False negative breast cancers on imaging and associated risk factors: a single institution six-year analysis.

PURPOSE: Mitigating false negative imaging studies remains an important issue given its association with worse morbidity and mortality in patients with breast cancer. We aimed to identify risk factors that predispose to false negative breast imaging exams. METHODS: In an IRB-approved, HIPAA compliant retrospective study, we identified all patients who were diagnosed with breast cancer within 365 days of a negative imaging study assessed as BI-RADS 1-3 between January 1, 2014 and January 31, 2020. A matched cohort based on mammographic breast density was created from randomly selected studies with BI-RADS 4-5 designation that yielded breast cancer at pathology within the same time frame. Patient and cancer characteristics, prior personal history of breast cancer and gene mutation status were collected from patient charts. Pearson chi-squared and Student's t-test on two independent groups with significance at < 0.05 was used for statistical analysis. RESULTS: We identified 155 false negative studies of 129 missed cancers and 128 breast density matched true positive cancers. False negative studies were screening mammograms in 57.42% (89/155), diagnostic mammograms in 29.68% (46/155), ultrasounds in 6.45% (10/155) and MRIs in 6.45% (10/155). Rates of personal (41.09% vs. 18.75%, p < 0.001) and family history of breast cancer (68.22% vs. 49.21%, p = 0.002) were higher in the false negative cohort and remained significant when asymptomatic MRI-detected cancers were removed. CONCLUSION: Our findings suggest that supplemental screening may be useful in breast cancer survivors.

An interactive web application for exploring human plasma and fibroblast metabolomics data from patients with inborn errors of metabolism.

Metabolomic profiling is instrumental in understanding the systemic and cellular impact of inborn errors of metabolism (IEMs), monogenic disorders caused by pathogenic genomic variants in genes involved in metabolism. This study encompasses untargeted metabolomics analysis of plasma from 474 individuals and fibroblasts from 67 subjects, incorporating healthy controls, patients with 65 different monogenic diseases, and numerous undiagnosed cases. We introduce a web application designed for the in-depth exploration of this extensive metabolomics database. The application offers a user-friendly interface for data review, download, and detailed analysis of metabolic deviations linked to IEMs at the level of individual patients or groups of patients with the same diagnosis. It also provides interactive tools for investigating metabolic relationships and offers comparative analyses of plasma and fibroblast profiles. This tool emphasizes the metabolic interplay within and across biological matrices, enriching our understanding of metabolic regulation in health and disease. As a resource, the application provides broad utility in research, offering novel insights into metabolic pathways and their alterations in various disorders.

Machine learning models to prognose 30-Day Mortality in Postoperative Disseminated Cancer Patients.

Patients with disseminated cancer at higher risk for postoperative mortality see improved outcomes with altered clinical management. Being able to risk stratify patients immediately after their index surgery to flag high risk patients for healthcare providers is vital. The combination of physician uncertainty and a demonstrated optimism bias often lead to an overestimation of patient life expectancy which can precent proper end of life counseling and lead to inadequate postoperative follow up. In this cohort study of 167,474 postoperative patients with multiple types of disseminated cancer, patients at high risk of 30-day postoperative mortality were accurately identified using our machine learning models based solely on clinical features and preoperative lab values. Extreme Gradient Boosting, Random Forest, and Logistic Regression machine learning models were developed on the cohort. Among 167,474 disseminated cancer patients, 50,669 (30.3%) died within 30 days of their index surgery; After preprocessing, 28 features were included in the model development. The cohort was randomly divided into 133,979 patients (80%) for training the models and 33,495 patients (20%) for testing. The extreme gradient boosting model had an AUC of 0.93 (95% CI: 0.926-0.931), the random forest model had an AUC of 0.93 (95% CI: 0.930-0.934), and the logistic regression model had an AUC of 0.90 (95% CI: 0.900-0.906 the index operation. Ultimately, Machine learning models were able to accurately predict short-term postoperative mortality among a heterogenous population of disseminated cancer patients using commonly accessible medical features. These models can be included in electronic health systems to guide clinical judgements that affect direct patient care, particularly in low-resource settings.

Machine learning methods to predict presence of residual cancer following hysterectomy.

Surgical management for gynecologic malignancies often involves hysterectomy, often constituting the most common gynecologic surgery worldwide. Despite maximal surgical and medical care, gynecologic malignancies have a high rate of recurrence following surgery. Current machine learning models use advanced pathology data that is often inaccessible within low-resource settings and are specific to singular cancer types. There is currently a need for machine learning models to predict non-clinically evident residual disease using only clinically available health data. Here we developed and tested multiple machine learning models to assess the risk of residual disease post-hysterectomy based on clinical and operative parameters. Data from 3656 hysterectomy patients from the NSQIP dataset over 14 years were used to develop models with a training set of 2925 patients and a validation set of 731 patients. Our models revealed the top postoperative predictors of residual disease were the initial presence of gross abdominal disease on the diaphragm, disease located on the bowel mesentery, located on the bowel serosa, and disease located within the adjacent pelvis prior to resection. There were no statistically significant differences in performances of the top three models. Extreme gradient Boosting, Random Forest, and Logistic Regression models had comparable AUC ROC (0.90) and accuracy metrics (87-88%). Using these models, physicians can identify gynecologic cancer patients post-hysterectomy that may benefit from additional treatment. For patients at high risk for disease recurrence despite adequate surgical intervention, machine learning models may lay the basis for potential prospective trials with prophylactic/adjuvant therapy for non-clinically evident residual disease, particularly in under-resourced settings.