A machine learning approach to identify predictive molecular markers for cisplatin chemosensitivity following surgical resection in ovarian cancer

Development of a two-layer machine learning model for the forensic application of legal and illegal poppy classification based on sequence data

Poppies are beneficial plants with a variety of applications, including medicinal, edible, ornamental, and industrial purposes. Some Papaver species are forensically significant plants because they contain opium, a narcotic substance. Internationally trafficked species of illegal poppies are being identified by DNA barcoding employing multiple markers in response to their forensic value. However, effective markers for precise species identification of legal and illegal poppies are still under discussion, with research on illegal poppies focusing on Papaver somniferum L., and species identification studies of Papaver bracteatum and Papaver setigerum DC. still lacking. As a result, in order to evaluate the performance of genetic markers and classify their DNA sequences in the genus Papaver, this study developed the first machine learning-based two-layer model, in which the first layer classifies legal and illegal poppies from the given sequence and the second layer identifies species of illegal poppies using their sequences. We constructed the dataset and investigated biological features from four markers, internal transcribed spacer 1 (ITS1), internal transcribed spacer 2 (ITS2), transfer RNA Leucine (trnL), transfer RNA Leucine - transfer RNA Phenylalanine intergenic spacer (trnL-trnF intergenic spacer) and their combination, using four machine learning algorithms, K-nearest neighbor (KNN), Naïve Bayes (NB), extreme gradient boost (XGBoost) and Random Forest (RF). According to our findings, for Layer 1 to classify legal and illegal poppies, KNN-based models using combined ITS region achieved the greatest performance of accuracy 0.846 and 0.889 using training and test sets, respectively. Additionally, for Layer 2 to identify illegal poppy species, KNN-based models using combined ITS region achieved the best performance of 0.833 and 1.000 for using training and test sets, respectively. To validate the model, the combined ITS region, which includes ITS 1 and 2 sequences, from blind poppy samples were used as a case study, with the Layer 1 correctly classifying legal and illegal poppies with over 0.830 accuracy. Layer 2 correctly identified P. setigerum DC., however, only one of the three P. somniferum L. species was accurately identified. Nevertheless, our research shows that machine learning can be used to classify and identify legal and illegal poppy species using DNA barcodes which can then be used as an efficient and effective forensic tool for improved law enforcement and a safer society.

Application of artificial intelligence in the diagnosis, treatment, and recurrence prediction of peritoneal carcinomatosis

Peritoneal carcinomatosis (PC) is a type of secondary cancer which is not sensitive to conventional intravenous chemotherapy. Treatment strategies for PC are usually palliative rather than curative. Recently, artificial intelligence (AI) has been widely used in the medical field, making the early diagnosis, individualized treatment, and accurate prognostic evaluation of various cancers, including mediastinal malignancies, colorectal cancer, lung cancer more feasible. As a branch of computer science, AI specializes in image recognition, speech recognition, automatic large-scale data extraction and output. AI technologies have also made breakthrough progress in the field of peritoneal carcinomatosis (PC) based on its powerful learning capacity and efficient computational power. AI has been successfully applied in various approaches in PC diagnosis, including imaging, blood tests, proteomics, and pathological diagnosis. Due to the automatic extraction function of the convolutional neural network and the learning model based on machine learning algorithms, AI-assisted diagnosis types are associated with a higher accuracy rate compared to conventional diagnosis methods. In addition, AI is also used in the treatment of peritoneal cancer, including surgical resection, intraperitoneal chemotherapy, systemic chemotherapy, which significantly improves the survival of patients with PC. In particular, the recurrence prediction and emotion evaluation of PC patients are also combined with AI technology, further improving the quality of life of patients. Here we have comprehensively reviewed and summarized the latest developments in the application of AI in PC, helping oncologists to comprehensively diagnose PC and provide more precise treatment strategies for patients with PC.

Enhancing precision medicine: a nomogram for predicting platinum resistance in epithelial ovarian cancer

BACKGROUND

This study aimed to develop a novel nomogram that can accurately estimate platinum resistance to enhance precision medicine in epithelial ovarian cancer(EOC).

METHODS

EOC patients who received primary therapy at the General Hospital of Ningxia Medical University between January 31, 2019, and June 30, 2021 were included. The LASSO analysis was utilized to screen the variables which contained clinical features and platinum-resistance gene immunohistochemistry scores. A nomogram was created after the logistic regression analysis to develop the prediction model. The consistency index (C-index), calibration curve, receiver operating characteristic (ROC) curve, and decision curve analysis (DCA) were used to assess the nomogram's performance.

RESULTS

The logistic regression analysis created a prediction model based on 11 factors filtered down by LASSO regression. As predictors, the immunohistochemical scores of CXLC1, CXCL2, IL6, ABCC1, LRP, BCL2, vascular tumor thrombus, ascites cancer cells, maximum tumor diameter, neoadjuvant chemotherapy, and HE4 were employed. The C-index of the nomogram was found to be 0.975. The nomogram's specificity is 95.35% and its sensitivity, with a cut-off value of 165.6, is 92.59%, as seen by the ROC curve. After the nomogram was externally validated in the test cohort, the coincidence rate was determined to be 84%, and the ROC curve indicated that the nomogram's AUC was 0.949.

CONCLUSION

A nomogram containing clinical characteristics and platinum gene IHC scores was developed and validated to predict the risk of EOC platinum resistance.

Predictive Value of Machine Learning for Platinum Chemotherapy Responses in Ovarian Cancer: Systematic Review and Meta-Analysis

BACKGROUND

Machine learning is a potentially effective method for predicting the response to platinum-based treatment for ovarian cancer. However, the predictive performance of various machine learning methods and variables is still a matter of controversy and debate.

OBJECTIVE

This study aims to systematically review relevant literature on the predictive value of machine learning for platinum-based chemotherapy responses in patients with ovarian cancer.

METHODS

Following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we systematically searched the PubMed, Embase, Web of Science, and Cochrane databases for relevant studies on predictive models for platinum-based therapies for the treatment of ovarian cancer published before April 26, 2023. The Prediction Model Risk of Bias Assessment tool was used to evaluate the risk of bias in the included articles. Concordance index (C-index), sensitivity, and specificity were used to evaluate the performance of the prediction models to investigate the predictive value of machine learning for platinum chemotherapy responses in patients with ovarian cancer.

RESULTS

A total of 1749 articles were examined, and 19 of them involving 39 models were eligible for this study. The most commonly used modeling methods were logistic regression (16/39, 41%), Extreme Gradient Boosting (4/39, 10%), and support vector machine (4/39, 10%). The training cohort reported C-index in 39 predictive models, with a pooled value of 0.806; the validation cohort reported C-index in 12 predictive models, with a pooled value of 0.831. Support vector machine performed well in both the training and validation cohorts, with a C-index of 0.942 and 0.879, respectively. The pooled sensitivity was 0.890, and the pooled specificity was 0.790 in the training cohort.

CONCLUSIONS

Machine learning can effectively predict how patients with ovarian cancer respond to platinum-based chemotherapy and may provide a reference for the development or updating of subsequent scoring systems.

Gynecological cancer prognosis using machine learning techniques: A systematic review of last three decades (1990–2022)

OBJECTIVE

Many Computer Aided Prognostic (CAP) systems based on machine learning techniques have been proposed in the field of oncology. The objective of this systematic review was to assess and critically appraise the methodologies and approaches used in predicting the prognosis of gynecological cancers using CAPs.

METHODS

Electronic databases were used to systematically search for studies utilizing machine learning methods in gynecological cancers. Study risk of bias (ROB) and applicability were assessed using the PROBAST tool. 139 studies met the inclusion criteria, of which 71 predicted outcomes for ovarian cancer patients, 41 predicted outcomes for cervical cancer patients, 28 predicted outcomes for uterine cancer patients, and 2 predicted outcomes for gynecological malignancies broadly.

RESULTS

Random forest (22.30 %) and support vector machine (21.58 %) classifiers were used most commonly. Use of clinicopathological, genomic and radiomic data as predictors was observed in 48.20 %, 51.08 % and 17.27 % of studies, respectively, with some studies using multiple modalities. 21.58 % of studies were externally validated. Twenty-three individual studies compared ML and non-ML methods. Study quality was highly variable and methodologies, statistical reporting and outcome measures were inconsistent, preventing generalized commentary or meta-analysis of performance outcomes.

CONCLUSION

There is significant variability in model development when prognosticating gynecological malignancies with respect to variable selection, machine learning (ML) methods and endpoint selection. This heterogeneity prevents meta-analysis and conclusions regarding the superiority of ML methods. Furthermore, PROBAST-mediated ROB and applicability analysis demonstrates concern for the translatability of existing models. This review identifies ways that this can be improved upon in future works to develop robust, clinically translatable models within this promising field.

A deep tabular data learning model predicting cisplatin sensitivity identifies BCL2L1 dependency in cancer

Cisplatin, a platinum-based chemotherapeutic agent, is widely used as a front-line treatment for several malignancies. However, treatment outcomes vary widely due to intrinsic and acquired resistance. In this study, cisplatin-perturbed gene expression and pathway enrichment were used to define a gene signature, which was further utilized to develop a cisplatin sensitivity prediction model using the TabNet algorithm. The TabNet model performed better (>80 % accuracy) than all other machine learning models when compared to a wide range of machine learning algorithms. Moreover, by using feature importance and comparing predicted ovarian cancer patient samples, BCL2L1 was identified as an important gene contributing to cisplatin resistance. Furthermore, the pharmacological inhibition of BCL2L1 was found to synergistically increase cisplatin efficacy. Collectively, this study developed a tool to predict cisplatin sensitivity using cisplatin-perturbed gene expression and pathway enrichment knowledge and identified BCL2L1 as an important gene in this setting.

Classification and diagnostic prediction of breast cancer metastasis on clinical data using machine learning algorithms

Metastatic Breast Cancer (MBC) is one of the primary causes of cancer-related deaths in women. Despite several limitations, histopathological information about the malignancy is used for the classification of cancer. The objective of our study is to develop a non-invasive breast cancer classification system for the diagnosis of cancer metastases. The anaconda-Jupyter notebook is used to develop various python programming modules for text mining, data processing, and Machine Learning (ML) methods. Utilizing classification model cross-validation criteria, including accuracy, AUC, and ROC, the prediction performance of the ML models is assessed. Welch Unpaired t-test was used to ascertain the statistical significance of the datasets. Text mining framework from the Electronic Medical Records (EMR) made it easier to separate the blood profile data and identify MBC patients. Monocytes revealed a noticeable mean difference between MBC patients as compared to healthy individuals. The accuracy of ML models was dramatically improved by removing outliers from the blood profile data. A Decision Tree (DT) classifier displayed an accuracy of 83% with an AUC of 0.87. Next, we deployed DT classifiers using Flask to create a web application for robust diagnosis of MBC patients. Taken together, we conclude that ML models based on blood profile data may assist physicians in selecting intensive-care MBC patients to enhance the overall survival outcome.

A machine learning approach applied to gynecological ultrasound to predict progression-free survival in ovarian cancer patients

In a growing number of social and clinical scenarios, machine learning (ML) is emerging as a promising tool for implementing complex multi-parametric decision-making algorithms. Regarding ovarian cancer (OC), despite the standardization of features that can support the discrimination of ovarian masses into benign and malignant, there is a lack of accurate predictive modeling based on ultrasound (US) examination for progression-free survival (PFS). This retrospective observational study analyzed patients with epithelial ovarian cancer (EOC) who were followed in a tertiary center from 2018 to 2019. Demographic features, clinical characteristics, information about the surgery and post-surgery histopathology were collected. Additionally, we recorded data about US examinations according to the International Ovarian Tumor Analysis (IOTA) classification. Our study aimed to realize a tool to predict 12 month PFS in patients with OC based on a ML algorithm applied to gynecological ultrasound assessment. Proper feature selection was used to determine an attribute core set. Three different machine learning algorithms, namely Logistic Regression (LR), Random Forest (RFF), and K-nearest neighbors (KNN), were then trained and validated with five-fold cross-validation to predict 12 month PFS. Our analysis included n. 64 patients and 12 month PFS was achieved by 46/64 patients (71.9%). The attribute core set used to train machine learning algorithms included age, menopause, CA-125 value, histotype, FIGO stage and US characteristics, such as major lesion diameter, side, echogenicity, color score, major solid component diameter, presence of carcinosis. RFF showed the best performance (accuracy 93.7%, precision 90%, recall 90%, area under receiver operating characteristic curve (AUROC) 0.92). We developed an accurate ML model to predict 12 month PFS.

Cisplatin for cancer therapy and overcoming chemoresistance

Cisplatin spearheads the anticancer chemotherapeutics in present-day use although acute toxicity is its primary impediment factor. Among a plethora of experimental medications, a drug as effective or surpassing the benefits of cisplatin has not been discovered yet. Although Oxaliplatin is considered more superior to cisplatin, the former has been better for colorectal cancer while cisplatin is widely used for treating gynaecological cancers. Carcinoma imposes a heavy toll on mortality rates worldwide despite the novel treatment strategies and detection methods that have been introduced; nanomedicine combined with precision medicine, immunotherapy, volume-regulated anion channels, and fluorodeoxyglucose-positron emission tomography. Millions of deaths occur annually from metastatic cancers which escape early detection and the concomitant diseases caused by highly toxic chemotherapy that causes organ damage. It continues due to insufficient knowledge of the debilitative mechanisms induced by cancer biology. To overcome chemoresistance and to attenuate the adverse effects of cisplatin therapy, both in vitro and in vivo models of cisplatin-treated cancers and a few multi-centred, multi-phasic, randomized clinical trials in pursuant with recent novel strategies have been tested. They include plant-based phytochemical compounds, de novo drug delivery systems, biochemical/immune pathways, 2D and 3D cell culture models using small molecule inhibitors and genetic/epigenetic mechanisms, that have contributed to further the understanding of cisplatin's role in modulating the tumour microenvironment. Cisplatin was beneficial in cancer therapy for modulating the putative cellular mechanisms; apoptosis, autophagy, cell cycle arrest and gene therapy of micro RNAs. Specific importance of drug influx, efflux, systemic circulatory toxicity, half-maximal inhibition, and the augmentation of host immunometabolism have been identified. This review offers a discourse on the recent anti-neoplastic treatment strategies to enhance cisplatin efficacy and to overcome chemoresistance, given its superiority among other tolerable chemotherapies.

Perspectives on Ovarian Cancer 1809 to 2022 and Beyond

Unlike many other malignancies, overall survival for women with epithelial ovarian cancer has improved only modestly over the last half-century. The perspectives presented here detail the views of a gynecologic oncologist looking back and the view of the academic editor looking forward. Surgical beginnings in 1809 are merged with genomics, surgical advances, and precision therapy at present and for the future. Presentations in this special issue focus on factors related to the diagnosis of ovarian cancer: (1) markers for the preoperative assessment of primary and metastatic ovarian tumors, (2) demonstrations of the presence of pelvic fluid in ultrasound studies of ovarian malignancies, (3) the effects of age, menopausal status, and body habitus on ovarian visualization, (4) the ability of OVA1 to detect ovarian cancers when Ca125 was not informative, (5) the detection of tumor-specific changes in cell adhesion molecules by tissue-based staining, (6) presentation of a high discrimination model for ovarian cancer using IOTA Simple Rules and CA125, (7) review of low-grade serous carcinoma of the ovary, and (8) a comprehensive case report on ovarian carcinosarcoma.

Establish of an Initial Platinum-Resistance Predictor in High-Grade Serous Ovarian Cancer Patients Regardless of Homologous Recombination Deficiency Status

Backgrounds

Ovarian cancer (OC) is still the leading aggressive and lethal disease of gynecological cancers, and platinum-based regimes are the standard treatments. However, nearly 20%–30% of patients with OC are initial platinum resistant (IPR), and there is a lack of valid tools to predict whether they will be primary platinum resistant or not prior to chemotherapy.

Methods

Transcriptome data from The Cancer Genome Atlas (TCGA) was downloaded as the training data, and transcriptome data of GSE15622, GSE102073, GSE19829, and GSE26712 were retrieved from Gene Expression Omnibus (GEO) as the validation cohorts. Differentially expressed genes (DEGs) were selected between platinum-sensitive and platinum-resistant patients from the training cohort, and multiple machine-learning algorithms [including random forest, XGboost, and least absolute shrinkage and selection operator (LASSO) regression] were utilized to determine the candidate genes from DEGs. Then, we applied logistic regression to establish the IPR signature based on the expression. Finally, comprehensive clinical, genomic, and survival feature were analyzed to understand the application value of the established IPR signature.

Results

A total of 532 DEGs were identified between platinum-resistant and platinum-sensitive samples, and 11 of them were shared by these three-machine learning algorithms and utilized to construct an IPR prediction signature. The area under receiver operating characteristic curve (AUC) was 0.841 and 0.796 in the training and validation cohorts, respectively. Notably, the prediction capacity of this signature was stable and robust regardless of the patients’ homologous recombination deficiency (HRD) and mutation burden status. Meanwhile, the genomic feature was concordant between samples with high- or low-IPR signature, except a significantly higher prevalence of gain at Chr19q.12 (regions including CCNE1) in the high-IPR signature samples. The efficacy of prediction of platinum resistance of IPR signature successfully transferred to the precise survival prediction, with the AUC of 0.71, 0.72, and 0.66 to predict 1-, 3-, and 5-year survival, respectively. At last, we found a significantly different tumor-infiltrated lymphocytes feature, including lower abundance of CD4+ naive T cells in the samples with high-IPR signature. A relatively lower tumor immune dysfunction and exclusion (TIDE) value and more sensitivity to multiple therapies including Gefitinib may suggest the potency to transfer from platinum-based therapy to immunotherapy or target therapies in patients with high-IPR signature.

Conclusion

Our study established an IPR signature based on the expression of 11 genes that could stably and robustly distinguish OC patients with IPR and/or poor outcomes, which may guide therapeutic regimes tailoring.

Clinical significance of metabolism-related genes and FAK activity in ovarian high-grade serous carcinoma

BACKGROUND

Administration of poly (ADP-ribose) polymerase (PARP) inhibitors after achieving a response to platinum-containing drugs significantly prolonged relapse-free survival compared to placebo administration. PARP inhibitors have been used in clinical practice. However, patients with platinum-resistant relapsed ovarian cancer still have a poor prognosis and there is an unmet need. The purpose of this study was to examine the clinical significance of metabolic genes and focal adhesion kinase (FAK) activity in advanced ovarian high-grade serous carcinoma (HGSC).

METHODS

The RNA sequencing (RNA-seq) data and clinical data of HGSC patients were obtained from the Genomic Data Commons (GDC) Data Portal and analysed ( https://portal.gdc.cancer.gov/ ). In addition, tumour tissue was sampled by laparotomy or screening laparoscopy prior to treatment initiation from patients diagnosed with stage IIIC ovarian cancer (International Federation of Gynecology and Obstetrics (FIGO) classification, 2014) at the Saitama Medical University International Medical Center, and among the patients diagnosed with HGSC, 16 cases of available cryopreserved specimens were included in this study. The present study was reviewed and approved by the Institutional Review Board of Saitama Medical University International Medical Center (Saitama, Japan). Among the 6307 variable genes detected in both The Cancer Genome Atlas-Ovarian (TCGA-OV) data and clinical specimen data, 35 genes related to metabolism and FAK activity were applied. RNA-seq data were analysed using the Subio Platform (Subio Inc, Japan). JMP 15 (SAS, USA) was used for statistical analysis and various types of machine learning. The Kaplan-Meier method was used for survival analysis, and the Wilcoxon test was used to analyse significant differences. P < 0.05 was considered significant.

RESULTS

In the TCGA-OV data, patients with stage IIIC with a residual tumour diameter of 1-10 mm were selected for K means clustering and classified into groups with significant prognostic correlations (p = 0.0444). These groups were significantly associated with platinum sensitivity/resistance in clinical cases (χ2 test, p = 0.0408) and showed significant relationships with progression-free survival (p = 0.0307).

CONCLUSION

In the TCGA-OV data, 2 groups classified by clustering focusing on metabolism-related genes and FAK activity were shown to be associated with platinum resistance and a poor prognosis.

Effect of HIPEC on Peritoneal Recurrence in Peritoneal Metastasis Treated With Cytoreductive Surgery: A Systematic Review

BACKGROUND

Peritoneal metastasis (PM) is a late-stage manifestation of intra-abdominal malignancies. The current standard of care indicates that cure can only be achieved with cytoreductive surgery (CRS) which is often indicated with concurrent adjuvant hyperthermic intraperitoneal chemotherapy (HIPEC). However, the utility of HIPEC within subsets of PM is not fully understood. We seek to compare the effectiveness of HIPEC in improving peritoneal recurrence rates in PM of different origins.

METHODS

We conducted a systematic review of trials on the PubMed, EMBASE, and Cochrane databases, last searched in August 2021. Biases were assessed using the Cochrane Collaboration's tool for assessing the risk of bias in randomized trials as well as the Methodological Index for Non-Randomized Studies (MINORS) framework.

RESULTS

7 gastric PM studies, 3 ovarian PM studies, and 3 colorectal PM studies were included. Recurrence-free survival was improved in the HIPEC + CRS cohort in 5 gastric trials but only 1 ovarian trial and none of colorectal origin.

DISCUSSION

Our findings indicate decent effectiveness of HIPEC in gastric PM, but limited utility in ovarian and colorectal PM. Limitations in the current literature are attributed to the paucity of data available, a lack of homogeneity and consideration of novel and personalised treatment regimens. We implore for further studies to be conducted with a focus on patient selection and stratification, and suggest a reframing of approach towards modern molecular and targeted therapeutic options in future studies of HIPEC.

SYSTEMATIC REVIEW REGISTRATION

https://www.researchregistry.com/browse-the-registry#registryofsystematicreviewsmeta-analyses/registryofsystematicreviewsmeta-analysesdetails/60c1ffff0c1b78001e8efbe3/, identifier reviewregistry1166.