Dermatology: Practical and Conceptual Dermatology Practical & Conceptual Research | Dermatol Pract Concept. 2021;11(3):e2021059 1 Dermoscopic Predictors of Tumor Thickness in Cutaneous Melanoma: A Retrospective Analysis of 245 Melanomas Enrique Rodríguez-Lomba1, Belén Lozano-Masdemont2, Lula María Nieto-Benito1, Elisa Hernández de la Torre1, Ricardo Suárez-Fernández1, José Antonio Avilés-Izquierdo1 1 Department of Dermatology, Hospital General Universitario Gregorio Marañón, Madrid, Spain 2 Department of Dermatology, Hospital Universitario de Móstoles, Madrid, Spain Key words: melanoma, Breslow index, tumor thickness, dermoscopy, dermatoscopy, epiluminescence microscopy Citation: Rodríguez-Lomba E, Lozano-Masdemont B, Nieto-Benito LM, Hernández de la Torre E, Suárez-Fernández R, Avilés-Izquierdo JA. Dermoscopic predictors of tumor thickness in cutaneous melanoma: a retrospective analysis of 245 melanomas. Dermatol Pract Concept. 2021;11(3):e2021059. DOI: https://doi.org/10.5826/dpc.1103a59 Accepted: December 23, 2020; Published: May 20, 2021 Copyright: ©2021 Rodríguez-Lomba et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License BY-NC-4.0, which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original authors and source are credited. Funding: None. Competing interests: The authors have no conflicts of interest to disclose. Authorship: All authors have contributed significantly to this publication. Corresponding author: Enrique Rodríguez-Lomba, MD, Department of Dermatology, Hospital General Universitario Gregorio Marañón, C/ Dr. Esquerdo, 46 – 28007 Madrid, Spain. Email: enriquerlomba@outlook.com Introduction: The literature regarding the association of dermoscopic structures with Breslow thick- ness in melanoma is scarce, limited to small case series, and mostly outdated. Objective: This study determined the dermoscopic patterns, colors and structures that are associated with melanoma in situ, thin melanomas (<0.8 mm) and thick melanomas potentially requiring sentinel lymph node biopsy according to current guidelines (≥0.8 mm). Methods: A retrospective evaluation of 245 dermoscopic images of primary cutaneous melanoma located on the trunk or limbs was performed by consensus of 2 dermoscopists. Results: Red-pink, blue-gray and white color, blue-white veil, shiny white streaks, irregular vessels, blue-black pigmentation, milky red areas, pseudolacunae, ulceration and rainbow pattern were asso- ciated with thickness ≥0.8 mm, whereas atypical pigmented network, regression and hypopigmented areas were significantly associated with early melanomas. Limitations: This is a retrospective study performed in a single institution. Melanomas of special sites were excluded from our evaluation. Dermoscopy is based on subjective evaluations that depend largely on the observers’ experience. Conclusions: The identification of certain dermoscopic structures and colors might help in the dis- crimination between thin and thick melanomas. ABSTRACT 2 Research | Dermatol Pract Concept. 2021;11(3):e2021059 Introduction The literature regarding the association of patterns, colors and dermoscopic structures with Breslow thickness in mel- anoma is scarce, limited to small case series, and mostly based on outdated nomenclature [1-5]. Certain specific dermoscopic structures could help in the early identification of melanomas with tumor thickness ≥0.8 mm, potentially eligible for sentinel lymph node biopsy according to cur- rent guidelines. On the other hand, one author [6] recently criticized the established recommendation of re-excision of clinical safety margins in thin melanomas that have already been completely excised. According to them, a “personal- ized excision” approach would be preferable and should be included in future guidelines [6]. Dermoscopy could be useful in the identification of sharply confined melanomas with a high probability of tumor thickness <0.8 mm, allowing us to excise them in a single surgical procedure with adjusted margins. The aim of this study was to determine the dermo- scopic colors and structures that are associated with both early melanomas (<0.8 mm) and thick melanomas (≥0.8 mm). Materials and Methods A retrospective evaluation of 245 dermoscopic images of primary cutaneous melanoma was performed. All the der- moscopic images were collected from the database of the Melanoma Unit in our department. These were obtained using a digital microscopy system comprising a DermLite Photo II Pro HR dermoscopy lens [3Gen] on an E-420 cam- era (Olympus). The lesion diameter had to be small enough to fit in the whole picture in order to qualify for inclusion. We excluded cases without histopathological confirmation, cases with melanoma metastases, as well as primary mela- nomas of special sites (facial, acral, nail, genital or mucosal melanoma). Images with thick hair density, blood or scales that impeded an adequate dermoscopic evaluation were also excluded. Clinical and histopathological data was obtained from patients’ records and included age at diagnosis, sex, anatomical location of the tumor, and tumor diameter, pal- pability, and Breslow thickness. A list of the dermoscopic criteria established by previous publications was evaluated by consensus of 2 blinded expert dermoscopists (J.A.A.I., E.R.L). The following dermoscopic features were analyzed: colors (light brown, dark brown, black, blue-gray, red-pink, white), asymmetry of color and structures, atypical pigmented network, irregular globules, streaks, irregular blotches, shiny white streaks, negative pig- ment network, blue-white veil, hypopigmented areas, prom- inent skin markings, structureless brown areas, blue-black pigmentation, milky red areas, rainbow pattern, pseudolacu- nae, ulceration, and irregular vessels. Data were analyzed using SPSS version 22.0 (32-bits edition). Univariate analysis for qualitative variables was performed using Pearson’s chi-square test and Fisher’s exact test. A P value less than .05 was considered statistically significant. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated using 2×2 contingency tables. The odds ratio (OR) was calculated for all variables with a confidence interval of 95% (95% CI). Results A total of 245 melanomas were analyzed. Three subtypes were initially defined: melanoma in situ (intraepidermal), thin melanomas (Breslow thickness <0.8 mm), and thick mel- anomas (Breslow thickness ≥0.8 mm). The number of cases in each of these groups was 52 (21.2%), 98 (40.0%), and 95 (38.7%), respectively. The mean Breslow thickness was 0.98 mm (range, 0-9.00 mm). The median Breslow thickness was 0.60 mm, and only 23 melanomas (9.3%) had a Breslow thickness larger than 1.5 mm. Clinical and epidemiological features are presented in Table 1. Tumor diameter and palpability were the only clinical variables that showed statistically significant differences (P = .001 for both). All in situ melanomas were non-palpable. Less than half of the <0.8 mm melanomas were clinically raised (n = 46; 46.9%), while most of the ≥0.8 mm melano- mas were palpable (n = 86; 90.5%). Thick melanomas were more frequently larger than 10 mm (73.6%) than were thin melanomas (50.0%) and melanoma in situ (48.1%). There were no significant differences in age at diagnosis (mean ages, 59.13, 57.55 and 59.19 years, respectively) or sex distribution among the groups. Melanoma in situ and thin melanomas were slightly more frequent in women (53.8% and 53.1%, respectively), while thick melanomas were more prevalent in men (53.7%). Most of the melanomas were located in areas of intermittent sun exposure (n = 225; 91.8%) due to the exclusion criteria in our study design. Table 2 summarizes the frequencies of colors and dermo- scopic structures within each group. Light brown color was more frequent in early melanomas (melanoma in situ and thin melanoma), while blue-gray, red-pink or white color was more commonly observed in thick melanomas (P ˂.05). The presence of three or more colors showed no significant difference between groups. Early melanomas presented a higher frequency of atypical pigmented network, regression and hypopigmented areas. Shiny white streaks, blue-white veil, blue-black pigmentation, milky red areas, rainbow pattern, pseudolacunae, ulceration, irregular vessels and polymorphous vascular pattern were more frequent in thick melanomas (P ˂ .05). No significant differences were observed in color or structure asymmetry between the three groups. Research | Dermatol Pract Concept. 2021;11(3):e2021059 3 Table 1. Clinical and Epidemiological Features, by Tumor Thickness Feature Melanoma in situ (n = 52) Melanoma ˂0.8 mm (n = 98) Melanoma ≥0.8 mm (n = 95) P Mean age, y 59.13 57.55 59.19 NS ≥65 y, n (%) 23 (44.2) 43 (43.9) 43 (45.3) NS Male:Female, n (%) 24:28 (46.2:53.8) 46:52 (46.9:53.1) 51:44 (53.7:46.3) NS Palpability, n (%) 0 (0) 46 (46.9) 86 (90.5) .001 Tumor diameter ≥10 mm, n (%) 25 (48.1) 49 (50.0) 70 (73.6) .001 Tumor location, n (%) NS Chronic exposure 5 (9.6) 7 (7.1) 2 (2.1) Intermittent exposure 45 (86.5) 89 (90.8) 91 (95.8) Non-exposed areas 2 (3.8) 2 (2.0) 2 (2.1) NS = Not significant. Table 2. Frequencies of Colors and Dermoscopic Structures, by Tumor Thickness Feature Melanoma in situ (n = 52) Melanoma ˂0.8 mm (n =98) Melanoma ≥0.8 mm (n = 95) P Color Light brown 50 (96.2) 93 (94.9) 78 (82.1) .003 Dark brown 49 (94.2) 92 (93.9) 79 (83.2) .240 Black 34 (65.4) 47 (48.0) 55 (57.9) .104 Blue-gray 22 (42.3) 58 (59.2) 77 (81.1) .000 Red-pink 6 (11.5) 22 (22.4) 49 (51.6) .000 White Three or more colors 6 (11.5) 42 (80.8) 23 (23.5) 81 (82.7) 41 (43.2) 84 (88.4) .000 .383 Atypical pigmented network 36 (69.2) 41 (41.8) 33 (34.7) .000 Irregular globules 22 (42.3) 51 (52.0) 37 (38.9) .172 Irregular blotches 22 (42.3) 54 (55.1) 40 (42.1) .139 Regression 21 (40.4) 47 (48.0) 24 (25.3) .004 Shiny white streaks 9 (17.3) 35 (35.7) 52 (54.7) .000 Hypopigmented areas 8 (15.4) 15 (15.3) 6 (6.3) .104 Streaks 7 (13.5) 23 (23.5) 18 (18.9) .333 Prominent skin markings 6 (11.5) 7 (7.1) 5 (5.3) .376 Structureless brown areas 5 (9.6) 13 (13.3) 15 (15.8) .576 Blue-white veil 4 (7.7) 17 (17.3) 55 (57.9) .000 Negative pigment network 4 (7.7) 7 (7.1) 7 (7.4) .992 Irregular vessels 4 (7.7) 21 (21.4) 41 (43.2) .000 Blue-black pigmentation 3 (5.8) 5 (5.1) 18 (18.9) .003 Milky red areas 2 (3.8) 14 (14.3) 34 (35.8) .000 Ulceration 0 (0) 4 (4.1) 35 (36.8) .000 Rainbow pattern 0 (0) 6 (6.1) 25 (26.3) .000 Pseudolacunae 0 (0) 3 (3.1) 23 (24.2) .000 Two cohorts were grouped for the calculation of sensi- tivity, specificity, PPV, NPV and OR: early melanomas (n = 150) and thick melanomas (n = 95). Results are presented in Table 3. The presence of three or more colors presented the highest sensitivity (88.4%) for a thickness ≥0.8 mm, although most of the criteria showed sensitivity values lower than 50%. Most of the criteria showed higher than 80% specificity for ≥0.8 mm tumor thickness. The highest spec- ificity was obtained for pseudolacunae (98.0%), ulceration (96.7%) and rainbow pattern (95.3%). PPV was highest for 4 Research | Dermatol Pract Concept. 2021;11(3):e2021059 ulceration (87.5%) and pseudolacunae (88.5%), whereas the other features showed low values. NPV was lower than 80% for all dermoscopic features. The following colors and struc- tures were associated with melanoma thickness ≥0.8 mm: red-pink (OR = 4.641), blue-gray (OR = 3.743), white (OR = 3.168), blue-white veil (OR = 8.446), shiny white streaks (OR = 2.913), irregular vessels (OR = 3.796), blue-black pigmentation (OR = 4.149), milky red areas (OR = 4.668), pseudolacunae (OR = 15.653), ulceration (OR = 16.917) and rainbow pattern (OR = 7.296). On the other hand, atypical pigmented network (OR = 5.505), regression (OR = 0.408) and hypopigmented areas (OR = 0.372) were significantly associated with early melanomas. Discussion Argenziano et al [1] reported for the first time in 1997 the differences of colors and dermoscopic structures in a case series of 72 melanomas (41 thin melanomas <0.75 mm, 31 thick melanomas ≥0.75 mm). The authors reported a higher frequency of pigment network, radial streaming and white scar-like areas in thin melanomas, whereas gray-blue areas, structural asymmetry and a vascular pattern were more fre- quent in thick melanomas. A significant association between the presence of a pigment network and thin melanomas was noted, as well as gray-blue areas and a vascular pattern in thick melanomas. These findings were confirmed in a study of 84 melanomas by Stante et al [2]. A few attempts have been made to reliably predict tumor thickness by dermoscopy, such as the clinical-dermoscopic algorithm published in 1999 by Argenziano et al [3]. The authors reported that the combination of palpability, tumor diameter ≥15 mm, pigment network, gray-blue areas, and atypical vascular pattern increased the prediction accuracy by 14% compared to palpability alone and 9% compared to dermoscopy alone. Others have tried to apply well-known Table 3. Diagnostic Accuracy of Colors and Dermoscopic Structures for Tumor Thickness ≥0.8 mm Feature Sens. Spec. PPV NPV OR (95% CI) Color Red-pink 51.6 81.3 63.6 72.6 4.641 (2.611-8.242) Blue-gray 46.7 81.1 49.0 79.5 3.743 (2.044-6.853) White 43.2 80.7 58.6 69.1 3.168 (1.785-5.626) Black 57.9 46.0 40.4 63.3 1.123 (0.632-1.965) Dark brown 83.2 6.0 35.9 36.0 0.315 (0.133-0.740) Light brown Three or more colors 82.1 88.4 4.7 18.0 35.3 40.6 29.2 71.1 0.225 (0.081-0.562) 1.676 (0.789-3.563) Asymmetry of colors 78.9 22.7 39.3 63.0 1.099 (0.589-2.051) Asymmetry of structures 78.9 16.0 37.3 54.5 0.714 (0.370-1.380) Ulceration 36.8 96.7 87.5 70.7 16.917 (6.326-45.265) Pseudolacunae 24.2 98.0 88.5 67.1 15.653 (4.549-53.854) Blue-white veil 57.9 86.0 72.4 76.3 8.446 (4.568-15.67) Rainbow pattern 26.3 95.3 78.1 67.1 7.296 (3.018-17.687) Milky red areas 35.8 89.3 68.0 68.7 4.668 (2.395-9.096) Blue-black pigmentation 18.9 94.7 69.2 64.8 4.149 (1.724-9.983) Irregular vessels 43.2 83.3 62.1 69.8 3.796 (2.102-6.855) Shiny white streaks 54.7 70.7 54.2 71.1 2.913 (1.702-4.973) Structureless brown areas 15.8 88.0 45.5 62.3 1.375 (0.656-2.880) Negative pigment network 7.4 92.7 38.9 61.2 1.005 (0.376-2.690) Streaks 18.9 80.0 37.5 60.9 0.935 (0.488-1.792) Irregular blotches 42.1 49.3 34.5 57.4 0.708 (0.422-1.189) Irregular globules 38.9 51.3 33.6 57.0 0.673 (0.399-1.134) Prominent skin markings 5.3 91.3 27.8 60.4 0.585 (0.202-1.699) Atypical pigmented network 34.7 48.7 30.0 54.1 0.505 (0.295-0.856) Regression 25.3 54.7 26.1 53.6 0.408 (0.238-0.717) Hypopigmented areas 6.3 84.7 20.7 58.8 0.372 (0.149-0.950) Sens. = sensitivity; Spec. = specificity; PPV = positive predictive value; NPV = negative predictive value; OR = odds ratio; CI = confidence interval. Research | Dermatol Pract Concept. 2021;11(3):e2021059 5 Table 4. Significant Associations Between Specific Dermoscopic Features and Breslow Thickness Melanoma ˂0.8 mm Melanoma ≥0.8 mm Colors Light brown Red-pink Blue-gray White Dermoscopic features Atypical pigmented network Regression Hypopigmented areas Blue-white veil Shiny white streaks Irregular vessels Blue-black pigmentation Milky red areas Rainbow pattern Pseudolacunae Ulceration dermoscopic algorithms to predict tumor thickness. The ABCD rule had an adequate performance in predicting tumor thickness in a series of 84 cutaneous melanomas when a total dermatoscopy score (TDS) cut-off of 6.80 was applied [4]. Despite the initially promising results of the preoperative assessment of melanoma thickness by dermoscopy, few stud- ies of this approach have been published since then. Nomenclature in dermoscopy has expanded over the past decade, and the impact of some dermoscopic structures has not yet been studied. Melanoma guidelines have been updated, and 0.8 mm is now considered the cut-off between T1a and T1b melanomas [7]. The present study expands on this literature and updates it with updated nomenclature. In addition, our study had a bigger sample than previous studies, and we investigated a larger variety of dermoscopic structures. The only clinical features that reached statistical signifi- cance were tumor size ≥10 mm, a slightly lower cut-off than 15 mm as previously reported [3], and palpability. Although in situ melanomas were all non-palpable and more than 90% of the >0.8 mm thick melanomas were clearly palpable, 46.9% of the early invasive melanomas (<0.8 mm) were also palpable. The analysis of this single parameter would have classified these thin melanomas as thick when they were not, and on the contrary, 9.5% of thick melanomas would have been wrongly classified as thin. Table 4 presents the colors and structures associated with melanomas <0.8 mm and ≥0.8 mm. Thin melanomas were associated with light brown color whereas thick melano- mas were associated not only with blue-gray color, but also red-pink and white. However, the clinical and dermoscopic evaluation of colors must take into account the area of exam- ination, both histologically and dermoscopically. It is not uncommon for melanomas, especially superficial-spreading subtypes, to present areas of different thickness throughout their width. Thick melanomas may show light brown areas in their periphery, and blue areas in the thicker center that must always be considered. No statistically significant differences were observed regarding asymmetry of color or structures, unlike a previous study based on digital analysis [8]. This discrepancy might be related to a more rigid interpretation of asymmetry by computerized algorithms than the human eye. Atypical pig- mented network, regression and hypopigmented areas were associated with early melanomas here (Figure 1) as in previ- ous reports [1, 2]. Ulceration, pseudolacunae, blue-white veil, milky red areas, irregular vessels, blue-black pigmentation and shiny white streaks were associated with a melanoma tumor thickness ≥0.8 mm (Figure 2). The latter has already been reported to be associated with thicker melanomas in a case series of 144 melanomas [9]. The evaluation of Figure 1. Dermoscopic predictors of thin melanoma (<0.8 mm tumor thickness). (A) Atypical pigmented network. (B) Regression. (C) Hypopigmented areas. 6 Research | Dermatol Pract Concept. 2021;11(3):e2021059 dermoscopic structures should consider the same principles mentioned previously in color evaluation. Mun et al [10] recently reported the dermoscopic differ- ences between thin and thick acral melanomas (<2 vs. >2 mm) in a cohort of 75 cases. The authors concluded that blue (OR = 7.09), white (OR = 5.04), atypical vessels (OR = 34.58), blue-white veil (OR = 9.60) and ulceration (OR = 5.08) were associated with thick acral melanomas. While our study spe- cifically excluded acral melanomas, most of our results are consistent with theirs. The combination of dermoscopy and other imaging tech- niques such as optical coherence tomography, multispectral imaging and high-frequency ultrasonography could further enhance the preoperative assessment of melanoma patients towards “personalized medicine”. Together with palpability and these specific dermoscopic findings, they could allow a fairly precise tumor thickness prediction. However, caution is recommended before the excision of early melanomas in a single surgical procedure. To date, clinical guidelines do not have a recommendation for this approach, and further evidence is required before standardization. It should be lim- ited to cases with a high malignancy suspicion, and not done in doubtful cases, in order to minimize the risk of causing unnecessary, large scars in benign lesions. The benefits and risks of performing a single-step surgery on a patient need to be adequately addressed before proceeding. Limitations of our study are its retrospective nature and single-institution design. Melanomas of special sites were excluded from our evaluation. Dermoscopy is based on subjective evaluations that depend largely on the observers’ experience. Larger prospective studies focusing on the reli- ability of the combination of certain colors and structures are required to confirm and validate our findings. Conclusions Certain dermoscopic structures and colors might help in the discrimination between thin and thick melanomas. Although none of them are entirely specific to either group, the combination of more than one of them in a single lesion increases the probability of an adequate tumor thickness pre- diction. For example, pigmented lesions presenting palpable blue-white veil, milky red areas and ulceration are unlikely to be early melanomas. On the other hand, lesions presenting an atypical pigmented network and small foci of regression without any other dermoscopic features are very unlikely to be >0.8 mm thickness. The reliability of the combination of certain dermoscopic colors and structures is to be determined, and should be the subject of future studies. References 1. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Delfino M. Epiluminescence microscopy: criteria of cutaneous melanoma progression. J Am Acad Dermatol. 1997;37(1):68-74. DOI: 10.1016/S0190-9622(97)70213-5. 2. Stante M, De Giorgi V, Cappugi P, Gianotti B, Carli P. Non-inva- sive analysis of melanoma thickness by means of dermoscopy: a retrospective study. Melanoma Res. 2001;11(2):147-152. DOI: 10.1097/00008390-200104000-00009. PMID: 11333124. 3. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Delfino M. Clinical and dermatoscopic criteria for the preoperative evalu- ation of cutaneous melanoma thickness. J Am Acad Dermatol. 1999;40(1):61-68. DOI: 10.1016/S0190-9622(99)70528-1. 4. Carli P, De Giorgi V, Palli D, Giannotti V, Giannotti B. Preop- erative assessment of melanoma thickness by ABCD score of dermatoscopy. J Am Acad Dermatol. 2000;43(3):459-466. DOI: 10.1067/mjd.2000.106518. PMID: 10954657. Figure 2. Dermoscopic predictors of thick melanoma (≥0.8 mm tumor thickness). (A) Blue-black pigmentation. (B) Blue-white veil. (C) Shiny white streaks. (D) Irregular vessels. (E) Rainbow pattern. (F) Milky red areas. (G) Ulceration. (F) Pseudolacunae. Research | Dermatol Pract Concept. 2021;11(3):e2021059 7 5. De Giorgi V, Carli P. Dermoscopy and preoperative evaluation of melanoma thickness. Clin Dermatol. 2002;20(3):305-308. DOI: 10.1016/S0738-081X(02)00224-9. 6. Weyers W. “Personalized excision” of malignant melanoma-need for a paradigm shift in the beginning era of personalized medi- cine. Am J Dermatopathol. 2019;41(12):884-896. DOI: 10.1097/ DAD.0000000000001450. PMID: 31490196. 7. Balch CM, Gershenwald JE, Soong S, et al. Final version of 2009 AJCC Melanoma staging and classification. J Clin Oncol. 2009;27(36):6199-6206. DOI: 10.1200/JCO.2009.23.4799. PMID: 19917835. 8. Rubegni P, Cevenini G, Sbano P, et al. Evaluation of cutaneous melanoma thickness by digital dermoscopy analysis: a retrospec- tive study. Melanoma Res. 2010;20(3):212-217. DOI: 10.1097/ CMR.0b013e328335a8ff. PMID: 20375922. 9. Verzi AE, Quan VL, Walton KE, et al. The diagnostic value and histologic correlate of distinct patterns of shiny white streaks for the diagnosis of melanoma: A retrospective, case-control study. J Am Acad Dermatol. 2018;78(5):913-919. DOI: 10.1016/j. jaad.2017.11.021. PMID: 29138058. 10. Mun JH, Jo G, Darmawan CC, et al. Association between Breslow thickness and dermoscopic findings in acral melanoma. J Am Acad Dermatol. 2018;79(5):831-835. DOI: 10.1016/j. jaad.2018.06.004. PMID: 29906546.