernstoff MS, Duray PH, Kirkwood JM. Comparison of nonfamilial and familial melanoma. Dermatology 1992; 184: 2–7. 5 Lucchina LC, Barnhill RL, Duke DM, Sober AJ. Familial cutaneous melanoma. Melanoma Res 1995; 5: 413–418. 6 Kopf AW, Hellman LJ, Rogers GS et al. Familial malignant melanoma. JAMA 1986; 256: 1915–1919. 7 Anderson DE. Clinical characteristics of the genetic variety of cutaneous melanoma in man. Cancer 1971; 28: 721–725. 8 Wallace DC, Beardmore GL, Exton LA. Familial malignant melanoma. Ann Surg 1973; 177: 15–20. 9 Florell SR, Boucher KM, Garibotti G et al. Population-based analysis of prognostic factors and survival in familial melanoma. J Clin Oncol 2005; 23: 7168–7177. 10 Grange F, Chompret A, Guilloud-Bataille M et al. Comparison between familial and nonfamilial melanoma in France. Arch Dermatol 1995; 131: 1154–1159. 11 Greene MH, Clark WH Jr, Tucker MA, Kramer KH, Elder DE, Fraser MC. High risk of malignant melanoma in melanoma-prone families with dysplastic nevi. Ann Intern Med 1985; 102: 458–465. 12 Whiteman DC, Parsons PG, Green AC. p53 expression and risk factors for cutaneous melanoma: a case-control study. Int J Cancer 1998; 77: 843–848.
936
13 Whiteman DC, Watt P, Purdie DM, Hughes MC, Hayward NK, Greene AC. Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous melanoma. J Natl Cancer Inst 2003; 95: 806–812. 14 Curtin JA, Fridlyand J, Kageshita T et al. Distinct sets of genetic alterations in melanoma. N Engl J Med 2005; 353: 2135–2147. 15 Gandini S, Sera F, Cattaruzza MS et al. Meta-analysis of risk factors for cutaneous melanoma: I. Common and atypical naevi. Eur J Cancer 2005; 41: 28–44. 16 Aitken JF, Youl P, Green A, MacLennan R, Martin NG. Accuracy of case-reported family history of melanoma in Queensland, Australia. Melanoma Res 1996; 6: 313–317. 17 Aitken JF, Duffy DL, Green A, Youl P, MacLennan R, Martin NG. Heterogeneity of melanoma risk in families of melanoma patients. Am J Epidemiol 1994; 140: 961–973. 18 Pollock PM, Weeraratna A, Trent JM. Genetics and molecular staging. In: Balch CM, Houghton AN, Sober AJ et al., eds. Cutaneous Melanoma, 4th edn. Quality Medical Publishing, Inc., St. Louis, MO, 2003: 687–712. 19 Newton JA, Bataille V, Griffiths K et al. How common is the atypical mole syndrome phenotype in apparently sporadic melanoma? J Am Acad Dermatol 1993; 29: 989–996. 20 Vasen HF, Gruis NA, Frants RR, van der Velden PA, Hille ET, Bergman W. Risk of developing pancreatic cancer in families with familial atypical multiple mole melanoma associated with a specific 19 deletion of p16 (p16-Leiden). Int J Cancer 2000; 87: 809–811. 21 Goldstein AM, Chan M, Harland M et al. Features associated with germline CDKN2A mutations: a GenoMEL study of melanoma-prone families from three continents. J Med Genet 2007; 44: 99–106. 22 Puig S, Malvehy J, Badenas C et al. Role of the CDKN2A locus in patients with multiple primary melanomas. J Clin Oncol 2005; 23: 3043–3051. 23 Newton JA. Familial melanoma. Clin Exp Dermatol 1993; 18: 5–11. 24 Garbe C, Buttner P, Weiss J et al. Associated factors in the prevalence of more than 50 common melanocytic nevi, atypical melanocytic nevi, and actinic lentigines: multicenter case-control study of the central malignant melanoma registry of the German Dermatological Society. J Invest Dermatol 1994; 102: 700–705.
© 2008 The Authors JEADV 2008, 22, 931– 936 Journal compilation © 2008 European Academy of Dermatology and Venereology
