Convolutional neural networks (CNNs) have outperformed dermatologists in classifying pigmented skin lesions under artificial conditions. We investigated, for the first time, the performance of three-dimensional (3D) and two-dimensional (2D) CNNs and dermatologists in the early detection of melanoma in a real-world setting. In this prospective study, 1690 melanocytic lesions in 143 patients with high-risk criteria for melanoma were evaluated by dermatologists, 2D-FotoFinder-ATBM and 3D-Vectra WB360 total body photography (TBP). Excision was based on the dermatologists' dichotomous decision, an elevated CNN risk score (study-specific malignancy cut-off: FotoFinder >0.5, Vectra >5.0) and/or the second dermatologist's assessment with CNN support. The diagnostic accuracy of the 2D and 3D CNN classification was compared with that of the dermatologists and the augmented intelligence based on histopathology and dermatologists' assessment. Secondary end-points included reproducibility of risk scores and naevus counts per patient by medical staff (gold standard) compared to automated 3D and 2D TBP CNN counts. The sensitivity, specificity, and receiver operating characteristics area under the curve (ROC-AUC) for risk-score-assessments compared to histopathology of 3D-CNN with 95% confidence intervals (CI) were 90.0%, 64.6% and 0.92 (CI 0.85-1.00), respectively. While dermatologists and augmented intelligence achieved the same sensitivity (90%) and comparable classification ROC-AUC (0.91 [CI 0.80-1.00], 0.88 [CI 0.77-1.00]) with 3D-CNN, their specificity was superior (92.3% and 86.2%, respectively). The 2D-CNN (sensitivity: 70%, specificity: 40%, ROC-AUC: 0.68 [CI 0.46-0.90]) was outperformed by 3D CNN and dermatologists. The 3D-CNN showed a higher correlation coefficient for repeated measurements of 246 lesions (R = 0.89) than the 2D-CNN (R = 0.79). The mean naevus count per patient varied significantly (gold standard: 210 lesions; 3D-CNN: 469; 2D-CNN: 1324; p < 0.0001). Our study emphasises the importance of validating the classification of CNNs in real life. The novel 3D-CNN device outperformed the 2D-CNN and achieved comparable sensitivity with dermatologists. The low specificity of CNNs and the lack of automated counting of TBP nevi currently limit the use of augmented intelligence in clinical practice.

The exponential increase in algorithm-based mobile health (mHealth) applications (apps) for melanoma screening is a reaction to a growing market. However, the performance of available apps remains to be investigated. In this prospective study, we investigated the diagnostic accuracy of a class 1 CE-certified smartphone app in melanoma risk stratification and its patient and dermatologist satisfaction. Pigmented skin lesions ≥ 3 mm and any suspicious smaller lesions were assessed by the smartphone app SkinVision^® (SkinVision^® B.V., Amsterdam, the Netherlands, App-Version 6.8.1), 2D FotoFinder ATBM^® master (FotoFinder ATBM^® Systems GmbH, Bad Birnbach, Germany, Version 3.3.1.0), 3D Vectra^® WB360 (Canfield Scientific, Parsippany, NJ, USA, Version 4.7.1) total body photography (TBP) devices, and dermatologists. The high-risk score of the smartphone app was compared with the two gold standards: histological diagnosis, or if not available, the combination of dermatologists’, 2D and 3D risk assessments. A total of 1204 lesions among 114 patients (mean age 59 years; 51% females (55 patients at high-risk for developing a melanoma, 59 melanoma patients)) were included. The smartphone app’s sensitivity, specificity, and area under the receiver operating characteristics (AUROC) varied between 41.3–83.3%, 60.0–82.9%, and 0.62–0.72% according to two study-defined reference standards. Additionally, all patients and dermatologists completed a newly created questionnaire for preference and trust of screening type. The smartphone app was rated as trustworthy by 36% (20/55) of patients at high-risk for melanoma, 49% (29/59) of melanoma patients, and 8.8% (10/114) of dermatologists. Most of the patients rated the 2D TBP imaging (93% (51/55) resp. 88% (52/59)) and the 3D TBP imaging (91% (50/55) resp. 90% (53/59)) as trustworthy. A skin cancer screening by combination of dermatologist and smartphone app was favored by only 1.8% (1/55) resp. 3.4% (2/59) of the patients; no patient preferred an assessment by a smartphone app alone. The diagnostic accuracy in clinical practice was not as reliable as previously advertised and the satisfaction with smartphone apps for melanoma risk stratification was scarce. MHealth apps might be a potential medium to increase awareness for melanoma screening in the lay population, but healthcare professionals and users should be alerted to the potential harm of over-detection and poor performance. In conclusion, we suggest further robust evidence-based evaluation before including market-approved apps in self-examination for public health benefits.

Cutaneous lichenoid drug eruptions (LDE) are adverse drug reactions (ADR) characterized by symmetric, erythematous, violaceous papules reminiscent but rarely fully characteristic of lichen planus (LP). We aimed to analyse the literature describing cases of LDE within the last 20 years to provide additional insight into culprit drugs, typical latency to onset of the eruption, the spectrum of clinical presentations, severity and management. A literature search was conducted in MEDLINE between January 2000 and 27 January 2021. The keywords ‘lichenoid drug rash’ and ‘lichenoid drug eruption’ were used. Cases were included if LDE diagnosis was made, and culprit drugs were identified. A total of 323 cases with LDE were identified from 163 published case reports and studies. The mean patient age was 58.5 years (1 month to 92 years), and 135 patients (41.8%) were female. Checkpoint inhibitors (CKI) were the most frequently reported culprit drugs (136 cases; 42.1%), followed by tyrosine kinase inhibitors (TKI) (39 cases; 12.0%) and anti‐TNF‐α‐monoclonal antibodies (13 cases; 4.0%). The latency between initiation of the drug and manifestation was 15.7 weeks (range: 0.1–208 weeks). After discontinuing the culprit drug, the median time to resolution was 14.2 weeks (range: 0.71–416 weeks). One hundred thirty‐six patients (42.1%) were treated with topical, and 54 patients (16.7%) with systemic glucocorticoids. Overall, we conclude that, albeit rare, LDE is challenging to diagnose ADR induced by mostly CKI, TKI, and biologics. Treatment modalities resemble that of lichen planus, and the culprit drugs had to be discontinued in only 26%, which is low compared with other types of adverse drug reactions. This is probably due to the low risk of aggravation (e.g. toxic epidermal necrolysis) if the drug is continued and the benefit/risk ratio favouring the drug, as is often the case in cancer therapy.

Vitiligo-like depigmentation (VLD) is an immune-related adverse event (irAE) of checkpoint-inhibitor (CPI) treatment, which has previously been associated with a favourable outcome. The aim of this study was to explore clinical, biological and prognostic features of melanoma patients with VLD under CPI-treatment and to explore whether they exhibit a characteristic immune response profile in peripheral blood. Melanoma patients developing VLD under CPI were included in a prospective observational single-center cohort study. We collected and analysed clinical parameters, photographs and serum from 28 VLD patients. They received pembrolizumab (36%), nivolumab (11%), ipilimumab/nivolumab (32%) or clinical trial medications (21%). We performed a high-throughput proteomics assay (Olink), in which we identified a distinct proteomic signature in VLD patients in comparison to non-VLD CPI patients. Our clinical assessments revealed that VLD lesions had a predominantly symmetrical distribution pattern, with mostly smaller “freckle-like” macules and a preferential distribution in UV-exposed areas. Patients with previous targeted therapy showed a significantly longer time lapse between CPI initiation and VLD onset compared to non-pre-treated patients (12.5 vs. 6.25 months). Therapy responders exhibited a distinct proteomic profile when compared with non-responders in VLD such as upregulation of EDAR and downregulation of LAG3. ITGA11 was elevated in the VLD-group when compared to non-VLD-CPI-treated melanoma patients. Our findings demonstrate that on a proteomic level, VLD is characterized by a distinct immune signature when compared to CPI-treated patients without VLD and that therapy responsiveness is reflected by a characteristic immune profile. The pathomechanisms underlying these findings and how they could relate to the antitumoral response in melanoma remain to be elucidated.

The incidence of cutaneous melanoma continues to rise despite the increased use of sunscreens within the last several decades. Some research even suggests that the use of sunscreen is associated with increased rates of melanoma. Given the aggressive, and often deadly, nature of cutaneous melanoma, the aim of this communication is to better elucidate the relationship between sunscreen use and melanoma development and if there are other preventative measures to be aware of. A search was performed to identify the studies that have investigated melanoma development in individuals who used sunscreen and those who did not. Study limitations and possible confounding variables were identified, which guided a subsequent search to determine what data were available to support that these limitations and confounding variables may explain the perplexing association between sunscreen use and melanoma development. Five hypotheses were generated, which were related to increased awareness and reporting, the relationship between sunscreen use and the duration of sun exposure, the importance of broad-spectrum protection, and the effect of sunscreen on reactive oxygen species formation. The main conclusion is that more recent studies that control for confounding variables are required to determine the true effect of adequate broad-spectrum sunscreen use today on the development of melanoma.