Baseline Dermoscopic Patterns Predict Long-Term Changes in Nevus Diameter and in Dermoscopic Features

Abstract

Background:

When monitoring melanocytic neoplasms, the pattern of change may distinguish nevi from melanoma. Anticipating the growth dynamics of nevi based on their dermoscopic pattern is important to make this distinction.

Objective:

The primary aim was to examine the association between nevus dermoscopic pattern at baseline and diameter change during long-term monitoring. The secondary aim was to examine the association between nevus dermoscopic pattern at baseline and changes in both dermoscopic pattern and color during long-term monitoring.

Methods:

The study included high-risk adult patients that underwent ≥2 total-body photography (TBP) sessions, with at least 14 years' time gap between first and last sessions. Nevi on the torso, with available dermoscopic images, were included. New and disappearing nevi were defined as nevi not appearing on the first and last TBP, respectively. Nevus diameter and color were assessed on clinical images of first and last TBP images. Dermoscopic images were analyzed for dermoscopic patterns and structures at baseline and follow-up.

Results:

877 nevi from 101 patients were included. Mean follow-up time between TBPs and between dermoscopic images was 16.7 and 11.5 years, respectively. Most nevi were reticular or structureless at baseline, but new nevi had a higher frequency of peripheral globules and smudgy patterns. Peripheral globules and diffuse negative network patterns as well as regression structures were associated with nevus diameter growth. 30% and 15% of new and existing nevi, respectively, demonstrated dermoscopic pattern change, mainly transforming into reticular and structureless patterns.

Conclusions:

Among high-risk patients, nevi showing peripheral globules or negative network are more likely to grow in diameter during long-term monitoring. Most nevi retain their overall dermoscopic pattern and those that change, mostly transform into reticular or structureless patterns.

Graphical Abstract

In the study, the authors have identified dermoscopic patterns that are associated with different stages in the life cycle of nevi. New nevi were associated with peripheral globules and smudgy patterns. Growing nevi were associated with peripheral globules and negative network patterns. Stable nevi were associated with reticular, structureless and central globules with peripheral network patterns.

Introduction

When monitoring melanocytic neoplasms with total body photography (TBP), history of change points at the diagnosis of melanoma over nevus.¹ However, all acquired nevi had appeared and changed at some point in life. Therefore, among changing lesions, the pattern of change is important for differentiating nevi from melanoma.^2,3 Dermoscopy has improved our ability to anticipate change among certain nevi, for example, nevi with a dermoscopic pattern of peripheral globules are likely to grow during monitoring.⁴ Notably, most nevi encountered during routine surveillance do not display peripheral globules, yet growth of nevi during long-term TBP monitoring is common. We previously observed that among patients at high-risk for melanoma, during long-term TBP- monitoring with sequential TBPs over >14 years, about half of all torso nevi grew in diameter.⁵

In addition to growth, nevi also display changes in dermoscopic pattern. In children, longitudinal studies over 1-4 years have demonstrated that about a third of nevi display change in the overall dermoscopic pattern, mostly from globular and peripheral globular patterns, to reticular and structureless patterns.^6-8 Studies in adults suggest that a similar fraction of nevi may exhibit banal changes, including symmetrical network growth⁸ and loss of dermoscopic structures⁹. However, most studies investigating dermoscopic changes in nevi performed follow-up for relatively short periods.

The primary aim of the present study is to examine the association between nevus dermoscopic pattern and diameter change during long-term monitoring of patients at high-risk of melanoma. The secondary aim is to examine the association between nevus dermoscopic pattern at baseline and dermoscopic pattern and color changes during monitoring.

Methods

Ethics

The study was approved by the Institutional Review Board at MSKCC under protocols #99-099 and #17-078.

Patients and nevi

This retrospective study included adult patients (age≥18 years) that underwent ≥2 total-body photography (TBP) sessions, with at least 14 years' time gap between first and last sessions. Notably, some patients underwent only two TBP sessions, while others underwent multiple TBP sessions. Demographic information and personal and family history of skin cancer were retrieved from the patients’ clinical charts. Included nevi were (1) located on the torso; the borders of the torso were upper anterior – clavicles, upper posterior - tip of shoulders, lower anterior and posterior borders - iliac crests, (2) with diameter ≥4mm (at any time point during follow-up) to avoid misclassification of lentigines as nevi,¹⁰ (3) with available dermoscopic image. Notably, there were no strict selection criteria for the inclusion of nevi for dermoscopic imaging, and not all torso nevi were dermoscopically imaged. Selection was based on real-time, clinical decision of the clinicians or medical photographer. As part of the study, about half of participants were randomly selected to have all their previously dermoscopically imaged nevi reimaged during their last TBP. Nevi that were excised during follow-up were excluded from the study.

Clinical image analysis

Overview clinical images of the torso, from the first and last TBP sessions were compared. The results of this clinical data analysis were previously reported.⁵ In brief, 'existing nevi' were defined as those present at first and last TBP, 'new nevi' as those absent at first TBP and present at last TBP, and 'disappearing nevi' as those present at first TBP and absent at last TBP. For each included nevus, we measured the largest diameter in millimeters from the TBP images. We denoted size change if the diameter at last image was different by at least 25% from the first diameter, otherwise the nevus was denoted as 'stable' in diameter.¹⁰ Additionally, we compared nevus color between first and last TBP sessions. This comparison was qualitative, based on clinical inspection of nevi from TBP images, and categorized as exhibiting no notable change, becoming notably lighter, or notably darker in color.

Dermoscopic image analysis

Two researchers (A.A.M and O.R.) performed a consensus reading and annotated the dermosocpic images at first and last dermoscopic imaging sessions for each included nevus. The researchers were blinded to all data except for the images. Images were analyzed for the following parameters: (1) overall dermoscopic pattern (Fig. 1): reticular; globular; reticuloglobular; smudgy (defined as a diffuse pattern that is not homogenous in pigmentation, displaying non-uniform pigmented areas or structures without clear demarcation, and lacking well-defined network or globules; this definition was coined by the authors and has not been used in prior publications); central structureless with peripheral network; central globules with peripheral structureless; central globules with peripheral network; central structureless with peripheral globules; central network with peripheral globules; structureless-homogenous; two component pattern; multicomponent; and diffuse negative network, (2) presence or absence of the following dermoscopy structures at baseline: network, typical; network, atypical; network, patchy; smudgy network-like; negative network; globules; dots; hypo-pigmented structureless area; hyper-pigmented structureless area; milia; regression structures seen as gray granularity; (3) changes observed in dermoscopic structures when performing side-by-side comparison of first and last dermoscopic images: network appeared; network faded; network became patchier; network became more atypical; negative network appeared; globules appeared; globules faded or disappeared; structureless area appeared; structureless area became lighter; shiny white structures appeared; halo appeared; regression structures as gray granularity appeared.

Figure-1. Images of selected dermoscopic patterns.

A. Smudgy pattern of a nevus at baseline. B. the same nevus eight years later demonstrating reticular pattern. C. diffuse negative network nevus at baseline. D. The same nevus one year later demonstrating growth in diameter. E. Structureless pattern nevus at baseline (to be compared with smudgy pattern in image A). F. peripheral globules nevus at baseline. G The same nevus one year later. H. The same nevus seven years later.

Statistical analysis

Baseline characteristics and categorical variables were summarized using descriptive statistics. To evaluate the association and likelihood of change in each baseline pattern a Chi-Squared test was performed. Further post hoc tests were conducted utilizing the Bonferonni correction method. As groups of nevi are nested within patients, and accounting for the lack of independence between nevi, nevus-level outcomes were analyzed using a generalized linear mixed model (GLMM) approach. The models were used to analyze the rate of change in nevi diameter across different patterns and structures. Similarly, the models were used to analyze the lightening or darkening of nevi across different patterns and structures, and adjusted odds ratios were derived. Multivariate models contained covariates for patients’ sex, age at first TBP and the number of years between the first and last TBP and between the first TBP and baseline dermoscopy image. A significance level of P<0.05 was used. The data were analyzed using SPSS, version 26.0 for Windows (SPSS, Inc).

Results

One-hundred and one patients were included in the analysis, of which 52% were male. Mean age at first TBP was 39.6 (±9.7 SD) years with a mean follow-up duration between first and last TBP of 16.7 (±1.5 SD) years. Of the patients, 67% were diagnosed with melanoma, either before undergoing their first TBP or after. A total of 877 nevi were included, of which 749 (85.4%) were classified as existing nevi, 122 (13.9%) as new nevi and 6 (0.7%) as disappearing nevi. Between first and last TBP, of 749 existing nevi, 335 (45%) grew by at least 25% in diameter, with a mean diameter increase of 1.1 (±1.9 SD) millimeters, and 42 (6%) nevi shrank by at least 25% in diameter, with a mean diameter decrease of 2.6 (±1.4 SD) millimeters. 372 (50%) nevi remained stable in their diameter. In addition, 136 (16%) nevi became notably lighter in color and 39 (4%) became notably darker.

Baseline dermoscopic features

Fifty-six percent (489) of nevi had their baseline dermoscopy image captured within one year after the first TBP, an additional 30% had theirs taken between 1 to 15 years post-first TBP, and the remainder between 16 to 20 years post-first TBP. The predominant baseline dermoscopic pattern among the 877 nevi was reticular (59%), followed by structureless (12%) and by central structureless with peripheral network pattern (10%, Fig. 2). Ninety percent of nevi exhibiting a pigment network demonstrated either a typical or patchy network, 4% displayed an atypical network, 4% – smudgy network-like, and 1% – negative network. Regression structures such as gray granularity were observed in 1.2% of all reticular nevi. The only other pattern showing regression structures as gray granularity was the multicomponent pattern, seen in 4.2% of these nevi.

Figure-2 – Baseline overall pattern of 877 nevi included in the study.

Of 122 new nevi, the majority (64.7%) exhibited a reticular pattern. The new reticular nevi accounted for 15.3% of all reticular nevi. In contrast, the new nevi with a central network and peripheral globules (n=11) accounted for 100% of nevi with that pattern, and new smudgy-patterned nevi (n=6) – for 21% of all nevi with that pattern (Fig. 3).

Figure-3 – Baseline overall patterns of new, stable and disappearing nevi included in the study.

Only six of the 877 nevi (0.7%) disappeared over follow-up – two were reticular and four were structureless (Fig. 3).

Clinical Follow-up: nevus diameter and color change

New and disappearing nevi were considered as zero milimieters in diameter at first and last TBP, respectively. In multivariate analysis, overall dermoscopic patterns that were significantly associated with nevus diameter growth included peripheral globules (P<0.001) and diffuse negative network (P<0.001). Conversely, baseline dermoscopic patterns least associated with nevus diameter growth were central globules with peripheral network, structureless, and two-component patterns (Table 1). Dermoscopic structures that were associated with nevus diameter growth, irrespective of the baseline overall dermoscopic pattern, included negative network, typical network and regression structures as gray granularity (Table 2). Notably, only 56% of baseline dermoscopic images were taken within a year post-first TBP. We therefore conducted a separate analysis on these nevi (data not shown), which yielded similar results, except for the typical network's lack of association with nevus growth. The multivariate analysis shown in tables 1&2 were controlled for number of years between first and last TBP, number of years between first TBP and baseline dermoscopy image, sex and patients’ age at first TBP.

Table 1.

Multivariate analysis for nevus diameter change association with different baseline overall dermoscopy pattern

Overall pattern (n = 877)	Overall change in diameter in milimeters (mean ± SD)	Growing nevi* N (%)	Shrinking nevi* N (%)	Stable nevi* N (%)	Rate of change of diameter difference** B (95% CI)	P-value
Reticular (n = 518)	2 ± 2.59	301 (58)	21 (4)	196 (38)	0 (referent)	N/A
Globular (n = 49)	1.34 ± 1.74	19 (39)	1 (2)	29 (59)	−0.16 (−0.82 – 0.49)	0.62
Reticuloglobular (n = 5)	3.1 ± 2.63	4 (80)	0 (0)	1 (20)	0.37 (−2.2 – 2.94)	0.78
Smudgy (n = 28)	1.55 ± 3.34	14 (50)	6 (21)	8 (29)	−0.1 (−1.23 – 1.04)	0.87
Central structureless and peripheral network (n = 85)	1.5 ± 2.57	39 (46)	5 (6)	41 (48)	−0.37 (−0.9 – 0.17)	0.18
Central globules and peripheral structureless (n = 2)	1 ± 0.42	0 (0)	0 (0)	2 (100)	0.12 (−0.81 – 1.06)	0.8
Central globules and peripheral network (n = 32)	1.04 ± 1.54	14 (44)	1 (3)	17 (53)	−0.76 (−1.34 – −0.18)	0.01
Central structureless and peripheral globules (n = 10)	1.46 ± 2.9	4 (40)	2 (20)	4 (40)	−0.44 (−2.15 – 1.27)	0.61
Structureless (n = 103)	0.57 ± 2.22	31 (30)	11 (11)	61 (59)	−0.68 (−1.24 – −0.12)	0.02
Central network and peripheral globules (n = 11)	7.45 ± 2.23	11 (100)	0 (0)	0 (0)	4.35 (3.5 – 5.19)	<0.001
Two component (n = 11)	0.84 ± 1.37	2 (29)	0 (0)	5 (71)	−0.6 (−1.5 – 0.31)	0.19
Multicomponent (n = 24)	2.69 ± 2.85	15 (63)	1 (4)	8 (33)	1.22 (0.16 – 2.28)	0.02
Negative network (n = 3)	8.53 ± 1.71	3 (100)	0 (0)	0 (0)	6.51 (5.1 – 7.92)	<0.001

^*A nevus was considered growing or shrinking if it’s diameter increased or decreased by at least 25%, respectively. Nevi with a change of less than 25% in their diameter were considered stable.

^**In this generalized linear mixed model, reticular pattern served as reference. Values represent the deviation from reference diameter change. Positive values represent increased diameter growth compared to reference, whereas negative values represent decreased diameter growth compared to reference. Analysis controlled for number of years between first and last TBP, number of years between first TBP and baseline dermoscopy image, sex and patients’ age at first TBP.

Table 2.

Multivariate analysis for association between nevus diameter growth during follow-up and the presence of dermoscopic structures at baseline

Dermoscopic Structure (n = 877)	Overall change in diameter in milimeters (mean ± SD)	Growing nevi* N (%)	Shrinking nevi* N (%)	Stable nevi* N (%)	Rate of change of diameter difference** B (95% CI)	P-value
Typical network (n = 593)	1.93 ± 2.56	331 (56)	24 (4)	238 (40)	0.43 (−0.04 – 0.91)	0.08
Atypical network (n = 28)	1.99 ± 2.17	19 (68)	1 (4)	8 (29)	−0.04 (−0.8 – 0.71)	0.91
Patchy network (n = 57)	2.42 ± 3.42	31 (54)	3 (5)	23 (40)	0.55 (−0.34 – 1.43)	0.23
Negative network (n = 8)	6.68 ± 2.39	8 (100)	0 (0)	0 (0)	4.48 (2.73 – 6.22)	<0.001
Smudgy area (n = 29)	1.72 ± 3.4	15 (52)	6 (21)	8 (28)	0.39 (−0.82 – 1.59)	0.53
Globules (n = 137)	2 ± 2.68	69 (50)	6 (4)	62 (45)	0.24 (−0.25 – 0.73)	0.33
Dots (n = 14)	2.73 ± 3.21	7 (50)	0 (0)	7 (50)	0.53 (−1.18 – 2.24)	0.54
hypopigmented structureless area (n = 75)	1.62 ± 2.51	36 (48)	5 (7)	34 (45)	−0.28 (−0.79 – 0.23)	0.28
Hyperpigmented structureless Area (n = 158)	1.03 ± 2.53	57 (36)	14 (9)	87 (55)	−0.33 (−0.77 – 0.12)	0.15
Milia (n = 41)	1.6 ± 2.24	19 (46)	1 (2)	21 (51)	0.44 (−0.17 – 1.04)	0.16
Regression (n = 7)	2.06 ± 1.02	6 (86)	0 (0)	1 (14)	1.12(0.3 – 1.95)	0.01

^*A nevus was considered growing or shrinking if it’s diameter increased or decreased by at least 25%, respectively. Nevi with a change of less than 25% in their diameter were considered stable.

^**In this generalized linear mixed model ,the nevus diameter change associated with each dermoscopic structure was compared to the diameter change in the absence of the structure, which served as reference. Values represent the deviation from reference diameter change. Positive values represent increased diameter growth compared to reference, whereas negative values represent decreased diameter growth compared to reference. Analysis controlled for number of years between first and last TBP, number of years between first TBP and baseline dermoscopy image, sex and patients’ age at first TBP.

We performed a multivariate analysis to investigate clinical color changes in nevi, with the overall dermoscopic pattern as the independent variable. Only baseline structureless pattern showed significant correlation with color change during follow-up; these structureless nevi tended to become lighter during follow-up (P<0.001; data not shown). Baseline dermoscopic structures that were associated with the lightening of nevi were typical network, hypo-pigmented structureless area, globules and milia. No association was found between baseline dermoscopic patterns or structures and the darkening of nevus color, however, only few nevi displayed darkening of color during follow-up.

Dermoscopic follow-up: dermoscopic patterns and structures change

Of the 877 nevi, 496 nevi (56.6%) underwent dermoscopic imaging at multiple time points, with an average follow-up period of 11.5 (± 8.5) years between baseline and final dermoscopic imaging sessions. Of the 496 dermoscopically-monitored nevi, 412 (83%) were existing nevi and 84 (17%) were new.

76 (15.3%) of the 496 dermoscopically followed-up nevi demonstrated change in their overall pattern during follow-up (Table 3). Of the 419 existing nevi with dermoscopic follow-up, 60 nevi (14.3%) demonstrated alterations in their dermoscopic pattern. Of the 77 new nevi that had dermoscopic follow-up, 23 (30%) demonstrated overall pattern change. Both existing and new nevi predominantly transitioned into reticular or structureless patterns. The baseline pattern least likely to change was reticular. Conversely, baseline patterns that were most likely to change were central network with peripheral globules and structureless with peripheral globules patterns, which transformed into reticular and structureless patterns, respectively. 47% of nevi with smudgy pattern exhibited change, primarily transforming into central structureless and peripheral network patterned nevi. Multicomponent nevi most commonly transitioned into reticular nevi.

Table 3.

Nevi baseline overall pattern versus overall pattern at last dermoscopy follow-up

	Pattern at last follow-up, N (row%)
Baseline pattern:	1	2	3	4	5	6	7	8	9	10	11	12	13
1 (n = 272)	256 (94)	0 (0)	0 (0)	1 (0.4)	4 (1.5)	0 (0)	0 (0)	0 (0)	10 (3.7)	0 (0)	1 (0.4)	0 (0)	0 (0)
2 (n = 27)	1 (3.7)	22 (81.5)	0 (0)	1 (3.7)	0 (0)	1 (3.7)	0 (0)	0 (0)	2 (7.4)	0 (0)	0 (0)	0 (0)	0 (0)
3 (n = 4)	3 (75)	0 (0)	1 (25)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
4 (n = 19)	2 (10.5)	2 (10.5)	0 (0)	10 (52.6)	3 (15.8)	0 (0)	0 (0)	0 (0)	1 (5.3)	0 (0)	0 (0)	1 (5.3)	0 (0)
5 (n = 67)	3 (4.5)	0 (0)	0 (0)	0 (0)	59 (88.1)	0 (0)	0 (0)	0 (0)	3 (4.5)	0 (0)	1 (1.5)	1 (1.5)	0 (0)
6 (n = 2)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	2 (100)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
7 (n = 22)	1 (4.5)	1 (4.5)	0 (0)	0 (0)	1 (4.5)	0 (0)	18 (81.8)	0 (0)	1 (4.5)	0 (0)	0 (0)	0 (0)	0 (0)
8 (n = 8)	2 (25)	1 (12.5)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	2 (25)	2 (25)	1 (12.5)	0 (0)	0 (0)	0 (0)
9 (n = 38)	2 (5.3)	0 (0)	0 (0)	1 (2.6)	4 (10.5)	0 (0)	0 (0)	0 (0)	31 (81.6)	0 (0)	0 (0)	0 (0)	0 (0)
10 (n = 11)	7 (63.6)	1 (9.1)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	3 (27.3)	0 (0)	0 (0)	0 (0)
11 (n = 6)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	2 (33.3)	0 (0)	4 (66.7)	0 (0)	0 (0)
12 (n = 18)	3 (16.7)	0 (0)	0 (0)	1 (5.6)	0 (0)	0 (0)	0 (0)	0 (0)	2 (11.1)	0 (0)	0 (0)	12 (66.7)	0 (0)
13 (n = 2)	0 (0)	0 (0)	0 (0)	1 (50)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	1 (50)	0 (0)
Total (n = 496)	280 (56.4)	27 (5.4)	1 (0.2)	15 3)	71 (14.3)	3 (0.6)	18 (3.6)	2 (0.4)	54 (10.9)	4 (0.8)	6 (1.2)	15 (3)	0 (0)

^1-Reticular

^2-Globular

^3-Reticuloglobular

^4-Smudgy

^5-Central structureless and peripheral network

^6-Central globules and peripheral structureless

^7-Central globules and peripheral network

^8-Central structureless and peripheral globules

^9-Structureless

^10-Central network and peripheral globules

^11-Two-component

^12-Multicomponent

^13-Negative network

Among new and existing nevi, the most frequent dermoscopic structural change was fading of pigment network – observed in 23.6% of all nevi, followed by the disappearance of globules – in 7.5% of all nevi (Table 4).

Table 4.

Changes in specific dermoscopy structures over follow-up

Network change	n (% out of 496)
Network appeared	25 (5)
Network patchier	25 (5.2)
Network faded	117 (23.6)
Negative network appeared	2 (0.4)
Network more atypical	13 (2.6)
Globules faded	37 (7.5)
Globules appeared	9 (1.8)
Structureless area faded	8 (1.6)
Structureless area appeared	2 (0.4)
Shiny-white structures appeared	1 (0.2)
Halo appeared	3 (0.6)
Regression appeared as grey granularity	4 (0.8)

Discussion

We report on the association of dermoscopic patterns with long-term dynamics of nevi on the torso, both clinical and dermoscopic. In line with prior publications,^11,12 the majority (74%) of nevi included in this study exhibited typical or patchy pigment networks, with infrequent instances of regression structures as gray granularity (0.7%). Fourteen percent (122) of nevi were clinically new. Most (65%) of the new nevi displayed a reticular pattern, and the proportion of reticular nevi among these new nevi was similar to their proportion among existing nevi. However, peripheral globular-, and smudgy- patterns were more frequently observed among the new nevi than among the existing nevi. Most existing nevi displayed limited size change, with an average change in diameter of 1.3 millimeters over 16.7 years of follow-up.

This study also elucidates the long-term dermoscopic changes, or lack thereof, observed in longitudinally monitored nevi on the torso over an average dermoscopic follow-up of 11.5 years. in alignment with findings from shorter follow-up studies,^6,7,12 we found that 85% (420) of the nevi remained stable in the overall dermoscopic pattern.

During the "active", dynamic or changing part of the curve in the life cycle of nevi, several dermoscopic pattern and structures were predictors of change. First, in line with previous studies,^2,13,14 the presence of peripheral globules, seen as central network with peripheral globules or as structureless with peripheral globules, predicted nevus diameter growth. As nevus diameter stabilizes, these peripheral globules tend to diminish and the pattern becomes diffuse network or structureless, respectively. In contrast, nevi with a pattern of central globules and a peripheral network were likely to remain stable in size and pattern. We reason that there is a fundamental difference in the globules between changing and stable nevi – the globules in changing nevi correlate with junctional nests of melanocytes and the globules in stable nevi correlate with dermal nests.¹⁵ Second, a pattern which we termed "smudgy” was also associated with change, being prevalent among new nevi. A third of nevi with smudgy pattern were likely to change into distinct reticular- or globular- pattern. Smudgy pattern is the blurring and indistinct appearance of dermoscopic structures. A smudgy appearance may be related to reduced contrast between structures in dermoscopic images, either due to lightly-colored pigmented structures that darken over time or to lack of sufficient spacing between pigmented structures during active growth. Third, a dermoscopic feature associated with nevus diameter growth was the presence of a negative network. Notably, negative network is also prevalent among Spitz/Reed nevi, which are frequently encountered during active growth.¹⁶ Finally, although a bit counterintuitive, the presence of regression as gray granularity was associated with nevus enlargement, which may correspond to an immune response implicated during active nevus growth.

The reticular and structureless patters were most common during the stable and senescent portion of the curve in the life cycle of nevi. Among nevi that underwent changes in their overall pattern, most transitioned into a reticular or a structureless patterns suggesting senescence. The most common dermoscopic structural alterations involved the fading of pigment networks and of globules, possibly accounting for the increasing frequency of structureless patterned nevi over time. Nevi that already exhibited a reticular or structureless pattern at baseline were more likely to retain their overall pattern. Structureless nevi were also associated with a stable diameter and with fading of color over time. Finally, nevi that involuted were either reticular or structureless at baseline.

In addition to describing the changes observed during the follow-up, it is crucial to highlight the alterations that were rarely seen among the nevi in this study. Firstly, no nevus developed new peripheral globules during the follow-up period; all peripheral globules were already present in the initial dermoscopic images and tended to fade over time. Therefore, beyond the asymmetrical distribution and variability in shape and size⁴, the appearance of new peripheral globules should also raise suspicion for melanoma. Secondly, in the current study, no nevus transitioned into a diffuse negative network pattern, and only two nevi (0.4%) developed a partial negative network. Therefore, in addition to irregular distribution and peripheral location of a negative network¹⁷, the emergence of a new negative network in a lesion should also raise suspicion for melanoma. Finally, it is known that nevi may exhibit regression as gray granularity, which is not considered a sign of melanoma if it involves up to 10% of the lesion’s surface.¹⁸ In the current study, only four nevi (0.8%) developed signs of regression, indicating that this change is rare in benign nevi and should be treated with caution.

Our research presents several limitations. First, the study data comes from a single medical center. Additionally, it is a retrospective study focusing on high-risk patients, and some findings may differ among low-risk patients. Different camera types and settings may impact color analysis, and changes in nevi might also be influenced by changes in patient weight and height, which were not accounted for. All nevi were preselected for digital dermoscopic imaging at baseline by either the clinician or the photographer. This selection process could potentially restrict the applicability of our results to atypical or outlier nevi. To mitigate this selection bias, in half of the cohort, we reimaged all previously imaged nevi during the follow-up session. Not all nevi had their baseline clinical and dermoscopic imaging at the same time. Therefore, we controlled all multivariate analyses for the time elapsed between TBP and dermoscopic imaging. None of the lesions included in the study underwent a biopsy. Instead, their benign nature was inferred from clinical and dermoscopic parameters and from prolonged monitoring of patterns during the study follow-up.

In conclusion, during extended dermoscopic and clinical follow-up, most nevi on the torso of high-risk patients demonstrated limited growth in diameter and retain their overall dermoscopic pattern. The presence of peripheral globules, a smudgy network-like structure or a negative network likely represent nevi that have not yet entered senescence. In contrast, the presence of diffuse reticular or structureless pattern is probably indicative of a stable or senescent nevus that it is nearing or has reached the conclusion of its growth phase.

Key Points.

Why was the study undertaken?

To investigate the dermoscopic patterns associated with new nevi, stable nevi, and growing nevi during long-term monitoring of patients at high-risk for melanoma.

What does this study add?

Nevi with peripheral globules, negative network, or regression structures predict nevus diameter growth. The pattern associated with new nevi included the peripheral globular pattern and smudgy pattern, which is being introduced as a new dermoscopic pattern.

What are the implications of this study for disease understanding and/or clinical care?

The observed dermoscopic patterns can help dermatologists anticipate growth of nevi, which may help in differentiating nevi from melanoma.

Data availability statement:

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.