Background: Multiple studies have compared the performance of artificial intelligence (AI)ebased models for automated skin cancer classification to human experts, thus setting the cornerstone for a successful translation of AI-based tools into clinicopathological practice.
Until recently, distant metastatic melanoma was considered refractory to systemic therapy. A better understanding of the interactions between tumors and the immune system and the mechanisms of regulation of T-cells led to the development of immune checkpoint inhibitors. This review summarizes the current novel data on the treatment of metastatic melanoma with anti-programmed cell death protein 1 (PD-1) antibodies and anti-PD-1-based combination regimens, including clinical trials presented at major conference meetings. Immune checkpoint inhibitors, in particular anti-PD-1 antibodies such as pembrolizumab and nivolumab and the combination of nivolumab with the anti-cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) antibody ipilimumab can achieve long-term survival for patients with metastatic melanoma. The anti-PD-1 antibodies nivolumab and pembrolizumab were also approved for adjuvant treatment of patients with resected metastatic melanoma. Anti-PD-1 antibodies appear to be well tolerated, and toxicity is manageable. Nivolumab combined with ipilimumab achieves a 5 year survival rate of more than 50% but at a cost of high toxicity. Ongoing clinical trials investigate novel immunotherapy combinations and strategies (e.g., Talimogene laherparepvec (T-VEC), Bempegaldesleukin (BEMPEG), incorporation or sequencing of targeted therapy, incorporation or sequencing of radiotherapy), and focus on poor prognosis groups (e.g., high tumor burden/LDH levels, anti-PD-1 refractory melanoma, and brain metastases).
Background: Clinicians and pathologists traditionally use patient data in addition to clinical examination to support their diagnoses. Objectives: We investigated whether a combination of histologic whole slides image (WSI) analysis based on convolutional neural networks (CNNs) and commonly available patient data (age, sex and anatomical site of the lesion) in a binary melanoma/nevus classification task could increase the performance compared with CNNs alone. Methods: We used 431 WSIs from two different laboratories and analysed the performance of classifiers that used the image or patient data individually or three common fusion techniques. Furthermore, we tested a naive combination of patient data and an image classifier: for cases interpreted as 'uncertain' (CNN output score <0.7), the decision of the CNN was replaced by the decision of the patient data classifier. Results: The CNN on its own achieved the best performance (mean AE standard deviation of five individual runs) with AUROC of 92.30% AE 0.23% and balanced accuracy of 83.17% AE 0.38%. While the classification performance was not significantly improved in general by any of the tested fusions, naive strategy of replacing the image classifier with the patient data classifier on slides with low output scores improved balanced accuracy to 86.72% AE 0.36%. Conclusion:In most cases, the CNN on its own was so accurate that patient data integration did not provide any benefit. However, incorporating patient data for lesions that were classified by the CNN with low 'confidence' improved balanced accuracy.
Considering the rising incidence, cutaneous squamous-cell carcinoma (cSCC) has a high clinical relevance. In patients with localized cSCC, complete surgical resection is indicated. Radiotherapy should be performed in patients with nonresectable tumours or in patients who are not suitable for surgery. Systemic therapy is reserved for cSCC that are neither surgically nor radiotherapeutically curable due to their extensive local spread and/or local or distant metastasis. In the absence of prospective randomized phase 3 trials to evaluate and compare the efficacy and safety of chemotherapeutics, epidermal growth factor receptor (EGFR) inhibitors and anti-PD-1 antibodies, no final recommendation for systemic therapy can be given for patients with locally advanced or metastatic cSCC. Anti-PD-1 antibodies currently show promising results with response rates of up to 50% in both locally advanced and metastatic cSCC. Anti-PD-1 antibodies appear to achieve higher response rates compared with EGFR inhibitors, and the duration of response appears to be superior to both chemotherapy and EGFR inhibitors. Compared with chemotherapy, the side effect profile of anti-PD-1 antibodies appears to be favourable. Altogether, PD-1 inhibitors are expected to become the new standard of care for patients with locally advanced and metastatic cSCC. Currently, placebo-controlled clinical trials are investigating the adjuvant use of cemiplimab and pembrolizumab in patients undergoing resection and radiotherapy of high-risk cSCC.Patients not eligible for anti-PD-1 treatment, e.g. in organ transplant recipients, or in patients refractory to anti-PD-1 may be offered EGFR inhibitors and/or chemotherapies. Chemotherapies appear to be superior to EGFR inhibitors in terms of response rates, whereas EGFR inhibitors have a more favourable toxicity profile. EGFR inhibitors are therefore more suitable for multimorbid and/or frail elderly patients. By combining EGFR inhibitors with local therapy such as surgery or radiotherapy, response rates and duration of response may be improved.
This decade has brought significant survival improvement in patients with metastatic melanoma with targeted therapies and immunotherapies. As our understanding of the mechanisms of action of these therapeutics evolves, even more impressive therapeutic success is being achieved through various combination strategies, including combinations of different immunotherapies as well as with other modalities. This review summarizes prospectively and retrospectively generated clinical evidence on modern melanoma therapy, focusing on immunotherapy and targeted therapy with BRAF kinase inhibitors and MEK kinase inhibitors (BRAF/MEK inhibitors), including recent data presented at major conference meetings. The combination of the anti-PD-1 directed monoclonal antibody nivolumab and of the CTLA-4 antagonist ipilimumab achieves unprecedented 5-year overall survival (OS) rates above 50%; however, toxicity is high. For PD-1 monotherapy (nivolumab or pembrolizumab), toxicities are in general well manageable. Today, novel combinations of such immune checkpoint inhibitors (ICIs) are under investigation, for example with cytokines and oncolytic viruses (i.e., pegylated interleukin-2, talimogene laherparepvec). Furthermore, current studies investigate the combined or sequential use of ICIs plus BRAF/MEK inhibitors. Several studies focus particularly on poor prognosis patients, as e.g., on anti-PD-1 refractory melanoma, patients with brain metastases, or uveal melanoma. It is hoped, on the road to cure, that these new approaches further improve long term survival in patients with advanced or metastatic melanoma.
Background: One prominent application for deep learningebased classifiers is skin cancer classification on dermoscopic images. However, classifier evaluation is often limited to holdout data which can mask common shortcomings such as susceptibility to confounding factors. To increase clinical applicability, it is necessary to thoroughly evaluate such classifiers on out-of-distribution (OOD) data. Objective: The objective of the study was to establish a dermoscopic skin cancer benchmark in which classifier robustness to OOD data can be measured. Methods: Using a proprietary dermoscopic image database and a set of image transformations, we create an OOD robustness benchmark and evaluate the robustness of four different convolutional neural network (CNN) architectures on it. Results: The benchmark contains three data setsdSkin Archive Munich (SAM), SAM-corrupted (SAM-C) and SAM-perturbed (SAM-P)dand is publicly available for download. To maintain the benchmark's OOD status, ground truth labels are not provided and test results should be sent to us for assessment. The SAM data set contains 319 unmodified and biopsy-verified dermoscopic melanoma (n Z 194) and nevus (n Z 125) images. SAM-C and SAM-P contain images from SAM which were artificially modified to test a classifier against low-quality inputs and to measure its prediction stability over small image changes, respectively. All four CNNs showed susceptibility to corruptions and perturbations. Conclusions: This benchmark provides three data sets which allow for OOD testing of binary skin cancer classifiers. Our classifier performance confirms the shortcomings of CNNs and provides a frame of reference. Altogether, this benchmark should facilitate a more thorough evaluation process and thereby enable the development of more robust skin cancer classifiers.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.