Congenital Nevocytic Nevi:Follow-Up of a Swedish Birth RegisterSample Regarding Etiologic Factors,Discomfort, and Removal Rate

Peter Berg, M.D. and Bernt Lindelof, M.D., Ph.D.

Department of Dermatology, Karolinska Hospital and Institute, Stockholm, Sweden

Abstract: Congenital nevi both small and large are frequently removed.We attempted to study the removal rate and etiologic aspects of congenitalnevi as well as their psychosocial effects through the use of a quality test. Aquestionnaire sent to a sample population of individuals with congenitalnevocytic nevi (n ¼ 192) collected from the Swedish Medical Birth Register(SMBR) was used as a test of the register’s quality and for collecting infor-mation on rate of removal and etiologic factors. The quality test indicatedthat only 85.3% of the nevi listed in the SMBR were true congenital nevocyticnevi. Of all true congenital nevocytic nevi reported in the questionnaire,39.8% had been removed and none of the rest had developed malignantmelanoma. The median time for follow-up was 14 years. The median age atremoval of the nevi was 9.7 years. Eight percent of respondents believed thatthe skin lesion caused taunting and changed their social activities. Noinfection or illness during pregnancy was specifically related to the devel-opment of congenital nevi. In conclusion, the larger the nevi, the morefrequently and earlier they are excised. With an excision rate of 40% ofcongenital nevocytic nevi, we found no malignant melanoma.

Melanocytic nevi are defects of development (1).A pigmented congenital nevocytic nevus (CNN) iscomposed of nevus cells and is present at birth. Thoseappearing later should not be considered congenital,even following a premature birth. There are no exacthistopathologic criteria for diagnosing CNN and there-fore nevi developing within the first months of lifeshould be regarded as tardiv nevi. They differ greatlyin size, from a few millimeters to a very large portion ofthe body surface (2–4).

The larger a CNN, the more common is surgicalintervention (such as excision and/or dermabrasion toreduce the number of nevocytic cells on the body surface(5–10). There is a correlationbetween largeCNNand therisk of malignant melanoma (MM) (11–16).

Normally when an acquired nevus is excised it isimportant to check the histopathologic specimen for freemargins to determine that nonevocytic cells remain. Thisis thought to reduce the risk of malignant transforma-tion, but it should be noted that MM can develop from

Address correspondence to Peter Berg, M.D., KarolinskaHospital and Institute, Department of Dermatology, S-171 76Solna, Sweden, or e-mail: peter.berg@ks.se.

293

Pediatric Dermatology Vol. 19 No. 4 293–297, 2002

residual nevus cells left after surgery (17). It seems to bevery common that even fairly small CNNs are not totallyexcised (10,18).

Today many methods used to remove nevi from achild’s skin do not result in total removal of all nevo-cytic cells. Therapeutic treatments such as dermabra-sion, curettage, and chemical peel only reduce thenumber of nevocytic cells. These techniques seemadequate, however, and so far none of the patients havedeveloped MM (19–22). However, the lack of totalelimination of nevocytic cells has not been evaluated asa risk factor for malignant transformation during alifetime.

Why CNN arise during fetal life is not clear, butgenetic or infectious factors may play a role. CNN mayarise because of untoward events during the pregnancy,such as influenza-like symptoms (23). Another possiblecause could be the presence of a c-met proto-oncogeneproduct in CNN in children with neurocutaneousmelanosis (24).

We performed a questionnaire survey of a sample ofCNN patients from the Swedish Medical Birth Register(SMBR) in order to determine how often excisions ofCNN are performed; when, where, how, and to whatextent are they removed;what effect these factors haveonthe chance of malignant transformation; if CNNs causesocial discomfort; and if there are any obvious etiologicfactors.

METHODS

Between 1973 and 1993, 3922 infants with congenitalnevi were registered in the SMBR. We sampled 1 in 20(192 persons). For seven of these persons we could nottrace a registration in the Swedish population register(SEMA group) because they had left the country, or forother unknown reasons. This left 185 persons. Wereceived answers from 150 subjects (81.1%), of which 83

were female (55%) and 67 were male (45%). To reducethe role of confounding factors, especially concerningmothers who smoked during pregnancy, we checked theSMBR records of smoking mothers and their newbornchildren registered between the years 1983 and 1997 withthe diagnosis of congenital nevi.

Questionnaire

A questionnaire was mailed to 185 persons registeredin the SMBR as having CNN. We used this to gatherinformation and as a quality test of the SMBR. Thequestionnaire had 11 questions with up to five parts.Four of the questions were related to visual analog scales(Table 1).

The questionnaire sought information about genderand age; when, where, how, and by whom any surgicalexcisions were performed; and whether the respondentstill had some CNN remaining. If so, the location, size,color, and structure of the skin lesion was documented.We also asked whether the skin lesion had caused anysocial or physical discomfort. We asked whether anyclose relatives of the respondents currently or previouslyhad CNN, and also whether the CNN had changed therespondent’s behavior with regard to sun exposure.Moreover, we sought information about whether thescars or remaining CNN caused any social discomfortwhich reduced possibilities for normal activities. Twoquestions concerning the mother were whether shesmoked or had any infections or illnesses during thepregnancy, or had any known present illness or disease.

The CNN were classified into three groups for jud-ging: definite, probable, and not probable. In the ques-tionnaireweusedexcisionwithhistopathologicdiagnosisas a criterion for the group ‘‘definite CNN,’’ ‘‘probableCNN’’was designatedon the basis of the clinical findingsbut without biopsy, and employment of vascular lasertreatment was used to indicate ‘‘not probable CNN.’’ In

TABLE 1. Items Included in the Questionnaire for Congenital Nevocytic Nevi

Categories Specific queries/reason for inclusion

Age, sex Relation to gender and ageSurgical removal Removal frequency, median age, technique used, by which specialist and where, follow-up treatmentCharacterization and localization Size, localization, color, hair frequency (VAS), and appearance of retained CNNPatient’s illness Any related diseaseSocial discomfort or bullying Cosmetic or other discomfort caused by the nevi (VAS)Changed social activities Current participation in sports and activities (VAS)Sun protection behavior Whether scar or nevi changed sun protection behavior (VAS)Relatives with CNN Hereditary factor in close relativesImpact of mother’s smoking Smoking during pregnancy as a risk factor for transformation of nevus cellsMother’s infections/illness A cause of CNN during pregnancyMother’s illness, now and earlier CNN’s connection with other diseases

VAS, visual analog scale.

294 Pediatric Dermatology Vol. 19 No. 4 July/August 2002

Swedenwedonot treatneviwith lasersbecauseof the riskof a diagnosis of pseudomelanoma in the treated areabeing made when examining histopathologic controls.Nevus classification was further based on the shape, hairgrowth, color, localization, and size of theCNN. In somecases the patient gave a diagnosis ‘‘not a CNN,’’ that is,hemangioma or cafe au lait spots.

The Swedish Medical Birth Register

The SMBR, covering all of Sweden, began in 1973.Registration was based initially on a medical birthregistration form, as part of a standardized record systemfor maternal health, obstetrics, child health, and neona-tal health care. A copy of the medical birth registrationform was sent to the National Board of Health andWelfare for all newborns, live as well as those stillborn.During 1982 the medical registration form was replacedwith copies of themedical records. For births in 1993 thedropout rate was about 1.1% of the total and about1.6% for perinatal deaths. About 80% of all births takeplace at specialized hospitals. Only 4% of infants areborn at hospitals with neither obstetric nor pediatricspecialty departments (25).

RESULTS

The median age of the patients with CNN at the time offilling out the questionnaire was 14 years (range 5–27years).

Quality Test

Weclassified the 150 patients from the information in thequestionnaire as having definite CNN (78.6%), probableCNN (6.7%), or probably not CNN (14.7%) (heman-giomas, cafe au lait spots, and other lesions that haddisappeared). Definite and probable CNN thusmade up85.3% of the total, which we defined as true CNN in thequality test.

Surgical Removal Rate

Fifty-one patients had undergone one or more opera-tions (dermabrasion or excision). Thus the removal ratewas 39.8% for true CNN. Ten of these had large CNNand 41 had small CNN.

In the group with definite CNN, the excision fre-quency was 43.2% (51/118). In the group with probableCNN, the excision frequency was 0% (0/10). In thegroup with not probable CNN, the excision frequencywas 0% (0/22), but 33% of the hemangiomas had beenlaser treated.

The median age at surgery was 9.7 years (range 0–25years). This represents the first excision, since somepatients underwent surgery many times. Almost 90% ofthe excisions were done by plastic surgeons or generalsurgeons, 5% by dermatologists, and the rest by otherdoctors. The larger the CNN, themore common and theearlier the excision.

Influence of Maternal Smoking

Of themothers of those patients whose nevi were excised(all defined as definite CNN), 39.2% were smokers dur-ing pregnancy. In the total group with definite CNN,only 25.4% of the mothers were smokers, and in thegroup with probably not CNN, 18% were smokers.However, in the SMBR statistical control sample of allliving newborn children with nevi registered during1983–1997 showed no significance concerning smokingmothers. Sixteen mothers had infections during preg-nancy and 21 had a current infection, but a specificinfection/illness was not identified.

Characterization and Distribution

CNNweremost common on the trunk, legs, head, arms,and neck, in that order. The head, however, was over-represented. Of all CNN reported, 29% had hairgrowth.

Social Behavior and Discomfort

About 30% of those sampled said they were very carefuland an additional 25% were moderately careful aboutsun exposure. Eight percent of those with true CNNdisliked the social reactions to their skin lesions, that is,they felt they were a target for bullying. This affectedtheir social activities, which were reduced. We wereunable to evaluate reported CNN in relatives because wecould not determine whether they were true CNN.

DISCUSSION

Since we were unable to review the cases of the 3922people with CNN in the SMBR, we used a sample ofdiagnosed CNN from the registry as a quality test. Thediagnosis of nevus is given at birth, mostly by pediatri-cians, so we felt it was important to do a quality test. Thequestionnairewas easy to fill out andwas kept as short aspossible so as to get a good response rate. Eighty percentof those queried returned it. To verify the effect of theremoval of CNN on the occurrence of malignant mel-anoma, a randomized study would be required, which ofcourse is ethically difficult.

Berg and Lindelof: Congenital Nevi Removal Rate 295

Why CNN arises during fetal life is unknown. Nohereditary factor has been found; neither has any envi-ronmental factor been proposed as important. Somehave argued that influenza-like symptoms could causefetal damage in pregnancy, giving rise to CNN (23). Theinvolvement of a c-met proto-oncogene product has alsobeen proposed (24).

We asked whether close relatives (parents or siblings)had a CNN. Evaluation of the responses showed a veryhigh frequency of CNN, probably because the patientswere unable to distinguish between CNN and acquirednevi, sowedecided that this question did not supply validinformation.

Only 8%of respondents said that their scars orCNNscaused reduced activities and social discomfort. Thisseems to indicate a good acceptance of the lesions.

Up to 25% were aware of the risk of sun exposurefor nevi, which indicates that the patients or the parentswere concerned aboutmalignant transformation.With aremoval rate of nearly 40%of all trueCNNwe found nomalignantmelanoma, although the observation period isshort. Probably most of the CNN are removed for cos-metic reasons, but there is indirect evidence that a certainremoval rate of CNNmay be relevant to decreasedMMdevelopment. When previous studies concerning MMdeveloping within CNN and acquired nevi are analyzed,the melanoma risk appears to be decreasing. However,most of the melanomas in large CNN arise duringchildhood, and in small CNN, MM is rare.

Various errorsmust be considered.Since thediagnosisCNN is given at birth and registered in the SMBR (25),there should be no question of the congenital nature ofthe lesions. However, since no biopsies have been taken,someerrors indiagnosismayoccur (13).Wedonotknowwhether the reason for the removal ofCNNwas cosmeticor because they seemed atypical, but we found it rea-sonable to believe that the lesions at highest risk wereremoved. The strength of the present study lies in theevaluation of CNN as definite or probable CNN in85.3% of the sampled subjects and in the quality testingof the SMBR.

In conclusion, with a removal rate of 40% of theCNN, no MM developed, so perhaps those lesions withthe highest risk were removed. However, there is nodirect evidence that MM are actually prevented by theexcision ofCNN. It is possible that the lower risk ofMMdevelopment in CNN at present compared to 20 yearsago may be due to improved surgical techniques. Fur-thermore, dermabrasion or other methods of reducingnevocytic cells seem adequate. Could there be a risk ofpseudomelanomas in partly treated CNN as well as inlaser-treated acquired nevi? None of our patients haddeveloped MM, but the follow-up time was limited. We

have foundno literature on the induction time forMMinpatients with CNN. The larger the nevi, the more com-mon is surgical removal. Probably many of the CNNwere removed for cosmetic reasons. It is notable that farfewer dermatologists than surgeons excised CNN.Lastly, we found no clear association between CNN andmaternal illness/infection during pregnancy. The CNNseem to have had limited effects on social life, but hadresulted in greater caution with regard to sun exposure.

ACKNOWLEDGMENTS

We thank Professor Sven Cnattingius, M.D. Ph.D. andPeter Herin, M.D. Ph.D. for valuable discussion andhelpful criticism.

REFERENCES

1. RookA.Naevi and other developmental defects. In: RookA, ed. Textbook of dermatology. London: BlackwellScientific, 1986:181–189.

2. Greeley PW, Middleton AG, Curtin JW. Incidence ofmalignancy in giant pigmented nevi. Plast Reconstr Surg1965;36:26–37.

3. Swerdlow AJ, English JSC, Qiao Z. The risk of melanomain patientswith congenital nevi: a cohort study. JAmAcadDermatol 1995;32:595–599.

4. Castilla EE, Dutra MD, Orioli-Parreiras IM. Epidemiol-ogy of congenital pigmented nevi: incidence rates andrelative frequencies. Br J Dermatol 1981;104:307–315.

5. Pers M. Naevus pigmentosus gigantigus: indikationer foroperativ behandling. Ugeskr Laeger 1963;125:613–619.

6. Keipert JA. Giant pigmented naevus: the frequency ofmalignant change and indications for treatment in prepu-bertal children. Australas J Dermatol 1985;26:81–85.

7. Fink-Puches R, Soyer HP, Pierer G, Kerl H, Happle R.Surgical treatment of child nevus. Eur J Dermatol2000;10:262–264.

8. Lawrence CM. Treatment options for giant congenitalnaevi. Clin Exp Dermatol 2000;25:7–11.

9. Michel JL, Laborde-Milaa Roux V, Chavrier Y, et al.Neonatal curettage of giant congenital nevi.AnnDermatolVenereol 2000;127:23–28.

10. Hopkins JD, SmithAW, Jackson IT.Adjunctive treatmentof congenital pigmented nevi with phenol chemical peel.Plast Reconstr Surg 2000;105:1–11.

11. Gari LM, Rivers JK, Kopf AW. Melanomas arising inlarge congenital nevocytic nevi: a prospective study.Pediatr Dermatol 1988;5:151–158.

12. Quaba AA, Wallace AF. The incidence of malignantmelanoma arising in large congenital nevocellular nevi.Plast Reconstr Surg 1986;78:174–179.

13. Marghoob A, Schoenbach S, Koph A, Orlow S, Nossa R,Bart R. Large congenital melanocytic nevi and the risk forthe development of malignant melanoma. Arch Dermatol1996;132:170–175.

14. LorentzenM, Pers M, Bretteville-Jensen G. The incidenceof malignant transformation in giant pigmented nevi.Scand J Plast Reconstr Surg 1977;11:163–167.

296 Pediatric Dermatology Vol. 19 No. 4 July/August 2002

15. Kopf AW, Bart RS, Hennessey P. Congenital nevocyticnevi and malignant melanomas. J Am Acad Dermatol1979;1:123–130.

16. Bittencourt FV, Marghoob AA, Kopf AW, et al. Largecongenital melanocytic nevi and the risk for developmentof malignant melanoma and neurocutaneous melanocyto-sis. Pediatrics 2000;106:736–741.

17. Rhodes AR, Wood WC, Sober AJ, Mihm MC Jr.Nonepidermal origin of malignant melanoma associatedwith a giant congenital nevocellular nevi. Plast ReconstrSurg 1981;67:782–790.

18. Walton RG, Jacobs AH, Cox AJ. Pigmented lesions innewborn infants. Br J Dermatol 1981;95:389–396.

19. Sandsmark M, Eskeland G, Ogaard AR, Abyholm F,Clausen OP. Treatment of large congenital nevi. A reviewand report of six cases. Scand J Plast Reconstr Surg HandSurg 1993;27:223–232.

20. Mollitt DL, Golladay ES. Management of congenitalnevocellular nevi. Am J Surg 1985;150:669–671.

21. Arons MS, Hurwitz S. Congenital nevocellular nevus. areview of the treatment controversy and a report of 46cases. Plast Reconstr Surg 1983;72:355–365.

22. Kruk-Jeromin J, Lewandowicz E, Rykala J. Surgicaltreatment of pigmented melanocytic nevi dependingupon their size and location. Acta Chir Plast 1999;41:20–24.

23. Dawson HA, Atherton DJ, Mayou B. A prospectivestudy of congenital melanocytic nevi: progress reportand evaluation after 6 years. Br J Dermatol 1996;134:617–623.

24. TakayamaH,NagashimaY,HaraM, et al. Immunohisto-chemical detection of the c-met proto-oncogene product inthe congenital melanocytic nevus of an infant withneurocutaneous melanosis. J Am Acad Dermatol2001;44:538–540.

25. Cnattingius S, Kallen B, Ericsson A, Gunnarskog J. Aquality study of a medical birth register. Scand J Soc Med1990;18:143–148.

Berg and Lindelof: Congenital Nevi Removal Rate 297