Colorectal Cancer Study of Austria (CORSA): A Population-Based Multicenter Study

Assessing Genotype-Phenotype Correlations with Deep Learning in Colorectal Cancer: A Multi-Centric Study

Background

Deep Learning (DL) has emerged as a powerful tool to predict genetic biomarkers directly from digitized Hematoxylin and Eosin (H&E) slides in colorectal cancer (CRC). However, few studies have systematically investigated the predictability of biomarkers beyond routinely available alterations such as microsatellite instability (MSI), and BRAF and KRAS mutations.

Methods

Our primary dataset comprised H&E slides of CRC tumors across five cohorts totaling 1,376 patients who underwent comprehensive panel sequencing, with an additional 536 patients from two public datasets for validation. We developed a DL model using a single transformer model to predict multiple genetic alterations directly from the slides. The model's performance was compared against conventional single-target models, and potential confounders were analyzed.

Findings

The multi-target model was able to predict numerous biomarkers from pathology slides, matching and partly exceeding single-target transformers. The Area Under the Receiver Operating Characteristic curve (AUROC, mean ± std) on the primary external validation cohorts was: BRAF (0·78 ± 0·01), hypermutation (0·88 ± 0·01), MSI (0·93 ± 0·01), RNF43 (0·86 ± 0·01); this biomarker predictability was mirrored across metrics and co-occurrence analyses. However, biomarkers with high AUROCs largely correlated with MSI, with model predictions depending considerably on MSI-associated morphology upon pathological examination.

Interpretation

Our study demonstrates that multi-target transformers can predict the biomarker status for numerous genetic alterations in CRC directly from H&E slides. However, their predictability is mainly associated with MSI phenotype, despite indications of slight biomarker-inherent contributions to a phenotype. Our findings underscore the need to analyze confounders in AI-based oncology biomarkers. To enable this, we developed a validated model applicable to other cancers and larger, diverse datasets.

Funding

The German Federal Ministry of Health, the Max-Eder-Programme of German Cancer Aid, the German Federal Ministry of Education and Research, the German Academic Exchange Service, and the EU.

Circulating tryptophan-kynurenine pathway metabolites are associated with all-cause mortality among patients with stage I-III colorectal cancer

Alterations within the tryptophan-kynurenine metabolic pathway have been linked to the etiology of colorectal cancer (CRC), but the relevance of this pathway for prognostic outcomes in CRC patients needs further elucidation. Therefore, we investigated associations between circulating concentrations of tryptophan-kynurenine pathway metabolites and all-cause mortality among CRC patients. This study utilizes data from 2102 stage I-III CRC patients participating in six prospective cohorts involved in the international FOCUS Consortium. Preoperative circulating concentrations of tryptophan, kynurenine, kynurenic acid (KA), 3-hydroxykynurenine (HK), xanthurenic acid (XA), 3-hydroxyanthranilic acid (HAA), anthranilic acid (AA), picolinic acid (PA), and quinolinic acid (QA) were measured by liquid chromatography-tandem mass spectrometry. Using Cox proportional hazards regression, we examined associations of above-mentioned metabolites with all-cause mortality, adjusted for potential confounders. During a median follow-up of 3.2 years (interquartile range: 2.2-4.9), 290 patients (13.8%) deceased. Higher blood concentrations of tryptophan, XA, and PA were associated with a lower risk of all-cause mortality (per doubling in concentrations: tryptophan: HR = 0.56; 95%CI:0.41,0.76, XA: HR = 0.74; 95%CI:0.64,0.85, PA: HR = 0.76; 95%CI:0.64,0.92), while higher concentrations of HK and QA were associated with an increased risk of death (per doubling in concentrations: HK: HR = 1.80; 95%CI:1.47,2.21, QA: HR = 1.31; 95%CI:1.05,1.63). A higher kynurenine-to-tryptophan ratio, a marker of cell-mediated immune activation, was associated with an increased risk of death (per doubling: HR = 2.07; 95%CI:1.52,2.83). In conclusion, tryptophan-kynurenine pathway metabolites may be prognostic markers of survival in CRC patients.

Beneficial microbiome and diet interplay in early-onset colorectal cancer

Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the second leading cause of cancer-related deaths worldwide. Although the risk of developing CRC increases with age, approximately 10% of newly diagnosed cases occur in individuals under the age of 50. Significant changes in dietary habits in young adults since industrialization create a favorable microenvironment for colorectal carcinogenesis. We aim here to shed light on the complex interplay between diet and gut microbiome in the pathogenesis and prevention of early-onset CRC (EO-CRC). We provide an overview of dietary risk factors associated with EO-CRC and contrast them with the general trends for CRC. We delve into gut bacteria, fungi, and phages with potential benefits against CRC and discuss the underlying molecular mechanisms. Furthermore, based on recent findings from human studies, we offer insights into how dietary modifications could potentially enhance gut microbiome composition to mitigate CRC risk. All together, we outline the current research landscape in this area and propose directions for future investigations that could pave the way for novel preventive and therapeutic strategies.

A Systematic Review of Cost-Effectiveness Analyses of Colorectal Cancer Screening in Europe: Have Studies Included Optimal Screening Intensities?

OBJECTIVE

To assess the range of strategies analysed in European cost-effectiveness analyses (CEAs) of colorectal cancer (CRC) screening with respect to the screening intervals, age ranges and test cut-offs used to define positivity, to examine how this might influence what strategies are found to be optimal, and compare them with the current screening policies with a focus on the screening interval.

METHODS

We searched PubMed, Web of Science and Scopus for peer-reviewed, model-based CEAs of CRC screening. We included studies on average-risk European populations using the guaiac faecal occult blood test (gFOBT) or faecal immunochemical test (FIT). We adapted Drummond's ten-point checklist to appraise study quality.

RESULTS

We included 39 studies that met the inclusion criteria. Biennial screening was the most frequently used interval which was analysed in 37 studies. Annual screening was assessed in 13 studies, all of which found it optimally cost-effective. Despite this, 25 of 26 European stool-based programmes use biennial screening. Many CEAs did not vary the age range, but the 14 that did generally found broader ranges optimal. Only 11 studies considered alternative FIT cut-offs, 9 of which found lower cut-offs superior. Conflicts between current policy and CEA evidence are less clear regarding age ranges and cut-offs.

CONCLUSIONS

The existing CEA evidence indicates that the widely adopted biennial frequency of stool-based testing in Europe is suboptimal. It is likely that many more lives could be saved throughout Europe if programmes could be offered with more intensive annual screening.

Higher Plasma Creatinine Is Associated with an Increased Risk of Death in Patients with Non-Metastatic Rectal but Not Colon Cancer: Results from an International Cohort Consortium

Colorectal cancer (CRC) is increasingly recognized as a heterogeneous disease. No studies have prospectively examined associations of blood metabolite concentrations with all-cause mortality in patients with colon and rectal cancer separately. Targeted metabolomics (Biocrates AbsoluteIDQ p180) and pathway analyses (MetaboAnalyst 4.0) were performed on pre-surgery collected plasma from 674 patients with non-metastasized (stage I-III) colon (n = 394) or rectal cancer (n = 283). Metabolomics data and covariate information were received from the international cohort consortium MetaboCCC. Cox proportional hazards models were computed to investigate associations of 148 metabolite levels with all-cause mortality adjusted for age, sex, tumor stage, tumor site (whenever applicable), and cohort; the false discovery rate (FDR) was used to account for multiple testing. A total of 93 patients (14%) were deceased after an average follow-up time of 4.4 years (60 patients with colon cancer and 33 patients with rectal cancer). After FDR adjustment, higher plasma creatinine was associated with a 39% increase in all-cause mortality in patients with rectal cancer. HR: 1.39, 95% CI 1.23-1.72, pFDR = 0.03; but not colon cancer: pFDR = 0.96. Creatinine is a breakdown product of creatine phosphate in muscle and may reflect changes in skeletal muscle mass. The starch and sucrose metabolisms were associated with increased all-cause mortality in colon cancer but not in rectal cancer. Genes in the starch and sucrose metabolism pathways were previously linked to worse clinical outcomes in CRC. In summary, our findings support the hypothesis that colon and rectal cancer have different etiological and clinical outcomes that need to be considered for targeted treatments.

Genome-Wide Association Study of Metachronous Colorectal Adenoma Risk among Participants in the Selenium Trial

Genetic variants related to colorectal adenoma may help identify those who are at highest risk of colorectal cancer development or illuminate potential chemopreventive strategies. The purpose of this genome-wide association study was to identify genetic variants that are associated with risk of developing a metachronous colorectal adenoma among 1,215 study participants of European descent from the Selenium Trial. Associations of variants were assessed with logistic regression analyses and validated in an independent case-control study population of 1,491 participants from the Colorectal Cancer Study of Austria (CORSA). No statistically significant genome-wide associations between any variant and metachronous adenoma were identified after correction for multiple comparisons. However, an intron variant of FAT3 gene, rs61901554, showed a suggestive association (P = 1.10 × 10-6) and was associated with advanced adenomas in CORSA (P = 0.04). Two intronic variants, rs12728998 and rs6699944 in NLRP3 were also observed to have suggestive associations with metachronous lesions (P = 2.00 × 10-6) in the Selenium Trial and were associated with advanced adenoma in CORSA (P = 0.03). Our results provide new areas of investigation for the genetic basis of the development of metachronous colorectal adenoma and support a role for FAT3 involvement in the Wnt/β-catenin pathway leading to colorectal neoplasia.Trial Registration number: NCT00078897 (ClinicalTrials.gov).

Use of Omics Technologies for the Detection of Colorectal Cancer Biomarkers

Colorectal cancer (CRC) is one of the most frequently diagnosed cancers with high mortality rates, especially when detected at later stages. Early detection of CRC can substantially raise the 5-year survival rate of patients, and different efforts are being put into developing enhanced CRC screening programs. Currently, the faecal immunochemical test with a follow-up colonoscopy is being implemented for CRC screening. However, there is still a medical need to describe biomarkers that help with CRC detection and monitor CRC patients. The use of omics techniques holds promise to detect new biomarkers for CRC. In this review, we discuss the use of omics in different types of samples, including breath, urine, stool, blood, bowel lavage fluid, or tumour tissue, and highlight some of the biomarkers that have been recently described with omics data. Finally, we also review the use of extracellular vesicles as an improved and promising instrument for biomarker detection.