Background: The diagnostic performance of convolutional neural networks (CNNs) for diagnosing several types of skin neoplasms has been demonstrated as comparable with that of dermatologists using clinical photography. However, the generalizability should be demonstrated using a large-scale external dataset that includes most types of skin neoplasms. In this study, the performance of a neural network algorithm was compared with that of dermatologists in both real-world practice and experimental settings. Methods and findings: To demonstrate generalizability, the skin cancer detection algorithm (https://rcnn.modelderm.com) developed in our previous study was used without modification. We conducted a retrospective study with all single lesion biopsied cases (43 disorders; 40,331 clinical images from 10,426 cases: 1,222 malignant cases and 9,204 benign cases); mean age (standard deviation [SD], 52.1 [18.3]; 4,701 men [45.1%]) were obtained from the Department of Dermatology, Severance Hospital in Seoul, Korea between January 1, 2008 and March 31, 2019. Using the external validation dataset, the predictions of the algorithm were compared with the clinical diagnoses of 65 attending physicians who had recorded the clinical diagnoses with thorough examinations in real-world practice. In addition, the results obtained by the algorithm for the data of randomly selected batches of 30 patients were compared with those obtained by 44 dermatologists in experimental settings; the dermatologists were only provided with multiple images of each lesion, without clinical information. With regard to the determination of malignancy, the area under the curve (AUC) achieved by the algorithm was 0.863 (95% confidence interval [CI] 0.852–0.875), when unprocessed clinical photographs were used. The sensitivity and specificity of the algorithm at the predefined high-specificity threshold were 62.7% (95% CI 59.9–65.1) and 90.0% (95% CI 89.4–90.6), respectively. Furthermore, the sensitivity and specificity of the first clinical impression of 65 attending physicians were 70.2% and 95.6%, respectively, which were superior to those of the algorithm (McNemar test; p < 0.0001). The positive and negative predictive values of the algorithm were 45.4% (CI 43.7–47.3) and 94.8% (CI 94.4–95.2), respectively, whereas those of the first clinical impression were 68.1% and 96.0%, respectively. In the reader test conducted using images corresponding to batches of 30 patients, the sensitivity and specificity of the algorithm at the predefined threshold were 66.9% (95% CI 57.7–76.0) and 87.4% (95% CI 82.5–92.2), respectively. Furthermore, the sensitivity and specificity derived from the first impression of 44 of the participants were 65.8% (95% CI 55.7–75.9) and 85.7% (95% CI 82.4–88.9), respectively, which are values comparable with those of the algorithm (Wilcoxon signed-rank test; p = 0.607 and 0.097). Limitations of this study include the exclusive use of high-quality clinical photographs taken in hospitals and the lack of ethnic diversity in the study population. Conclusions: Our algorithm could diagnose skin tumors with nearly the same accuracy as a dermatologist when the diagnosis was performed solely with photographs. However, as a result of limited data relevancy, the performance was inferior to that of actual medical examination. To achieve more accurate predictive diagnoses, clinical information should be integrated with imaging information.

Asymmetric small interfering RNAs (asiRNAs) that mediate RNA interference have been investigated for therapeutic use in various tissues, including skin tissue. Androgenetic alopecia (AGA) is caused by a combination of genetic factors, resulting in sensitivity to dihydrotestosterone (DHT), which binds to the androgen receptor (AR) to mediate a series of biomolecular changes leading to hair loss. This study aimed to evaluate the therapeutic potential of a cell-penetrating, AR-targeting asiRNA (cp-asiAR) for AGA treatment, which was designed to silence the AR gene. AGA mouse models were developed by stimulation with DHT, and ex vivo human scalp tissues were also used for analysis. Cp-asiAR-mediated changes in mRNA expression and protein levels of AR were assessed along with the examination of phenotypic improvements in mouse model of AGA. We also assessed downstream signaling associated with AR in primary human dermal papilla (DP) cells. Several cp-asiARs were screened for selecting the optimal sequence of AR using cell lines in vitro. A cholesterol-conjugated, chemically modified cp-asiAR candidate was optimized under passive uptake conditions in vitro. Intradermal cp-asiAR injection efficiently reduced mRNA and protein levels corresponding to AR in mouse models. Moreover, cp-asiAR injection promoted hair growth in mouse models with DHT-induced AGA. In ex vivo human hair follicle culture, the proportion of telogen hair decreased, and the mean hair bulb diameter increased in the cp-asiAR-treated group. In isolated primary human DP cells, AR expression was effectively downregulated by cp-asiAR. Furthermore, cp-asiAR attenuated DHT-mediated increases in interleukin-6, transforming growth factor-β1, and dickkopf-1 levels. No significant toxicity was observed in DP cells after cp-asiAR treatment. Cp-asiAR treatment showed effective downregulation of AR expression and prevention of DHT-mediated alterations in the hair cycle and hair diameter, indicating its potential as a novel therapeutic option for AGA.

Glutathione in its reduced form (GSH) is an antioxidant and also is involved in pheomelanin formation. Thus, it has been long believed that GSH has a skin whitening effect. However, its actual or direct effect is unproven. We evaluated the anti-melanogenic effects of GSH and its derivatives in vitro. We examined change of melanogenesis and its related proteins by GSH itself and its derivatives, including GSH monoethyl ester (GSH-MEE), GSH diethyl ester (GSH-DEE) and GSH monoisopropyl ester (GSH-MIPE) in Melan-A cells, Mel-Ab cells, and B16F10 cells. GSH and GSH-MEE did not display cytotoxic activity, but GSH-MIPE and GSH-DEE did. Intriguingly, GSH itself had no inhibitory effect on melanin production or intracellular tyrosinase activity. Rather, it was GSH-MEE and GSH-MIPE that profoundly reduced the amount of melanin and intracellular tyrosinase activity. Thus, GSH-MEE was selected as a suitable candidate skin-whitening agent and it did not alter melanogenesis-associated proteins such as microphthalmia-associated transcription factor (MITF), tyrosinase, tyrosinase-related protein (TRP)-1, and TRP-2, but it did increase the amount of suggested pheomelanin and suggested pheomelanin/eumelanin ratio. GSH-MEE was effective for anti-melanogenesis, whereas GSH itself was not. GSH-MEE could be developed as a safe and efficient agent for the treatment of hyperpigmentation skin disorders.

Catecholamines function via G protein-coupled receptors, triggering an increase in intracellular levels of 3′,5′-cyclic adenosine monophosphate (cAMP) in various cells. Catecholamine biosynthesis and the β-adrenergic receptor exist in melanocytes; thus, catecholamines may play critical roles in skin pigmentation. However, their action and mechanisms mediating melanogenesis in human skin have not yet been investigated. Therefore, we examined the potential anti-melanogenetic effect of carvedilol, a nonselective β-blocker with weak α1-blocking activities. Carvedilol reduced melanin content and cellular tyrosinase activity without compromising cellular viability in normal human melanocytes as well as in mel-Ab immortalized mouse melanocytes. Carvedilol downregulated microphthalmia-associated transcription factor (MITF), tyrosinase, tyrosinase-related protein (TRP)-1, and TRP-2. Carvedilol treatment led to the downregulation of phosphor-cAMP response element-binding protein (CREB). Moreover, the increase in cAMP levels upon treatment with forskolin reversed the anti-melanogenic action of carvedilol. In addition, carvedilol remarkably reduced the melanin index in ultraviolet-irradiated human skin cultures. Taken together, our results indicate that carvedilol effectively suppresses melanogenesis in human melanocytes and ex vivo human skin by inhibiting cAMP/protein kinase A/CREB signaling. The anti-melanogenic effects of carvedilol have potential significance for skin whitening agents.