Metabolomic Biomarkers for the Detection of Obesity-Driven Endometrial Cancer

Integrating proteomics and explainable artificial intelligence: a comprehensive analysis of protein biomarkers for endometrial cancer diagnosis and prognosis

Endometrial cancer, which is the most common gynaecological cancer in women after breast, colorectal and lung cancer, can be diagnosed at an early stage. The first aim of this study is to classify age, tumor grade, myometrial invasion and tumor size, which play an important role in the diagnosis and prognosis of endometrial cancer, with machine learning methods combined with explainable artificial intelligence. 20 endometrial cancer patients proteomic data obtained from tumor biopsies taken from different regions of EC tissue were used. The data obtained were then classified according to age, tumor size, tumor grade and myometrial invasion. Then, by using three different machine learning methods, explainable artificial intelligence was applied to the model that best classifies these groups and possible protein biomarkers that can be used in endometrial prognosis were evaluated. The optimal model for age classification was XGBoost with AUC (98.8%), for tumor grade classification was XGBoost with AUC (98.6%), for myometrial invasion classification was LightGBM with AUC (95.1%), and finally for tumor size classification was XGBoost with AUC (94.8%). By combining the optimal models and the SHAP approach, possible protein biomarkers and their expressions were obtained for classification. Finally, EWRS1 protein was found to be common in three groups (age, myometrial invasion, tumor size). This article's findings indicate that models have been developed that can accurately classify factors including age, tumor grade, and myometrial invasion all of which are critical for determining the prognosis of endometrial cancer as well as potential protein biomarkers associated with these factors. Furthermore, we were able to provide an analysis of how the quantities of the proteins suggested as biomarkers varied throughout the classes by combining the SHAP values with these ideal models.

Postmenopausal endometrial non-benign lesion risk classification through a clinical parameter-based machine learning model

OBJECTIVE

This study aimed to develop and evaluate a machine learning model utilizing non-invasive clinical parameters for the classification of endometrial non-benign lesions, specifically atypical hyperplasia (AH) and endometrioid carcinoma (EC), in postmenopausal women.

METHODS

Our study collected clinical parameters from a cohort of 999 patients with postmenopausal endometrial lesions and conducted preprocessing to identify 57 relevant characteristics from these irregular clinical data. To predict the presence of postmenopausal endometrial non-benign lesions, including atypical hyperplasia and endometrial cancer, we employed various models such as eXtreme Gradient Boosting (XGBoost), Random Forest (RF), Logistic Regression (LR), Support Vector Machine (SVM), Back Propagation Neural Network (BPNN), as well as two ensemble models. Additionally, a test set was performed on an independent dataset consisting of 152 patients. The performance evaluation of all models was based on metrics including the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, precision, and F1 score.

RESULTS

The RF model demonstrated superior recognition capabilities for patients with non-benign lesions compared to other models. In the test set, it attained a sensitivity of 88.1% and an AUC of 0.93, surpassing all alternative models evaluated in this study. Furthermore, we have integrated this model into our hospital's Clinical Decision Support System (CDSS) and implemented it within the outpatient electronic medical record system to continuously validate and optimize its performance.

CONCLUSIONS

We have trained a model and deployed a system with high discriminatory power that may provide a novel approach to identify patients at higher risk of postmenopausal endometrial non-benign lesions who may benefit from more tailored screening and clinical intervention.

Urine and serum metabolomic analysis of endometrial cancer diagnosis and classification based on ultra-performance liquid chromatography mass spectrometry

OBJECTIVE

This study aimed to reveal the urinary and serum metabolic pattern of endometrial cancer (EC) and establish diagnostic models to identify EC from controls, high-risk from low-risk EC, and type II from type I EC.

METHOD

This study included 146 EC patients (comprising 79 low-risk and 67 high-risk patients, including 124 type I and 22 type II) and 59 controls. The serum and urine samples were analyzed using ultraperformance liquid chromatography mass spectrometry. Analysis was used to elucidate the distinct metabolites and altered metabolic pathways. Receiver operating characteristic (ROC) analyses were employed to discover and validate the potential biomarker models.

RESULTS

Serum and urine metabolomes displayed significant differences between EC and controls, with metabolites related to amino acid and nicotinamide metabolisms. The serum and urine panels distinguished these two groups with Area Under the Curve (AUC) of 0.821 and 0.902, respectively. The panel consisting of serum and urine metabolites demonstrated the best predictive ability (AUC = 0.953 and 0.976 in discovering and validation group). In comparing high-risk and low risk EC, differential metabolites were enriched in purine and glutamine metabolism. The AUC values for serum and urine panels were 0.818, and 0.843, respectively. The combined panel exhibited better predictive accuracy (0.881 in discovering group and 0.936 in external validation). In the comparison between type I and type II group, altered folic acid metabolism was identified. The serum, urine and combined panels discriminated these two groups with the AUC of 0.829, 0.913 and 0.922, respectively.

CONCLUSION

The combined urine and serum metabolome effectively revealed the metabolic patterns in EC patients, offering valuable diagnostic models for EC diagnosis and classification.

Quantitative SWATH-based proteomic profiling of urine for the identification of endometrial cancer biomarkers in symptomatic women

BACKGROUND

A non-invasive endometrial cancer detection tool that can accurately triage symptomatic women for definitive testing would improve patient care. Urine is an attractive biofluid for cancer detection due to its simplicity and ease of collection. The aim of this study was to identify urine-based proteomic signatures that can discriminate endometrial cancer patients from symptomatic controls.

METHODS

This was a prospective case-control study of symptomatic post-menopausal women (50 cancers, 54 controls). Voided self-collected urine samples were processed for mass spectrometry and run using sequential window acquisition of all theoretical mass spectra (SWATH-MS). Machine learning techniques were used to identify important discriminatory proteins, which were subsequently combined in multi-marker panels using logistic regression.

RESULTS

The top discriminatory proteins individually showed moderate accuracy (AUC > 0.70) for endometrial cancer detection. However, algorithms combining the most discriminatory proteins performed well with AUCs > 0.90. The best performing diagnostic model was a 10-marker panel combining SPRR1B, CRNN, CALML3, TXN, FABP5, C1RL, MMP9, ECM1, S100A7 and CFI and predicted endometrial cancer with an AUC of 0.92 (0.96-0.97). Urine-based protein signatures showed good accuracy for the detection of early-stage cancers (AUC 0.92 (0.86-0.9)).

CONCLUSION

A patient-friendly, urine-based test could offer a non-invasive endometrial cancer detection tool in symptomatic women. Validation in a larger independent cohort is warranted.

Endometrial cancer diagnostic and prognostic algorithms based on proteomics, metabolomics, and clinical data: a systematic review

Endometrial cancer is the most common gynaecological malignancy in developed countries. Over 382,000 new cases were diagnosed worldwide in 2018, and its incidence and mortality are constantly rising due to longer life expectancy and life style factors including obesity. Two major improvements are needed in the management of patients with endometrial cancer, i.e., the development of non/minimally invasive tools for diagnostics and prognostics, which are currently missing. Diagnostic tools are needed to manage the increasing number of women at risk of developing the disease. Prognostic tools are necessary to stratify patients according to their risk of recurrence pre-preoperatively, to advise and plan the most appropriate treatment and avoid over/under-treatment. Biomarkers derived from proteomics and metabolomics, especially when derived from non/minimally-invasively collected body fluids, can serve to develop such prognostic and diagnostic tools, and the purpose of the present review is to explore the current research in this topic. We first provide a brief description of the technologies, the computational pipelines for data analyses and then we provide a systematic review of all published studies using proteomics and/or metabolomics for diagnostic and prognostic biomarker discovery in endometrial cancer. Finally, conclusions and recommendations for future studies are also given.

Potential serum metabolites and long-chain noncoding RNA biomarkers for endometrial cancer tissue

BACKGROUND

Endometrial carcinoma (EC) is one of the most common tumors in the female reproductive system. There are nearly 200 000 new cases every year. It is the third most common gynecological malignant tumor leading to female death. The incidence rate is closely related to lifestyle, and the incidence rate varies in different regions. The incidence rate of EC is ranking the first in the female reproductive system cancer just second only to breast, lung, and colorectal cancer in North America and Europe and the incidence rate of EC is only second, followed by breast cancer and cervical cancer in China.

PURPOSE

The potential metabolic markers of endometrial cancer were screened by liquid chromatograph mass spectrometer (LC-MS), and the tissues of patients with hysteromyoma and endometrial cancer were sequenced to explore the relationship between the disease and change in the content of long-chain noncoding RNA (lncRNA).

METHODS

Serum and tissue samples were collected from patients with endometrial dysplasia, endometrial cancer stage I, and endometrial cancer stage III. The metabolites in all serum samples were extracted, and the metabolites in all samples were detected by LC-MS/MS technology. The Pareto-scaling method was used for normalization, and the MetaboAnalyst 4.0 software was used for different analyses. The T test between groups showed that p ≤ 0.05 was regarded as the metabolite with a difference. Further, the function of differential metabolites was determined by metabolite function enrichment and co-expression analysis. Meanwhile, the differentially expressed lncRNA was detected by Illumina second-generation high-throughput sequencing technology, and the expression was analyzed by DEGseq software. Different lncRNA were screened according to p < 0.05. LncRNA with significant differences were screened by p < 0.01, q < 0.001, fold change ≥2, and false discovery rate (FDR) ≤0.001.

RESULTS

Through synthesis of T test, cluster heatmap, and ROC curve analysis, five biomarkers with potential diagnostic ability were obtained, including 2,3-Pyridinedicarboxylic acid (area under the curve (AUC) = 0.69), Hematommic acid, ethyl ester (AUC = 0.69), Maltitol (AUC = 0.69), 13(S)-HODE (AUC = 0.88), and D-Mannitol (AUC = 0.69) had potential diagnostic ability between EC phase I versus EC phase III. At the same time, lncRNA sequencing results showed that when endometrial atypical hyperplasia continued to change, including LINC00511, PVT1, and IQCH-AS1 (downregulated), and only changed significantly in the endometrial dysplasia group, including MALAT1, CARMN (downregulated) and LINC00648, BISPR, LINC01534, and LINC00930 (upregulated). Moreover, both differential metabolites and differential lncRNA were annotated to the lipid metabolism pathway, suggesting that this pathway played an important role in the occurrence and development of endometrial carcinoma.

CONCLUSIONS

It can combine the results of metabolomics and lncRNA sequencing to assist in the early diagnosis of endometrial precancerous lesions and endometrial cancer patients, to enhance the sensitivity and specificity of diagnosis, which has a certain clinical application prospect.

A Label-Free Proteomic Approach for the Identification of Biomarkers in the Exosome of Endometrial Cancer Serum

Endometrial cancers (ECs) are mostly adenocarcinomas arising from the inner part of the uterus. The identification of serum biomarkers, either soluble or carried in the exosome, may be useful in making an early diagnosis. We used label-free quantification mass spectrometry (LFQ-MS)-based proteomics to investigate the proteome of exosomes in the albumin-depleted serum from 12 patients with EC, as compared to 12 healthy controls. After quantification and statistical analysis, we found significant changes in the abundance (p < 0.05) of 33 proteins in EC vs. control samples, with a fold change of ≥1.5 or ≤0.6. Validation using Western blotting analysis in 36 patients with EC as compared to 36 healthy individuals confirmed the upregulation of APOA1, HBB, CA1, HBD, LPA, SAA4, PF4V1, and APOE. A multivariate logistic regression model based on the abundance of these proteins was able to separate the controls from the EC patients with excellent sensitivity levels, particularly for stage 1 ECs. The results show that using LFQ-MS to explore the specific proteome of serum exosomes allows for the identification of biomarkers in EC. These observations suggest that PF4V1, CA1, HBD, and APOE represent biomarkers that are able to reach the clinical stage, after a validation phase.

Diagnostic Accuracy of Liquid Biomarkers for the Non-Invasive Diagnosis of Endometrial Cancer: A Systematic Review and Meta-Analysis

Endometrial cancer rates are increasing annually due to an aging population and rising rates of obesity. Currently there is no widely available, accurate, non-invasive test that can be used to triage women for diagnostic biopsy whilst safely reassuring healthy women without the need for invasive assessment. The aim of this systematic review and meta-analysis is to evaluate studies assessing blood and urine-based biomarkers as a replacement test for endometrial biopsy or as a triage test in symptomatic women. For each primary study, the diagnostic accuracy of different biomarkers was assessed by sensitivity, specificity, likelihood ratio and area under ROC curve. Forest plots of summary statistics were constructed for biomarkers which were assessed by multiple studies using data from a random-effect models. All but one study was of blood-based biomarkers. In total, 15 studies reported 29 different exosomal biomarkers; 34 studies reported 47 different proteomic biomarkers. Summary statistic meta-analysis was reported for micro-RNAs, cancer antigens, hormones, and other proteomic markers. Metabolites and circulating tumor materials were also summarized. For the majority of biomarkers, no meta-analysis was possible. There was a low number of small, heterogeneous studies for the majority of evaluated index tests. This may undermine the reliability of summary estimates from the meta-analyses. At present there is no liquid biopsy that is ready to be used as a replacement test for endometrial biopsy. However, to the best of our knowledge this is the first study to report and meta-analyze the diagnostic accuracy of different classes of blood and urine biomarkers for detection of endometrial cancer. This review may thus provide a reference guide for those wishing to explore candidate biomarkers for further research.

Multi-Omic Profiling of Multi-Biosamples Reveals the Role of Amino Acid and Nucleotide Metabolism in Endometrial Cancer

Background

Endometrial cancer (EC) is one of the most common gynecological cancers. The traditional diagnosis of EC relies on histopathology, which, however, is invasive and may arouse tumor spread. There have been many studies aiming to find the metabolomic biomarkers of EC to improve the early diagnosis of cancer in a non-invasive or minimally invasive way, which can also provide valuable information for understanding the disease. However, most of these studies only analyze a single type of sample by metabolomics, and cannot provide a comprehensive view of the altered metabolism in EC patients. Our study tries to gain a pathway-based view of multiple types of samples for understanding metabolomic disorders in EC by combining metabolomics and proteomics.

Methods

Forty-four EC patients and forty-three controls were recruited for the research. We collected endometrial tissue, urine, and intrauterine brushing samples. Untargeted metabolomics and untargeted proteomics were both performed on the endometrial tissue samples, while only untargeted metabolomics was performed on the urine and intrauterine brushing samples.

Results

By integrating the differential metabolites and proteins between EC patients and controls detected in the endometrial tissue samples, we identified several EC-related significant pathways, such as amino acid metabolism and nucleotide metabolism. The significance of these pathways and the potential of metabolite biomarker-based diagnosis were then further verified by using urine and intrauterine brushing samples. It was found that the regulation of metabolites involved in the significant pathways showed similar trends in the intrauterine brushings and the endometrial tissue samples, while opposite trends in the urine and the endometrial tissue samples.

Conclusions

With multi-omics characterization of multi-biosamples, the metabolomic changes related to EC are illustrated in a pathway-based way. The network of altered metabolites and related proteins provides a comprehensive view of altered metabolism in the endometrial tissue samples. The verification of these critical pathways by using urine and intrauterine brushing samples provides evidence for the possible non-invasive or minimally invasive biopsy for EC diagnosis in the future.

Prevention of endometrial cancer through lifestyle Interventions: A systematic review and synthesis

Endometrial cancer is the most common gynaecological malignancy in Australia. Epidemiological studies have widely documented the association of endometrial cancer with modifiable lifestyle risk factors, most notably obesity. However, preventative strategies for endometrial cancer have not been well reported. The objective of this systematic review was to identify interventions targeted towards modifiable lifestyle risk factors that may reduce the risk of endometrial cancer. Literature published in the past ten years (January 2010 - January 2021) was retrieved from PubMed, Embase and Web of Science literature databases. Of 593 studies potentially eligible, 41 were assessed in full-text, and nine studies were included in the systematic review and synthesis without meta-analysis following the SWiM guidelines. The included studies were highly heterogenous with respect to the type of interventions implemented and the outcomes measured. We identified that diet and physical activity interventions, surgical weight loss interventions, and hormonal interventions were associated with changes in endometrial cancer biomarkers including circulating hormones and tissue markers. We identified a reduction in endometrial proliferation following lifestyle intervention as measured by the ki-67 proliferation index. Furthermore we identified an increase in adiponectin (a circulating biomarker of adiposity) following lifestyle intervention and a reduction in circulating insulin levels following lifestyle intervention. This review highlighted that the area of endometrial cancer prevention research is in its infancy and that further investigation of diet and physical activity interventions, surgical weight loss interventions, and hormonal interventions should be undertaken due to promising preliminary evidence.

Special Issue "Gynaecological Cancers Risk: Breast Cancer, Ovarian Cancer and Endometrial Cancer"

Over the last decade there have been significant advances and developments in our understanding of factors affecting women's cancer risk, our ability to identify individuals at increased risk and risk stratify populations, as well as implement and evaluate strategies for screening and prevention [...].

New Trends in the Detection of Gynecological Precancerous Lesions and Early-Stage Cancers

The prevention and early diagnostics of precancerous stages are key aspects of contemporary oncology. In cervical cancer, well-organized screening and vaccination programs, especially in developed countries, are responsible for the dramatic decline of invasive cancer incidence and mortality. Cytological screening has a long and successful history, and the ongoing implementation of HPV triage with increased sensitivity can further decrease mortality. On the other hand, endometrial and ovarian cancers are characterized by a poor accessibility to specimen collection, which represents a major complication for early diagnostics. Therefore, despite relatively promising data from evaluating the combined effects of genetic variants, population screening does not exist, and the implementation of new biomarkers is, thus, necessary. The introduction of various circulating biomarkers is of potential interest due to the considerable heterogeneity of cancer, as highlighted in this review, which focuses exclusively on the most common tumors of the genital tract, namely, cervical, endometrial, and ovarian cancers. However, it is clearly shown that these malignancies represent different entities that evolve in different ways, and it is therefore necessary to use different methods for their diagnosis and treatment.

Molecular Basis of Endometriosis and Endometrial Cancer: Current Knowledge and Future Perspectives

The human endometrium is a unique tissue undergoing important changes through the menstrual cycle. Under the exposure of different risk factors in a woman's lifetime, normal endometrial tissue can give rise to multiple pathologic conditions, including endometriosis and endometrial cancer. Etiology and pathophysiologic changes behind such conditions remain largely unclear. This review summarizes the current knowledge of the pathophysiology of endometriosis and its potential role in the development of endometrial cancer from a molecular perspective. A better understanding of the molecular basis of endometriosis and its role in the development of endometrial pathology will improve the approach to clinical management.

Comprehensive Library Generation for Identification and Quantification of Endometrial Cancer Protein Biomarkers in Cervico-Vaginal Fluid

Simple Summary

Endometrial cancer is the most common cancer of the female reproductive tract, and its incidence is rising. Early diagnosis has the potential to improve survival as women can receive care at the earliest possible stage when curative treatment is likely. Current tests for endometrial cancer diagnosis are sequentially invasive with low patient acceptability. A detection tool based on minimally invasive samples such as cervico-vaginal fluid would be a major advance in the field. This study focuses on the potential of detecting endometrial cancer based on the proteins and peptides expressed in cervico-vaginal fluid. Using Sequential window acquisition of all theoretical mass spectra (SWATH-MS), we present a spectral library of thousands of proteins in the cervico-vaginal fluid of women with or at risk of endometrial cancer. This important resource will enable the identification of endometrial cancer biomarkers in cervico-vaginal fluid and advances our knowledge of the role of proteomics in endometrial cancer detection.

Abstract

Endometrial cancer is the most common gynaecological malignancy in high-income countries and its incidence is rising. Early detection, aided by highly sensitive and specific biomarkers, has the potential to improve outcomes as treatment can be provided when it is most likely to effect a cure. Sequential window acquisition of all theoretical mass spectra (SWATH-MS), an accurate and reproducible platform for analysing biological samples, offers a technological advance for biomarker discovery due to its reproducibility, sensitivity and potential for data re-interrogation. SWATH-MS requires a spectral library in order to identify and quantify peptides from multiplexed mass spectrometry data. Here we present a bespoke spectral library of 154,206 transitions identifying 19,394 peptides and 2425 proteins in the cervico-vaginal fluid of postmenopausal women with, or at risk of, endometrial cancer. We have combined these data with a library of over 6000 proteins generated based on mass spectrometric analysis of two endometrial cancer cell lines. This unique resource enables the study of protein biomarkers for endometrial cancer detection in cervico-vaginal fluid. Data are available via ProteomeXchange with unique identifier PXD025925.