Can Natural Language Processing Improve Adnexal Torsion Predictions?

STUDY OBJECTIVE: To create a decision support tool based on machine learning algorithms and natural language processing (NLP) technology, to augment clinicians' ability to predict cases of suspected adnexal torsion. DESIGN: Retrospective cohort study SETTING: Gynecology department, university-affiliated teaching medical center, 2014-2022. PATIENTS: This study assessed risk-factors for adnexal torsion among women managed surgically for suspected adnexal torsion based on clinical and sonographic data. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The dataset included demographic, clinical, sonographic, and surgical information obtained from electronic medical records. NLP was used to extract insights from unstructured free text and unlock them for automated reasoning. The machine learning model was a CatBoost classifier that utilizes gradient boosting on decision trees. The study cohort included 433 women who met inclusion criteria and underwent laparoscopy. Among them, 320 (74%) had adnexal torsion diagnosed during laparoscopy, and 113 (26%) did not. The model developed improved prediction of adnexal torsion to 84%, with a recall of 95%. The model ranked several parameters as important for prediction. Age, difference in size between ovaries, and the size of each ovary were the most significant. The precision for the "no torsion" class was 77%, with a recall of 45%. CONCLUSIONS: Using machine learning algorithms and NLP technology as a decision-support tool for the diagnosis of adnexal torsion is feasible. It improved true prediction of adnexal torsion to 84% and decreased cases of unnecessary laparoscopy.

IL-6 contributes to metastatic switch via the differentiation of monocytic-dendritic progenitors into prometastatic immune cells.

BACKGROUND: Metastasis is the major cause of death in patients with cancer. Myeloid skewing of hematopoietic cells is a prominent promoter of metastasis. However, the reservoir of these cells in the bone marrow (BM) compartment and their differentiation pattern from hematopoietic stem and progenitor cells (HSPCs) have not been explored. METHODS: We used a unique model system consisting of tumor cell clones with low metastatic potential or high metastatic potential (met-low and met-high, respectively) to investigate the fate of HSPC differentiation using murine melanoma and breast carcinoma. Single-cell RNA sequencing (scRNA-seq) analysis was performed on HSPC obtained from the BM of met-low and met-high tumors. A proteomic screen of tumor-conditioned medium integrated with the scRNA-seq data analysis was performed to analyze the potential cross talk between cancer cells and HSPCs. Adoptive transfer of tumor-educated HSPC subsets obtained from green fluorescent protein (GFP)+ tagged mice was then carried out to identify the contribution of committed HSPCs to tumor spread. Peripheral mononuclear cells obtained from patients with breast and lung cancer were analyzed for HSPC subsets. RESULTS: Mice bearing met-high tumors exhibited a significant increase in the percentage of HSPCs in the BM in comparison with tumor-free mice or mice bearing met-low tumors. ScRNA-seq analysis of these HSPCs revealed that met-high tumors enriched the monocyte-dendritic progenitors (MDPs) but not granulocyte-monocyte progenitors (GMPs). A proteomic screen of tumor- conditioned medium integrated with the scRNA-seq data analysis revealed that the interleukin 6 (IL-6)-IL-6 receptor axis is highly active in HSPC-derived MDP cells. Consequently, loss of function and gain of function of IL-6 in tumor cells resulted in decreased and increased metastasis and corresponding MDP levels, respectively. Importantly, IL-6-educated MDPs induce metastasis within mice bearing met-low tumors-through further differentiation into immunosuppressive macrophages and not dendritic cells. Consistently, MDP but not GMP levels in peripheral blood of breast and lung cancer patients are correlated with tumor aggressiveness. CONCLUSIONS: Our study reveals a new role for tumor-derived IL-6 in hijacking the HSPC differentiation program toward prometastatic MDPs that functionally differentiate into immunosuppressive monocytes to support the metastatic switch.

A clinically meaningful metric of immune age derived from high-dimensional longitudinal monitoring.

Immune responses generally decline with age. However, the dynamics of this process at the individual level have not been characterized, hindering quantification of an individual's immune age. Here, we use multiple 'omics' technologies to capture population- and individual-level changes in the human immune system of 135 healthy adult individuals of different ages sampled longitudinally over a nine-year period. We observed high inter-individual variability in the rates of change of cellular frequencies that was dictated by their baseline values, allowing identification of steady-state levels toward which a cell subset converged and the ordered convergence of multiple cell subsets toward an older adult homeostasis. These data form a high-dimensional trajectory of immune aging (IMM-AGE) that describes a person's immune status better than chronological age. We show that the IMM-AGE score predicted all-cause mortality beyond well-established risk factors in the Framingham Heart Study, establishing its potential use in clinics for identification of patients at risk.

Immune-centric network of cytokines and cells in disease context identified by computational mining of PubMed.

Cytokines are signaling molecules secreted and sensed by immune and other cell types, enabling dynamic intercellular communication. Although a vast amount of data on these interactions exists, this information is not compiled, integrated or easily searchable. Here we report immuneXpresso, a text-mining engine that structures and standardizes knowledge of immune intercellular communication. We applied immuneXpresso to PubMed to identify relationships between 340 cell types and 140 cytokines across thousands of diseases. The method is able to distinguish between incoming and outgoing interactions, and it includes the effect of the interaction and the cellular function involved. These factors are assigned a confidence score and linked to the disease. By leveraging the breadth of this network, we predicted and experimentally verified previously unappreciated cell-cytokine interactions. We also built a global immune-centric view of diseases and used it to predict cytokine-disease associations. This standardized knowledgebase (http://www.immunexpresso.org) opens up new directions for interpretation of immune data and model-driven systems immunology.

Social network architecture of human immune cells unveiled by quantitative proteomics.

The immune system is unique in its dynamic interplay between numerous cell types. However, a system-wide view of how immune cells communicate to protect against disease has not yet been established. We applied high-resolution mass-spectrometry-based proteomics to characterize 28 primary human hematopoietic cell populations in steady and activated states at a depth of >10,000 proteins in total. Protein copy numbers revealed a specialization of immune cells for ligand and receptor expression, thereby connecting distinct immune functions. By integrating total and secreted proteomes, we discovered fundamental intercellular communication structures and previously unknown connections between cell types. Our publicly accessible (http://www.immprot.org/) proteomic resource provides a framework for the orchestration of cellular interplay and a reference for altered communication associated with pathology.