Comparative analysis of high- and low-level deep learning approaches in microsatellite instability prediction

Area under the curve may hide poor generalisation to external datasets

Objectives: To develop two different radiomic models based on preoperative contrast-enhanced computed tomography (PP CT) to predict microsatellite instability (MSI) in patients with colorectal cancer (CRC) before surgery. Methods: PP CT scans of 115 CC patients were segmented using 3DSlicer (v5.6.1). Model I included images from three different scanners (GE, Siemens, Philips), while Model II used only one scanner (GE). For Model I, 80 patients were used for training and 35 for internal validation; for Model II, 46 and 24 patients were used, respectively. Data on sex, age, tumour location, and MSI genomic status were collected. A total of 107 radiomic features (RFs) were extracted, and 30 and 35 RFs were identified as relevant for Models I and II, respectively, using the t-test or Mann-Whitney test (p < 0.05). The most robust RFs were selected using the LASSO regression method. Both models were internally validated. Results: Model I, based on 2 RFs and 1 clinical feature (LOCATION) achieved an AUC of 0.76 (95% CI: 0.65-0.87) in the training cohort and 0.74 (95% CI: 0.56-0.92) in the validation cohort. Model II, based on 3 RFs, achieved an AUC of 0.85 (95% CI: 0.73-0.96) in the training cohort and 0.72 (95% CI: 0.50-0.94) in the validation cohort. Conclusions: Both radiomic models showed good performance in distinguishing between MSI and non-MSI tumours, potentially reducing the need for invasive histological testing and improving treatment timing. Despite achieving a higher AUC, Model II showed signs of overfitting when compared to Model I, which incorporated two RFs and one clinical feature (LOCATION). Radiomics may function as a non-invasive preoperative screening tool to inform decisions regarding MSI testing and treatment. Building radiomic models on larger, more diverse datasets is preferable to enhance generalizability and reduce overfitting.

Evaluating morphological features for predicting microsatellite instability status in colorectal cancer

Immunotherapy deals with therapeutic interventions to arrest the progression of tumours using the immune system. These include checkpoint inhibitors, T-cell manipulation, cytokines, oncolytic viruses and tumour vaccines. In this paper, we present a survey of the latest developments on immunotherapy in colorectal cancer (CRC) and the role of artificial intelligence (AI) in this context. Among these, microsatellite instability (MSI) is perhaps the most popular IO biomarker globally. We first discuss the MSI status of tumours, its implications for patient management, and its relationship to immune response. In recent years, several aspiring studies have used AI to predict the MSI status of patients from digital whole-slide images (WSIs) of routine diagnostic slides. We present a survey of AI literature on the prediction of MSI and tumour mutation burden from digitised WSIs of haematoxylin and eosin-stained diagnostic slides. We discuss AI approaches in detail and elaborate their contributions, limitations and key takeaways to drive future research. We further expand this survey to other IO-related biomarkers like immune cell infiltrates and alternate data modalities like immunohistochemistry and gene expression. Finally, we underline possible future directions in immunotherapy for CRC and promise of AI to accelerate this exploration for patient benefits.

Prediction of Microsatellite Instability in Colorectal Cancer Based on Deep Learning

Multimodal deep learning model for predicting microsatellite instability in colorectal cancer by contrast-enhanced computed tomography and histopathology.

BackgroundCancer susceptibility germline variants generally require somatic alteration of the remaining allele to drive oncogenesis and, in some cases, tumor mutational profiles. Whether combined germline and somatic bi-allelic alterations are universally required for germline variation to influence tumor mutational profile is unclear. Here, we performed an exome-wide analysis of the frequency and functional effect of bi-allelic alterations in The Cancer Genome Atlas (TCGA).MethodsWe integrated germline variant, somatic mutation, somatic methylation, and somatic copy number loss data from 7790 individuals from TCGA to identify germline and somatic bi-allelic alterations in all coding genes. We used linear models to test for association between mono- and bi-allelic alterations and somatic microsatellite instability (MSI) and somatic mutational signatures.ResultsWe discovered significant enrichment of bi-allelic alterations in mismatch repair (MMR) genes and identified six bi-allelic carriers with elevated MSI, consistent with Lynch syndrome. In contrast, we find little evidence of an effect of mono-allelic germline variation on MSI. Using MSI burden and bi-allelic alteration status, we reclassify two variants of unknown significance in MSH6 as potentially pathogenic for Lynch syndrome. Extending our analysis of MSI to a set of 127 DNA damage repair (DDR) genes, we identified a novel association between methylation of SHPRH and MSI burden.ConclusionsWe find that bi-allelic alterations are infrequent in TCGA but most frequently occur in BRCA1/2 and MMR genes. Our results support the idea that bi-allelic alteration is required for germline variation to influence tumor mutational profile. Overall, we demonstrate that integrating germline, somatic, and epigenetic alterations provides new understanding of somatic mutational profiles.

Combining molecular subtype and traditional histo-pathologic tumor classifications to predict survival in patients with endometrioid or serous endometrial cancers (221)

Performance of Next-Generation Sequencing for the Detection of Microsatellite Instability in Colorectal Cancer With Deficient DNA Mismatch Repair

Remote sensing is a key tool to derive glacier surface velocities but existing mapping methods, such as cross-correlation techniques, can fail where surface properties change temporally or where large velocity variations occur spatially. High-resolution datasets, such as UAV imagery, offer a promising solution to tackle these issues and to study small-scale glacier dynamics, but new workflows are required to handle such data. Therefore, we tested the potential of new deep learning-based image-matching algorithms for deriving glacier surface velocities across the ablation area of a glacier with strong spatial variability in surface velocities (&amp;lt;5 m/yr to &amp;gt;100 m/yr) and substantial changes in surface properties between image acquisitions. For a thorough comparison of state-of-the-art methods and sensors, we applied three different techniques (cross-correlation using geoCosiCorr3D, feature tracking with ORB using SeaIceDrift and the new deep learning-based method using ICEpy4D) and three different platforms (Sentinel-2, PlanetScope, UAVs) to estimate glacier surface velocities. Results showed lowest errors for velocities derived with the deep learning-based approach applied to UAV imagery (RMSE = 2.17 m/yr, R2 = 0.99), followed by cross-correlation using Sentinel-2 images (RMSE = 21.0 m/yr, R2 = 0.59) and the deep learning-based approach with PlanetScope data (RMSE = 21.28 m/yr, R2 = 0.36). Cross-correlation with geoCosiCorr3D resulted in comparably high errors with the UAV dataset (RMSE = 36.22 m/yr, R2 = 0.24), whereas ORB-based feature tacking showed lowest performance with all sensors. Spatial patterns of computed velocities indicate that applying existing cross-correlation methods for areas with regular displacements or low glacier velocities yields suitable results on UAV data, but innovative deep learning-based approaches are required for resolving rapid changes in velocities or in surface properties. This novel method benefits from improved keypoint detection and matching through training using neural networks and data characterized by challenging geometries, outlier minimization and more robust descriptors by applying cross-attention layers. We conclude that continued development of deep learning-based feature tracking approaches for glacier velocity computations may substantially improve UAV-based velocity derivations applied to challenging situations. This method is able to deliver reliable displacement data in situations where traditional methods fail, which implies a new level of detail in understanding and interpreting glacier dynamics.

Genetic markers for distant metastatic disease in patients with colorectal cancer (CRC) are not well defined. Identification of genetic alterations associated with metastatic CRC could help to guide systemic and local treatment strategies. We evaluated the association of tumor necrosis factor receptor superfamily member 10C (TNFRSF10C) copy number variation (CNV) with distant metastatic disease in patients with CRC using The Cancer Genome Atlas (TCGA). Genetic sequencing data and clinical characteristics were obtained from TCGA for all available patients with CRC. There were 515 CRC patient samples with CNV and clinical outcome data, including a subset of 144 rectal adenocarcinoma patient samples. Using the TCGA CRC dataset, CNV of TNFRSF10C was evaluated for association with distant metastatic disease (M1 vs. M0). Multivariate logistic regression analysis with odds ratio (OR) using a 95% confidence interval (CI) was performed adjusting for age, T stage, N stage, adjuvant chemotherapy, gender, microsatellite instability (MSI), location, and surgical margin status. TNFRSF10C CNV in patients with CRC was associated with distant metastatic disease [OR 4.81 (95% CI, 2.13-10.85) P<0.001] and positive lymph nodes [OR 18.83 (95% CI, 8.42-42.09)]; P<0.001) but not MSI (OR P=0.799). On multivariate analysis, after adjusting for pathologic T stage, N stage, adjuvant chemotherapy, gender, and MSI, TNFRSF10C CNV remained significantly associated with distant metastatic disease (OR P=0.018). Subset analysis revealed that TNFRSF10C CNV was also significantly associated with distant metastatic disease in patients with rectal adenocarcinoma (OR P=0.016). TNFRSF10C CNV in patients with CRC is associated with distant metastatic disease. With further validation, such genetic profiles could be used clinically to support optimal systemic treatment strategies versus more aggressive local therapies in patients with CRC, including radiation therapy for rectal adenocarcinoma.

Approximately 15% of Colon Cancers are Microsatellite Instable (MSI). Frameshift Peptides (FPs) formed in MSI Colon Cancer are potential targets for immunotherapeutic strategies. Here we comprehensively characterize the mutational landscape of 71 MSI Colon Cancer patients from the cancer genome atlas (TCGA). We confirm that the mutations in MSI Colon Cancers are frequently frameshift deletions (23% in MSI; 1% in microsatellite stable), We find that these mutations cluster at specific locations in the genome which are mutated in up to 41% of the patients. We filter these for an adequate variant allele frequency, a sufficient mean mRNA level and the formation of a Super Neo Open Reading Frame (SNORF). Finally, we check the influence of Nonsense Mediated Decay (MMD) by comparing RNA and DNA sequencing results. Thereby we identify a set of 20 NMD-escaping Public FPs (PFPs) that cover over 90% of MSI Colon, 62.2% of MSI Endometrial and 58.8% of MSI Stomach cancer patients and 3 out of 4 Lynch patients in the TCGA-COAD. This underlines the potential for PFP directed immunotherapy, both in a therapeutic and a prophylactic setting in multiple types of MSI cancers.

The ability to live migrate virtual machines (VMs) between physical servers without any perceivable service interruption is pivotal for building more energy efficient Cloud Computing infrastructures in the future. Nevertheless, energy efficiency is not worth the effort if quality metrics (e.g., QoS, QoE) are severely decreased by, e.g., dynamic consolidation using live migration. In this work, we present results for a low level approach by patching KVM to implement Xen's live migration stop conditions, and a high level approach by monitoring the progress and service level state of a live migration and estimating the impact on the service level during the remaining migration time. We compare these approaches with an unpatched, vanilla KVM version and different sets of parameters used for live migration. Our service level management approaches offer superior QoS during migration. Especially, they allow to migrate also highly utilized VMs with comparably small influence on QoS.

Simple SummaryBesides the diagnosis of endometrial carcinoma (EC), the identification of EC subtypes is an important precondition for the effective treatment of the disease. Molecular factors have an important role in this context. Biomedical analysis is already very possible in experimental approaches. However, for clinical practice the procedures are often too complex, too expensive, and too time-consuming. In the present study, very good evaluated molecular markers were detected using an alternative method and compared with the original method. Analyses showed very good results. The important advantage of the new procedure is that the required molecular data can be obtained exclusively by sequencing. This may greatly simplify EC subtype classification and could be more easily incorporated into routine clinical diagnosis.Background: The Cancer Genome Atlas (TCGA) network (United States National Cancer Institute) identified four molecular endometrial cancer (EC) subtypes using an extensive multi-method approach. The aim of this study was to determine the four TCGA EC molecular subtypes using a single-method whole-exome sequencing (WES)-based approach provided by MH Guide (Molecular Health, Heidelberg, Germany). Methods: WES and clinical data of n = 232 EC patients were obtained from TCGA. The four TCGA EC molecular subtypes designated as (i) Mutated Polymerase ε (POLE), (ii) Microsatellite Instability (MSI), (iii) Copy Number (CN) low and, (iv) CN-high were determined using the MH Guide software. The prognostic value of the subtypes determined by MH Guide were compared with the TCGA classification. Results: Analysis of WES data using the MH Guide software led to the precise identification of the four EC molecular subtypes analogous to the TCGA classification. Both approaches displayed high concordance in terms of prognostic significance. Conclusions: The multi-method-based TCGA EC molecular subtypes can reliably be reproduced by the single-method-based MH Guide approach. The easy-to-implement single-method MH Guide approach represents a promising diagnostic tool.

The integration of artificial intelligence (AI) and magnetic resonance imaging in rectal cancer has the potential to enhance diagnostic accuracy by identifying subtle patterns and aiding tumor delineation and lymph node assessment. According to our systematic review focusing on convolutional neural networks, AI-driven tumor staging and the prediction of treatment response facilitate tailored treat-ment strategies for patients with rectal cancer. This paper summarizes the current landscape of AI in the imaging field of rectal cancer, emphasizing the performance reporting design based on the quality of the dataset, model performance, and external validation. AI-driven tumor segmentation has demonstrated promising results using various convolutional neural network models. AI-based predictions of staging and treatment response have exhibited potential as auxiliary tools for personalized treatment strategies. Some studies have indicated superior performance than conventional models in predicting microsatellite instability and KRAS status, offer-ing noninvasive and cost-effective alternatives for identifying genetic mutations. Image-based AI studies for rectal can-cer have shown acceptable diagnostic performance but face several challenges, including limited dataset sizes with standardized data, the need for multicenter studies, and the absence of oncologic relevance and external validation for clinical implantation. Overcoming these pitfalls and hurdles is essential for the feasible integration of AI models in clinical settings for rectal cancer, warranting further research.