A basic requirement for artificial intelligence (AI)-based image analysis systems, which are to be integrated into clinical practice, is a high robustness. Minor changes in how those images are acquired, for example, during routine skin cancer screening, should not change the diagnosis of such assistance systems.To quantify to what extent minor image perturbations affect the convolutional neural network (CNN)-mediated skin lesion classification and to evaluate three possible solutions for this problem (additional data augmentation, test-time augmentation, anti-aliasing).We trained three commonly used CNN architectures to differentiate between dermoscopic melanoma and nevus images. Subsequently, their performance and susceptibility to minor changes ('brittleness') was tested on two distinct test sets with multiple images per lesion. For the first set, image changes, such as rotations or zooms, were generated artificially. The second set contained natural changes that stemmed from multiple photographs taken of the same lesions.All architectures exhibited brittleness on the artificial and natural test set. The three reviewed methods were able to decrease brittleness to varying degrees while still maintaining performance. The observed improvement was greater for the artificial than for the natural test set, where enhancements were minor.Minor image changes, relatively inconspicuous for humans, can have an effect on the robustness of CNNs differentiating skin lesions. By the methods tested here, this effect can be reduced, but not fully eliminated. Thus, further research to sustain the performance of AI classifiers is needed to facilitate the translation of such systems into the clinic.
9555 Background: For the treatment of advanced BRAF V600 -mutated melanoma, targeted therapy (BRAF/MEK-inhibition) is a standard of care. Encorafenib + binimetinib (EB) were approved in the EU in Sep 2018 and in Switzerland in Nov 2019, based on positive results from COLUMBUS (NCT01909453), with a median progression-free survival (PFS) of 14.9 mo (4-year PFS: 26%) and overall survival (OS) of 33.6 mo (4-year OS: 39%). As data from controlled trials are based on selected populations, BERING MELANOMA investigates the use of EB under real-world conditions in a broader population. Methods: BERING MELANOMA is an ongoing, multi-national, multi-center, prospective, longitudinal, non-interventional study. It analyzes the effectiveness, quality of life and tolerability of EB-treatment under real-world conditions (primary endpoint: 1-year PFS-rate), focusing on the first- (1L) and second-line setting and including an evaluation of the impact of prognostic factors. The project aims to enroll up to 750 patients (pts) in a total of 80 German, Austrian and Swiss sites with a study duration of 8 yrs. So far, from Oct 2019 to Jan 2021, 153 pts have been included. Pts with prior BRAF-/MEK-inhibition (except adjuvant use completed > 6 mo) and > 1 prior treatment line were excluded. Results: Here we present the first planned interim analysis based on the initial 100 enrolled pts (91 treated / 89 evaluable; median follow-up: 8.1 mo). This analysis set shows a median age of 63.0 yrs (range 29.0-88.0), 52% of pts were female. 81% presented with distant metastases (brain: 31%), with an involvement of ≥3 organ systems in 51% and an elevated LDH in 42%. 54% of pts underwent prior systemic therapy (adjuvant: 28%; 1L: 24%, with ipilimumab + nivolumab as main 1L-treatment: 52%). EB was mainly administered in the 1L-setting (65%). Main reasons for EB-selection were: physician's preference (37%), efficacy (34%), quality of life (21%). Median estimated EB treatment duration was 12.7 mo (95%CI 8.3-NE), median relative dose intensity was 100% for both drugs. Treatment adaptations were required in 34% of pts. Adverse events (AE) were reported in 76% of pts (grade 3/4: 26%). Main AE (≥10%, all grades) were: nausea (18%), diarrhea (17%), CK increase (15%), fatigue (11%), gamma-GT increase (11%). No fatal toxicities were observed. Conclusions: This first interim analysis of BERING MELANOMA shows an investigation of EB in a real-world population with advanced disease. Despite the poorer prognosis configuration as compared to the pivotal study, the observed treatment duration and tolerability profile are largely consistent with data derived from COLUMBUS without any new safety signals. The second interim analysis will be performed after enrollment of 200 pts and will include an initial analysis of effectiveness data. Clinical trial information: NCT04045691.
Recent years have been witnessing a substantial improvement in the accuracy of skin cancer classification using convolutional neural networks (CNNs). CNNs perform on par with or better than dermatologists with respect to the classification tasks of single images. However, in clinical practice, dermatologists also use other patient data beyond the visual aspects present in a digitized image, further increasing their diagnostic accuracy. Several pilot studies have recently investigated the effects of integrating different subtypes of patient data into CNN-based skin cancer classifiers.
As most chemotherapeutic drugs are ineffective in the treatment of chondrosarcoma, we studied the expression pattern and function of SOX9, the master transcription factor for chondrogenesis, in chondrosarcoma, to understand the basic molecular principles needed for engineering new targeted therapies. Our study shows an increase in SOX9 expression in chondrosarcoma compared to normal cartilage, but a decrease when the tumors are finally defined as dedifferentiated chondrosarcoma (DDCS). In DDCS, SOX9 is almost completely absent in the non-chondroid, dedifferentiated compartments. CRISPR/Cas9-mediated knockout of SOX9 in a human chondrosarcoma cell line (HTB94) results in reduced proliferation, clonogenicity and migration, accompanied by an inability to activate MMP13. In contrast, adhesion, apoptosis and polyploidy formation are favored after SOX9 deletion, probably involving BCL2 and survivin. The siRNA-mediated SOX9 knockdown partially confirmed these results, suggesting the need for a certain SOX9 threshold for particular cancer-related events. To increase the efficacy of chondrosarcoma therapies, potential therapeutic approaches were analyzed in SOX9 knockout cells. Here, we found an increased impact of doxorubicin, but a reduced sensitivity for oncolytic virus treatment. Our observations present novel insight into the role of SOX9 in chondrosarcoma biology and could thereby help to overcome the obstacle of drug resistance and limited therapy options.
(1) Background: Rare skin cancers include epithelial, neuroendocrine, and hematopoietic neoplasias as well as cutaneous sarcomas. Ultraviolet (UV) radiation and sunburns are important drivers for the incidence of certain cutaneous sarcomas; however, the pathogenetic role of UV light is less clear in rare skin cancers compared to keratinocyte cancer and melanoma. In this study, we compared the degree of actinic elastosis (AE) as a surrogate for lifetime UV exposure among selected rare skin cancers (atypical fibroxanthoma [AFX], pleomorphic dermal sarcoma [PDS], dermatofibrosarcoma protuberans [DFSP], Kaposi sarcoma [KS], Merkel cell carcinoma [MCC], and leiomyosarcoma [LMS]) while taking into account relevant clinical variables (age, sex, and body site). (2) Methods: We newly established a semi-quantitative score for the degree of AE ranging from 0 = none to 3 = total loss of elastic fibers (basophilic degeneration) and multiplied it by the perilesional vertical extent (depth), measured histometrically (tumor-associated elastosis grade (TEG)). We matched the TEG of
