Refine
Has Fulltext
- yes (4)
Is part of the Bibliography
- yes (4)
Document Type
- Journal article (3)
- Doctoral Thesis (1)
Language
- English (4) (remove)
Keywords
- skin cancer (4) (remove)
Institute
EU-Project number / Contract (GA) number
- 277775 (1)
Recent research revealed the superiority of artificial intelligence over dermatologists to diagnose melanoma from images. However, 30–50% of all melanomas and more than half of those in young patients evolve from initially benign lesions. Despite its high relevance for melanoma screening, neither clinicians nor computers are yet able to reliably predict a nevus’ oncologic transformation. The cause of this lies in the static nature of lesion presentation in the current standard of care, both for clinicians and algorithms. The status quo makes it difficult to train algorithms (and clinicians) to precisely assess the likelihood of a benign skin lesion to transform into melanoma. In addition, it inhibits the precision of current algorithms since ‘evolution’ image features may not be part of their decision. The current literature reveals certain types of melanocytic nevi (i.e. ‘spitzoid’ or ‘dysplastic’ nevi) and criteria (i.e. visible vasculature) that, in general, appear to have a higher chance to transform into melanoma. However, owing to the cumulative nature of oncogenic mutations in melanoma, a more fine-grained early morphologic footprint is likely to be detectable by an algorithm. In this perspective article, the concept of melanoma prediction is further explored by the discussion of the evolution of melanoma, the concept for training of such a nevi classifier and the implications of early melanoma prediction for clinical practice. In conclusion, the authors believe that artificial intelligence trained on prospective image data could be transformative for skin cancer diagnostics by (a) predicting melanoma before it occurs (i.e. pre-in situ) and (b) further enhancing the accuracy of current melanoma classifiers. Necessary prospective images for this research are obtained via free mole-monitoring mobile apps.
Merkel cell carcinoma (MCC) is a virally associated cancer characterized by its aggressive behavior and strong immunogenicity. Both viral infection and malignant transformation induce expression of MHC class I chain-related protein (MIC) A and B, which signal stress to cells of the immune system via Natural Killer group 2D (NKG2D) resulting in elimination of target cells. However, despite transformation and the continued presence of virally-encoded proteins, MICs are only expressed in a minority of MCC tumors in situ and are completely absent on MCC cell lines in vitro. This lack of MIC expression was due to epigenetic silencing via MIC promoter hypo-acetylation; indeed, MIC expression was re-induced by pharmacological inhibition of histone deacetylases (HDACs) both in vitro and in vivo. This re-induction of MICs rendered MCC cells more sensitive to immune-mediated lysis. Thus, epigenetic silencing of MICs is an important immune escape mechanism of MCCs.
Melanoma is the most aggressive skin cancer with very limited treatment options. Upon appearance of metastases chemotherapeutics are used to either kill or slow down the growth of cancer cells by inducing apoptosis or senescence, respectively. With melanomas originating from melanocytes, it is vital to elucidate the mechanisms that distinguish senescence induction from proliferation and tumourigenicity. Xmrk (Xiphophorus melanoma receptor kinase), the fish orthologue of the human epidermal growth factor receptor (EGFR), causes highly aggressive melanoma in fish. Using an inducible variant, HERmrk, I showed that high receptor levels result in melanocyte senescence, whereas low and medium expression allows for cell proliferation and tumourigenicity. Mechanistically, HERmrk leads to increased reactive oxygen species (ROS) levels, which trigger a DNA damage response. Consequently, multinucleated, senescent cells develop by both endomitosis and fusion. Furthermore, oncogenic N‐RAS (N-‐RAS61K) induces a similar multinucleated phenotype in melanocytes. In addition, I found that both overexpression of C‐MYC and the knockdown of miz‐1 (Myc‐interacting zinc finger protein 1) diminished HERmrk‐induced senescence entry. C‐MYC prevent ROS induction, DNA damage and senescence, while acting synergistically with HERmrk in conveying tumourigenic features to melanocytes. Further analyses identified cystathionase (CTH) as a novel target gene of Myc and Miz-1 crucial for senescence prevention. CTH encodes an enzyme involved in the synthesis of cysteine from methionine, thereby allowing for increased ROS detoxification. Even though senescence was thought to be irreversible and hence tumour protective, I demonstrated that prolonged expression of the melanoma oncogene N‐RAS61K in pigment cells overcomes initial OIS by triggering the emergence of tumour‐initiating, mononucleated stem‐like cells from multinucleated senescent cells. This progeny is dedifferentiated, highly proliferative, anoikis‐resistant and induces fast‐growing, metastatic tumours upon transplantation into nude mice. Our data demonstrate that induction of OIS is not only a cellular failsafe mechanism, but also carries the potential to provide a source for highly aggressive, tumour‐initiating cells.
Background
Melanoma is the most dangerous type of skin cancer but is curable if detected early. Recent publications demonstrated that artificial intelligence is capable in classifying images of benign nevi and melanoma with dermatologist-level precision. However, a statistically significant improvement compared with dermatologist classification has not been reported to date.
Methods
For this comparative study, 4204 biopsy-proven images of melanoma and nevi (1:1) were used for the training of a convolutional neural network (CNN). New techniques of deep learning were integrated. For the experiment, an additional 804 biopsy-proven dermoscopic images of melanoma and nevi (1:1) were randomly presented to dermatologists of nine German university hospitals, who evaluated the quality of each image and stated their recommended treatment (19,296 recommendations in total). Three McNemar's tests comparing the results of the CNN's test runs in terms of sensitivity, specificity and overall correctness were predefined as the main outcomes.
Findings
The respective sensitivity and specificity of lesion classification by the dermatologists were 67.2% (95% confidence interval [CI]: 62.6%–71.7%) and 62.2% (95% CI: 57.6%–66.9%). In comparison, the trained CNN achieved a higher sensitivity of 82.3% (95% CI: 78.3%–85.7%) and a higher specificity of 77.9% (95% CI: 73.8%–81.8%). The three McNemar's tests in 2 × 2 tables all reached a significance level of p < 0.001. This significance level was sustained for both subgroups.
Interpretation
For the first time, automated dermoscopic melanoma image classification was shown to be significantly superior to both junior and board-certified dermatologists (p < 0.001).