Results of surveillance programme aimed at early diagnosis of cutaneous melanoma in high risk Mediterranean subjects

ARTICLE

Auteur(s) : Paolo CARLI¹, Daniela BALZI², Vincenzo DE GIORGI¹, Daniela MASSI³, Domenico PALLI⁴, Alessandra CHIARUGI¹, Paolo NARDINI¹, Benvenuto GIANNOTTI¹

¹ Department of Dermatology, University of Florence, ² Epidemiology Unit, ASL Florence, ³ Department of Human Pathology and Oncology, University of Florence, ⁴ Epidemiology Unit, Centro per lo Studio e la Prevenzione Oncologica (CSPO), Via degli Alfani, 37, 50121 Florence ‐‐ Italy

Article accepted on 25\7\03 Key words: Risk factors for cutaneous melanoma have been mainly investigated in fair‐skinned populations of northern European origin [1, 2], while little is known about predictors of melanoma risk in Mediterranean populations. Although not predominantly fair‐skinned, people living southern Europe are nevertheless not spared by the so‐called "melanoma epidemic", with an average increase of melanoma incidence rates of about 5 % per year [3, 4]. Many case‐control studies indicated that a high number of nevi and the presence of atypical nevi are the strongest predictor of melanoma risk [1,2]; this has been recently confirmed also in Mediterranean populations [5, 6]. Little is known however about the impact of prevention programmes based on specialized periodical examination of at risk individuals on melanoma diagnosis. Our aim was to investigate the occurrence of new melanomas and the microstaging of disease at diagnosis in an a priori defined high‐risk cohort actively followed over time. Moreover, in order to evaluate the excess risk associated with a specific pattern of nevi, the number of tumors observed in this cohort was compared with the number of melanomas expected according to age‐ and sex‐specific incidence rates in the general population, provided by the local cancer registry. In addition, we also enrolled a group of intermediate‐risk subjects in a similar programme of counselling and specialized examination, in order to have a comparison group.

Materials and methods

Since 1989, when an educational campaign, directed at the general population and family doctors, was launched in the Florence area, the Pigmented Lesion Clinic (PLC) of the Department of Dermatology University Hospital has been the referral center for the early diagnosis of melanoma in this area. According to the campaign protocol, residents with self‐ or GP‐detected suspicious pigmented skin lesions as well as subjects with risk factors for melanoma requiring periodic examination were invited to undergo dermatological examination at the local PLC.

Standardized examination of subjects at PLC

At each PLC examination, the subjects were visited by dermatologists with experience in the diagnosis of cutaneous pigmented lesions (P.C., V.D.G.). All the body surface except the scalp and genitalia were systematically examined and the number of common and atypical nevi counted and recorded in a digitalized file, in agreement with a published protocol [6]. The genitalia and scalp were examined when the subjects were referred for a potentially suspicious pigmented skin lesion at these sites. Clinical definitions of common and atypical nevi have been reported elsewhere [5]. Briefly, acquired melanocytic nevi were defined as pigmented lesions larger than 2 mm. in diameter clinically different from freckles, lentigines, blue nevi, seborrheic keratoses, and pigmented basal cell carcinoma. Acquired melanocytic nevi larger than 6 mm. in diameter that showed at least two of the following criteria were defined as atypical: irregular or ill‐defined borders, variegated or irregularly distributed pigmentation, background erythema, or accentuated skin markings. The eye color was evaluated according to a four‐category scale (black‐brown, hazel, green‐gray, blue). Phototype according to Fitzpatrick was also recorded (category from I to IV) [5].

Criteria for entry into the cohort

The present cohort study included all the consecutively observed subjects showing more than 30 common nevi and 3 or more atypical nevi (here considered at high‐risk), who had at least two dermatological examinations at PLC with a minimum interval of 1 year between the two consultations and who were resident in the district of Florence (about 1,150,000 inhabitants). A group of subjects with less than 30 common nevi and 3 atypical nevi were also enrolled (intermediate‐risk).

The number of nevi chosen to identify the high‐risk group was defined a priori, in accordance with the results of a previous population‐based case‐control study performed by our group in the Florence area [5]. According to this study, subjects with more than 30 nevi showed a relative risk of melanoma, as expressed by the odds ratio, of 22.3 compared to subjects with less than 10 nevi [5]. The presence of at least three atypical nevi as an additional factor for entry into the study was chosen in accordance with recent studies carried out in a predominantly fair skinned population [7]. No definite estimates of risk of melanoma associated with number of atypical nevi are available for Mediterranean populations as yet.

Only subjects resident in the Florence district (municipality of Florence and surrounding towns) were considered eligible for the study in order to compare the observed cases of melanoma in the cohort to that expected according to age‐ and sex‐ specific incidence rates in the general population residing in the same geographically defined area.

Subjects with personal or family history of melanoma or with a diagnosis of melanoma at the time of the first PLC observation were excluded.

The enrollment period (first PLC consultation) started in 1990 and finished in 1996.

In the series of 2,479 consecutive subjects without melanoma who had a first visit at the PLC in the enrollment period, 455 individuals, all Caucasians, fulfilled the above mentioned inclusion criteria. Therefore the study population included 218 high risk subjects (96 males, 122 females, mean age 27.8 years, range 18‐61); the intermediate risk group included 237 subjects (80 males, 157 females, mean age 30.7, range 19‐73).

The major reasons for which the remaining subjects were not eligible were: i) the lack of a second consultation, ii) residence outside the Florence district or iii) family history of melanoma.

When a subject included in the PLC file was referred to the PLC for further examinations, the examination outcome, i.e., no suspicious lesion or suspicious lesion to be excised, was recorded in the file. At the end of the study, the list of cohort subjects for whom surgical excision of a pigmented lesion was recommended were retrieved from the PLC data base. In these cases the files were examined and histopathologic materials reviewed by an independent pathologist with expertise in melanoma to finalize the diagnosis (D.M.). The date of detection of a histologically confirmed melanoma was considered the time of exit from the cohort for that subject.

Registry‐based incidence rate of in situ and invasive melanoma in the Florence District

Melanoma incidence rates were provided by the local cancer registry (Tuscany Cancer Registry), which produces, since 1985, incidence data in accordance with the quality standard requested by the International Agency for Research on Cancer [4]. The proportion of histological verification for melanoma cases was 96 % in 1997. The proportion of Death Certification Only (DCO) cases, ie. melanoma cases for which the registry does not have clinical documentation and which represent the "unsuccessful" case as far as completeness of collection is concerned, is 0.4 %.

For the purpose of our study, age‐ and sex‐specific incidence rates of melanoma were recalculated to include in situ lesions, collected but not routinely included in the incidence rates. Thus, both digitalized and hard copies of histopathologic certifications of all in situ melanomas as well as atypical melanocytic hyperplasias were retrieved. Lesions classified with an equivocal terminology, such as "atypical melanocytic hyperplasia", were reviewed by an experienced pathologist (D.M.) to confirm the diagnosis of in situ melanoma according to currently used histopathological criteria [8]. In particular, a diagnosis of in situ melanoma was made when there were increased numbers of atypical melanocytes, both in solitary units and\or in nests at the dermo‐epidermal junction and at all levels of the epidermis, with possible extension of melanocytes far down into adnexal structures. Cytologically, melanocytes show marked variations in size and shape and display enlarged, pleomorphic nuclei.

Since the incidence of melanoma is increasing over time, the local cancer registry data base analyzed as raw data for calculation of the overall incidence rate of melanoma (invasive plus in situ forms) included the period 1995‐1998, ie, the most recent available data about melanoma incidence in the Florence area.

The 218 high‐risk subjects have been followed for an average of 3.4 years (range 1‐6.5), for an overall total of 741.7 person‐years. The mean number of examinations was 2.7 + \ ‐‐ SD 2.0 for females (range 1‐10) and 2.8 + \ ‐‐ SD 1.9 (range 1‐8) for males. No statistically significant difference in the number of examinations was found according to the number of common nevi or atypical nevi.

Intermediate risk subjects have been followed for an average of 3.1 years (range 1‐7), with an overall total of 658.3 person‐years; the mean number of examinations was 2.00 (range 1‐6). Characteristics of the study population are shown in Table I‐II.

Results

At consultations during the follow up of the high risk group, 27 pigmented skin lesions in 24 subjects were defined as suspicious and referred for surgery. Histopathological examination identified a cutaneous melanoma in 4 and a melanocytic nevus in the remaining 23 lesions (18\23 histologically classified as nevi with architectural disorder and cytologic atypia). The malignant \benign ratio (melanomas detected\benign lesions excised) resulted 1: 5.7.

In the intermediate‐risk group, 6 pigmented lesions were considered suspicious or equivocal in 5 subjects during follow up examination and excised. Histologic diagnoses were melanocytic nevi in 4 cases and non melanocytic lesion in 2 cases (1 pigmented seborrheic keratosis, 1 basal cell carcinoma).

Thus, during the study period, 4 new melanomas were detected in subjects with more than 30 common nevi and three or more atypical nevi while no melanoma has been found in the intermediate‐risk group.

All detected melanomas were in situ (Clark level I) (Table III). The average interval between the entry in the cohort and the date of melanoma diagnosis was 4.6 years (range 1.4‐6.5). Three out of 4 cases that developed a melanoma spontaneously reported a change in the index lesion to the PLC dermatologist, and the lesion subsequently was diagnosed as a melanoma. In one case, the subject did not notice any change in the lesion over time and the diagnosis was based on purely morphologic criteria adopted by the clinician during the follow up PLC visit. The mean age at enrollment into the study was 35.9 years for the 4 subjects who later developed a melanoma and 27.6 for the remaining 214 subjects who did not. The mean number of follow up examination was 2.2 for subjects developing melanoma and 2.7 for subjects without melanoma. The average number of common nevi was 63.5 and 72.3 in subjects developing melanoma or not, respectively. The mean number of atypical nevi was 4.5 and 9.7 in subjects with or without melanoma, respectively.

Discussion

Recent case‐control studies indicate that the number of melanocytic nevi and the presence of atypical nevi are the strongest risk factors for non familial melanoma in Mediterranean populations [5‐6]. Up to now, however, no demonstration of a suitable prevention program based on follow up of subjects presenting the above mentioned conditions has been reported. From our data, four new melanomas developed in 218 high risk subjects in a mean follow up period of 4.6 years. All these tumors were detected when still in situ, therefore associated with a favourable prognosis. This finding suggests clinically relevant effects of the prevention programme in term of earliness of diagnosis. The number of melanomas was higher than that expected in the general population, confirming the role played by the large number of nevi as strong risk factor for melanoma. The frequency of new melanomas resulted about 40‐fold higher than expected in the age‐ and sex‐matched general population. A word of caution is however needed since high risk individuals were examined within an intensive control programme which rendered these subjects more prompt to notice the suspicious lesion or to have this lesion diagnosed compared to general population. Nevertheless, the significance of the predictor of risk associated with large number of nevi is given by the fact that no melanoma was found in the intermediate‐risk subjects submitted to the same follow up protocol. Regarding possible histologic misdiagnosis of in situ melanoma, all the precautions to avoid this risk were taken: first, diagnosis came from a pathologist blinded as to the risk factor category of the subjects, in order to avoid any overestimation of atypia of the melanocytic lesions; second, histologic slides were reviewed by an independent pathologist with great expertise in melanoma and melanocytic lesions who provided the same diagnosis; third, all these lesions showed clinical and dermoscopical features clearly suggestive of melanoma, with, therefore, strong correlation between clinical and histologic diagnosis of melanoma.

Counseling of at risk individuals is probably important to improve the self‐detection of suspicious lesions [1,2]: in our program during the first dermatological consultation (entry in the cohort), all subjects were given an educational leaflet to inform them of the importance of self‐skin examination to detect any new suspicious lesions (ABCDE rule, "changing mole") [9]. Actually in our study, 3 of the 4 subjects who developed a melanoma alerted the PLC dermatologist to the presence of a suspicious lesion, thus showing that they had performed the recommended self‐skin examination correctly.

Periodic examination of subjects coupled with recommendation of self‐skin examination, led in all cases to discovery of the tumor when still in situ. This scenario would probably change with a higher number of patients since the possible occurrence of nodular melanoma histotypes, which lack the in situ phase since they show more aggressive behaviour (vertical growth phase ab initio). Therefore, the detection of all melanomas when still in situ should be considered a contingent, although promising result.

It has sometimes been pointed out that after screening campaigns the increased workload of subjects with atypical lesions seeking consultation results in a reduction in the ratio between the number of MM and number of nevi excised compared to pre‐campaign values (from 1:11 to 1: 16) [15]. This is due to the fact that the worry about the risk of false negative diagnosis prompts clinicians to lower the treatment threshold and consequently to prescribe excision of a larger numbers of benign lesions. Even if a number of benign lesions was excised to rule out the suspicion of a melanoma the melanoma\nevi ratio observed in the high risk group (4:23, ie. 1: 5.7) can be considered acceptable to guarantee the timely excision of all the tumors in these subjects.

Some authors have suggested using photographic surveillance (baseline photography and follow up) to enable early diagnosis of melanoma [16]. This is a very controversial point that requires more reliable support from randomized studies [17]. In our clinic, photographing the body surface in subjects with many nevi has been abandoned over time because of doubtful utility ‐‐ and the fact that it is very time‐consuming ‐‐ when compared to the good diagnostic performance obtained by a skilled eye using the new method for non‐invasive diagnosis of pigmented skin lesions, ie, dermoscopy [18]. Since atypical nevi, by definition, undergo modification over time, total body photographs could force the dermatologist to remove any changing or new moles in these patients, leading to unnecessary surgical removal of many nevi [17]. Very recently, follow up of atypical lesions by means of digital dermoscopy ‐‐ instead of immediate removal ‐‐ has been suggested [19,20]. However, a study of Kittler et al. alerted dermatologists to the risk of an increased number of false negative diagnoses (melanomas left unexcised) associated with this procedure [21].

In conclusion, a follow up program of high risk individuals classified on the basis of number of nevi proved feasible and should be recommended as a part of prevention programs in southern European populations.

References

1 . Garbe C, Buttner P, Weis J, et al. Risk factors for developing cutaneous melanoma and criteria for identifying persons at risk: multicenter case‐control study of the Central Malignant Melanoma Registry of the German Dermatological Society. J Invest Dermatol 1994; 102: 695‐9.

2 . Roush Ordlund JJ, Forget B, Gruber SB, Kirkwood JM. Independence of dysplastic nevi from total nevi in determining risk for non familial melanoma. Prev Med 1988; 17: 273‐9.

3 . Vinceti M, Bergomi M, Borciani N, Serra L, Vivoli G. Rising melanoma incidence in an Italian community from 1986 to 1997. Melanoma Research 1999; 9: 97‐103.

4 . Buiatti E, Balzi D, Barchielli A, Carli S, Crocetti E, Giovannetti L. Registro Tumori Toscano in Cancer Incidence in Five Continents vol. VII, Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J (Eds.), IARC Sci. Publication n. 143, pgg. 526‐529, Lyon.

5 . Carli P, Biggeri A, Giannotti B. Malignant melanoma in Italy: risks associated with common and clinically atypical melanocytic nevi. J Am Acad Dermatol 1995; 32: 734‐9.

6 . Naldi L, Imberti GL, Parazzini F, Gallus S, LA Vecchia C. Pigmentary traits, modalities of sun reaction, history of sunburns, and melanocytic nevi as risk factors for cutaneous malignant melanoma in the italian population. Cancer 2000

; 88: 2703‐10.

7 . Tucker MA, Halpern A, Holly EA, et al. Clinically recognized dysplastic nevi. A central risk factor for cutaneous melanoma. JAMA 1997; 277: 1439‐44.

8 . Elder DE, Murphy GF. Atlas of tumor pathology: Melanocytic tumors of the skin. Third series. Washington D.C.: Armed Forces Institute of Pathology; Rosai J ed., 1991.

9 . Rhodes AR. Public education and cancer of the skin. What do people need to know about melanoma and nonmelanma skin cancer¿ Cancer 1995; 75: 613‐36.

10 . MacKie RM, McHenry P, Hole D. Accelerated detection with prospective surveillance for cutaneous malignant melanoma in high‐risk groups. Lancet 1993; 341: 1618‐20.

11 . Masri GD, Clark WH, Guerry D 4th, Halpern A, Thompson CJ, Elder DE. Screening and surveillance of patients at high risk for malignant melanoma results in detection of earlier disease. J Am Acad Dermatol 1990; 22: 1042‐8.

12 . Kelly JW, Yeatman JM, Regalia C, Mason G, Henham AP. A high incidence of melanoma found in patients with multiple dysplastic naevi by photographic surveillance. Med J Aust 1997; 167: 191‐4.

13 . Schneider JS, Moore DH, Sagebiel RW. Risk factors for melanoma incidence in prospective follow up. The importance of atypical (dyasplastic) nevi. Arch Dermatol 1994; 130: 1002‐7.

14 . Halpern AC, Guerry D 4th, Elder DE, Trock B, Synnestvedt M. A cohort study of melanoma in patients with dysplastic nevi. J Invest Dermatol 1993; 100: 346S‐9S.

15 . Burton RC, Coates MS, Hersey P, et al. An analysis of a melanoma epidemic. Int J Cancer 1993; 55: 765‐70.

16 . Slue W, Kopf AW, Rivers JK. Total‐body photographs of dysplastic nevi. Arch Dermatol 1988; 124: 1239‐43.

17 . Kanzler MH, Mraz‐Gernhard. Primary cutaneous malignant melanoma and its precursor lesions: diagnostic and therapeutic overview. J Am Acad Dermatol 2001; 45: 260‐76.

18 . Pehamberger H, Binder M, Steiner A, et al. In vivo epiluminescence microscopy: improvement of early diagnosis of melanoma. J Invest Dermatol 1993; 100: 356S‐62S.

19 . Stolz W, Schiffner R, Plillet L et al. Improvement of monitoring of melanocytic skin lesions with the use of a computerized acquisition and surveillance unit with a skin surface microscopic television camera. J Am Acad Dermatol 1996; 35: 202‐7.

20 . Kittler H, Pehamberger H, Wolff K, Binder M. Follow up of melanocytic skin lesions with digital epiluminescence microscopy: patterns of modifications observed in early melanoma,atypical nevi and common nevi. J Am Acad Dermatol 2000; 43: 467‐76.

21 . Kittler H, Binder M. Risks and benefits of sequential imaging of melanocytic skin lesions in patients with multiple atypical nevi. Arch Dermatol 2001; 137: 1590‐5.