Search

Background: Some of the most common cutaneous neoplasms are Bowen’s disease and seborrheic keratosis, a malignant and a benign proliferation, respectively. These entities represent a significant fraction of a dermatopathologists’ workload, and in some cases, histological differentiation may be challenging. The potential of deep learning networks to distinguish these diseases is assessed. Methods: In total, 1935 whole-slide images from three institutions were scanned on two different slide scanners. A U-Net-based segmentation deep learning algorithm was trained on data from one of the centers to differentiate Bowen’s disease, seborrheic keratosis, and normal tissue, learning from annotations performed by dermatopathologists. Optimal thresholds for the class distinction of diagnoses were extracted and assessed on a test set with data from all three institutions. Results: We aimed to diagnose Bowen’s diseases with the highest sensitivity. A good performance was observed across all three centers, underlining the model’s robustness. In one of the centers, the distinction between Bowen’s disease and all other diagnoses was achieved with an AUC of 0.9858 and a sensitivity of 0.9511. Seborrheic keratosis was detected with an AUC of 0.9764 and a sensitivity of 0.9394. Nevertheless, distinguishing irritated seborrheic keratosis from Bowen’s disease remained challenging. Conclusions: Bowen’s disease and seborrheic keratosis could be correctly identified by the evaluated deep learning model on test sets from three different centers, two of which were not involved in training, and AUC scores > 0.97 were obtained. The method proved robust to changes in the staining solution and scanner model. We believe this demonstrates that deep learning algorithms can aid in clinical routine; however, the results should be confirmed by qualified histopathologists.

(1) Background: Melanoma has the highest mortality of all cutaneous tumors, despite recent treatment advances. Many relevant genetic events have been identified in the last decade, including recurrent ARID1A mutations, which in various tumors have been associated with improved outcomes to immunotherapy. (2) Methods: Retrospective analysis of 116 melanoma samples harboring ARID1A mutations. Assessment of clinical and genetic characteristics was performed as well as correlations with treatment outcome applying Kaplan-Meier (log-rank test), Fisher's exact and Chi-squared tests. (3) Results: The majority of ARID1A mutations were in cutaneous and occult melanoma. ARID1A mutated samples had a higher number of mutations than ARID1A wild-type samples and harbored UV-mutations. A male predominance was observed. Many samples also harbored NF1 mutations. No apparent differences were noted between samples harboring genetically inactivating (frame-shift or nonsense) mutations and samples with other mutations. No differences in survival or response to immunotherapy of patients with ARID1A mutant melanoma were observed. (4) Conclusions: ARID1A mutations primarily occur in cutaneous melanomas with a higher mutation burden. In contrast to findings in other tumors, our data does not support ARID1A mutations being a biomarker of favorable response to immunotherapies in melanoma. Larger prospective studies would still be warranted.

PMCA4 is a critical regulator of Ca 2+ homeostasis in mammalian cells. While its biological and prognostic relevance in several cancer types has already been demonstrated, only preclinical investigations suggested a metastasis suppressor function in melanoma. Therefore, we studied the expression pattern of PMCA4 in human skin, nevus, as well as in primary and metastatic melanoma using immunohistochemistry. Furthermore, we analyzed the prognostic power of PMCA4 mRNA levels in cutaneous melanoma both at the non-metastatic stage as well as after PD-1 blockade in advanced disease. PMCA4 localizes to the plasma membrane in a differentiation dependent manner in human skin and mucosa, while nevus cells showed no plasma membrane staining. In contrast, primary cutaneous, choroidal and conjunctival melanoma cells showed specific plasma membrane localization of PMCA4 with a wide range of intensities. Analyzing the TCGA cohort, PMCA4 mRNA levels showed a gender specific prognostic impact in stage I-III melanoma. Female patients with high transcript levels had a significantly longer progression-free survival. Melanoma cell specific PMCA4 protein expression is associated with anaplasticity in melanoma lung metastasis but had no impact on survival after lung metastasectomy. Importantly, high PMCA4 transcript levels derived from RNA-seq of cutaneous melanoma are associated with significantly longer overall survival after PD-1 blockade. In summary, we demonstrated that human melanoma cells express PMCA4 and PMCA4 transcript levels carry prognostic information in a gender specific manner.

Acral naevi are benign melanocytic tumors occurring at acral sites. Occasionally they can progress to become malignant tumors (melanomas). The genetics of acral naevi have not been assessed in larger studies. In our study, a large cohort of 130 acral naevi was screened for gene mutations known to be important in other naevi and melanoma subtypes by targeted next-generation sequencing. Mutation status was correlated with clinicopathological parameters. Frequent mutations in genes activating the MAP kinase pathway were identified, including n = 87 (67%) BRAF , n = 24 (18%) NRAS , and one (1%) MAP2K1 mutations. BRAF mutations were almost exclusively V600E ( n = 86, 99%) and primarily found in junctional and compound naevi. NRAS mutations were either Q61K or Q61R and frequently identified in dermal naevi. Recurrent non-V600E BRAF , KIT , NF1 , and TERT promoter mutations, present in acral melanoma, were not identified. Our study identifies BRAF and NRAS mutations as the primary pathogenic event in acral naevi, however, distributed differently to those in non-acral naevi. The mutational profile of acral naevi is distinct from acral melanoma, which may be of diagnostic value in distinguishing these entities.