The clinical relevance of comprehensive molecular analysis in rare cancers is not established. We analyzed the molecular profiles and clinical outcomes of 1,310 patients (rare cancers, 75.5%) enrolled in a prospective observational study by the German Cancer Consortium that applies whole-genome/exome and RNA sequencing to inform the care of adults with incurable cancers. On the basis of 472 single and six composite biomarkers, a cross-institutional molecular tumor board provided evidence-based management recommendations, including diagnostic reevaluation, genetic counseling, and experimental treatment, in 88% of cases. Recommended therapies were administered in 362 of 1,138 patients (31.8%) and resulted in significantly improved overall response and disease control rates (23.9% and 55.3%) compared with previous therapies, translating into a progression-free survival ratio >1.3 in 35.7% of patients. These data demonstrate the benefit of molecular stratification in rare cancers and represent a resource that may promote clinical trial access and drug approvals in this underserved patient population.Significance: Rare cancers are difficult to treat; in particular, molecular pathogenesis–oriented medical therapies are often lacking. This study shows that whole-genome/exome and RNA sequencing enables molecularly informed treatments that lead to clinical benefit in a substantial proportion of patients with advanced rare cancers and paves the way for future clinical trials.See related commentary by Eggermont et al., p. 2677.This article is highlighted in the In This Issue feature, p. 2659

Multiple studies have compared the performance of artificial intelligence (AI)-based models for automated skin cancer classification to human experts, thus setting the cornerstone for a successful translation of AI-based tools into clinicopathological practice. The objective of the study was to systematically analyse the current state of research on reader studies involving melanoma and to assess their potential clinical relevance by evaluating three main aspects: test set characteristics (holdout/out-of-distribution data set, composition), test setting (experimental/clinical, inclusion of metadata) and representativeness of participating clinicians. PubMed, Medline and ScienceDirect were screened for peer-reviewed studies published between 2017 and 2021 and dealing with AI-based skin cancer classification involving melanoma. The search terms skin cancer classification, deep learning, convolutional neural network (CNN), melanoma (detection), digital biomarkers, histopathology and whole slide imaging were combined. Based on the search results, only studies that considered direct comparison of AI results with clinicians and had a diagnostic classification as their main objective were included. A total of 19 reader studies fulfilled the inclusion criteria. Of these, 11 CNN-based approaches addressed the classification of dermoscopic images; 6 concentrated on the classification of clinical images, whereas 2 dermatopathological studies utilised digitised histopathological whole slide images. All 19 included studies demonstrated superior or at least equivalent performance of CNN-based classifiers compared with clinicians. However, almost all studies were conducted in highly artificial settings based exclusively on single images of the suspicious lesions. Moreover, test sets mainly consisted of holdout images and did not represent the full range of patient populations and melanoma subtypes encountered in clinical practice.

Due to their ability to solve complex problems, deep neural networks (DNNs) are becoming increasingly popular in medical applications. However, decision-making by such algorithms is essentially a black-box process that renders it difficult for physicians to judge whether the decisions are reliable. The use of explainable artificial intelligence (XAI) is often suggested as a solution to this problem. We investigate how XAI is used for skin cancer detection: how is it used during the development of new DNNs? What kinds of visualisations are commonly used? Are there systematic evaluations of XAI with dermatologists or dermatopathologists? Google Scholar, PubMed, IEEE Explore, Science Direct and Scopus were searched for peer-reviewed studies published between January 2017 and October 2021 applying XAI to dermatological images: the search terms histopathological image, whole-slide image, clinical image, dermoscopic image, skin, dermatology, explainable, interpretable and XAI were used in various combinations. Only studies concerned with skin cancer were included. 37 publications fulfilled our inclusion criteria. Most studies (19/37) simply applied existing XAI methods to their classifier to interpret its decision-making. Some studies (4/37) proposed new XAI methods or improved upon existing techniques. 14/37 studies addressed specific questions such as bias detection and impact of XAI on man-machine-interactions. However, only three of them evaluated the performance and confidence of humans using CAD systems with XAI. XAI is commonly applied during the development of DNNs for skin cancer detection. However, a systematic and rigorous evaluation of its usefulness in this scenario is lacking.

Purpose: To identify novel genomic regions of interest in acute myeloid leukemia (AML) with complex karyotypes, we applied comparative genomic hybridization to microarrays (array-CGH), allowing high-resolution genome-wide screening of genomic imbalances. Patients and Methods: Sixty AML cases with complex karyotypes were analyzed using array-CGH; parallel analysis of gene expression was performed in a subset of cases. Results: Genomic losses were found more frequently than gains. The most frequent losses affected 5q (77%), 17p (55%), and 7q (45%), and the most frequent genomic gains 11q (40%) and 8q (38%). Critical segments could be delineated to genomic fragments of only 0.8 to a few megabase-pairs of DNA. In lost/gained regions, gene expression profiling detected a gene dosage effect with significant lower/higher average gene expression levels across the genes located in the respective regions. Furthermore, high-level DNA amplifications were identified in several regions: 11q23.3-q24.1 (n = 7), 21q22 (n = 6), 11q23.3 (n = 5), 13q12 (n = 3), 8q24 (n = 3), 9p24 (n = 2), 12p13 (n = 2), and 20q11 (n = 2). Parallel analysis of gene expression in critical amplicons displayed overexpressed candidate genes (eg, C8FW and MYC in 8q24). Conclusion: In conclusion, a large spectrum of genomic imbalances, including novel recurring changes in AML with complex karyotypes, was identified using array-CGH. In addition, the combined analysis of array-CGH data with gene expression profiles allowed the detection of candidate genes involved in the pathogenesis of AML.

Juvenile myelomonocytic leukemia (JMML) is an aggressive myeloproliferative disorder of early childhood characterized by mutations activating RAS signaling. Established clinical and genetic markers fail to fully recapitulate the clinical and biological heterogeneity of this disease. Here we report DNA methylome analysis and mutation profiling of 167 JMML samples. We identify three JMML subgroups with unique molecular and clinical characteristics. The high methylation group (HM) is characterized by somatic PTPN11 mutations and poor clinical outcome. The low methylation group is enriched for somatic NRAS and CBL mutations, as well as for Noonan patients, and has a good prognosis. The intermediate methylation group (IM) shows enrichment for monosomy 7 and somatic KRAS mutations. Hypermethylation is associated with repressed chromatin, genes regulated by RAS signaling, frequent co-occurrence of RAS pathway mutations and upregulation of DNMT1 and DNMT3B, suggesting a link between activation of the DNA methylation machinery and mutational patterns in JMML.

Clinicians and pathologists traditionally use patient data in addition to clinical examination to support their diagnoses. 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. 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. The CNN on its own achieved the best performance (mean ± standard deviation of five individual runs) with AUROC of 92.30% ± 0.23% and balanced accuracy of 83.17% ± 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% ± 0.36%. 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.