Prenatal recognition of cerebriform sebaceous naevus.

Cerebriform sebaceous naevus (CSN) is a rare morphological sebaceous naevus variant and challenging to diagnose prenatally due to its flat, smooth and waxy appearance and lack of association with extracutaneous manifestations.A multigravida was referred to our tertiary obstetric unit at 24 weeks of gestation for evaluation of fetal auricular lesions. We were able to further characterise the lesions via serial obstetric ultrasound imaging with the aid of three-dimensional (3D) technology. Although the precise diagnosis prenatally was uncertain, the use of 3D technology allowed the reconstruction of the fetal cutaneous lesions for multidisciplinary assessment to facilitate the development of a neonatal management plan. The diagnosis of CSN was made postnatally on biopsy.

PRAME Expression in Melanocytic Proliferations in Special Sites.

The subset of nevi occurring at special sites (eg, acral skin, anogeni-tal region, breast, ear, flexural surfaces) have normal histologic variations that preclude the use of routinely used diagnostic criteria for malignancy. Suggested criteria for differentiating malignant special-site lesions from benign lesions have been described, but there is an unmet need for a validated test aiding in the delineation of benign and malignant lesions at special sites. Preferentially expressed antigen of melanoma (PRAME) expression has been characterized as a relatively specific marker of melanoma, but not within the specific population of special-site lesions. This study aimed to determine if PRAME may serve as a specific marker of melanoma within the population of special-sites lesions.

BRAF Mutated and Morphologically Spitzoid Tumors, a Subgroup of Melanocytic Neoplasms Difficult to Distinguish From True Spitz Neoplasms.

Drivers of Spitz neoplasms include activating point mutations in HRAS and Spitz-associated genomic fusions. It has become evident that some BRAF -mutated melanocytic neoplasms can morphologically mimic Spitz tumors (STs). These have been termed BRAF mutated and morphologically spitzoid (BAMS). In this study, 17 experts from the International Melanoma Pathology Study Group assessed 54 cases which included 40 BAMS and 14 true STs. The participants reviewed the cases blinded to the genomic data and selected among several diagnostic options, including BAMS, ST, melanoma, and other. A total of 38% of all diagnostic selections in the BAMS cases were for BAMS, whereas 32% were for ST. In 22 of the BAMS cases, the favored diagnosis was BAMS, whereas in 17 of the BAMS cases, the favored diagnosis was ST. Among the 20 cases in the total group of 54 with the highest number of votes for ST, half were BAMS. Of BAMS, 75% had a number of votes for ST that was within the SD of votes for ST seen among true ST cases. There was poor interobserver agreement for the precise diagnosis of the BAMS (kappa = 0.16) but good agreement that these cases were not melanoma (kappa = 0.7). BAMS nevi/tumors can closely mimic Spitz neoplasms. Expert melanoma pathologists in this study favored a diagnosis of ST in nearly half of the BAMS cases. There are BAMS cases that even experts cannot morphologically distinguish from true Spitz neoplasms.

Federated Learning for Decentralized Artificial Intelligence in Melanoma Diagnostics.

Importance: The development of artificial intelligence (AI)-based melanoma classifiers typically calls for large, centralized datasets, requiring hospitals to give away their patient data, which raises serious privacy concerns. To address this concern, decentralized federated learning has been proposed, where classifier development is distributed across hospitals. Objective: To investigate whether a more privacy-preserving federated learning approach can achieve comparable diagnostic performance to a classical centralized (ie, single-model) and ensemble learning approach for AI-based melanoma diagnostics. Design, Setting, and Participants: This multicentric, single-arm diagnostic study developed a federated model for melanoma-nevus classification using histopathological whole-slide images prospectively acquired at 6 German university hospitals between April 2021 and February 2023 and benchmarked it using both a holdout and an external test dataset. Data analysis was performed from February to April 2023. Exposures: All whole-slide images were retrospectively analyzed by an AI-based classifier without influencing routine clinical care. Main Outcomes and Measures: The area under the receiver operating characteristic curve (AUROC) served as the primary end point for evaluating the diagnostic performance. Secondary end points included balanced accuracy, sensitivity, and specificity. Results: The study included 1025 whole-slide images of clinically melanoma-suspicious skin lesions from 923 patients, consisting of 388 histopathologically confirmed invasive melanomas and 637 nevi. The median (range) age at diagnosis was 58 (18-95) years for the training set, 57 (18-93) years for the holdout test dataset, and 61 (18-95) years for the external test dataset; the median (range) Breslow thickness was 0.70 (0.10-34.00) mm, 0.70 (0.20-14.40) mm, and 0.80 (0.30-20.00) mm, respectively. The federated approach (0.8579; 95% CI, 0.7693-0.9299) performed significantly worse than the classical centralized approach (0.9024; 95% CI, 0.8379-0.9565) in terms of AUROC on a holdout test dataset (pairwise Wilcoxon signed-rank, P < .001) but performed significantly better (0.9126; 95% CI, 0.8810-0.9412) than the classical centralized approach (0.9045; 95% CI, 0.8701-0.9331) on an external test dataset (pairwise Wilcoxon signed-rank, P < .001). Notably, the federated approach performed significantly worse than the ensemble approach on both the holdout (0.8867; 95% CI, 0.8103-0.9481) and external test dataset (0.9227; 95% CI, 0.8941-0.9479). Conclusions and Relevance: The findings of this diagnostic study suggest that federated learning is a viable approach for the binary classification of invasive melanomas and nevi on a clinically representative distributed dataset. Federated learning can improve privacy protection in AI-based melanoma diagnostics while simultaneously promoting collaboration across institutions and countries. Moreover, it may have the potential to be extended to other image classification tasks in digital cancer histopathology and beyond.

Fusion between an Algorithm Based on the Characterization of Melanocytic Lesions' Asymmetry with an Ensemble of Convolutional Neural Networks for Melanoma Detection.

Melanoma is still a major health problem worldwide. Early diagnosis is the first step toward reducing its mortality, but it remains a challenge even for experienced dermatologists. Although computer-aided systems have been developed to help diagnosis, the lack of insight into their predictions is still a significant limitation toward acceptance by the medical community. To tackle this issue, we designed handcrafted expert features representing color asymmetry within the lesions, which are parts of the approach used by dermatologists in their daily practice. These features are given to an artificial neural network classifying between nevi and melanoma. We compare our results with an ensemble of 7 state-of-the-art convolutional neural networks and merge the 2 approaches by computing the average prediction. Our experiments are done on a subset of the International Skin Imaging Collaboration 2019 dataset (6296 nevi, 1361 melanomas). The artificial neural network based on asymmetry achieved an area under the curve of 0.873, sensitivity of 90%, and specificity of 67%; the convolutional neural network approach achieved an area under the curve of 0.938, sensitivity of 91%, and specificity of 82%; and the fusion of both approaches achieved an area under the curve of 0.942, sensitivity of 92%, and specificity of 82%. Merging the knowledge of dermatologists with convolutional neural networks showed high performance for melanoma detection, encouraging collaboration between computer science and medical fields.

Dermoscopic Changes in Melanocytic Lesions in 368 Patients With Atypical Nevus Syndrome and Their Association With Melanoma Incidence: A Cohort Study. / Cambios en la dermatoscopia digital en lesiones melanocíticas en 368 pacientes con diagnóstico de síndrome de nevus atípicos y su asociación con el desarrollo de melanoma: estudio de cohorte.

BACKGROUND AND OBJECTIVE: Atypical nevus syndrome has been described as one of the main risk factors for melanoma. The aim of this study was to analyze dermoscopic changes observed in melanocytic lesions over a follow-up period of 5 years in patients with atypical nevus syndrome. MATERIAL AND METHODS: We conducted a retrospective follow-up study of a cohort of patients seen at a specialized skin cancer and digital body mapping clinic in Medellin, Colombia, between January 2017 and December 2022. We analyzed the dermoscopic changes observed during this period and explored their association with newly diagnosed melanoma. RESULTS: A total of 368 patients (187 women) with a median (interquartile range) age of 43 (37-51) years were included. The dermoscopic features observed at 5 years were an atypical network (222 patients, 60.3%), asymmetric globules (163, 44.2%), white-gray regression areas (105, 28.5%), lesion regression (72, 19.5%), a negative pigment network (59, 16%), asymmetric eccentric pigmentation (28, 7.6%), asymmetric projections (21, 5.7%), and asymmetric vascular patterns (8, 2.1%). Melanoma was diagnosed in 12.2% of patients during follow-up. Features significantly associated with a shorter time to melanoma onset were grayish-white areas (P <.001), asymmetric globules (P=.011), asymmetric eccentric pigmentation (P=.047), and a negative pigment network (P=.001). CONCLUSIONS: The main dermoscopic features of melanocytic lesions in patients with atypical nevus syndrome associated with progression to melanoma were grayish-white areas, asymmetric globules, asymmetric spots, and a negative pigment network.

[Translated article] Dermoscopic Changes in Melanocytic Lesions in 368 Patients With Atypical Nevus Syndrome and Their Association With Melanoma Incidence: A Cohort Study.

BACKGROUND AND OBJECTIVE: Atypical nevus syndrome has been described as one of the main risk factors for melanoma. The aim of this study was to analyze dermoscopic changes observed in melanocytic lesions over a follow-up period of 5 years in patients with atypical nevus syndrome. MATERIAL AND METHODS: We conducted a retrospective follow-up study of a cohort of patients seen at a specialized skin cancer and digital body mapping clinic in Medellin, Colombia, between January 2017 and December 2022. We analyzed the dermoscopic changes observed during this period and explored their association with newly diagnosed melanoma. RESULTS: A total of 368 patients (187 women) with a median (interquartile range) age of 43 (37-51) years were included. The dermoscopic features observed at 5 years were an atypical network (222 patients, 60.3%), asymmetric globules (163, 44.2%), white-gray regression areas (105, 28.5%), lesion regression (72, 19.5%), a negative pigment network (59, 16%), asymmetric eccentric pigmentation (28, 7.6%), asymmetric projections (21, 5.7%), and asymmetric vascular patterns (8, 2.1%). Melanoma was diagnosed in 12.2% of patients during follow-up. Features significantly associated with a shorter time to melanoma onset were grayish-white areas (P<.001), asymmetric globules (P=.011), asymmetric eccentric pigmentation (P=.047), and a negative pigment network (P=.001). CONCLUSIONS: The main dermoscopic features of melanocytic lesions in patients with atypical nevus syndrome associated with progression to melanoma were grayish-white areas, asymmetric globules, asymmetric spots, and a negative pigment network.

Unusual conjunctival melanocytic proliferations. Report of five cases and review of the literature.

Indeterminate melanocytic proliferations of the conjunctiva have both benign and malignant features that previously made these lesions nearly impossible to categorize in existing classification schemes. With the evolution of immunohistochemistry and molecular genetics, however, subclassifications have emerged that allow for a more tailored diagnosis and management. These conjunctival melanocytic proliferations include deep penetrating nevus, granular cell nevus, and nevoid melanoma. There remains a small subset of conjunctival melanocytic proliferations that defy precise characterization as nevi, primary acquired melanosis, or melanomas despite currently available ancillary diagnostic modalities and remain indeterminate. We highlight these unusual types of nevi and melanomas, with an update on their morphologic, immunohistochemical, and molecular genetic characteristics.

Lipid fingerprint-based histology accurately classifies nevus, primary melanoma, and metastatic melanoma samples.

Probably, the most important factor for the survival of a melanoma patient is early detection and precise diagnosis. Although in most cases these tasks are readily carried out by pathologists and dermatologists, there are still difficult cases in which no consensus among experts is achieved. To deal with such cases, new methodologies are required. Following this motivation, we explore here the use of lipid imaging mass spectrometry as a complementary tool for the aid in the diagnosis. Thus, 53 samples (15 nevus, 24 primary melanomas, and 14 metastasis) were explored with the aid of a mass spectrometer, using negative polarity. The rich lipid fingerprint obtained from the samples allowed us to set up an artificial intelligence-based classification model that achieved 100% of specificity and precision both in training and validation data sets. A deeper analysis of the image data shows that the technique reports important information on the tumor microenvironment that may give invaluable insights in the prognosis of the lesion, with the correct interpretation.

The role of R21 expression in differential diagnosis of melanocytic lesions.

Background: Cyclic adenosine monophosphate (cAMP) is an intracellular signal transmitter involved in the regulation of melanocyte growth, proliferation, and melanogenesis. R21 is a monoclonal antibody against the soluble adenylyl cyclase (sAC) protein. Various nuclear and cytoplasmic R21 expression patterns in melanocytic lesions have been previously reported. Pan-nuclear staining was defined as specific for melanoma and was found supportive in the assessment of surgical margins. Aims: The aim of this study is to evaluate the different expression patterns of R21 immunostain and investigate its effectiveness in the differential diagnosis of cutaneous malignant and benign melanocytic lesions. Settings and Design: Fifty invasive cutaneous melanoma and 50 benign melanocytic proliferation were included in the study. Materials and Methods: Paraffin blocks that best reflected tumor morphology were studied via immunohistochemical staining for R21. For all patterns, the cases showing staining in 25% or more tumor cells were considered as positive. Statistical Analysis used: Yates' Chi-square, Pearson Chi-square exact test, Spearman correlation were used. Results and Conclusions: Dot-like Golgi staining was characteristic for nevi (12/50) and seen only in one melanoma. Pan-nuclear staining was striking for melanoma (36/50). This pattern was observed in 2 dysplastic and 3 common melanocytic nevi too. None of the Spitz nevi expressed R21 in pan-nuclear pattern. For the diagnosis of melanoma, sensitivity and specificity of the pan-nuclear expression were 72% and 90%, respectively. Positive and negative predictive values were found as 87% and 76%. R21, a second-generation immunohistochemical marker, can be used in the differential diagnosis of benign and malignant melanocytic lesions.

Generation of a Melanoma and Nevus Data Set From Unstandardized Clinical Photographs on the Internet.

Importance: Artificial intelligence (AI) training for diagnosing dermatologic images requires large amounts of clean data. Dermatologic images have different compositions, and many are inaccessible due to privacy concerns, which hinder the development of AI. Objective: To build a training data set for discriminative and generative AI from unstandardized internet images of melanoma and nevus. Design, Setting, and Participants: In this diagnostic study, a total of 5619 (CAN5600 data set) and 2006 (CAN2000 data set; a manually revised subset of CAN5600) cropped lesion images of either melanoma or nevus were semiautomatically annotated from approximately 500 000 photographs on the internet using convolutional neural networks (CNNs), region-based CNNs, and large mask inpainting. For unsupervised pretraining, 132 673 possible lesions (LESION130k data set) were also created with diversity by collecting images from 18 482 websites in approximately 80 countries. A total of 5000 synthetic images (GAN5000 data set) were generated using the generative adversarial network (StyleGAN2-ADA; training, CAN2000 data set; pretraining, LESION130k data set). Main Outcomes and Measures: The area under the receiver operating characteristic curve (AUROC) for determining malignant neoplasms was analyzed. In each test, 1 of the 7 preexisting public data sets (total of 2312 images; including Edinburgh, an SNU subset, Asan test, Waterloo, 7-point criteria evaluation, PAD-UFES-20, and MED-NODE) was used as the test data set. Subsequently, a comparative study was conducted between the performance of the EfficientNet Lite0 CNN on the proposed data set and that trained on the remaining 6 preexisting data sets. Results: The EfficientNet Lite0 CNN trained on the annotated or synthetic images achieved higher or equivalent mean (SD) AUROCs to the EfficientNet Lite0 trained using the pathologically confirmed public data sets, including CAN5600 (0.874 [0.042]; P = .02), CAN2000 (0.848 [0.027]; P = .08), and GAN5000 (0.838 [0.040]; P = .31 [Wilcoxon signed rank test]) and the preexisting data sets combined (0.809 [0.063]) by the benefits of increased size of the training data set. Conclusions and Relevance: The synthetic data set in this diagnostic study was created using various AI technologies from internet images. A neural network trained on the created data set (CAN5600) performed better than the same network trained on preexisting data sets combined. Both the annotated (CAN5600 and LESION130k) and synthetic (GAN5000) data sets could be shared for AI training and consensus between physicians.

Immunohistochemistry for PRAME in Dermatopathology.

ABSTRACT: Preferentially expressed antigen in melanoma (PRAME) is a tumor-associated antigen first identified in a melanoma patient and found to be expressed in most melanomas as well as in variable levels in other malignant neoplasms of epithelial, mesenchymal, or hematolymphoid lineage. Detection of PRAME expression in formalin-fixed paraffin-embedded tissue is possible by immunohistochemistry (IHC) with commercially available monoclonal antibodies. In situ and invasive melanoma frequently show a diffuse pattern of nuclear PRAME immunoreactivity which contrasts with the infrequent and typically nondiffuse staining seen in nevi. In many challenging melanocytic tumors, results of PRAME IHC and other ancillary tests correlate well, but not always: The tests are not interchangeable. Most metastatic melanomas are positive for PRAME, whereas nodal nevi are not. Numerous studies on PRAME IHC have become available in the past few years with results supporting the value of PRAME IHC as an ancillary tool in the evaluation of melanocytic lesions and providing insights into limitations in sensitivity and specificity as well as possible pitfalls that need to be kept in mind by practicing pathologists.

Refining the application of PRAME-a useful marker in high CSD and acral melanoma subtypes.

Pathologic discordance affecting patient management may approach 20% in melanocytic cases following specialist review. The diagnostic utility of PRAME has been highlighted in several studies but interpretative challenges exist including its use in severely dysplastic compound nevi showing progression to melanoma in situ, nevoid melanoma, and coexisting nevi with melanoma. We examine the PRAME status of a broad spectrum of melanocytic lesions including challenging, dysplastic nevi with severe atypia from a large Irish patient cohort. Retrospective review of the dermatopathology database was conducted to evaluate the PRAME staining characteristics of two hundred and twenty-one melanocytic lesions using a commercially available PRAME antibody (EPR20330). The proportion of nuclear labeling and intensity of staining was recorded. The sensitivity and specificity of PRAME for in situ and malignant melanocytic lesions was 77% and 100%, respectively. Virtually all of our melanoma in situ from high-cumulative sun damaged (CSD) skin (22/23) and all acral lentiginous melanoma (5/5) were PRAME positive while 80% (8/10) of our lentigo maligna melanoma showed diffuse expression. None of our benign subgroup showed diffuse immunoexpression (0/82), including thirty-seven moderate or severely dysplastic nevi. In all cases of melanoma in situ arising in association with a dysplastic compound nevus (0/10), no immunoexpression was observed in the nevic component while in five cases of melanoma in situ with coexistent, intradermal nevus immunostaining was confined to the in situ component. A total of 100% (2/2) of desmoplastic melanomas and 50% (4/8) of nodular melanomas were PRAME positive. PRAME is a sensitive and highly specific immunostain in the diagnosis of in situ and invasive melanoma and we emphasize its application in the evaluation of high CSD and acral melanoma subtypes as well as in challenging threshold cases.

Genetic and phenotypic diversities of nevus spilus phenotypes: Case series and a proposed diagnostic algorithm.

Nevus spilus (NS) is composed of multiple types that characterized by a congenital hyperpigmented patch within variable even superimposed lesions originating from melanocytic lineage cells. The molecular mechanism and classification of diverse NS phenotypes remain unclear. Five children with a phenotype of NS were genotyped by the panel based on next-generation sequencing in this study. DNA from biopsies, blood samples and hair follicle were sequenced to confirm the presence of a somatic mutation. Sequencing results indicated somatic mutation in the gene of NRAS or HRAS in all biopsies from the nevi, and the pathogenic variants were not detected in the samples of blood and hair follicle. This study successfully identified the somatic mutation in five unrelated children with diverse NS phenotypes. Moreover, it provided typical images and differential diagnoses between variable NS phenotypes in clinical, pathological, and genetic features, and first proposed a clinical diagnostic algorithm that contributed to simplifying and optimizing the diagnoses and management of these overlapped diseases.

PRAME and LEF1 in Combined Deep Penetrating Nevus and Combined Blue Nevus: Utility and Pitfalls.

ABSTRACT: Deep penetrating nevi (DPN), particularly those showing combined features, or combined deep penetrating nevi (CDPN), may show histopathological resemblance to blue nevus (BN) and melanoma. Preferentially Expressed Antigen in MElanoma (PRAME) is a marker that helps distinguish melanoma from benign melanocytic lesions. Lymphoid enhancer-binding factor 1 (LEF1) has been proposed to be used in conjunction with ß-catenin for diagnosis of DPN. The immunohistochemical expression of PRAME and LEF1 was evaluated in 10 DPN (including 6 CDPN and 2 DPN-like proliferations with atypical features), 16 BN (including combined and cellular BN), and 2 melanomas with features of DPN or BN. PRAME was negative in most DPN (n = 10/10, n = 9/10, one case with discrepancy between readers) and all BN (n = 16/16), while the 2 melanomas included were positive (n = 2/2). All DPN were positive for LEF1 (n = 9/9) while only a subset of BN were positive (n = 6/16, P = 0.0028; n = 5/16, P = 0.001, per both readers). LEF1 seemed to be easier to interpret than ß-catenin because of its nuclear pattern of expression. The expression of LEF1 in the regular nevus component of combined BN presents a potential pitfall in practice because it may lead to misinterpretation of LEF1 as positive in the BN component of the lesion. However, a subset (approximately one-third) of combined BN seemed to show true LEF1 expression. Taking into account pitfalls in interpretation, the combinatorial panel of PRAME and LEF1, in addition to conventional histopathological features, may be useful to distinguish CDPN from combined BN and other benign and malignant mimics.