Anogenital lichen sclerosus et atrophicus lesions in a case series of cancer patients on immunotherapy.

Lichenoid reactions are one of the most frequently observed toxicities with anticancer agents and, recently, a rapid emergence of immunotherapies in oncology has hastened the need to better characterize their unique toxicity profiles, particularly for less common skin toxicities, including anogenital lichen sclerosus et atrophicus (LSA). This case series describes four patients with advanced cancer (one melanoma, two lung cancers, and one kidney tumor) developing LSA lesions while receiving an immunotherapy. Medical records from 2017 to 2020 were retrospectively reviewed. Two patients received pembrolizumab, anti-programmed cell death-1 (PD-1), one nivolumab, anti-programmed cell death-1 (PD-1), and one ipilimumab, an immune checkpoint inhibitor. LSA emerged after a median of 3 months (range, 2-4 months) from starting immunotherapy. All LSA cases were grade 2. Three cases occurred on the penis and one case on the anus. All patients improved after a specific treatment for LSA, and no LSA-related antineoplastic treatment interruption/life-threatening condition were reported. To date, this is the first case series of LSA lesions associated with immunotherapy. Early LSA recognition and management is helpful in cancer patients on immunotherapy allowing a long survival and treatment response.

Melanoma in children and adolescents: analysis of susceptibility genes in 123 Italian patients.

BACKGROUND: A polygenic inheritance involving high, medium and low penetrance genes has been suggested for melanoma susceptibility in adults, but genetic information is scarce for paediatric patients. OBJECTIVE: We aim to analyse the major high and intermediate melanoma risk genes, CDKN2A, CDK4, POT1, MITF and MC1R, in a large multicentre cohort of Italian children and adolescents in order to explore the genetic context of paediatric melanoma and to reveal potential differences in heritability between children and adolescents. METHODS: One-hundred-twenty-three patients (<21 years) from nine Italian centres were analysed for the CDKN2A, CDK4, POT1, MITF, and MC1R melanoma predisposing genes. The rate of gene variants was compared between sporadic, familial and multiple melanoma patients and between children and adolescents, and their association with clinico-pathological characteristics was evaluated. RESULTS: Most patients carried MC1R variants (67%), while CDKN2A pathogenic variants were found in 9% of the cases, the MITF E318K in 2% of patients and none carried CDK4 or the POT1 S270N pathogenic variant. Sporadic melanoma patients significantly differed from familial and multiple cases for the young age at diagnosis, infrequent red hair colour, low number of nevi, low frequency of CDKN2A pathogenic variants and of the MC1R R160W variant. Melanoma in children (≤12 years) had more frequently spitzoid histotype, were located on the head/neck and upper limbs and had higher Breslow thickness. The MC1R V92M variant was more common in children than in adolescents. CDKN2A common polymorphisms and MC1R variants were associated with a high number of nevi. CONCLUSION: Our results confirm the scarce involvement of the major high-risk susceptibility genes in paediatric melanoma and suggest the implication of MC1R gene variants especially in the children population.

The association between pesticide use and cutaneous melanoma: a systematic review and meta-analysis.

BACKGROUND: The incidence of cutaneous melanoma (CM), the deadliest form of skin cancer, has gradually increased in the last decades among populations of European origin. Epidemiological studies suggested that farmers and agricultural workers are at an increased risk of CM because they were exposed to pesticides. However, little is known about the relationship between pesticides and CM. OBJECTIVES: To investigate the association between exposure to pesticides and CM by systematically reviewing the literature. Secondary aim was to determine the categories of pesticides mainly involved in CM development. METHODS: A systematic review of the literature was performed up to September 2018 using MEDLINE, Embase and Web of Science. Studies assessing CM risk in licensed pesticide applicators were considered. Strict criteria were established to select independent studies and risk estimates; random effect models, taking into account heterogeneity, were applied. A pooled risk estimate for CM was calculated for the use of each type of pesticide and type of exposure. Between-study and estimate heterogeneity was assessed and publication bias investigated. RESULTS: A total of nine studies (two case-controls and seven cohorts) comprising 184 389 unique subjects were included. The summary relative risks for the categories 'herbicides - ever exposure', 'insecticides - ever exposure', 'any pesticide - ever exposure' and 'any pesticide - high exposure' resulted 1.85 [95% confidence interval (CI): 1.01, 3.36], 1.57 (95% CI: 0.58, 4.25), 1.31 (95% CI: 0.85, 2.04) and 2.17 (95% CI: 0.45, 10.36), respectively. Herbicides and insecticides had no between-study heterogeneity (I2  = 0%), while a significant heterogeneity (I2  > 50%) was detected for the high exposure to any pesticide. No indication for publication bias was found. CONCLUSIONS: Individuals exposed to herbicides are at an increased risk of CM. Future properly designed observational studies are required to confirm this finding.

Evaluation of the agreement between TNM 7th and 8th in a population-based series of cutaneous melanoma.

BACKGROUND: The 8th edition of TNM has introduced new rules for staging cutaneous melanoma. OBJECTIVE: To compare TNM 7th and 8th editions in defining pathological stages of melanoma. METHODS: A population-based series of 1847 skin melanoma from Romagna cancer registry (Italy) incident during 2003-2012 has been used to measure the agreement (with Cohen's kappa) between TNM 8th and 7th editions in defining melanoma stage. Disease-specific survival has been computed for each stage according to TNM 7th and 8th. RESULTS: The agreement between the two TNM editions was quite good when considered on average (kappa = 70.7%), moderate for stage I (61.5%), nearly perfect for stage II (95.0%), but extremely poor for stage III (8.1%). The overall melanoma-specific observed survival was 90.8% at 5 year and 88.9% at 10 year with a strong prognostic effect of stage. CONCLUSION: TNM 8th edition introduces several changes which do not seem really helpful in addressing the care of stage I melanoma and may complicate the definition and comparability of stage III.

A comparative dermoscopic and reflectance confocal microscopy study of naevi and melanoma with negative pigment network.

BACKGROUND: Negative pigment network (NPN) is a dermoscopic structure observed more frequently among melanomas than naevi. Precise tissue correlates of NPN are still elusive. OBJECTIVE: To describe the reflectance confocal microscopy (RCM) findings underlying NPN in melanocytic neoplasms. METHODS: We retrospectively identified all melanocytic neoplasms displaying dermoscopic NPN that were imaged with RCM and subsequently biopsied between 2011 and 2015. Images from study lesions (n = 50) were evaluated for dermoscopic and RCM Criteria. Histopathological correlational study was performed in a subset of cases (n = 15). RESULTS: The study data set consisted of 21 melanomas (42%) and 29 naevi (58%). Melanomas showed more frequently irregularly shaped globules than naevi (62% vs. 28%, P = 0.03); NPN also tended to be more asymmetrically located among melanomas (86%) than naevi (62%), albeit not significant (P = 0.06). Under RCM, we observed three patterns of dermal papillae (DP): (i) 'Dark DP' - whereby DP were devoid of nests and often surrounded by a junctional proliferation as thick-Rings - this pattern was less common among melanomas (n = 10, 48%) than naevi (n = 23, 79%, P = 0.02); (ii) 'Bulging DP' - whereby junctional nests of melanocytes protrude into the DP, often in association with junctional proliferation as Meshwork - with comparable frequency among melanomas (n = 12, 57%) and naevi (n = 23, 79%, P = 0.09) and (iii) 'Expanded DP' - whereby junctional and/or dermal nests filled and expanded the DP, often in association with dermal-epidermal junction (DEJ) Clod pattern - seen more commonly among melanomas (n = 15, 71%) than naevi (n = 6, 21%, P < 0.001). Dermoscopy-RCM correlation and comparison to histopathological findings show that the hypo-pigmented lines of NPN correlate with broadened epidermal retes, which often show overlying surface dells and wedge-shaped hypergranulosis, while the pigmented globules of NPN correlate with a predominantly-junctiona of melanocytes along and between the elongated retes. CONCLUSIONS: Dermoscopic NPN correlates with three DEJ RCM patterns with differing frequency between naevi and melanomas.

Morphological classification of melanoma metastasis with reflectance confocal microscopy.

BACKGROUND: Cutaneous malignant melanoma metastases differential diagnosis is challenging, as clinical and dermoscopic features can simulate primary melanoma or other benign or malignant skin neoplasms, and in-vivo reflectance confocal microscopy could assist. Our aim was to identify specific reflectance confocal microscopy features for cutaneous malignant melanoma metastases, and epidermal and dermal involvement. METHODS: A retrospective, multicentre observational study of lesions with proven cutaneous malignant melanoma metastases diagnosis between January 2005 and December 2016. Lesions were retrospectively assessed according to morphological features observed at reflectance confocal microscopy. Potential homogeneous subgroups of epidermal or dermal involvement were investigated with cluster analysis. RESULTS: Cutaneous malignant melanoma metastases (51 lesions in 29 patients) exhibited different frequencies of features according to metastasis dermoscopy patterns. Lesions classified at dermoscopy with nevus-like globular and non-globular patterns were more likely to be epidermotropic, showing characteristics of epidermal and dermal involvement at reflectance confocal microscopy. Other dermoscopy pattern classifications were more likely to be dermotropic, showing characteristics od dermal involvement at reflectance confocal microscopy. Distinguishing features at reflectance confocal microscopy included irregular (78%) and altered (63%) epidermis, pagetoid infiltration (51%), disarranged junctional architecture (63%), non-edged papillae (76%), dense and sparse, and cerebriform nests in the upper dermis (74%), and vascularity (51%). Cluster analysis identified three groups, which were retrospectively correlated with histopathological diagnoses of dermotropic and epidermotropic diagnoses (P < 0.001). The third cluster represents lesions with deep dermis morphological changes, which were too deep for evaluation with reflectance confocal microscopy. CONCLUSIONS: Specific reflectance confocal microscopy features of cutaneous malignant melanoma metastases for correct diagnosis, and subtype diagnosis, seem achievable in most cases where morphological alterations are located above the deep dermis.

Integration of reflectance confocal microscopy in sequential dermoscopy follow-up improves melanoma detection accuracy.

BACKGROUND: Successful treatment of melanoma depends on early diagnosis, but its varied clinical presentation means that no single noninvasive method or criterion can provide reliable detection in all cases. OBJECTIVES: To determine whether combining sequential dermoscopy imaging with reflectance confocal microscopy (RCM) can improve melanoma detection and reduce the burden of unnecessary excisions. METHODS: We conducted a retrospective study with median follow-up of 25 months. We included equivocal pigmented lesions that lacked clear dermoscopy criteria for melanoma at baseline but were excised subsequently because of changes during digital monitoring. RCM imaging was performed before excision. Main melanoma dermoscopy features, seven-point checklist score at baseline, and changes in structure and/or colour, and development of new melanoma-specific criteria at follow-up (scored as major, moderate or minor) were considered. Main melanoma RCM criteria were evaluated and diagnosis was made. Histopathological diagnosis was the reference standard for defining parameter frequency and diagnostic accuracy. RESULTS: Seventy lesions were included. Major changes were more frequently correlated with melanoma diagnosis, although one-third (four of 12) of melanomas showed moderate or minor changes. Cytological atypia and architectural disarrangement on RCM were correlated with melanoma diagnosis. A correct melanoma diagnosis was achieved with RCM in almost all cases (11 of 12, 92%). Referring for excision only those lesions with RCM-positive features and/or presenting major changes at digital dermoscopy follow-up, theoretically 27 of 58 naevi could be saved from surgery. CONCLUSIONS: Integration of RCM in the clinical and instrumental strategy for managing difficult pigmented lesions provided additional diagnostic information useful in the decision-making process.

Pigmented nodular melanoma: the predictive value of dermoscopic features using multivariate analysis.

BACKGROUND: Nodular melanoma (NM), representing 10-30% of all melanomas, plays a major role in global mortality related to melanoma. Nonetheless, the literature on dermoscopy of NM is scanty. OBJECTIVES: To assess odds ratios (ORs) to quantify dermoscopic features of pigmented NM vs. pigmented superficial spreading melanoma (SSM), and pigmented nodular nonmelanocytic and benign melanocytic lesions. METHODS: To assess the presence or absence of global patterns and dermoscopic criteria, digitized images of 457 pigmented skin lesions from patients with a histopathological diagnosis of NM (n = 75), SSM (n = 93), and nodular nonmelanocytic and benign melanocytic lesions (n = 289; namely, 39 basal cell carcinomas, 85 seborrhoeic keratoses, 81 blue naevi, and 84 compound/dermal naevi) were retrospectively collected and blindly evaluated by three observers. RESULTS: Multivariate analysis showed that ulceration (OR 4.07), homogeneous disorganized pattern (OR 10.76), and homogeneous blue pigmented structureless areas (OR 2.37) were significantly independent prognostic factors for NM vs. SSM. Multivariate analysis of dermoscopic features of NM vs. nonmelanocytic and benign melanocytic lesions showed that the positive correlating features leading to a significantly increased risk of NM were asymmetric pigmentation (OR 6.70), blue-black pigmented areas (OR 7.15), homogeneous disorganized pattern (OR 9.62), a combination of polymorphous vessels and milky-red globules/areas (OR 23.65), and polymorphous vessels combined with homogeneous red areas (OR 33.88). CONCLUSIONS: Dermoscopy may be helpful in improving the recognition of pigmented NM by revealing asymmetric pigmentation, blue-black pigmented areas, homogeneous disorganized pattern and abnormal vascular structures, including polymorphous vessels, milky-red globules/areas and homogeneous red areas.

Morphological features of naevoid melanoma: results of a multicentre study of the International Dermoscopy Society.

BACKGROUND: Naevoid melanoma (NeM), a rare variant of melanoma, can be difficult to detect as its clinical and histopathological morphology can simulate a naevus. OBJECTIVES: To describe the clinical and dermoscopic features associated with NeM. METHODS: Lesions with a histopathological diagnosis of NeM were collected via an e-mail request sent to all members of the International Dermoscopy Society. All lesions were histopathologically reviewed and only lesions fulfilling a set of predefined histopathological criteria were included in the study and analysed for their clinical and dermoscopic features. RESULTS: Twenty-seven of 58 cases (47%) fulfilled the predefined histopathological criteria for NeM and were included in the study. Clinically, 16 of the 27 NeMs presented as a nodular lesion (59%), eight (30%) as plaque type and three (11%) as papular. Analysis of the global dermoscopic pattern identified three types of NeM. The first were naevus-like tumours (n = 13, 48%), typified by a papillomatous surface resembling a dermal naevus. In these lesions local dermoscopic features included irregular dots/globules (46%), multiple milia-like cysts (38%) and atypical vascular structures (46%). The second type were amelanotic tumours (n = 8, 30%), typified by an atypical vascular pattern (75%). The third type consisted of tumours displaying a multicomponent pattern (n = 4, 15%), characterized by classical local melanoma-specific criteria. Two lesions (7%) were classified as mixed-pattern tumours as they did not manifest any of the aforementioned patterns. CONCLUSIONS: While NeMs may be clinically difficult to differentiate from naevi, any papillomatous lesion displaying dermoscopically atypical vessels and/or irregular dots/globules should prompt consideration for the possible diagnosis of NeM.

CDKN2A mutations could influence the dermoscopic pattern of presentation of multiple primary melanoma: a clinical dermoscopic genetic study.

BACKGROUND: Patients who develop cutaneous melanoma are at increased risk of developing a second primary melanoma. There are many aetiological reasons by which the risk of a second melanoma increases. Among others, genetic factors may contribute to modulating this risk. The risk of identifying a CDKN2A germline mutation increases with the number of primary melanomas and with the presence of familial history of melanoma. Patients with melanoma are especially encouraged to have regular follow-up visits with their dermatologist to perform clinical and dermatoscopic examination. In particular, dermoscopy could be very useful in multiple primary melanoma (MPM) patients. OBJECTIVES: To analyse the clinical and dermatoscopic features of multiple melanomas, focusing on those features that are more frequently found in the same patient to recognize them earlier and understand whether they appear with the similar peculiar dermatoscopic features, especially in CDKN2A carriers. METHODS: Medical records of MPM patients were selected from a database including 1065 patients with histopathologically proven melanoma diagnosis, all treated at the dermatology clinic of the University of Florence from 2000 to 2013. Pictures of melanoma were independently and blindly administered to three dermatologist experts in dermoscopy to evaluate the presence or absence of ABCD criteria for each clinical image, and the main pattern for the dermoscopic images. The results were then analyzed and crossed to rate the clinical and dermoscopic features of MPM. RESULTS: Seventy five (7.0%) of 1065 patients included in our database were found to carry an MPM disease. Among them, we selected 12 (16%) patients with three or more MPMs. The presence of the CDKN2A melanoma susceptibility gene was observed in 4/12 (33.33%) patients; two patients presented the C500G and c.5 + 1delG polymorphisms in the CDKN2A gene. In CDKN2A carriers, each patient showed a similar and specific dermatoscopic pattern in their lesions. CONCLUSIONS: Even being aware of the limitations of this study, according to hereditary characters and their modes of transmissions, we could speculate that for each patient with a CDKN2A germline mutation, it is possible to find the same kind of dermoscopical pattern among their melanocytic tumours.

Dermoscopy and reflectance confocal microscopy of pigmented actinic keratoses: a morphological study.

BACKGROUND: Actinic keratoses (AKs) are very common lesions on sun damaged skin and, when pigmented, represent a challenge in the differential diagnosis with early melanoma. Non-invasive diagnostic methods, such as dermoscopy and reflectance confocal microscopy (RCM) have been shown to improve the diagnostic accuracy of melanoma and non-melanoma skin cancer, however, only one case report described confocal findings of pigmented AKs up to now. OBJECTIVES: The aim of our retrospective morphological study was to analyse dermoscopic and confocal images of a series of histopathologically proven pigmented AKs, located on the face and other body sites, to define peculiar features of these "difficult to diagnose" lesions. METHODS: Clinical, dermoscopic and RCM images of 17 histopathologically confirmed pigmented AKs were retrospectively collected from the databases of four skin lesion clinics in Italy and USA. Dermoscopic and RCM images were analysed for prevalent morphological features. RESULTS: The majority of the lesions were located on the face (n = 8); followed by scalp (n = 4) and trunk (n = 4); and one lesion was located on the lower limbs. On dermoscopy the majority of lesions were characterized by grey dots/globules/granularity and structureless brown pigmentation. The main RCM feature of pigmented AKs was as follows: (i) the presence of epidermal changes (atypical keratinocytes, parakeratosis, scaling); (ii) increased epidermal thickness; (iii) bright, small, dermal papillae with enlarged interpapillary space; and (iv) intraepidermal dendritic cells referrable to Langherans cells. Features suggestive of melanocytic lesions, such as nesting, meshwork pattern or atypical cells infiltrating the junction, were never detected in our case series at the dermal epidermal junction (DEJ) level. CONCLUSION: Larger case series with adequate control population are warranted to validate these findings and to test their value in clinical setting.

Melanoma detection in Italian pigmented lesion clinics.

AIM: Accuracy in melanoma detection is important to recognize early curable melanomas and to minimize the unnecessary excision of benign lesions. The aim of this paper was to evaluate melanoma screening accuracy of Italian pigmented lesion clinics in terms of number needed to excise (NNE), melanoma thickness, and number of melanomas diagnosed during patient follow-up. METHODS: Information on all skin tumors excised in 2011 were extracted from the databases of the participating centers. Information whether the lesion was excised at the baseline examination or during patient follow-up was recorded, as well as the overall number of patients examined in each center in 2011. RESULTS: After e-mail solicitation, 22 of 40 centers agreed to participate. A total of 8229 excised lesions were collected. The overall number of examined patients was 86.564, thus 9.5% of screened patients had a lesion removed. Of the excised lesions, 866 were diagnosed as melanoma (1% of examined patients) and 5311 (88.9%) were melanocytic nevi. Three NNE were calculated giving values of 7.9 excised lesions to find 1 melanoma, 7.1 melanocytic lesions to find 1 melanoma, and 3.7 lesions to find 1 skin malignancy. The median melanoma thickness was 0.6 mm, with only 15.1% of melanomas ≥ 1 mm of thickness. Melanomas detected over time were 96 (11.1%; mean thickness, 0.3 mm), with 15.6% of lesions excised after short-term follow-up and 84.4% after long-term follow-up. CONCLUSION: The NNE values comparable to those achieved in specialized clinical settings and the high number of early melanomas diagnosed at the baseline examination or during patient follow-up indicate a high level of accuracy in melanoma screening achieved by Italian pigmented lesion clinics.

Reconsidering S100B: Why it is time to abandon its use in detecting disease recurrence in stage III melanoma patients.

BACKGROUND: Especially in the era of successful systemic therapy, there is an urgent need to detect early disease recurrence in stage III melanoma patients. This study investigates if serum S100 calcium-binding protein B (S100B) can detect disease recurrence in stage III melanoma patients. METHODS: A retrospective cohort study was conducted at the University Medical Center Groningen (UMCG). Adult AJCC 8th stage III melanoma patients in whom serum S100B was measured as part of follow-up from January 2010 until April 2023 were included. The association between serum S100B and disease recurrence was evaluated using standard definitions for sensitivity and positive predictive value (PPV). RESULTS: Overall, 147 patients were included (mean age was 60.4 years, 53.1 % were female). Most patients were classified as stage IIIB (39, 26.5 %) and IIIC (73, 49.7 %). During median follow-up of 56 months, 69 (46.9 %) patients experienced disease recurrence. Seventeen out of 18 patients with elevated serum S100B (≥0.15 µg/L) experienced disease recurrence (PPV of 94.4 %). However, 52 out of 69 patients with disease recurrence had normal serum S100B (sensitivity of 24.6 %). Eight out of 17 (47.1 %) patients were asymptomatic (P = 0.608), twelve (70.6 %) patients had at least four distant metastases (P < 0.001). CONCLUSION: The clinical value of serum S100B to detect disease recurrence in stage III melanoma patients is negligible since only one out of four patients with disease recurrence have elevated serum S100B. Furthermore, half of stage III melanoma patients with elevated S100B experienced symptoms, and most patients already have multiple distant metastases.

Sunbed use among subjects at high risk of melanoma: an Italian survey after the ban.

BACKGROUND: The World Health Organization classified the entire ultraviolet (UV) spectrum and artificial UV tanning devices as carcinogenic to humans. Italian law has prohibited the use of tanning equipment by children under 18 years of age and by high-risk populations. OBJECTIVES: The present large survey aimed to determine the prevalence of current sunbed use in Italy and to identify user characteristics. This study identifies starting points for future national interventions to reduce the health risks of exposure to artificial UV radiation. METHODS: In 2011 we conducted a survey of 4703 people in an area on the sunny Mediterranean coast in Italy. Through multivariate logistic models we investigated the associations of sunbed use with phenotypical factors. RESULTS: Overall prevalence of sunbed use was 20%, higher among women (22% vs. 16%; P < 0·0001), and young (22% vs. 17% for age < 35 years; P < 0·0001) and highly educated people (22% vs. 14%; P < 0·0001). Subjects at high risk of melanoma used sunbeds significantly more; i.e. people with freckles (25% vs. 18%; P < 0·0001), red hair (30% vs. 19%; P = 0·01) or fair eyes (22% vs. 19%; P = 0·006). Associations were confirmed in multivariate models. CONCLUSIONS: More skin cancer monitoring is needed at tanning centres, and educational campaigns should be promoted, especially for young women and subjects at high risk of melanoma.

Dermoscopy of acral melanoma: a multicenter study on behalf of the international dermoscopy society.

BACKGROUND: Most studies on dermoscopy of acral lesions were conducted in Asian populations. In this study, we analyzed these features in a predominantly Caucasian population. OBJECTIVE: Estimate the prevalence of dermoscopic features in acral lesions, and assess their level of agreement between observers. METHODS: In this retrospective multicenter study, 167 acral lesions (66 melanomas) were evaluated for 13 dermoscopic patterns by 26 physicians, via a secured Internet platform. RESULTS: Parallel furrow pattern, bizarre pattern, and diffuse pigmentation with variable shades of brown had the highest prevalence. The agreement for lesion patterns between physicians was variable. Agreement was dependent on the level of diagnostic difficulty. CONCLUSION: Lesions with a diameter >1 cm were more likely to be melanoma. We found as well that a benign pattern can be seen in parts of melanomas. For this reason one should evaluate an acral lesion for the presence of malignant patterns first.

Changes observed in slow-growing melanomas during long-term dermoscopic monitoring.

BACKGROUND: Melanomas vary in growth rate from rapidly developing nodular melanomas to slow-growing melanomas (SGM) that hardly change over years. OBJECTIVES: To evaluate longitudinal changes in dermoscopic findings of SGM. METHODS: We retrospectively analysed a dermoscopic image dataset from 15 pigmented lesion clinics, of SGM that were followed sequentially by digital dermoscopy for at least 1 year. We evaluated baseline and follow-up images for changes in global pattern, organization, colours, structure and size. RESULTS: Our series consisted of 92 SGM. On follow-up, these melanomas developed the following dermoscopic findings: more homogeneous and less reticular global dermoscopic pattern; more frequent disorganization of pattern (baseline, 67% vs. follow-up, 79%); decreased prominence of light brown colour, increased prominence of dark brown colour, and increased frequency of the colours red, white, grey, blue and black (baseline: 29%, 3%, 18%, 6% and 33% vs. follow-up: 41%, 10%, 31%, 13% and 45%, respectively); decrease in prominence of dermoscopic structure of pigmented network, with a concomitant increase in prominence of structureless areas; and increased prominence or new appearance of melanoma-specific dermoscopic structures, including negative network, blue-white structures and blotches. The majority of lesions (75%) remained the same size or grew by < 2 mm in diameter. An increase in lesion size was associated with change in the total number of colours and structures (χ(2) = 14·3, P = 0·027) at follow-up. CONCLUSIONS: While their diameter changed minimally over time, most SGM became more disorganized, revealed loss of network in favour of structureless areas, and developed new colours.