I I I II II I II III II II I I I I

Frequency of moles as a key to melanoma incidence? F. H. J, Rampen, M.D., H. L. M. van der Meeren, M.D., and J. B. M. Boezeman, Ph.D. Nijmegen, The Netherlands

Subjects with light skin complexion are especially prone to develop melanoma. Furthermore, individuals with large numbers of moles are also at increased risk of developing melanoma. We studied the association between these two markers of melanoma in a group of healthy white people: 116 children 6 to 9 years of age, 78 children 10 to 13 years of age, and 133 medical and nursing students 18 to 30 years of age. Moles were counted on the chest, back, and lower legs. Skin complexion was established with the use of a scoring system based on burning/tanning ability, eye and hair color, and freckling tendency. A multi_factorial statistical analysis was performed. Average mole counts increased with age (p < 0.0001). Male subjects had more nevi on the trunk than female subjects, whereas counts on the lower legs were higher in female than in male subjects (p = 0.0001). Skin complexion was an important denominator of mole proneness; subjects with a light complexion had higher mean counts than those with a dark complexion (p = 0.0001). The present study shows that a significant association exists between skin complexion phenotype and numbers of moles. The pattern of the appearance of pigmented nevi roughly correlates with the sex and site distribution of melanoma. (J AM ACAD DERMATOL 15:1200-1203, 1986.)

The precise role of sunlight in the etiology of cutaneous melanoma remains a debatable issue. Recent interest in the topic suggests that the ha- bitual and intermittent sun exposure during rec- reational activities, especially when accompanied by painful or blistering sunburns, appears a dom- inant indicator of risk of melanoma. 1-4 A second red thread through the literature is skin complex- ion. People who bum easily rather than tan are particularly prone to develop melanoma, s9 A third factor that has gained little attention until lately is the association between common acquired moles and melanoma risk. The number of palpable nevi

From the Department of Dermatology, University of Nijmegen. Accepted for publication June 11, 1986. Reprint requests to: Dr, Frans H. J. Rampen, Department of Der-

matology, University ofNijmegen, Javastraat 104, 6524 MJ Nijme- gen, The Netherlands.

1200

on the arms 9'i~ and the total number of nevi on the body 6,11 appear to be important denominators. Be- cause skin complexion and number of nevi are both phenotypic parameters associated with me lanoma risk, we wondered whether these factors are in- terrelated. We hypothesized that mole counts would be higher in people with a light skin com- plexion than in those with a dark complexion. Therefore, we investigated the frequency o f moles in children 6 to 13 years of age and in young adults 18 to 30 years of age to see whether such an as- sociation exists.

MATERIALS AND METHODS

Three groups of healthy white persons were included: (1) 116 children 6 to 9 years of age, (2) 78 children 10 to 13 years of age, and (3) 133 medical and nursing students 18 to 30 years of age.

Skin complexion could not be determined according

Volume 15 Number 6 December, 1986

Frequency of moles key to melanoma incidence? 1201

T a b l e I. Scoring index of four biologic parameters correlated with skin type

Score = 1 ] Score = 2

Complexion constituent 1. Burning/tanning ability Bum rather than tan Tan rather than burn 2. Eye color Blue-green Gray-hazel-brown 3. Hair color Red-blond Brown-black 4. Freckling tendency Present Absent

to the Fitzpatrick classification.~2 This classification dis- tinguishes between four subtypes of Caucasian skin ac- cording to the acute and chronic reactions to sunlight from "always burn, never tan" = type I, to "never burn, always tan" = type IV. Self-reported tendency tO burn or tan is unreliable in children. Therefore, we used a modified grading system based on four variables that are correlated directly with skin complexion: burn- ing/tanning ability, eye and hair color, and freckling tendency (Table I). Thus, a scoring index was obtained ranging from 4-8. Subjects with a score of 4-6 were classified as having a "light" complexion, whereas a score of 7-8 indicated a"dark" complexion. The burn- ing/tanning ability of the adult group was obtained by the subject's own assessment of the acute and chronic reactions to sun exposure. The pigmentary traits of the children were judged by one observer by physical ex- amination. All children were examined between June 5 and June 12, 1985, immediately after the first Sunny period of the Dutch 1985 summer season, in order to enable an objective recording of the burning/tanning reactions present at that time.

Moles were counted on three sites: (1) the back above the iliac crest up to the seventh cervical vertebra, (2) the chest and upper aspect of the abdomen above the umbilicus up to the clavicles, and (3) the lower legs from the ankles up to the patella. These areas were chosen because they represent the predilection sites of melanoma. All clinically apparent moles, the smallest of which were about 1 mm across, were included. If in doubt, pigmentary spots were not considered.

A multifactorial statistical analysis of variance was performed to establish the relative importance of the four class variables (complexion, age, sex, and loca- tion) in relation to average mole counts.

RESULTS

All biologic parameters studied proved to be relevant as to their effect on mean mole counts (Table II). The results according to the various subgroups are shown in Table III. The excess of

Table II, Relationship between average mole counts and biologic characteristics

Biologic parameters ...... Significance (p value)

Age p < 0.0001 Skin complexion p = 0.0001 Body site p = 0.0001 Sex p = 0.005

average mole counts in the light complexion groups is highly significant irrespective of site, sex, and age (P = 0.0001). The preponderance of moles in the light complexion categories was most conspicuous in the age group of t8 to 30 years, irrespective of sex and body site (P = 0.0003).

The mean number of moles per subject in- creased with age (P < 0.0001). There were marked sex differences with regard to mole counts for the various body sites (P = 0.000i). Regarding the leg, counts were higher in female than in male subjects. B y contrast, male subjects had higher average counts on the chest and back than female subjects. Counts on the chest were consistently smaller than those on the back (ratio in all subgroups, approximately 2: 3).

DISCUSSION

When interpreting these results we should keep in mind that possible sources of bias may have been introduced in this study. The assessment of skin complexion is subject to several variations in interpretation by the observer and by the person examined. The Fitzpatrick classification 12 is based on the reaction of the skin to acute and chronic exposure to sunlight (burning/tanning history) as gauged by the subject interviewed. This classifi- cation system has a merely theoretical basis. Burn-

Journal of the

1202 Rampen et al American Academy of Dermatology

Table 11I. Mean number of moles according to localization, sex, age group, and skin complexion

Light complexion Dark complexion

Age group No. of Mean number of No. of Mean number of Body site Sex (yr) cases moles ( ± SD) cases moles (± SD)

Lower leg Male 6-9 39 2.2 (2.2) 19 2.3 (2.0) 10-13 23 3.1 (2.9) 18 2.8 (2.8) 18-30 34 10.3 (9.7) 42 6.2 (5.8)

Female 6-9 33 3.6 (2.7) 25 2.7 (2.6) 10-13 21 6.5 (7.0) 16 4.2 (4.8) 18-30 32 14.5 (13.3) 25 6.5 (7.9)

Chest Male 6-9 39 4.7 (2.6) 19 4.7 (3.8) 10-13 23 8.5 (6.3) 18 7.2 (5.2) 18-30 34 15.8 (13.4) 42 12.9 (10.7)

Female 6-9 33 5.1 (4.3) 25 3.8 (3.9) 10-13 21 6.0 (5.2) 16 7.3 (3.2) !8-30 32 11.5 (7.8) 25 6.9 (6.4)

Back Male 6-9 39 8.0 (4.6) 19 7.2 (4.8) 10-13 23 13.0 (10.2) 18 11.8 (7.1) 18-30 34 24.9 (21.3) 42 19.3 (14.1)

Female 6-9 33 7.2 (5.8) 25 6.2 (6.2) 10-13 21 9.9 (7.0) 16 10.6 (8.0) 18-30 32 17.1 (11.9) 25 10.8 (7.2)

SD: Standard deviation,

ing/tanning histories are rather subjective accord- ing to the attitude of the individual with respect to his or her habits of solar exposure. Moreover, such histories are unreliable in children. There- fore, we employed an arbitrary scoring system us- ing burning/tanning ability, eye and hair color, and freckling tendency as the ingredients. These bio- logic parameters bear a strong mutual association. We believe that this scoring system is a more re- liable index of skin typing than the burning/tan- ning ability only.

The discrimination between freckles and moles (or other pigmentary disturbances) can be difficult. Because we included moles of all sizes, some of the smaller ones might have been freckles. Nev- ertheless, we believe that it is not justified to ex- clude from this kind of investigations nevi with a diameter less than an arbitrary minimum, for ex- ample 2 or 3 ram. 1°'13 Likewise, the inclusion of palpable nevi only, excluding macular ones, possibly gives a blurred portrait of the actual situation. 9

The frequency of pigmented nevi in a nonme-

lanoma population has been studied in several in- dependent surveys. 13-16 In only one of these studies mole counts were compared with skin complex- ion. ~ In the latter study relating total mole counts to eye and hair color and freckling tendency failed to reveal any significant association. Our data are in sharp contrast with these findings.

Other authors have studied mole counts in me- lanoma patients as compared with sex- and age- matched controls. 6'9-11 From these reports it is clear that mole counts are higher in melanoma patients than in controls. It has been demonstrated that the number of palpable nevi on the arms bears the strongest association with melanoma risk when compared with other complexion characteristics like acute and chronic reaction to sunlight and hair color. 9

Our results on mole counts concur with what is known about the incidence of cutaneous melanoma in white people. In female subjects there is a pre- ponderance of melanomas on the legs, whereas in male subjects the preferred site is the back. From our study it is obvious that mole counts on the

Volume 15 Number 6 December, 1986

Frequency of moles key to me lanoma incidence? 1203

lower legs are m u c h higher in female than in male

subjects. On the trunk, however , counts are higher

in male than in female subjects. Thus, the sex differences with regard to the anatomic distribution

o f me lanoma c losely fol lows the incidence o f moles on the cor responding sites. Furthermore,

mole counts are considerably higher in the light complex ion ca tegory than in darker subjects. Also,

this is in agreement with epidemiologic data on

melanoma. It is suggested that me lanoma inci- dence adheres to a mathemat ica l or probabilist ic

model accord ing to the occurrence of nevi.

REFERENCES

1. MacKie RM, Aitchison T: Severe sunburn and subse- quent risk of primary cutaneous malignant melanoma in Scotland. Br J Cancer 46:955-960, 1982.

2. Lew RA, Sober AJ, Cook N, et al: Sun exposure habits in patients with cutaneous melanoma: A case control study. J Dermatol Surg Oncol 9:981-986, 1983.

3. Elwood JM, Gallagher RP, Hill GB, Pearson JCG: Cu- taneous melanoma in relation to intermittent and constant sun exposure; the Western Canada Melanoma Study. Int J Cancer 35:427-433, 1985.

4. Green A, Siskind V, Bain C, Alexander J: Sunburn and malignant melanoma. Br J Cancer 51:393-397, 1985.

5. Beitner H, Ringborg U, Wennersten G, LagerlGf B: Fur- ther evidence for increased light sensitivity in patients with malignant melanoma. Br J Dermatol 104:289-294, 1981.

6. Beral V, Evans S, Shaw H, Milton G: Cutaneous factors

related to the risk of malignant melanoma. Br J Dermatol 109:165-172, 1983.

7. Elwood JM, Gallagher RP, Hill GB, et al: Pigmentation and skin reaction to sun as risk factors for cutaneous melanoma: Western Canada Melanoma Study. Br Med J 288:99-102, 1984.

8. Elwood JM, Gallagher RP, Davison J, Hill GB: Sunburn, suntan and the risk of cutaneous malignant melanoma; The Western Canada Melanoma Study. Br J Cancer 51:543-549, 1985.

9. Holman CDJ, Armstrong BK: Pigmentary traits, ethnic origin, benign nevi and family history as risk factors for cutaneous malignant melanoma. JNCI 72:257-266, 1984.

10. Green, A, MacLennan R, Siskind V: Common acquired naevi and the risk of malignant melanoma. Int J Cancer 35:297-300, 1985.

11. Swerdlow A J, English J, MacKie RM, et ah Benign naevi associated with high risk of melanoma. Lancet 2:168, 1984.

12. Melsky JW, Tanenbaum L, Parrish JA, et al: Oral me- thoxsalen photochemotherapy for the treatment of pso- riasis. A cooperative clinical trial. J Invest Dermatol 68:328-335, 1977.

13. MaeKie RM, English J, Aitchison TC, et al: The number and distribution of benign pigmented moles (melanocytic naevi) in a healthy British population. Br J Dermatol 113:167-174, 1985.

14. Pack GT, Lenson N, Gerber DM: Regional distribution of moles and melanomas. Arch Surg 65:862-870, 1952.

15. Nicholls EM: Developmeiat and elimination of pigmented moles and the anatomical distribution of primary malig- nant melanoma. Cancer 32:191-195, 1973.

16. Cooke KR, Spears GFS, Skegg DCG: Frequency of moles in a defined population. J Epidemiol Community Health 39:48-52, 1985.

ABSTRACTS

Comparison of the in vitro and in vivo activity of the bis-triazole derivative UK 49,858 with that of amphotericin B against Histoplasma capsulatum

Kobayashi GS, Travis S, Medoff G: Antimicrob Agents Chemother 29:660-662, 1986

The well-known Dr. Kobayashi, mycologist at Washington Uni- versity in St. Louis, MO, presents animal trials with a new antibiotic, comparable to amphotericin-B, but not so toxic.

Philip C. Anderson, M.D.

Filamentous Histoplasma capsulatum involving a ventriculoatrial shunt

Schwartz JG, Tic FO, Fetchick RJ: Neurosurgery 18:487- 490, 1986

Histoplasmosis may show the organism in the filamentous form more often than textbooks admit. Shunts tertd to become infected rather easily; somehow they interfere with local defenses.

Philip C. Anderson, M.D.

Drug eruptions from phenylbutazone in Jamu

Giam YC, Tham SN, Tan T, et al: Ann Acad Med Singapore 15:118-121, 1986

There is some irony here for the herbalists. Supposedly indigenous bona fide folk medicines were found mixed in secret with diazepam or phenylbutaZone. Adverse cutaneous reactions occui" to all com- ponents. Often, even the sincere and honest herbalists lack some scientific rigor in their practices.

Philip C. Anderson, M.D.

Thrombotic thrombocytopenic purpura and systemic lupus erythematosus

Fox DA, Faix JD, Coblyn J, et al: Ann Rheum Dis 45:319- 322, 1986

One cause of cutaneous purpura is lupus erythematosus. Philip C. Anderson, M.D.