Dermoscopy of Small Melanomas: Just Miniaturized Dermoscopy?
Results
Overall, 352 MMs (73·0%) showed a 6·01–20-mm diameter, 79 lesions (16·4%) were ≤ 6 mm and 51 lesions (10·6%) were > 20 mm.
Twenty-two MMs (4·6% of the total) were ≤ 4 mm, showing a mean diameter of 3·4 mm. The mean age of patients with micromelanoma was 42·18 ± 14·93 years, significantly lower than that of the patients in total (55·52 ± 16·15 years; P = 0·000). Thirteen of these lesions (59%) were in women, whereas in the total population of patients with MMs both sexes were equally represented. Seven lesions (32%) were localized on the trunk, three (14%) on the upper limbs, and the remaining 12 on the lower limbs. The frequency of micromelanomas on the lower limbs was significantly higher in these patients vs. the whole group (12 lesions, 55% vs. 134 lesions, 27·8%; P = 0·014). The mean thickness of micromelanomas was 0·22 ±0·34 mm, and 12 lesions (55%) were in situ. These values differed significantly from those of the whole group (0·9 ± 1·63 mm, P = 0·049; and 28·0%, P = 0·015, respectively). Three micromelanomas (14%) had arisen on a naevus, whereas 126 (27·4%) of larger MMs had an ex naevo origin. Two of the 22 patients (9%) with micromelanomas had shown a previous MM, vs. 34 subjects (7·4%) with larger MMs (seven had previously developed more than one malignant melanocytic skin lesion).
Table 1 illustrates the frequency of the 7-point and ABCD criteria and scores in 482 melanomas of different diameters. Of the 409 MMs (84·9% of the total) with diameter 4·01–20 mm, 61–84·6% (depending on size) showed two-axis asymmetry, with mean asymmetry values ranging from 1·51 to 1·84. By contrast, only 36% of micromelanomas (eight lesions) were asymmetric, whereas 48 lesions > 20 mm in size (94%) showed two-axis asymmetry. Thus, lesion asymmetry increased according to diameter.
An inverse trend was observed for sharp demarcation of lesion borders. Eight borders were recorded in five lesions of ≤ 4 mm (23%), in 12·5–15·3% of MMs of 4·01–20 mm and in five MMs > 20 mm (10%).
The number of colours increased according to lesion diameter. With respect to micromelanomas, colours did not significantly differ in 4·01–6-mm MMs, but were significantly higher in MMs > 6 mm (significant values are marked in Table 1). Conversely, the mean number of dermoscopic structures was similar in all MM categories.
In micromelanomas, an atypical network was present in 77% of lesions (n = 17), a blue–whitish veil in 5% (one lesion), atypical vessels in 9% (two lesions), irregular pigmentation in 50% (11 lesions), irregular dots/globules in 59% (13 lesions), irregular streaks in 36% (eight lesions) and regression in 32% of cases (seven lesions).
Four parameters of the 7-point checklist (blue–whitish veil, atypical vessels, irregular globules/dots and regression) showed values increasing according to MM diameter. On the contrary, an atypical network and irregular pigmentation were more frequently observed in micromelanomas than in larger ones. In the former group, TDS and 7-point score were significantly lower with respect to MMs > 6 mm (P-values ranging from 0·045 to 0·000).
Table 2 describes the clinical, histological and dermoscopic aspects of the 22 micromelanomas. The smallest MM was 2·4 mm in size. Only six lesions (27%) were ≤ 3 mm wide. Twelve lesions (55%) were in situ, whereas the other 10 (45%) were 0·2–2 mm thick. Only three micromelanomas (14%) had arisen on a naevus. From a clinical point of view, 12 lesions (55%) were eye catching, in spite of their small size, because of colour variegation (n = 1) or because of their dark colour (n = 11). They were often associated with abrupt borders (n = 9). Five lesions (23%) were clinically asymmetric, 7 (32%) were light in colour and inconspicuous, and 4 (18%) were palpable. The latter corresponded to invasive lesions.
As regards the dermoscopic patterns and features, most micromelanomas showed a bi- or multicomponent pattern. The three reticular lesions (14%) presented with an atypical network, which was also observed in one island, four bicomponent, seven multicomponent and two spitzoid lesions (5%, 18%, 32% and 9%, respectively). Four lesions displayed a spitzoid pattern with peripheral structures (18%). One lesion was globular and another was a dermoscopic island (5%). Irregular pigmentation was found in seven bicomponent or multicomponent lesions (32%); irregular globules were observable in 12 MMs (55%) and irregular streaks in 8 (36%). Regression, consisting of grey–blue areas or peppering, was found in seven cases (32%). Atypical vessels were observable in one case, as was a blue–whitish veil.
When subdividing micromelanomas into two groups according to Breslow thickness (12 in situ and 10 invasive MMs), we observed that the mean patients' age was higher (but not significantly) in patients with in situ MM vs. invasive MM (45 ± 15·75 years vs. 38·8 ± 11·49 years) and the mean diameter was similar (3·32 ± 0·51 mm vs. 3·55 ±0·43 mm, respectively). Women prevailed in the invasive MM group (70%), whereas in the in situ group men and women were equally represented. Six lesions were located on the lower limbs and had an eye-catching appearance in both groups. Multicomponent and bicomponent patterns were found in both groups with a similar frequency, whereas a spitzoid pattern was observable in three invasive MMs (30%) and in one in situ MM (8%).