1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. References


Naevogenesis is a process known to occur throughout life. To date, investigators have made conclusions about new naevi in adults based on results of cross-sectional studies.


To determine the incidence of new naevus development in adults and to describe the dermoscopic morphology of new naevi.


A cohort of 182 patients seen at the outpatient dermatology clinic at Memorial Sloan-Kettering Cancer Center between 2000 and 2009 was evaluated with baseline total body photographs. The patients were aged 17 years or older and had presented for routine follow-up surveillance examination at least 3 months after baseline total body photographs. The number of new naevi and the dermoscopic morphology of these naevi were recorded.


Of the 182 patients evaluated, 50 (27%) developed at least one new naevus during follow-up. The incidence of new naevi was 202 per 1000 person-years of follow-up. The most common types of naevi were reticular (47·1%), followed by the homogeneous (22·1%) and complex (reticuloglobular) patterns (15·4%).


Our results provide support for the theory that there are two distinct pathways of naevogenesis, a dynamic process occurring throughout life. This study demonstrates that the predominant dermoscopic morphology of newly acquired naevi in adults is reticular.

Based on cross-sectional studies, it has been shown that the formation of new naevi starts during the first years of life and increases during adolescence and early adulthood, resulting in a maximum naevus count by approximately 40 years of age. It has been postulated that, later in life, development of new naevi diminishes and existing naevi begin to involute.[1, 2] However, results from recent longitudinal studies suggest that naevus development is a dynamic process that continues throughout the lifetime of an individual, with the rate of new naevus formation being influenced by exogenous factors such as ultraviolet (UV) exposure, and endogenous factors such as skin phototype and other hereditary factors.[3-6] These new insights provide an alternate explanation to the observations made in cross-sectional studies. That is, differences in the development of naevi observed over different ages may simply be a reflection of the relative rates of naevus formation and involution. While in youth proportionally more naevi are developing than regressing, in older ages proportionally more naevi are regressing than developing.

The observed differences in the dermoscopic morphology of congenital vs. acquired naevi[4] and youth vs. adulthood[5] have led to the concept of a dual pathway for naevogenesis, with an endogenous pathway (i.e. not directly influenced by UV exposure) responsible for naevi with a predominantly globular morphology, and an exogenous pathway (i.e. directly influenced by UV exposure) responsible for naevi with a predominantly reticular morphology.[6-9] While it is true that most naevi in youth are globular and most naevi in adulthood are reticular, it remains unknown whether this is due to globular naevi transforming into reticular naevi over time or whether new naevi in adults begin life manifesting a predominantly reticular pattern. Results from the Study of Naevi in Children (SONIC) have shown that it is extremely unlikely for globular naevi to transform to reticular naevi in children during the 5th to 8th grades (age 10–14 years); 69% of the naevi observed during a 3-year period had no change in dermoscopic pattern and only 3–4% showed crossover from reticular to globular or globular to reticular pattern.[10] Thus, if transformation of globular to reticular naevi is not common, it stands to reason that new naevi developing in adults must be manifesting a predominantly reticular pattern from inception.

At the Memorial Sloan-Kettering Cancer Center (MSK), the routine acquisition of high-resolution standardized total body photographs (TBPs) is offered to patients at risk for developing skin cancer. This photodocumentation allows clinicians to identify new and changing lesions. Using TBP allowed us to determine the incidence of new naevus development and assess the dermoscopic patterns manifested by these naevi in a cohort of high-risk adult patients.

Patients and methods

  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. References


We identified a sample of 182 patients aged 17 years or older who presented for follow-up skin cancer surveillance and had had TBP at least 3 months earlier, from the outpatient dermatology clinic at MSK between 2000 and 2009. Patients with TBP are considered individuals at risk for developing melanoma, with many having three or more atypical naevi, a high naevus count and/or a personal or family history of melanoma. These patients are monitored every 3–12 months depending on their presumed risk of melanoma.

Data collection

As part of the monitoring of high-risk patients, TBP is performed routinely at MSK.[3] The images are maintained in the MIRROR™ Body Mapping image system, using DermaGraphiX software (Canfield Imaging Systems, Fairfield, NJ, U.S.A.). A Hasselblad camera with a PhaseOne P65 digital back (Hasselblad, Gothenburg, Sweden) is used to capture TBPs. Supplementary clinical and dermoscopic images are acquired with a Nikon D80/90 camera (Nikon U.S.A. Inc., Melville, NY, U.S.A.) with an Epiflash attachment (Canfield Imaging Systems). Patients presenting for follow-up are evaluated by comparison with their baseline images. We utilized these data for the eligible recruited patients.

The attending dermatologist (A.A.M.) determined the individual dermoscopic naevus patterns of the four largest naevi for each study patient, which did not include the new naevus. In addition, the overall predominant dermoscopic naevus patterns of 10 naevi were described. The 10 naevi of interest were those in close proximity to the new naevus, or if no new naevus had developed the 10 largest naevi were used.

Individual naevus dermoscopic typing was used to classify each naevus as having a simple type or complex type. Naevi with a simple dermoscopic type included those with a reticular, globular or homogeneous pattern; the complex-type naevi were defined as having a combination of two patterns (i.e. reticuloglobular; Fig. 1). For each patient, the predominant dermoscopic pattern was categorized as simple or complex. The simple predominant pattern included naevi that had the same dermoscopic pattern (i.e. reticular, globular or homogeneous), whereas a complex predominant pattern included multiple naevi patterns expressed by the same patient, including reticular, globular and/or homogeneous. Each patient was assigned to one of four dermoscopic categories, firstly by their individual naevus type and then by their overall predominant naevus pattern (i.e. simple naevus type–simple predominant pattern, simple naevus type–complex predominant pattern, complex naevus type–simple predominant pattern and complex naevus type–complex predominant pattern; Table 1).

Table 1. Classification of dermoscopic patterns
Individual naevus typeOverall predominant pattern
 Simple predominant patternComplex predominant pattern
Simple naevus typeSimple naevus type–simple predominant patternSimple naevus type–complex predominant pattern
Complex naevus typeComplex naevus type–simple predominant patternComplex naevus type–complex predominant pattern

Figure 1. Individual naevus dermoscopic patterns recorded for the patients: (a) simple: reticular pattern; (b) simple: globular pattern; (c) simple: homogeneous pattern; (d) complex: reticuloglobular pattern.

Download figure to PowerPoint

Demographic information was obtained for the study patients and included age, sex, date of last total body skin examination, date of TBP, total body naevus count at last skin examination and relevant past medical history. This included history of melanoma, basal cell carcinoma and squamous cell carcinoma, actinic damage and high-risk naevus phenotype (HRNP). Total body naevus count was categorized into three groups (≤ 50, 51–100 or > 100 naevi) and the predominant dermoscopic naevus pattern was categorized. HRNP was defined as patients having at least three clinically atypical naevi. HRNP was further classified into mild, moderate or severe, with total naevus counts of < 50, 50–100 and > 100, respectively. The anatomical location, size and dermoscopic pattern were recorded for all new naevi > 2 mm.

Statistical analysis

Descriptive statistics were used to characterize demographic, clinical and dermoscopic variables. The number and proportion of patients with new naevi were calculated and stratified by age and total body naevus count. The incidence of new naevi during the follow-up period was estimated as the total number of new naevi divided by the total number of years of follow-up. Zero-inflated Poisson regression was used to assess the relationship between demographic, clinical and dermoscopic variables and the study outcome (naevus count). Statistical analyses were performed using Stata 10.1 (StataCorp, College Station, TX, U.S.A.).


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. References

In total 182 patients were identified and recruited to participate in the study. The demographic and clinical characteristics of the study population are presented in Table 2. The median age of the patients was 45·5 years. A history of nonmelanoma skin cancer and photodamaged skin was not a hallmark of the patients. Some 21% reported a history of nonmelanoma skin cancer and 33% reported actinic skin damage. A majority of the sample had more than 50 naevi (76%), dysplastic naevi (83%) and/or a previous history of melanoma (73%). The median length of time between baseline TBP and the most recent follow-up was 24·5 months (range 1–127 months). In total 109 new naevi developed over the course of follow-up. The incidence of new naevi was 202 per 1000 person-years of follow-up. Of the 182 patients evaluated, 50 (27%) developed at least one new naevus between the TBP session and the most recent follow-up (median time for patients with no new naevi was 20 months and for patients with at least one new naevus 49 months). Of those 50 patients, 28 (56%) developed one new naevus, nine (18%) developed two new naevi, four (8%) developed three new naevi and nine developed four or more new naevi (Fig. 2). Of the 109 new naevi documented, 18 were biopsied, revealing three new primary melanomas, one vascular metastatic lesion, 10 dysplastic naevi, two junctional naevi and two compound naevi.

Table 2. Demographic and clinical characteristics of the patients (= 182)
 PatientsPresence of new naevi, n
n %
  1. BCC, basal cell carcinoma; SCC, squamous cell carcinoma. Some categories do not total 100% due to missing data.

Age (years)
< 30179·311
≥ 7063·33
History of melanoma
History of BCC/SCC
Actinic damaged skin
Naevus count
≤ 50 naevi4224·19
51–100 naevi5833·319
> 100 naevi7442·520
High-risk naevus phenotype

Figure 2. Histogram of the number of new naevi per patient in the cohort (= 109).

Download figure to PowerPoint

Table 3 presents the association between the patient and clinical characteristics and development of new naevi. The incidence rate ratio (IRR) for naevus development in the youngest quartile compared with the oldest quartile was 4·3 [95% confidence interval (CI) 2·3–8·1]. Interestingly, the more naevi that a participant had, the less likely they were to have a new naevus [IRR 0·47 (95% CI 0·25–0·89) for patients with > 100 naevi compared with < 50 naevi]. Participants with HRNP had an IRR of 2·7 (95% CI 1·0–7·1) vs. those with no HRNP.

Table 3. The association between total new naevi and total body naevus count, age, sex, melanoma history and presence of high-risk naevus phenotype (HRNP)
 Adjusted IRR95% CIP-value
  1. IRR, incidence rate ratio; CI, confidence interval.

Total body naevus counts
≤ 501·00
> 1000·470·25–0·890·02
Age (quartiles)
Q1 – youngest4·332·31–8·11< 0·001
Q4 – oldest1·00
Melanoma history
Presence of HRNP

The most common dermoscopic pattern observed was the simple naevus type–simple predominant pattern, comprising 56% of the lesions, followed by the complex naevus type–simple predominant pattern and the complex naevus type–complex predominant pattern. The reticular pattern was the most common dermoscopic type of new naevus (47·1%), followed by the homogeneous (22·1%) and complex naevus types (15·4%). Most new naevi presented on the trunk (82·7%; Table 4). No new globular naevi were seen in patients over the age of 40 years, while reticular new naevi were present in all age ranges.

Table 4. Anatomical location of new naevi, by dermoscopic pattern (50 patients, 104 new naevi)a
Dermoscopic patternLocation
Extremity, n (%)Head/neck, n (%)Trunk, n (%)Total, n (%)
  1. a

    Dermoscopic pattern was not available for two of the 109 new naevi, and location was not available for three.

Complex1 (6)0 (0)15 (94)16 (15·4)
Globular0 (0)0 (0)10 (100)10 (9·6)
Homogeneous2 (9)0 (0)21 (91)23 (22·1)
Other0 (0)0 (0)2 (100)2 (1·9)
Peripheral globules2 (50)0 (0)2 (50)4 (3·8)
Reticular12 (24)1 (2)36 (73)49 (47·1)
Total17 (16·3)1 (0·9)86 (82·7)104 (100)


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. References

Our results support the theory that naevogenesis is a dynamic process occurring throughout life, with new naevi continuing to develop into adulthood. Of the 182 patients evaluated, 50 (27%) developed at least one new naevus during follow-up. Thirty-five of the 50 patients who developed new naevi were under the age of 50 years, which is not surprising given that it is well known that the count of naevi increases until the fifth decade of life.[11] We observed an incidence rate of new naevus development of 202 per 1000 person-years. In a similar study, Banky et al.[12] reported an incidence rate of 277 per 1000 person-years. The lower rate observed in our data may be attributable to differences in the study population and the methods used to ascertain new naevi. One major difference is the geographical location of the two study populations, with Banky et al. examining an Australian population and this study conducted in New York. The study population of Banky et al. included a higher proportion of patients with > 100 naevi at baseline (82·5%), compared with the 40·7% in our study. Their study included patients aged 16–74 years, with a median age of 38 years, whereas our patients were aged 16·8–77·0 years, with a median age of 45·5 years. Additionally, the patients of Banky et al. had a median length of follow-up of 34 months, which is longer than our mean follow-up of 24·5 months. Our results indicate that longer follow-up leads to the detection of more new naevi, as the average time to follow-up was 20 months in the group with no new naevi compared with 49 months in those developing at least one new naevus. Both studies do corroborate the finding that new naevus development in adults is not an entirely uncommon event. The present study provides the additional information that most of the new naevi developing in adults manifest a reticular pattern, which is in sharp contrast to the observation that most naevi in children manifest a globular pattern.

One surprising result of this study was that those with fewer naevi were found to be more likely to develop a new naevus than those with more naevi. This could be because in individuals with numerous naevi, those naevi destined to manifest throughout the course of an individual's lifetime have already appeared.[13] Another possible explanation for this is observer limitations, despite the aid of TBP. A singular new naevus may be more easily detected on a patient with fewer naevi, where it is contrasted by a less distracting background. A singular new naevus on a person with > 50 naevi may be inadvertently missed, either by background visual interference or by the increased probability that the border overlaps with an existing naevus.

Change is sensitive for detection of malignant lesions, but not very specific. In fact, < 3% of all changing lesions will prove to be melanoma.[12, 14] While new naevi in patients with a history of melanoma should alert the physician to closer examination of the lesion, experience shows that given the presence of a benign dermoscopic pattern, one may adequately monitor new lesions rather than conduct excision. Consistent with these findings, of the 109 new naevi identified, only four, or approximately 3·7%, proved to be melanoma.

Results of a recent embryogenic study have elucidated two distinct pathways of melanocyte precursor cell migration, a deep ventral pathway and a superficial–dorsolateral pathway.[15] The ventral migration of melanocytes is linked to migration along nerve trunks, and this pathway may result in the development of predominantly intradermal or compound naevi. We speculate that naevi arising via this pathway are determined in utero, are not directly influenced by UV radiation, and give rise to predominantly globular naevi. Most of these naevi will become clinically manifest during childhood, accounting for most naevi in children being of the globular type.

The SONIC study has revealed that globular naevi do not transform into reticular naevi. A possible explanation may relate to the pathways of melanocyte migration during embryogenesis. The superficial–dorsolateral migration of melanocyte precursors may be primarily responsible for the deposition of melanocytes along the dermoepidermal junction. Perhaps UV radiation can more readily reach these melanocytes, resulting in alterations that may trigger the formation of naevi with a junctional component, which often manifest a reticular pattern under dermoscopy. Thus, many of these reticular naevi may truly be considered acquired naevi. In contrast, many of the intradermal naevi, even if appearing after childhood, may represent congenital naevi or tardive congenital naevi.

The study is limited in its ability to define the dynamic nature of the natural history of naevogenesis completely by the fact that no data were collected on changing or regressing naevi. Further investigation into these aspects of naevogenesis is warranted. Additionally, this study was limited in its collection of data regarding the presence or absence of scalp naevi, buttock naevi, dorsal foot naevi and iris naevi, which could have been used to define a subpopulation with the dysplastic naevus syndrome defined by Newton and supported by the World Health Organization.[16]

This study provides support for the theory that naevogenesis is a dynamic process occurring throughout life, with new naevi continuing to develop into adulthood, with an incidence of 202 new naevi per 1000 person-years of follow-up. The IRR for naevus development is higher in the youngest age quartile and in patients with HRNP. Newly acquired naevi tend to develop on the trunk, and show a predominant dermoscopic reticular pattern.


  1. Top of page
  2. Summary
  3. Patients and methods
  4. Results
  5. Discussion
  6. References