Videodermoscopic pattern analysis of acral melanocytic nevi
Tugba R. Ekmekci, M.D., Ihlamurdere Cad. No. 99/19, Besiktas, Istanbul, Turkey. Email: email@example.com
Three major and three minor types of acral melanocytic lesions have been described: parallel furrow, lattice-like, fibrillar, reticular, homogeneous and globular. To examine the dermoscopic patterns of acral melanocytic lesions and to investigate the association of these patterns with clinical features. This study included 419 acral melanocytic lesions of 295 patients. Dermoscopic images were stored in a standardized digital system and assessed. The incidence rates of the patterns were as follows: parallel furrow (62.5%), reticular (12.2%), globular (7.6%), fibrillar (5.0%), lattice-like (4.3%), homogeneous (3.8%), non-typical (3.8%), globulostreak-like (0.5%) and parallel ridge (0.2%). There was a significant association between pattern and diameter; nevi showing globular pattern were significantly smaller than those showing lattice-like pattern. There was a significant correlation between pattern and localization. The fibrillar pattern was observed more commonly on the feet, and the lattice-like pattern was seen almost entirely on the hands. Homogeneous and globular patterns were proportionately more common on the feet than on the hands. A significant association was found between patterns and colors. Compared with the lattice-like pattern, parallel furrow, fibrillar and homogeneous patterns were lighter in color; the globular pattern was darker than the homogeneous pattern. There was no significant correlations between pattern and clinical presentation. In conclusion, unlike all other studies, in the present study, the two most common patterns following the parallel furrow pattern were reticular and globular, which are minor patterns. Besides the differences in pattern distribution and inherent variability in individual assessment, the age of the study group and location of nevi influence the pattern, possibly in association with race.
Acral melanocytic nevi, compared with rarely seen acral melanoma, are relatively more common.1 While melanocytic nevi located on the palms or soles of the feet are seen in 42% of black and 23% of white people,1 in Japanese people, 8% of melanocytic lesions are located on the soles of the feet.2,3
It is sometimes difficult to distinguish early malignant melanoma from acral nevi, even for experienced dermatologists.2,3 Therefore, dermoscopy, a non-invasive technique that allows for in vivo diagnosis, has been widely applied to the diagnosis of melanocytic lesions.2,4,5 Dermoscopy has been shown to significantly enhance the diagnostic accuracy for these lesions.2,5
Depending upon where acral melanocytic lesions are located, certain dermoscopic features specific to certain regions are seen.4 Three major patterns have been established for benign melanocytic lesions located in the region: (i) parallel furrow pattern (PFP) (parallel linear pigmentation along the sulci of skin markings); (ii) lattice-like pattern (pigmented lines that follow and cross the sulci of skin markings); and (iii) fibrillar pattern (fine pigmented lines that cross skin markings).2,6 In addition to these patterns, several micropatterns have been identified, including homogeneous (structureless, light brown pigmentation), reticular (network pigmentation), globular (composed of dots and/or globules distributed in a non-parallel fashion), non-typical (characterized by dermoscopic features that do not conform to any other typical pattern and that cannot be classified as a multi-component pattern), transition (showing a combination parallel furrow/lattice-like pigment network pattern in one area and the pattern in another area), and globulostreak-like (dark brown globules attached to brown linear or curvilinear streak-like structures).3,7,8 The parallel ridge pattern, which shows band-like pigmentation on the ridges of the skin, is the dermoscopic characteristic indicator of acral malignant melanoma.6,9 The other sign of malignancy is the multi-component pattern (existence of more than two patterns). The most commonly seen PFP variants have been described as double dotted-line, double-line, single dotted-line, single-line, single-line PFP associated with a fine reticulated background, single-line PFP associated with dots, and single-line PFP associated with homogeneous pigmentation.7
A number of studies on the dermoscopic characteristics of acral melanocytic lesions in white individuals have been conducted during the past 20 years. However, only one such study from Turkey exists in the published work, and it was performed with a small number of patients. The aim of the present study was to examine the dermoscopic pattern of acral melanocytic lesions in large patient populations and to investigate the association of these patterns with demographic data and clinical characteristics.
This study included 295 patients with pigmented lesions located in the acral region who applied to the dermatology department of our hospital. All patients gave their informed consent for participation. The study was carried out according to the Declaration of Helsinki adopted in 1964 and revised in 2004 and the principles of the 3rd Istanbul Clinical Research Ethics Committee, and it received the approval of the Clinical Research Ethics Board.
Lentigo simplex was considered as melanocytic nevus. All congenital and acquired lesions and nevi located in the transition line between hairless and hairy skin were included in the study. Nevi located in the dorsal and subungual regions were excluded from the study. Nevus diameter, number, anatomical location (palms of hands, soles of feet), clinical description (maculae, papules) and color (light brown, dark brown, black, variegated, red, blue) for all patients were recorded.
Images of all patients were stored in a standardized digital system (Molemax Plus I; Derma Instruments LP, Vienna, Austria). While microimages were being taken at a magnification of ×30 in such a way as to maintain a maximum field of 1 cm, careful attention was paid to avoid shifting the probe of the dermoscope. Macroscopic and dermoscopic images of the lesion region were taken by adjusting the standard color and brightness and were stored.
According to the criteria introduced by Saida et al.2,9–12 and revised by Altamura,8 Malvehy and Puig,5 and Özdemir,7 the lesions were evaluated in terms of properties of patterns such as the PFP and its variants: lattice-like, fibrillar, homogeneous, globular, reticular, non-typical, parallel ridge, globulostreak-like and multi-component.
The decision for excision was made for lesions suspected of malignancy (parallel ridge pattern, multi-component pattern, irregular diffuse pigmentation, irregular or peripheral dots and/or globules, abrupt edges, atypical streaks, blue-white veil or regression structures) and for lesions of more than 7 mm based on the criteria by Saida3,5,7,8 without taking the pattern into account.
Data were recorded with the use of SPSS ver. 17.0 for Windows. In comparisons of patterns with demographic data (excluding age and diameter) and clinical characteristics, the χ2-test was applied to the analysis of binary or multiple groups. In the binary comparison of colors and patterns, P < 0.0002 was considered significant, with Bonferroni correction as a post-hoc test. Kruskal–Wallis variance analysis was used in the comparison of age, diameter and median values. To reduce type I error arising from the influence of multiple comparisons, the Mann–Whitney U-test was used with Bonferroni correction in post-hoc analysis in determination of the source of basic effects found significant at P = 0.002. The results were evaluated with the use of a 95% confidence interval; significance was assessed at P < 0.05.
The study included 419 lesions from 295 patients (200 women [67.8%] and 95 men [32.2%]; mean age, 31.8 ± 14.5 years; median age, 29 years; age range, 5–86 years).
The average nevus size of the lesion was 2.66 ± 1.61 mm (range, 0.68–20 mm), and the median value was 2.31. A maximum of five acral melanocytic nevi were observed in one patient. Assessment was performed based on the number of lesions. In 210 patients, one lesion was found; in 61 patients, two lesions; in 13 patients, three lesions; in seven patients, four lesions; and in four patients, five lesions.
Of 419 lesions, 317 (75.7%) nevi were on the palms of the hands and 102 (24.3%) were on the soles of the feet. A total of 410 maculae and nine papular lesions were detected. When they were examined based on color, 203 (48.4%) were found to be light brown, 205 (48.9%) dark brown, 10 (2.4%) variegated and one (0.2%) red.
Pattern distributions in the 419 nevi are shown in Table 1.
Table 1. Distribution of patterns in 419 nevi
| Double dotted-line PFP||60 (14.3)|
| Double-line PFP||8 (1.9)|
| Single dotted-line PFP||84 (20.0)|
| Single-line PFP||46 (11.0)|
| Single-line PFP associated with a fine reticulated background||34 (8.1)|
| Single-line PFP associated with dots||17 (4.0)|
| Single-line PFP associated with homogeneous pigmentation||13 (3.1)|
| Single-line PFP total||110 (26.2)|
| Parallel furrow total||262 (62.5)|
|Parallel ridge||1 (0.2)|
Of five nevi with the double dotted-line PFP, three were associated with a homogeneous pattern and one was associated with reticular dots; of eight nevi with the single dotted-line PFP, five were associated with a homogeneous pattern, two with dots, and one with non-typical and reticular structures. Two nevi with the fibrillar pattern were associated with dots. Again, one nevi with the lattice-like pattern was associated with dots and globules. Lesions with the lattice-like pattern were associated with 27.8% of PFP nevi. Reticular pigmentation was observed in the background of eight of the lesions with lattice-like patterns.
Three lesions with a globulostreak-like pattern and parallel ridge pattern were not appropriate for statistical analysis because they were very rarely seen in the study group compared the other patterns, and they were excluded from the evaluation. Statistical analysis was performed based on a total of 416 nevi (Table 2).
Table 2. Comparison of patterns in terms of diameter, location, clinical pattern and color
|Parallel furrow||262 (63.0)||2.24||207 (65.7)/55 (54.5)||258 (63.4)/4 (44.4)||138 (68.0)/123 (60)|
|Reticular|| 51 (12.3)||2.38|| 39 (12.4)/12 (11.9)|| 49 (12.0)/2 (22.2)|| 24 (11.8)/26 (12.7)|
|Globular|| 32 (7.7)||1.74|| 22 (7.0)/10 (9.9)|| 32 (7.9)/0 (0)|| 8 (3.9)/22 (10.7)|
|Fibrillar|| 21 (5. 0)||2.39|| 10 (3.2)/11 (10.9)|| 21 (5.2)/0 (0)|| 13 (6.4)/7 (3.4)|
|Lattice-like|| 18 (4.3)||2.92|| 17 (5.4)/1 (1.0)|| 17 (4.2)/1 (11.1)|| 2 (1.0)/14 (6.8)|
|Homogeneous|| 16 (3.8)||2.27|| 9 (2.9)/7 (6.9)|| 15 (3.7)/1 (11.1)|| 13 (6.4)/3 (1.4)|
|Non-typical|| 16 (3.8)||2.63|| 11 (3.5)/5 (5.0)|| 15 (3.7)/1 (11.1)|| 5 (2.5)/10 (4.9)|
|Total||416 (100)||2.29||315 (100)/101 (100)||407 (100)/9 (100)||203 (100)/205 (100)*|
There was a statistically significant difference between median diameter and pattern (P = 0.028). The median diameter of each pattern was compared with that of other patterns separately. The diameters of nevi with a globular pattern were significantly smaller than those of nevi with a lattice-like pattern (P = 0.001).
A significant difference was determined in terms of pattern and localization (hands or feet) (P = 0.006). However, because the number of nevi located on feet was fewer, we did not perform statistical interpattern comparison. The fibrillar pattern was more often observed in feet. The homogeneous pattern and globular pattern were proportionally more frequent in feet than in hands. In contrast, while the lattice-like pattern was almost always observed on the hands, the reticular pattern showed a proportionally equal distribution in both hands and feet. The PFP was proportionally more common in the hand.
No significant difference was present between pattern and clinical description (P = 0.432).
Because other colors (variegated, red) showed a lower incidence in statistical analysis, they were ruled out. Only in light brown and dark brown nevi was a comparison drawn between pattern and color, and 406 lesions were included in the assessment. A significant difference was identified between pattern and color (P = 0.0001). The color of each pattern was compared with that of other patterns separately. Compared with the lattice-like pattern, the PFP, fibrillar and homogeneous patterns were significantly lighter brown in color (P = 0.002, P = 0.002, and P = 0.0001, respectively). Compared with the homogeneous pattern, the globular pattern was significantly darker brown in color (P = 0.001).
Eight lesions (four that were >7 mm, three with blue-white veils, one with a parallel ridge pattern) were excised. Five lesions proved to be consistent with compound nevi and three with intradermal nevi.
The major dermoscopic patterns seen in acral melanocytic nevi are PFP, fibrillar and lattice-like.2,5,9,10,12 The proportions of patterns in our study to those of other studies are given in Table 3. The PFP was detected at a slightly higher rate (62.5%) in our study than in other studies. Single-line PFP is the PFP prototype.7 Ozdemir et al.7 found this variant at a rate of 51.8% among PFP in conjunction with its other subtypes. In descending frequency, this ratio also includes homogeneous pigmentation PFP (26.3%), dot PFP (21.0%) and single-line PFP associated with a fine reticulated background (10.5%). In the present study, the rate of single-line PFP was 41.9%. Subtypes within this pattern included single-line PFP associated with a fine reticulated background (30.9%), single-line PFP associated with homogeneous pigmentation (15.4%) and single-line PFP associated with dots (11.8%). In addition, we detected homogeneous pigmentation, dots and non-typical structures on the ground of variants of double dotted-line and single dotted-line PFP.
Table 3. Comparison of our study with other studies
|Mean age (years)||–||34||24.9||29||29|
|Palmar %/plantar %||–||21.4/78.6||9.3/70.5‡||52.7/47.3||75.5/24.5|
|Parallel furrow, n (%)||40 (42)||111 (52.9)||304 (42.1)||110 (58.5)||262 (62.5)|
|Lattice-like, n (%)||13 (13)||26 (12.4)||108 (14.9)||12 (6.4)||18 (4.3)|
|Fibrillar, n (%)||20 (21)||13 (6.2)||78 (10.8)||23 (12.2)||21 (5.0)|
|Homogeneous, n (%)||2 (2)||15 (7.1)||67 (9.3)||12 (6.4)||16 (3.8)|
|Globular, n (%)||5 (5)||11 (5.2)||39 (5.4)||4 (2.1)||32 (7.6)|
|Reticular, n (%)||3 (3)||5 (2.4)||15 (2.1)||8 (4.3)||51 (12.2)|
|Transition, n (%)||NE||NE||13 (1.8)||NE||0|
|Globulostreak-like, n (%)||NE||NE||NE||10 (5.3)||2 (0.5)|
|Non-typical, n (%)||14 (14)||29 (13.8)||99 (13.7)||6 (3.2)||16 (3.8)|
Malvehy and Puig5 found the rate of the reticular pattern characterized by a pigmented network to be 2.4% with no relationship to any other pattern. Saida et al.2 reported that the reticular background was accompanied by PFP at a rate of 5.8%. Altamura et al.8 found the rate of the reticular pattern to be 2.1%. The rate of reticular pigmentation associated with the PFP and lattice-like pattern was determined to be 1.8%, and they called this a transition pattern because these nevi were located in the transition between hairy and hairless areas. Ozdemir et al.7 reported that 4.3% of lesions had a reticular pattern. While 75% of lesions were located in the skin folds, all were located on the hands. PFP was observed together with a “crista dotted pattern” 25% of the time. Ozdemir et al.7 pointed out that the only anatomical location of the reticular pattern was the hand, and they reported that a link might be present between pattern and location. In the present study, the rate of the reticular pattern was as high as 12.2%. Of these lesions, two (4%) were located on the skin fold and one (2%) was on the transition area between hairy and hairless skin; none were associated with any other pattern. Furthermore, while the rate of the single-line PFP was 8.1% on the reticular background, a number of nevi with the double dotted-line PFP and lattice-like pattern were also associated with the reticular background. In addition, and in contrast to the study by Ozdemir et al.,7 we established that nevi with the reticular pattern were equally distributed on the hands and feet, which was in accordance with the study by Altamura et al.8
In this study, the globular pattern was detected at a slightly higher rate (7.6%) than in other studies (Table 3). It was also significantly smaller (diameter, optical density: 1.74) than the lattice-like pattern. The globulostreak-like pattern described by Ozdemir et al.7 ranged 1–3 mm in diameter. The globular pattern was detected at a higher rate in the present study compared with other studies, especially that of Ozdemir et al. The reason for this is that we may have assigned a globular pattern to the nevi, while they assigned a globulostreak-like pattern at a rate of 5.3%. The rate of the globulostreak-like pattern was only 0.5% in the current study. The association between the globular pattern and anatomical location has not been previously shown,8 but in the present study, it was proportionally much higher on the sole of the foot. Ozdemir et al.7 reported that in 66.6% of nevi with a globular pattern, there were dots and globules in a parallel sequence as described by Akasu et al.11 and Saida et al.1 Globules had a parallel sequence at a rate of 31.3% in the present study.
The fibrillar pattern is believed to be a PFP variant and occurs with the slope of the keratinized layer. On moving the probe of the dermoscope horizontally, the fibrillar pattern appeared as the PFP.13 Again, it was demonstrated that over time the fibrillar pattern transformed into the PFP or a homogeneous pattern.5,7 Miyazaki et al.6 showed that the fibrillar pattern was located more frequently on the foot. Malvehy and Puig5 and Altamura et al.8 established that the fibrillar pattern had a significantly higher incidence on the sole of the foot. In the present study, the rate of the fibrillar pattern was considerably low (5%) compared with that in other studies (Table 3). The lower rate can be explained by race and the rate of nevi located on the soles of the feet (24.5%). In contrast, in the studies by Malvehy and Puig,5 Altamura et al.8 and Ozdemir et al.,7 the rates were 78.6%, 70.5% and 47.3%, respectively. Similar to the findings in the study by Saida,1 there were several fibrillar pattern lesions associated with dots or globules in the study by Özdemir.7 Moreover, in the present study, dots were associated with this pattern in two lesions.
The lattice-like pattern was located mostly on the hand.6 Conversely, it was reported that this pattern was noted significantly more often on the sole of the foot.8 Although nevi were located on the hands at a rate of 75.5% in the present study, the rate of the lattice-like pattern was quite low (4.3%) compared with that in other studies (Table 3). The value was closer to that found by Ozdemir et al.7 from Turkey than that detected by studies involving white and Japanese individuals.2,8,9,12,14 Ozdemir et al.7 reported that the PFP was associated with 41.7% of lesions in which the lattice-like pattern was detected. In the present study, the rate was 27.8%. Based on these percentages, it is likely that we interpreted the lattice-like pattern associated with the PFP mostly as the PFP, while Ozdemir et al.7 interpreted it mostly as the lattice-like pattern. This is acceptable because dermoscopic assessments can vary across individuals, and it supports the reports that the lattice-like pattern is actually a PFP subtype.2,9,12 Saida stated that this pattern could be associated with dots and globules.1,2 We also encountered dots and globules in some melanocytic lesions with the lattice-like pattern.
Because of its high prevalence rate, Ozdemir et al.7 proposed that the homogeneous pattern should be recognized as one of the major dermoscopic patterns. The homogeneous pattern was seen at a rate as low as 3.8% and was determined to be mostly located on the foot in the present study. Altamura et al.8 also found the homogeneous pattern to have a far more significant incidence on the soles of the feet. This low rate may be because the rate of localization of nevi on the sole was 24.5% in the current study. Instead, this rate was 78.6%, 70.5% and 47.3% in other studies. This pattern reportedly shows clinical and dermoscopic involution in acral melanocytic nevi.15
The non-typical pattern has been explained by the normal development process of mature nevi, the presence of mature islands of a more deep-seated nevus component, and the presence of fibrosis or histological atypia.5 The non-typical pattern was reportedly between 3.2% and 14% in a number of studies.2,3,5,7,8 In the present study, the rate was 3.8%. Ozdemir et al. also found this rate to be low, and they ascribed it to the globulostreak-like pattern, which they had recently identified. In the present study, because all previously described patterns were considered, the lower rate of the non-typical pattern was not unexpected. Malvehy and Puig5 found the non-typical pattern on the palm of the hand, while Altamura et al.8 found it significantly more often on the soles of the feet. The present study did not find a significant relationship between localization and the non-typical pattern.
In addition to regional discrepancies, ethnic differences affect pattern distribution. However, Saida and Koga3 attributed the main reason for the difference in pattern distribution to individual differences in assessments. In terms of the combination of the two patterns not infrequently seen, dermatologists make an arbitrary decision regarding which pattern to choose, which provides variable results.3
In conclusion, unlike other studies, the present study ranked the reticular and globular patterns, also called minor patterns, as the two most common patterns following the PFP. The age of the study group, the anatomical location of nevi, race and inherent variability in individual assessment also affect differences in pattern distribution.