Non-cancer Cases Research Articles

P1073 Clinical characteristics and outcomes of patients with Inflammatory Bowel Disease under long term (>10years) anti-TNFa treatment

Abstract Background There is limited information in the literature regarding long term effectiveness and safety of anti-TNFa agents in patients with Inflammatory Bowel Disease (IBD). We aimed to study the effectiveness and safety of anti-TNFa agents in IBD patients receiving this treatment beyond 10 years . Methods Consecutive IBD patients under anti-TNFa treatment >10 years were prospectively studied. Demographic and clinical characteristics as well as treatment details and outcomes were recorded based on an updated IBD registry. Outcomes included need for treatment optimization [dose escalation or immunosuppressant (IMM) addition], disease related hospitalizations and surgeries, serious infections and malignancies. Drug survival of anti-TNF treatment >10 years was assessed (at 13 and 15 years). Results Out of 395 patients being initiated with any anti-TNFa in our department, 56 with at least 10 years of anti-TNFa use [males 32(57%), mean age (±SD) 53.3(±13) years, mean disease duration (±SD) 22.8 (±1.15) years were included. Patients were 82%(46/56) with CD, mean duration of anti-TNFa use (±SD) 162 (±32.93) months, 39 (70%) under infliximab, 17 under adalimumab, while anti-TNFa represented the first biologic in 88% of them. More than half of the patients [54%(30/56)] were under combination treatment with IMM (24 azathioprine, 6 methotrexate) and 45% (25/56) needed dose escalation. During follow up 75% (42/56) remained on the index anti-TNFa, precisely 84% (31/37) at 13 and 55% (16/29) at 15 years after drug initiation respectively (figure 1). Main causes of discontinuation were loss of response (8/14), patient preference 1, malignancy 2, osteonecrosis 1, death of other cause 1 and deep remission 1. During follow up, a total of 30 disease related hospitalizations in 18 patients, 12 serious infections, 6 malignancies (non Hodgkin lymphoma of the liver, prostate carcinoma, thyroid gland carcinoma, hepatocellular carcinoma and two cases of non melanoma skin cancer) and 4 disease related operations were recorded. Conclusion More than half of the patients under long-term anti-TNFa treatment (>10 years) continue to receive the agent for an additional 5 years. Rates of serious infections and malignancies are no negligible and close monitoring is needed in these patients.

Traditional versus modern approaches to screening mammography: a comparison of computer-assisted detection for synthetic 2D mammography versus an artificial intelligence algorithm for digital breast tomosynthesis.

Traditional computer-assisted detection (CADe) algorithms were developed for 2D mammography, while modern artificial intelligence (AI) algorithms can be applied to 2D mammography and/or digital breast tomosynthesis (DBT). The objective is to compare the performance of a traditional machine learning CADe algorithm for synthetic 2D mammography to a deep learning-based AI algorithm for DBT on the same mammograms. Mammographic examinations from 764 patients (mean age 58years ± 11) with 106 biopsy-proven cancers and 658 cancer-negative cases were analyzed by a CADe algorithm (ImageChecker v10.0, Hologic, Inc.) and an AI algorithm (Genius AI Detection v2.0, Hologic, Inc.). Synthetic 2D images were used for CADe analysis, and DBT images were used for AI analysis. For each algorithm, an overall case score was defined as the highest score of all lesion marks, which was used to determine the area under the receiver operating characteristic curve (AUC). The overall AUC was higher for 3D AI than 2D CADe (0.873 versus 0.693, P < 0.001). Lesion-specific sensitivity of 3D AI was higher than 2D CADe (94.3 versus 72.6%, P = 0.002). Specificity of 3D AI was higher than 2D CADe (54.3 versus 16.7%, P < 0.001), and the rate of false marks on non-cancer cases was lower for 3D AI than 2D CADe (0.91 versus 3.24 per exam, P < 0.001). A deep learning-based AI algorithm applied to DBT images significantly outperformed a traditional machine learning CADe algorithm applied to synthetic 2D mammographic images, with regard to AUC, sensitivity, and specificity.

Engrailed-2 (EN2) protein in cervical mucus: a novel biomarker for endometrial carcinoma.

This study aims to demonstrate that the EN2 protein in cervical mucus may serve as a novel biomarker for screening endometrial cancer. This study included 133 patients who were treated at Beijing Obstetrics and Gynecology Hospital. According to the pathological results of hysteroscopy endometrial biopsy, the patients were divided into endometrial cancer group (n = 55), endometrial atypical hyperplasia group (n = 16), benign lesion group (n = 28), and control group (n = 34). All patients collected cervical mucus before hysteroscopy, and the level of EN2 protein in cervical mucus was detected by ELISA in the four groups of patients. Nine patients who underwent surgical treatment for endometrial cancer were included, and endometrial cancer lesions and adjacent tissues without endometrial cancer infiltration were taken from the endometrial cancer lesions and cervical tissues. Fifteen patients who underwent hysterectomy for benign lesions were included, and endometrial and cervical tissues were taken from the endometrial cancer lesions and adjacent tissues without endometrial cancer infiltration. PT-PCR, immunohistochemistry, and WB were used to verify the expression differences of EN2 protein in cancerous lesions and endometrial tissues without endometrial infiltration, cervical tissues, and endometrial tissues without endometrial lesions. In addition, HEC1B, KLE, and HeLa cell lines were used to characterize EN2 overexpression in endometrial cancer cells. Immunohistochemistry, RT-PCR, and confocal analysis were used to detect the expression of EN2 protein in different cell lines. In different groups, the average concentration of EN2 protein in cervical mucus in the EC group was significantly higher than that in the benign group and the normal control group (573.9 ± 123.4ng/mL vs 153.5 ± 106.2ng/mL and 153.0 ± 107.5ng/mL, P < 0.001). The concentration of EN2 protein in EIN3 case specimens was significantly higher than that in the normal control group (P < 0.001). ROC analysis showed that the optimal threshold for diagnosing endometrial cancer in cervical mucus was 321.1ng/ml, with a sensitivity of 92.6% and specificity of 95.4% for distinguishing EINIII, EC, and non-cancerous cases. In clinical specimens, the expression level of EN2 protein in endometrial cancer tissues was significantly higher than that in endometrial tissues and cervical tissues without endometrial cancer infiltration. In addition, PT-PCR, immunohistochemistry, suggest that EN2 protein is overexpressed in endometrial cancer cell lines compared to cervical adenocarcinoma cells (P < 0.01). The concentration of EN2 protein in cervical mucus can effectively identify endometrial cancer and precancerous lesions. The increased expression of EN2 protein in endometrial cancer lesions leads to an increase in the concentration of EN2 protein in the cervical mucus of endometrial cancer patients.This is a small single-center study, and larger studies are needed to determine the role of EN2 protein in the diagnosis of endometrial cancer.

Identifying central dimensions of quality of life including life-related values, preferences and functional health in older patients with cancer: a scoping review protocol.

Older patients with cancer already represent the largest proportion of cancer survivors which will further increase in the upcoming years. However, older patients are highly underrepresented in clinical research, leading to a detrimental knowledge gap. Research on important aspects of quality of life (QoL) and associated factors for older patients with cancer is insufficient to date. The objective of this scoping review therefore is to investigate the dimensions of QoL including functional health, life-relevant values and preferences in older adults with cancer across all tumor entities and health care settings. It will further identify medical, sociodemographic, psychosocial and geriatric aspects associated with QoL in the elderly and compare these with younger cancer patients and older non-cancer cases. Published articles investigating QoL dimensions and associated factors in older patients with cancer, i.e., exclusively patients ≥65 years or mean/median age ≥ 70 years for age-mixed samples, or that compare results of older with younger cancer patients or with older non-cancer cases will be considered for this scoping review. Older patients with cancer across all tumor entities, disease stages and health care setting will be included. PubMed and PsychINFO databases will be searched for relevant articles. Abstracts and titles will be screened for basic inclusion, and two independent reviewers will conduct a full text screening to evaluate the age criteria and decide on the final inclusion of the study. Data on study and participant characteristics, QoL dimensions and geriatric factors will be extracted using a data extraction sheet. Results will be summarized descriptively to address the objectives of this review. The findings of this scoping review will provide valuable insights into central dimensions of QoL, including values, preferences and functional health in older adults with cancer, and help to improve targeted interventions and healthcare planning.

Cancer information and population health resource: a resource for catchment area data and cancer outcomes research.

The University of North Carolina at Chapel Hill Lineberger Comprehensive Cancer Center has developed a novel data resource, the Cancer Information and Population Health Resource (CIPHR), for conducting catchment area evaluation and cancer population health research that links the North Carolina Central Cancer Registry (NCCCR) to medical and pharmacy claims data from Medicare, Medicaid, and private plans operating within North Carolina. This study's aim was to describe the CIPHR data and provide examples of potential cohorts available in those data. We present the underlying populations included in the NCCCR and claims data before linkage and demonstrate estimated sample sizes when these data are linked and commonly used insurance enrollment criteria are applied. Data for the years 2003-2020 are present in CIPHR and include 947 977 cancer cases from the NCCCR and 21.6 million enrollees in public and private health insurance (cancer and noncancer cases). When limited to first or only cancers (n = 672 377), 86% could be linked to insurance enrollment for at least 1 month during 2003-2020 (n = 582 638), with 62% of individuals linking to enrollment during the month of cancer diagnosis. Among all registry cancer cases, 47% (n = 317 898) had continuous insurance enrollment for at least 12 months before and after cancer diagnosis. CIPHR illustrates the utility of establishing and maintaining a statewide, comprehensive cancer population health database. This resource serves to characterize the cancer center catchment area and aids in tracking cancer outcomes and trends in care delivery as well as identifying disparities that require intervention and policy focus.

Improving Computer-aided Detection for Digital Breast Tomosynthesis by Incorporating Temporal Change.

Purpose To develop a deep learning algorithm that uses temporal information to improve the performance of a previously published framework of cancer lesion detection for digital breast tomosynthesis. Materials and Methods This retrospective study analyzed the current and the 1-year-prior Hologic digital breast tomosynthesis screening examinations from eight different institutions between 2016 and 2020. The dataset contained 973 cancer and 7123 noncancer cases. The front end of this algorithm was an existing deep learning framework that performed single-view lesion detection followed by ipsilateral view matching. For this study, PriorNet was implemented as a cascaded deep learning module that used the additional growth information to refine the final probability of malignancy. Data from seven of the eight sites were used for training and validation, while the eighth site was reserved for external testing. Model performance was evaluated using localization receiver operating characteristic curves. Results On the validation set, PriorNet showed an area under the receiver operating characteristic curve (AUC) of 0.931 (95% CI: 0.930, 0.931), which outperformed both baseline models using single-view detection (AUC, 0.892 [95% CI: 0.891, 0.892]; P < .001) and ipsilateral matching (AUC, 0.915 [95% CI: 0.914, 0.915]; P < .001). On the external test set, PriorNet achieved an AUC of 0.896 (95% CI: 0.885, 0.896), outperforming both baselines (AUC, 0.846 [95% CI: 0.846, 0.847]; P < .001 and AUC, 0.865 [95% CI: 0.865, 0.866]; P < .001, respectively). In the high sensitivity range of 0.9 to 1.0, the partial AUC of PriorNet was significantly higher (P < .001) relative to both baselines. Conclusion PriorNet using temporal information further improved the breast cancer detection performance of an existing digital breast tomosynthesis cancer detection framework. Keywords: Digital Breast Tomosynthesis, Computer-aided Detection, Breast Cancer, Deep Learning © RSNA, 2024 See also commentary by Lee in this issue.

Optimal Cut-Point Selection Methods Under Binary Classification When Subclasses Are Involved.

In practice, we often encounter binary classification problems where both main classes consist of multiple subclasses. For example, in an ovarian cancer study where biomarkers were evaluated for their accuracy of distinguishing noncancer cases from cancer cases, the noncancer class consists of healthy subjects and benign cases, while the cancer class consists of subjects at both early and late stages. This article aims to provide a large number of optimal cut-point selection methods for such setting. Furthermore, we also study confidence interval estimation of the optimal cut-points. Simulation studies are carried out to explore the performance of the proposed cut-point selection methods as well as confidence interval estimation methods. A real ovarian cancer data set is analyzed using the proposed methods.

Artificial Intelligence in Early Detection of Cervical Intraepithelial Neoplasia

Artificial Intelligence (AI) is a quickly evolving field of technology used to develop intelligent machines capable of performing tasks such as problem solving, decision making , perception, language processing, and learning. This paper explores the application of AI in the field of gynecological oncology, specifically in the diagnosis of cervical cancer. The paper proposes a hybrid AI model that uses a Gaussian mixture model and a deep learning model to segment and classifies colposcope images. The model performed with satisfactory segmentation metrics of sensitivity, specificity, dice index, and Jaccard index of 0.976, 0.989, 0.954, and 0.856, respectively. This model aims to accurately classify cancer and non-cancer cases from a colposcope image. The results showed that this method could effectively segment the colposcopy images and extract the cervix region. This can be a valuable tool for automated cancer diagnosis and can help improve the diagnosis's accuracy.

The prognostic and clinical value of genes associate with immunity and amino acid Metabolism in Lung Adenocarcinoma

BackgroundLung adenocarcinoma (LUAD) is the commonest subtype of primary lung cancer. A comprehensive analysis of the association of immunity with amino acid metabolism in LUAD is critical for understanding the disease. MethodsThe present study examined LUAD and noncancerous cases from the TCGA database. Differentially expressed genes (DEGs) between LUAD and noncancerous tissues were detected by analyzing processed expression profiles. We cross-referenced the up-regulated DEGs with Immune and Amino Acid Metabolism-related genes (I&AAMGs), resulting in Immune and Amino Acid Metabolism related differentially expressed genes (IAAAMRDEGs). The STRING database was employed to analyze PPI on IAAAMRDEGs, obtaining excavated hub genes, whose biological processes, molecular functions and cellular components were examined with GO/KEGG. Potential mechanisms related to LUAD were investigated by GSEA and GSVA. A prognostic model was built by LASSO-COX analysis, taking into consideration risk scores and prognostic factors to determine biomarkers affecting LUAD occurrence and prognosis. ResultsTotally 377 genes were detected at the intersection of upregulated DEGs and I&AAMGs. Analysis of PPI on these 377 IAAAMRDEGs yielded 17 hub genes. A LASSO regression analysis was utilized to assess the prognostic values of the 17 hub genes. Validation using the combined dataset confirmed 4 genes, e.g., polo-like kinase (PLK1), Ribonucleotide Reductase Subunit M2 (RRM2), Thyroid Hormone Receptor Interactor 13 (TRIP13), and Hyaluronan-Mediated Motility Receptor (HHMR). The model's accuracy was further assessed by ROC curve analysis and the COX model. In addition, immunohistochemical staining obtained from the HPA database, revealed enhanced PLK1 expression in LUAD samples. ConclusionLUAD pathogenesis is highly associated with immunity and amino acid metabolism. The PLK1, RRM2, TRIP13, and HMMR genes have prognostic values for LUAD. PLK1 upregulation in LUAD might be involved in tumorigenesis by modulating the cell cycle and represents a potential prognostic factor in clinic.

Abstract PO4-07-03: Machine learning breast cancer risk prediction using sequential past mammograms - a pilot study

Abstract Background and Aim Some studies have reported the use of mammogram AI deep learning algorithms to accurately predict the risk of future breast cancer development in women. In these studies, random and unrelated mammograms were independently used for training of the AI model. We propose a novel deep learning system using sequential past mammograms instead of random studies and aimed to determine if it can further improve risk prediction in a pilot study. Methods Pre-training of the backbone of a Swin-transformer machine learning model was performed by applying region-of-interest lesional masks of benign and malignant on a training set of 1644 mammograms derived from the open-access Curated Breast Imaging Subset of Digital Database for Screening Mammography dataset. To further develop the model, deep learning training and validation were performed on the Cohort of Screen-age Women (CSAW) open-access screening mammogram set which was derived from the long term data of 873 women who developed breast cancer for the first time and 7850 non-cancer women. We then compared the diagnostic accuracy of a model using sequential past mammograms for training against the standard, non-sequential model. For the sequential model, imaging segmentation of the potential cancer site on mammography was performed on all past, pre-cancer mammograms obtained within 60 to 2555 days of cancer diagnosis in the pre-processing stage for breast cancer cases. AI training was then carried out by combining the pre-cancer mammograms belonging to the same woman in a temporal sequence using a transformer encoder. The learning rate started from 1e-6 and decayed 10 times every 10 epochs. For the non-sequential model, we removed the temporal encoder and used random, unrelated mammogram studies for training. Validation was performed on a dataset consisting of 75 pre-cancer mammograms from breast cancer cases and 69 mammograms from non-cancer cases. Results Sensitivity, specificity and overall accuracy on the validation set were 77.3% (58/75), 68.1% (47/69) and (105/144) 72.9% respectively for the sequential model compared to 69.3% (52/75), 69.6% (48/69) and 69.4% (100/144) respectively for the non-sequential model. Conclusion The study showed that deep learning breast cancer risk prediction can be further improved by using sequential past mammograms instead of random mammograms for AI model training. This can also potentially enhance other AI risk prediction models that employ combined mammogram and traditional breast cancer clinical risk factors. A larger validation study will be useful. Contact information: email Lester Leong at lester.leong.c.h@singhealth.com.sg ROC curve for breast cancer risk prediction using AI deep learning training with sequential past mammograms. Citation Format: Lester Leong, Mingjie Xu, Weimin Huang, Erli Zhang, Engracia Loh, Sze Yiun Teo, Geok Hoon Lim, Veronique Kiak Mien Tan, Yirong Sim, Fredrik Strand, Ryan Tan. Machine learning breast cancer risk prediction using sequential past mammograms - a pilot study [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-07-03.

Abstract PO5-07-08: A Serum Biomarker for the Early Detection of Breast cancer

Abstract Background: After decades of use, clinicians and medical societies are still debating on the clinical utility of mammography, the standard of care for breast cancer (BC) screening, because of its limitations such as subjective visual interpretation, high rate of false-positive and false-negative (in dense breast) and concerns over radiation exposure. There are 55 million women, and about 45% of them considered dense breast, in need of an accurate, easy to use test for the screening of BC regardless of breast tissue complexities to differentiate between BC and benign findings such as calcification and fibrosis. To this date there is no blood protein biomarker (BM) in clinical use for the early detection of BC. We have developed a novel BC specific protein biomarker -based blood test to detect patients with early-stage BC and established herein the diagnostic performance of our BM in discriminating BC patients from healthy individuals and benign breast cases. Methods: A total of 433 serum samples (318 primary BC, 32 benign breast and 83 normal breast) were obtained from the Cooperative Human Tissue Network and from the IRCCS San Raffaele Pisana Research Center (blind samples). 247 BC samples were provided as blind samples and were used to validate the screening utility. Serum samples were tested with our BC specific BM by sandwich ELISA. Receiver operating characteristic (ROC) curves were calculated to evaluate the diagnostic performance, by determining sensitivity (SE), specificity (SP), area under the curve (AUC), and confidence interval (CI). Statistical analysis, ROC curves and scatter plots were performed with GraphPad Prism 7.0. Results: BM elevation in early cancer cases (stage I-II, Ductal Carcinoma) was found to be statistically significant (p&amp;lt; 0.0001) as compared to healthy and benign cases. Our test discriminated early BC patients from non-cancer cases with 87.3% SE at 85.2% SP and was further validated in a blind set including 247 BC (Ductal &amp; Lobular Carcinoma, stage I-III), 30 benign and 83 normal breast cases (88%SE, 86%SP). The sensitivity for early-stage (I-II) and late stage (III) BC was 86.8% and 94.3%, respectively. While lobular carcinoma is particularly not well detected by mammography with a sensitivity as low as 57% and much lower for women with dense breast, it was detected with high sensitivity (90%) in our test. 80% of fibrocystic changes (FCC), the most common form of benign breast disease in women over 50 years of age were negative with our test. Although it is a benign condition, FCC may be misdiagnosed as a malignant lesion by physical examination and imaging. Conclusion: Our results support the utility of our blood-based test for the early detection of breast cancer which presents improved features over mammography such as higher lobular cancer sensitivity, lower detection rate of common form of benign condition often misdiagnosed by imaging, absence of radiation and expected high compliance. Further testing is needed to confirm accuracy of our test in the dense breast population. Citation Format: Christine Chavany, Jeffrey Dea, Monica Guevara, Rafael Hernandez-Gonzalez, Patrizia Ferroni, Fiorella Guadagni, Rosaura Valle, Moncef Jendoubi. A Serum Biomarker for the Early Detection of Breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-07-08.

Abstract 1269: Large-scale validaton studies of a blood-based effective and affordable test for multicancer early detection

Abstract Cancer early detection aims at reducing cancer deaths. Unfortunately, many established cancer screening methods are not suitable for use in low- and middle-income countries (LMICs) due to cost, complexity, and dependency on extensive medical infrastructure. Nearly 10,000 participants (2003 cancer cases and 7888 non-cancer cases) were divided into one training and five independent validation cohorts across different races, sample types and platforms. One tube of peripheral blood was collected from each participant and quantified using a panel of seven protein tumor markers (PTMs) consisting of AFP, CA125, CA15-3, CA19-9, CA72-4, CEA and CYFRA 21-1 by common clinical immunoassay analyzers. An algorithm named OncoSeek was established using artificial intelligence (AI) to distinguish cancer cases from non-cancer cases by calculating the probability of cancer (POC) index based on the quantification of the seven PTMs and clinical information including sex and age, and to predict the possible affected tissue of origin (TOO). The conventional clinical method that relied only on a single threshold for each PTM would make a big problem when combining the results of those markers as the false positive rate would accumulate as the number of markers increased. Nevertheless, OncoSeek was empowered by AI to significantly reduce the false positive rate, increasing the specificity from 54.0% to 93.0%. The overall sensitivity of OncoSeek was 51.7%, resulting in 84.6% accuracy. The performance was consistent in the training and the five validation cohorts from three countries (Brazil, China and United States) including two sample types (plasma and serum) and three different platforms (Roche, Luminex and ELISA). The sensitivities ranged from 39.0% to 77.6% for the detection of the nine common cancer types (breast, colorectum, liver, lung, lymphoma, oesophagus, ovary, pancreas and stomach), which account for 59.2% of global cancer deaths annually. Furthermore, it has shown excellent sensitivity in several high-mortality cancer types for which there are lacking routine screening tests in the clinic, such as the sensitivity of pancreatic cancer was 77.6%. The overall accuracy of TOO prediction in the true positives was 65.4%, which could assist the clinical diagnostic workup. OncoSeek significantly outperforms the conventional clinical method, representing a novel blood-based test for multicancer early detection (MCED) that is non-invasive, easy, efficient and robust. Moreover, the accuracy of TOO facilitates the follow-up diagnostic workup. OncoSeek is affordable (~$20) and accessible requiring nothing more than a blood draw at the screening sites, which makes it adoptable in LMICs. Citation Format: Mao Mao, Bing Wei, Qingxia Xu, Yong Shen, Raphael Brandão, Shiyong Li, Wei Wu, Pingping Xing, Yinyin Chang, Dandan Zhu. Large-scale validaton studies of a blood-based effective and affordable test for multicancer early detection [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1269.

Revolutionizing gastric cancer diagnosis through advanced machine learning approaches

&lt;p class="Keywords"&gt;Early detection of gastric cancer through a Computer-Aided Detection (CAD) system has the potential to significantly reduce the mortality rate associated with this disease. This study aims to investigate the effects of class imbalance on the performance of machine learning classifiers in this context. Using a dataset of 145,787 screening records from NHS Liverpool Hospital, we employed stratified sampling to create balanced and unbalanced datasets and evaluated the performance of four machine learning algorithms—Logistic Regression, Support Vector Machine, Naive Bayes, and Multilayer Perceptron—under five different test conditions. The study’s novelty lies in its detailed examination of class imbalance in gastric cancer diagnosis, emphasizing the crucial role of balanced datasets in machine learning-based early detection systems. For the MLP model under 10-fold cross-validation, the Class 0 sensitivity (non-cancer cases) of the unbalanced dataset was 0.968, higher than the balanced dataset’s 0.902. However, the Class 1 sensitivity (cancer cases) and Positive Predictive Value (PPV) of the unbalanced dataset were much lower (0.383 and 0.527) than those of the balanced dataset (0.959 and 0.907), indicating a significant improvement in identifying true positive cases when using a balanced dataset. These findings highlight the negative effect of class imbalance on prediction accuracy for positive cancer cases and underscore the importance of addressing this imbalance for more reliable and accurate predictions in medical diagnosis and screening. This approach has the potential to improve patient outcomes and may contribute to strategies aimed at reducing the mortality rate associated with gastric cancer.&lt;/p&gt;

Non-Invasive Cancer Detection Using Blood Test and Predictive Modeling Approach.

The incidence of cancer, which is a serious public health concern, is increasing. A predictive analysis driven by machine learning was integrated with haematology parameters to create a method for the simultaneous diagnosis of several malignancies at different stages. We analysed a newly collected dataset from various hospitals in Jordan comprising 19,537 laboratory reports (6,280 cancer and 13,257 noncancer cases). To clean and obtain the data ready for modelling, preprocessing steps such as feature standardization and missing value removal were used. Several cutting-edge classifiers were employed for the prediction analysis. In addition, we experimented with the dataset's missing values using the histogram gradient boosting (HGB) model. The feature ranking method demonstrated the ability to distinguish cancer patients from healthy individuals based on hematological features such as WBCs, red blood cell (RBC) counts, and platelet (PLT) counts, in addition to age and creatinine level. The random forest (RF) classifier, followed by linear discriminant analysis (LDA) and support vector machine (SVM), achieved the highest prediction accuracy (ranging from 0.69 to 0.72 depending on the scenario and method investigated), reliably distinguishing between malignant and benign conditions. The HGB model showed improved performance on the dataset. After investigating a number of machine learning methods, an efficient screening platform for non-invasive cancer detection is provided by the integration of haematological indicators with proper analytical data. Exploring deep learning methods in the future work, could provide insights into more complex patterns within the dataset, potentially improving the accuracy and robustness of the predictions.

The relationship between HYDIN and fallopian tubal cilia loss in patients with epithelial ovarian cancer.

Primary cilia play an important role in the development of cancer by regulating signaling pathways. Several studies have demonstrated that women with BRCA mutations have, on average, 50% fewer ciliated cells compared with general women. However, the role of tubal cilia loss in the development of epithelial ovarian cancer (EOC) remains unclear. Few specific studies have been found in linking HYDIN, a ciliary defect associated gene that encodes HYDIN axonemal central pair apparatus protein, which is involved in the transduction of Hedgehog (Hh) signal and is considered a cancer associated antigen, to ovarian cancer. Therefore, our study aimed to investigate the correlation between HYDIN gene mutations and tubal cilia loss in EOC. A whole exome sequencing (WES), immunohistochemistry (IHC), western blot, and reverse transcription quantitative (RT q) PCR were performed in 80 patients with EOC and 50 cases of non ovarian cancer to detect the mutations and expression of tubal ciliary marker, ciliary morphology, and abnormal rate. We found that the incidence of tubal cilia loss was higher in EOC group with decreased expression of HYDIN compared with the control group (P<0.05). This study suggests that tubal ciliary loss is evident in epithelial fallopian tube carcinoma, and ciliary cells may be involved in the occurrence and development of EOC, and cilia-related gene HYDIN is expected to be a tumor marker for epithelial ovarian cancer.

Association of Overweight and Inflammatory Indicators with Breast Cancer: A Cross-Sectional Study in Chinese Women.

This cross-sectional study aimed to explore the association of overweight and inflammatory indicators with breast cancer risk in Chinese patients. Weight, height, and peripheral blood inflammatory indicators, including white blood cell count (WBC), neutrophil count (NE), lymphocyte count (LY), platelet count (PLT) and the concentration of hypersensitivity C-reactive protein (hsCRP), were collected in 383 patients with benign breast lumps (non-cancer) and 358 patients with malignant breast tumors (cancer) at the First Affiliated Hospital of Soochow University, China, from March 2018 to July 2020. Body mass index (BMI), neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR) and systemic immune-inflammation index (SII) were determined according to the ratio equation. The correlations among overweight, inflammatory indicators, and the proportion of non-cancer or cancer cases were analyzed. BMI is associated with an increased breast cancer risk. Compared with non-cancer patients, the average WBC count, NE count, NLR, and level of hsCRP were significantly higher in cancer patients. The level of hsCRP was closely associated with the size of malignant breast tumors. We conclude that overweight and high levels of hsCRP may serve as putative risk factors for malignant breast tumors in Chinese women.

Role of liquid biopsy in non small cell lung cancer

Lung cancer is the leading cause of morbidity and mortality worldwide. It is usually diagnosed in advance stage. miRNA present in serum and pleural fluid can be studied for early diagnosis of lung cancer. Present study was carried out to evaluate whether miRNA can be used as biomarkers in diagnosis of non small cell lung cancer. The study was intended to find the non-invasive tumour biomarkers for presence of lung malignancy with the intent of instituting early diagnosis to reduce lung cancer related mortality. The aim of the study was to evaluate circulating microRNA expression in adenocarcinoma and squamous cell carcinoma lung in comparison with age and sex matched healthy controls. The expression of these miRNA was correlated with histopathology and/or immunohistochemistry. The circulating miRNA expression in age and sex matched non-smoking healthy controls was also noted. It was a Prospective observational study in which 50 cases of non small cell lung cancer was included. 50 healthy non smoker volunteers (control group, well adjusted to the patients according to the age and sex) were also included in the study. About 5 ml of serum and wherever possible pleural fluid was collected in the sterile container. The sample was allowed to stand at room temperature for one hour, and then samples were centrifuged at 1300g for ten minutes at room temperature.RNA was extracted using miRNeasy mini kit (Cat no. 217004) and quantative PCR was done. The patients age, sex, histopathological results, clinical staging, immunohistochemistry, presence of pleural effusion. Expression of mi RNA (miRNA 21, miRNA 17-92 cluster, miRNA 221/222, miRNA Let- 7, miRNA 34 and miRNA 200) were studied. Out of 50 patients of suspected lung cancer 17 were females (34%) and 33 (66%) were males. Mean age of presentation was 63.26 years. 37 patients gave history of smoking (74%) while 13 patients were non Smokers (26%). 29 patients (58%) showed histomorphological features suggestive of adenocarcinoma whereas 21 patients (42%) showed histomorphological features of squamous cell carcinoma. EGFR mutation was seen in 10 patients (34%). Pleural effusion was present in 20 cases.Statistically significant correlation was found between the expression of miRNA in healthy controls and in lung cancer patients. All the tested miRNAs were significantly correlated with the corresponding expression in the healthy control. As compared to healthy controls that let-7, miR-34 and miR-200 were downregulated in lung cancer patients whereas miRNA-221, miRNA 17-92, miRNA-21 were upregulated in lung cancer patients. miR 34, miR 200 and let 7 was detected in healthy controls also. No statically significant correlation of miRNA with age, sex, smoking, histopathological type, grade of tumor, stage of disease, EGFR mutation and IHC was found. Stastically significant correlation was found between miRNA 200 and pleural effusion patients. Present study concludes that miRNA can be a potential biomarker for diagnosing lung cancer. To date, there is convincing evidence supporting the potential role of miRNAs as biomarkers for lung cancer diagnosis and prognosis. However, further research is required in this aspect.

GENETIC ALGORITHM AND RANDOM FOREST CLASSIFIER FUSION: A CUTTING-EDGE APPROACH FOR BREAST CANCER DIAGNOSIS

Breast cancer was a significant cause of mortality in women worldwide, highlighting the importance of early detection in improving patient survival rates. Although machine learning algorithms had shown effectiveness in diagnosing breast cancer, there was still room for improvement. This paper introduced a ground-breaking method that combined Genetic Algorithms (GAs) with Random Forest Classifiers (RFCs) for breast cancer diagnosis. GA’s were used to select the most informative features from the breast cancer dataset, while RFCs were employed to classify the data into cancerous and non-cancerous cases. The proposed approach was evaluated on a publicly available breast cancer dataset, and the results demonstrated a remarkable accuracy of 79.31%, surpassing the accuracy of RFCs without GA-based feature selection (77.58%). This innovative approach held great promise in improving the accuracy of early diagnosis and potentially saving lives. By leveraging the strengths of GAs and RFCs, this novel approach offered an effective means of diagnosing breast cancer and had the potential to revolutionize early detection practices.

Abstract A030: Blood-based screening panel for lung cancer based on clonal hematopoietic mutations in tumor infiltrating immune cells

Abstract Early detection can significantly reduce mortality due to lung cancer. However, the high cost of the currently approved screening protocol has limited its uptake. Presented here is a blood-based screening panel based on clonal hematopoietic mutations. Mutations in tumor cells that inhibit immune destruction have been extensively studied. However, mutations in immune cells that may prevent an effective anti-tumor immune response remain relatively unstudied. Animal model studies suggest that clonal hematopoietic mutations in tumor infiltrating immune cells can modulate cancer progression, representing potential predictive biomarkers. The goal of this study was to determine if the clonal expansion of these mutations in blood samples could predict the occurrence of lung cancer. We identified a set of 98 potentially pathogenic clonal hematopoietic mutations in tumor infiltrating immune cells. A logistic regression machine learning model based on these mutations correctly classified lung cancer and non-cancer blood samples with 94.12% sensitivity (95% Confidence Interval: 92.20-96.04%) and 85.96% specificity (95% Confidence Interval: 82.98-88.95%) in a test set of 578 lung cancer and 545 non-cancer samples from 18 different cohorts. More importantly, 89.98% of the cancer cases were unambiguously predicted with probability of cancer &amp;gt;0.90 and 74.86% of the non-cancer cases with probability of cancer &amp;lt;0.10. These results suggest that it may be possible to develop an accurate blood-based lung cancer screening panel. Unlike most other “liquid biopsies” currently under development, the assay presented here is based on standard sequencing protocols and uses a relatively small number of rationally selected mutations as predictors. Citation Format: Ramu Anandakrishnan, Ryan Shahidi, Andrew Dai, Veneeth Antony. Blood-based screening panel for lung cancer based on clonal hematopoietic mutations in tumor infiltrating immune cells [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Tumor Immunology and Immunotherapy; 2023 Oct 1-4; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Cancer Immunol Res 2023;11(12 Suppl):Abstract nr A030.

Breast cancer early detection in TP53 SNP protein sequences based on a new Convolutional Neural Network model

INTRODUCTION: Breast cancer (BC) is the most commonly occurring cancer and the second leading cause for women’s disease death. The BC cases are associated with genital mutations which are inherited from older generations or acquired overtime. If the diagnosis is done at the first stage, effects associated with certain treatments can be limited, costs can be saved and the diagnostic time can be minimized. This can also help specialists target the best treatment to increase the rate of cures. Nevertheless, its discovery in patients is very challenging due to silent symptoms aside from the fact the routine screening is not recommended for women under 40 years old.OBJECTIVES: Several efforts are aimed at the BC early detection using machine and deep learning systems. The proposed algorithms use different data types to distinguish between cancerous and non-cancerous cases; as: mammography, ultrasound and MRI (magnetic resonance imaging) images. Then, different learning tools were applied on this data for the classification task. Despite the classification rates which exceed 90%, the major drawback of all these methods is that they are applicable only after the appearance of the cancerous tumors, which reduces the cure rates.METHODS: We propose a new technique for early breast cancer screening. For the data, we focus on cancerous and non-cancerous SNP (Single Nucleotide Polymorphism) protein sequences of the TP53 gene in chromosome 17. This gene is shown to be linked to different single amino acid mutations on which we will shed light here. The method we propose transforms SNP textual sequences into digital vectors via coding. Then, RGB scalogram images are generated using the continuous wavelet transform. A pretreatment of color coefficients is applied to scalograms aiming at creating four different databases. Finally, a CNN deep learning network is used for the binary classification of cancerous and non-cancerous images.RESULTS: During the validation process, we reached good performance with specificity of 97.84%, sensitivity of 96.45%, an overall accuracy of 95.29% and an equal run time of 12 minutes 3 seconds. These values ensure the efficiency of our method.To enhance more these results, we used the ORB feature detection technique. Consequently, the classification rates have been improved to reach 95.9% as accuracyCONCLUSION: Our method will allow significant savings time and lives by detecting the disease in patients whose genetic mutations are beginning to appear.