Temporal Changes in Size and Dermoscopic Patterns of New and Existing Nevi in Adolescents

To the Editor:

Cross-sectional studies have shown significant differences in the size and patterns of nevi by age(Zalaudek et al., 2006a, Zalaudek et al., 2011), which may have important implications in the clinical evaluation of skin lesions and in understanding melanocytic growth of benign and malignant lesions. The principal dermoscopic nevus pattern in childhood is globular, whereas a reticular pattern predominates in adulthood.(Zalaudek et al., 2006a, Zalaudek et al., 2006b, Zalaudek et al., 2011) Multiple factors may contribute to the observed age-related shift in dermoscopic pattern, but few temporal studies have monitored individual nevi over time. The objective of the present study was to document temporal changes in dermoscopic pattern and size of nevi during childhood and adolescents.

The Study of Nevi in Children (SONIC) was approved by the Institutional Review Boards at Boston and Harvard Universities. Recruitment and methodology were previously detailed.(Oliveria et al., 2014, Oliveria et al., 2013) Three cohorts of participants originated from schools in Framingham, Massachusetts. We performed longitudinal follow-up of selected nevi on the back and legs of participants. Assessments, including skin examination, nevus photography, and questionnaires, were completed in the 8^th or 9^th grade (baseline) and again in the 11^th grade (follow-up). One of the three cohorts (n=366) had undergone a previous assessment of back nevi in the 5^th grade. Photography methodology, definitions for anatomical sites, nevus selection methodology, nevus surface area calculations, and dermoscopic pattern classification have been previously published (Supplementary Materials).(Fonseca et al., 2015, Geller et al., 2007, Oliveria et al., 2014, Oliveria et al., 2013) Notably, a complex nevus dermoscopic pattern included nevi with both network and globules, nevi with a distinct peripheral rim of globules, and nevi with a starburst pattern.

Baseline assessment in the 8^th or 9^th grade was performed in 569 participants (consent rate for study participation of 32%). Of these, 395 participants (69%), of which 60% were males and 89% Caucasians, underwent repeat evaluation in 11^th grade (see demographics table in Supplementary Materials). A total of 1,375 nevi on the back (n=1054, 77%) or legs (n=321, 23%) were serially imaged with dermoscopy (Table 1), of which, 69% (n=944) demonstrated no change in dermoscopic classification. We observed that 60% of complex, 37% of homogenous, 31% of globular, and 22% of reticular nevi demonstrated dermoscopic pattern change at follow-up. Moreover, 21 of 22 nevi (95%) with peripheral globules lost their rim of globules, and were most frequently classified as a homogeneous (n=9) or reticular (n=6) at follow-up. There was minimal shift in dermoscopic pattern from reticular to globular [1% (n=8)] or globular to reticular [4% (n=8)]. The net change in the proportion of dermoscopic patterns was +1.2% for complex, −5.8% for homogenous, +1.5% for globular, and +3.1% for reticular. Four percent of nevi disappeared during the study. No differences in rate of nevus disappearance were found by dermoscopic pattern (chi-square 5.03, p=0.17).

Table 1:

Dermoscopic pattern changes of nevi at baseline and follow-up (n=395 students).

Follow-up (11th grade)
		Complex	Homogenous	Globular	Reticular	Disappear	Total
Baseline (8th or 9th grade)	Complex	23	12	9	10	2	56
	Homogenous	10	358	54	123	31	576
	Globular	15	29	128	8	6	186
	Reticular	21	77	8	435	16	557
	Total	69	476	199	576	55	1,375

From 366 participants imaged at 5^th grade, 189 new back nevi were imaged in 8^th grade; of these, 44% were reticular, 21% globular, 19% homogenous, and 16% complex. New back nevi (n=218) were imaged in 112 of 395 participants imaged at 11^th grade follow-up: 44% were reticular, 23% homogenous, 17% complex, and 16% globular. New leg nevi (n=88) were also imaged in in 44 of 395 participants in 11^th grade: 54% were reticular, 43% homogenous, 2% complex, and none were globular.

The mean (SD) surface area for nevi at baseline (8^th/9^th grade) and follow-up (11^th grade) was 6.59±6.89 mm² and 7.70±8.79 mm², respectively. The average percent increase in surface area for back nevi was 7.33±0.57%/year, leg nevi was 4.67±0.84%/year, and all nevi was 6.75±0.48%/year. Nevi with a complex pattern demonstrated the most growth (+15.3±3.8%/year), followed by globular (+8.44±1.1%/year), reticular (+6.51±0.72%/year), and homogenous (+5.59±0.76%/year). The subgroup of complex nevi with peripheral globules showed the largest growth (+25.6±9.6%/year).

A convenience sample of 353 nevi (293 back and 60 leg nevi) from 263 students were used to assess nevus growth, while accounting for constitutional changes in the perilesional body surface area (i.e., overall growth of the participant), using high-resolution three-dimensional stereophotogrammetry-based imaging as previously described.(Chung et al., 2016) Since the mean area of all nevi increased by 8.3±0.8%/year and perilesional surface area increased by 6.2±0.2%/year, the calculated mean nevus-growth surface area increase was 2.1±0.8%/year.

Over 2–3 years of monitoring in late adolescence, the majority of nevi on the back and legs retained their dermoscopic pattern; this is consistent with previous data documenting the dermoscopic stability of individual back nevi in younger children.(Scope et al., 2008) For nevi that did exhibit a change, we observed the greatest net shift in dermoscopic pattern to the reticular type, a trend also seen in our same cohort from 5^th to 8^th grade.(Scope et al., 2011) In addition, nevi with reticular and globular patterns infrequently changed from one pattern to the other. A notable exception was nevi with a peripheral rim of globules, which are known to ultimately stop enlarging, lose their peripheral globules, and retain the dermoscopic pattern present at their center.(Bajaj et al., 2015) Of new nevi found at follow-up in the 8^th grade and 11^th grades, we found that as our cohort aged, fewer new nevi were globular, while new reticular nevi were most prevalent; this, along with the net shift toward a reticular pattern in changing existing nevi, appear to be contributing factors to the observation of reticular patterned nevi becoming the most common pattern in adulthood. A study examining the dermoscopic patterns of new nevi in adults is consistent with our findings.(Oliveria et al., 2013)

Significant changes in nevus surface area were observed. Size change is a sensitive but nonspecific marker of malignancy. A better understanding of normal nevus growth patterns may allow physicians to better differentiate abnormal growth and reduce unnecessary biopsies. For example, nevi with a peripheral rim of globules exhibited marked size changes. Additionally, up to 75% of surface area changes in nevi could be attributed to regional body growth. We suspect that accelerated growth in adolescence likely influences the observed growth of nevi during this age range. Hence, the ability to recognize disproportionate nevus size change may have clinical utility for identifying suspicious lesions.(Korotkov et al., 2015)

A predominance of males and Caucasians among participants limits the generalizability of our study. Imaging was also restricted to select nevi of the back and legs. Finally, participant dropout risks unrecognized selection bias. Further studies in larger cohorts with longer follow-up are needed to validate our observations.