Stratification Of Patients Research Articles

Improving the performance and interpretability on medical datasets using graphical ensemble feature selection.

A major hindrance towards using Machine Learning (ML) on medical datasets is the discrepancy between a large number of variables and small sample sizes. While multiple feature selection techniques have been proposed to avoid the resulting overfitting, overall ensemble techniques offer the best selection robustness. Yet, current methods designed to combine different algorithms generally fail to leverage the dependencies identified by their components. Here, we propose Graphical Ensembling (GE), a graph-theory-based ensemble feature selection technique designed to improve the stability and relevance of the selected features. Relying on four datasets, we show that GE increases classification performance with fewer selected features. For example, on rheumatoid arthritis patient stratification, GE outperforms the baseline methods by 9% Balanced Accuracy while relying on fewer features. We use data on sub-cellular networks to show that the selected features (proteins) are closer to the known disease genes, and the uncovered biological mechanisms are more diversified. By successfully tackling the complex correlations between biological variables, we anticipate that GE will improve the medical applications of ML. https://github.com/ebattistella/auto_machine_learning.

Genome-wide in vivo CRISPR screen identifies TGFβ3 as actionable biomarker of palbociclib resistance in triple negative breast cancer

Triple negative breast cancer (TNBC) remains exceptionally challenging to treat. While CDK4/6 inhibitors have revolutionized HR + breast cancer therapy, there is limited understanding of their efficacy in TNBC and meaningful predictors of response and resistance to these drugs remain scarce. We conducted an in vivo genome-wide CRISPR screen using palbociclib as a selection pressure in TNBC. Hits were prioritized using microarray data from a large panel of breast cancer cell lines to identify top palbociclib sensitizers. Our study defines TGFβ3 as an actionable determinant of palbociclib sensitivity that potentiates its anti-tumor effects. Mechanistically, we show that chronic palbociclib exposure depletes p21 levels, contributing to acquired resistance, and that TGFβ3 treatment can overcome this. This study defines TGFβ3 as an actionable biomarker that can be used to improve patient stratification for palbociclib treatment and exploits the synergistic interaction between CDK4/6 and TGFβ3 to propose a new combinatorial treatment for TNBC.

PDAC heterogeneity resolved spatially at the single-cell level: new biological answers, new questions on optimal translation†.

Pancreatic cancer is a highly aggressive disease. Developing new strategies and using powerful methodologies for its early detection, coupled with in-depth comprehension of the mechanisms governing subtype evolution, will not only help to stratify PDAC patients' prognosis but also prevent unfavourable subtype plasticity upon treatment with chemotherapy. Michiels et al have developed a new approach to better capture PDAC heterogeneity at the single tumour duct spatial resolution level, leveraging detection of transcripts for mutant KRAS and multiple subtype markers. Their study sheds light on the association of mutant KRAS and PDAC phenotypic heterogeneity. The findings support functional cooperation of plastic tumour cells and opens new challenges towards PDAC patient stratification and therapeutic intervention. Pathology-based tools will be of prime importance to address these issues in a clinically meaningful manner. © 2024 The Pathological Society of Great Britain and Ireland.

Improving Prediction of Survival and Progression in Metastatic Non-Small Cell Lung Cancer After Immunotherapy Through Machine Learning of Circulating Tumor DNA.

To use modern machine learning approaches to enhance and automate the feature extraction from the longitudinal circulating tumor DNA (ctDNA) data and to improve the prediction of survival and disease progression, risk stratification, and treatment strategies for patients with 1L non-small cell lung cancer (NSCLC). Using IMpower150 trial data on patients with untreated metastatic NSCLC treated with atezolizumab and chemotherapies, we developed a machine learning algorithm to extract predictive features from ctDNA kinetics, improving survival and progression prediction. We analyzed kinetic data from 17 ctDNA summary markers, including cell-free DNA concentration, allele frequency, tumor molecules in plasma, and mutation counts. Three hundred and ninety-eight patients with ctDNA data (206 in training and 192 in validation) were analyzed. Our models outperformed existing workflow using conventional temporal ctDNA features, raising overall survival (OS) concordance index to 0.72 and 0.71 from 0.67 and 0.63 for C3D1 and C4D1, respectively, and substantially improving progression-free survival (PFS) to approximately 0.65 from the previous 0.54-0.58, a 12%-20% increase. Additionally, they enhanced risk stratification for patients with NSCLC, achieving clear OS and PFS separation. Distinct patterns of ctDNA kinetic characteristics (eg, baseline ctDNA markers, depth of ctDNA responses, and timing of ctDNA clearance, etc) were revealed across the risk groups. Rapid and complete ctDNA clearance appears essential for long-term clinical benefit. Our machine learning approach offers a novel tool for analyzing ctDNA kinetics, extracting critical features from longitudinal data, improving our understanding of the link between ctDNA kinetics and progression/mortality risks, and optimizing personalized immunotherapies for 1L NSCLC.

Overall survival prediction and risk stratification in patients with neuroblastoma through machine learning in the large multi-institutional PRIMAGE cohort.

10054 Background: Neuroblastoma (NB) is the most prevalent solid cancer in childhood in which imaging plays a pivotal role at every step of the patient's journey. This investigation aimed to develop a machine learning model using clinical, molecular (MycN amplification), and magnetic resonance (MR) radiomics features at diagnosis to predict patient’s overall survival (OS) and improve their risk stratification. Methods: A database comprising clinical, molecular, and International Neuroblastoma Risk Group (INRG) staging on 513 patients with accessible MR imaging (discovery cohort) was employed for model training, validation, and testing. An additional 22 patients from non-discovery cohort hospitals served as an external validation group. Tumor segmentations of NB were manually and semi-automatically performed on corresponding T2-weighted MR images by an experienced radiologist. In total, 107 radiomics features were extracted and harmonized across MR scan manufacturers and magnetic field strengths using the nested ComBat methodology to correct both batch effects. These radiomics features, combined with clinical and molecular data, were utilized as input for the models. A nested cross-validation approach was employed for model development to determine the optimal preprocessing, machine learning algorithm, and model configuration. Results: The discovery cohort yielded a C-index of 0.788 ± 0.049 in the test partitions, with a random survival forest (RSF) exhibiting the best performance. In the validation cohort, a C-index of 0.934 was achieved. Interpretability analysis identified lesion heterogeneity, size, and molecular variables as crucial factors in OS prediction. The model demonstrated superior predictive performance and patient stratification compared to conventional staging systems in both cohorts. Conclusions: The RSF predictive model exhibited high performance, emphasizing the significant contribution of radiomics features and alignment with established clinical and molecular variables. The model stratified NB patients into low-, intermediate-, and high-risk categories and suggests that radiomics features potentially could improve current risk stratification systems. Additional external validation is warranted as this may present new evidence for enhancing patient care and clinical decision-making for NB patients.

Homologous recombination deficiency score served as an independent prognosis factor in esophageal squamous cell carcinoma.

e16080 Background: Homologous recombination deficiency (HRD) is a functional impairment of the homologous recombination repair (HRR) pathway, pivotal for mending DNA double-strand breaks, contributing significantly to cancer-associated genomic instability. While BRCA mutations traditionally underlie HRR defects, recent studies highlight varied responses to poly (ADP-ribose) polymerase (PARP) inhibitors based on specific HRR-related gene alterations. Despite HRD's relevance in various cancers, its role in esophageal squamous cell carcinoma (ESCC) remains underexplored. Methods: We conducted a retrospective analysis of HRD scores in 96 ESCC patients, assessing correlations with clinical characteristics and survival outcomes. Genomic sequencing was performed using a customized superHRD NGS panel. HRD scores were calculated based on 54,000 single-nucleotide polymorphisms, employing Kruskal-Wallis ranksum tests and two cut-off points for quantitative analysis. Kaplan-Meier survival and Cox regression model were employed to evaluate disease-free survival (DFS) and overall survival (OS). Results: Higher HRD scores correlated with advanced tumor stages, recurrence, and specific gene mutations ( TP53 and ABCB1). APC mutations were associated with lower HRD scores. Patients with high HRD scores demonstrated significantly shorter disease-free survival (DFS) and a trend towards shorter overall survival (OS), especially those who did not receive adjuvant therapy, emphasizing HRD's prognostic value. Conversely, HRD-high patients undergoing adjuvant therapy showed a tendency towards extended OS. Multivariate analysis identified HRD as an independent prognostic factor (HR = 2.814 for recurrence, HR = 1.772 for death). Validation with The Cancer Genome Atlas (TCGA) ESCC dataset supported our findings. Conclusions: This study unveils associations between HRD scores, clinical characteristics, and genomic mutations in ESCC, emphasizing HRD's potential as a prognostic biomarker. HRD assessment may aid in patient stratification and personalized treatment strategies. Future investigations should explore underlying molecular mechanisms and therapeutic implications of HRD-associated pathways in ESCC management.

Advancing rare cancer research by MALDI mass spectrometry imaging: Applications, challenges, and future perspectives in sarcoma.

MALDI mass spectrometry imaging (MALDI imaging) uniquely advances cancer research, by measuring spatial distribution of endogenous and exogenous molecules directly from tissue sections. These molecular maps provide valuable insights into basic and translational cancer research, including tumor biology, tumor microenvironment, biomarker identification, drug treatment, and patient stratification. Despite its advantages, MALDI imaging is underutilized in studying rare cancers. Sarcomas, a group of malignant mesenchymal tumors, pose unique challenges in medical research due to their complex heterogeneity and low incidence, resulting in understudied subtypes with suboptimal management and outcomes. In this review, we explore the applicability of MALDI imaging in sarcoma research, showcasing its value in understanding this highly heterogeneous and challenging rare cancer. We summarize all MALDI imaging studies in sarcoma to date, highlight their impact on key research fields, including molecular signatures, cancer heterogeneity, and drug studies. We address specific challenges encountered when employing MALDI imaging for sarcomas, and propose solutions, such as using formalin-fixed paraffin-embedded tissues, and multiplexed experiments, and considerations for multi-site studies and digital data sharing practices. Through this review, we aim to spark collaboration between MALDI imaging researchers and clinical colleagues, to deploy the unique capabilities of MALDI imaging in the context of sarcoma.

Gene expression correlates of adagrasib response in KRASG12C mutated non-small cell lung cancer (NSCLC).

8529 Background: KRASG12C inhibitors such as adagrasib and sotorasib have shown clinical promise in targeting KRASG12C-mutated lung cancers; however, most patients develop primary or secondary resistance. A biomarker of response to KRASG12C inhibitor is needed for better patient stratification and to understand the resistance mechanism. Methods: We analyzed transcriptional correlates of adagrasib treatment outcome in 68 patients in the KRYSTAL-1 trial, a phase 1/2 clinical trial of adagrasib monotherapy in the second line and beyond treatment for NSCLC. We also treated KRAS-mutated lung cancer mouse models and organoids with a KRAS inhibitor for the long term and characterized the resistant tumors’ transcriptional profile to identify resistance mechanisms. We also performed serial gene expression analysis of KrasG12D mutated lung organoids undergoing squamous transformation to identify transcription factor involved in the resistance process. Results: In patients with lung adenocarcinoma with KRASG12C and STK11/ LKB1 co-mutations, we find an enrichment of the squamous cell carcinoma gene signature in pre-treatment biopsies correlates with a poor response to adagrasib. Studies of Lkb1-deficient KRASG12C and KrasG12D lung cancer mouse models and organoids, treated with KRAS inhibitors adagrasib and MRTX1133, respectively, reveal that tumors invoke a lineage plasticity program, adeno-to-squamous transition (AST), that enables resistance to KRAS inhibition. We identify TP63 to be a transcription factor whose expression correlated with squamous transformation. The analysis of lineage plasticity program, adagrasib resistant tumors, and p63 regulon revealed KRT6A to be a common biomarker whose expression correlated with overall survival in the KRYSTAL-1 cohort. Conclusions: KRASG12C mutated lung adenocarcinoma patients with a higher expression of squamous cell carcinoma gene expression signature respond poorly to adagrasib treatment. Expression of the AST plasticity signature and KRT6A at baseline correlates with poor adagrasib responses. These data indicate the role of AST in KRAS inhibitor resistance and provide predictive biomarkers for KRAS-targeted therapies in lung cancer.

Development and Standardization of an Osteoradionecrosis Classification System in Head and Neck Cancer: Implementation of a Risk-Based Model.

Osteoradionecrosis of the jaw (ORN) can manifest in varying severity. The aim of this study is to identify ORN risk factors and develop a novel classification to depict the severity of ORN. Consecutive patients with head and neck cancer (HNC) treated with curative-intent intensity-modulated radiation therapy (IMRT) (≥45 Gy) from 2011 to 2017 were included. Occurrence of ORN was identified from in-house prospective dental and clinical databases and charts. Multivariable logistic regression model was used to identify risk factors and stratify patients into high-risk and low-risk groups. A novel ORN classification system was developed to depict ORN severity by modifying existing systems and incorporating expert opinion. The performance of the novel system was compared with 15 existing systems for their ability to identify and predict serious ORN event (jaw fracture or requiring jaw resection). ORN was identified in 219 of 2,732 (8%) consecutive patients with HNC. Factors associated with high risk of ORN were oral cavity or oropharyngeal primaries, received IMRT dose ≥60 Gy, current/ex-smokers, and/or stage III to IV periodontal condition. The ORN rate for high-risk versus low-risk patients was 12.7% versus 3.1% (P < .001) with an AUC of 0.71. Existing ORN systems overclassified serious ORN events and failed to recognize maxillary ORN. A novel ORN classification system, ClinRad, was proposed on the basis of vertical extent of bone necrosis and presence/absence of exposed bone/fistula. This system detected serious ORN events in 5.7% of patients and statistically outperformed existing systems. We identified risk factors for ORN and proposed a novel ORN classification system on the basis of vertical extent of bone necrosis and presence/absence of exposed bone/fistula. It outperformed existing systems in depicting the seriousness of ORN and may facilitate clinical care and clinical trials.

UBE2R2-AS1, as a prognostic marker of gastric cancer, promotes the malignant phenotype of gastric cancer cells.

This study aimed to unveil the potential of UBE2R2-AS1 dysregulation in gastric cancer. In addition, its biological function was assessed. UBE2R2-AS1 expression was predicted in the ENCORI database. Paired gastric cancer and noncancerous tissues were collected. UBE2R2-AS1 expression was confirmed using RT-qPCR in our patient set. The association of UBE2R2-AS1 with the clinical data of patients was analyzed. Evaluation of the prognostic value of UBE2R2-AS1 was via Kaplan-Meier and Univariate/Multivariate Cox analyses. The effect of UBE2R2-AS1 on the cancer cell malignant phenotype was investigated. Gastric cancer tissues and cells significantly overexpressed UBE2R2-AS1. UBE2R2-AS1 was significantly more abundant in unfavorable clinical pathology, including advanced TNM stage and lymph node metastasis. High expression of UBE2R2-AS1 predicted a poor prognosis with a hazard ratio (HR) of 3.041 and 2.805 after Univariate and Multivariate Cox analysis, respectively. UBE2R2-AS1 can act as a sponge for miR-302b-5p to promote cell proliferation, migration, and invasion of gastric cancer. The expression of UBE2R2-AS1 allowed the prognostic stratification of gastric cancer patients. UBE2R2-AS1 may accelerate the progression of gastric cancer via miR-302b-5p.

Revolutionizing outcome prediction for immunotherapy with vision transformer-based pathomics signature in esophageal squamous cell cancer.

e16049 Background: Immune checkpoint inhibitors (ICIs) has notably enhanced the survival prospects for patients with esophageal squamous cell carcinoma (ESCC), however the clinical benefit was limited to only a small portion of patients. This study aimed to perform a deep learning signature based on H&amp;E-stained pathological specimens for the purpose of identifying effective markers that can accurately predict the clinical benefit of ICIs among patients with ESCC. Methods: ESCC patients receiving PD-1 inhibitors from Shandong Cancer Hospital were included. Whole slide images (WSIs) of H&amp;E-stained histological specimens of included patients were collected, and subsequently randomly allocated into two sets: a training set comprising 70% of the samples, and a validation set consisting of the remaining 30%. The labels of images were defined by the progression-free survival (PFS) with the interval of 4 months. The pretrained ViT model was used for patch-level model training, and all patches were projected into probabilities after linear classifier. Then the most predictive patches were passed to RNN for final patient-level prediction to construct ESCC-pathomics signature (ESCC-PS). Accuracy rate and survival analysis were performed to evaluate the performance of ViT-RNN survival model in validation cohort. Results: 163 ESCC patients receiving PD-1 inhibitors were included for model training. There were 486,188 patches of 1024*1024 pixels from 324 WSI images of H&amp;E-stained histological specimens after image pre-processing. The training cohort consisted of 120 patients with a total of 227 images, while the validation cohort comprised 43 patients and featured 97 images. Notably, both groups exhibited balanced baseline characteristics between them. The ESCC-PS achieved an accuracy of 84.5% in the validation cohort, and could distinguish patients into three risk groups with the median PFS of 2.7, 4.8 and 16.7 months (P &lt; 0.001). The multivariate cox analysis revealed ESCC-PS could act as an independent predictor of survival from PD-1 inhibitors (P &lt; 0.001). A combined signature incorporating ESCC-PS and expression of PD-L1 shows significantly improved accuracy in outcome prediction of PD-1 inhibitors compared to ESCC-PS and PD-L1 anlone, with the area under curve value of 0.952, 0.924, 0.648 for 6-month PFS and C-index of 0.816, 0.806, 0.66, respectively. Conclusions: The outcome supervised pathomics signature based on deep learning has the potential to enable superior prognostic stratification of ESCC patients receiving PD-1 inhibitors, which convert the images pixels to an effective and labour-saving tool to optimize clinical management of ESCC patients.

Biomarker-Driven Oncology Trial Design and Subgroup Characterization: Challenges and Potential Solutions.

In oncology drug development, using biomarkers to select a study population more likely to benefit from a therapeutic effect is critical to increase the efficiency of a clinical trial in demonstrating effectiveness. This perspective delves into therapeutic product approvals that were tested in pivotal trials with all-comers populations, but ultimately received US Food and Drug Administration approval for use within specific patient subgroups identified by biomarkers. Despite initial designs for efficacy and safety assessments in overall populations, a favorable benefit-risk assessment was primarily established in biomarker-positive subgroups. Analyzing these cases, we summarize key considerations pivotal to totality of evidence for regulatory benefit-risk assessments for biomarker-defined subgroup versus all-comers approvals, including biological and clinical rationales, biomarker prevalence, safety data, overall trial design, and subgroup efficacy characterization. Furthermore, a decision tree is proposed to guide optimal clinical trial design, delineating between patient enrichment and stratification, accounting for key factors including biological and clinical rationale, marker type (discreate or continuous), prevalence, assay readiness, and turnaround times for marker assessment. Finally, a recommended approach for subgroup characterization involves prespecifying magnitude of improvement that would be considered clinically meaningful in the biomarker-negative subgroup, which can be supplemented with methodologies such as Bayesian to incorporate evidence from similar studies when available. In summary, this perspective underscores the importance of clinical trial innovations, statistical methodologies and regulatory considerations, to optimize biomarker-driven drug development for patients with cancer.

Development and evaluation of a disulfidoptosis-related lncRNA index for prognostication in clear cell renal cell carcinoma

BackgroundThis study introduces a novel prognostic tool, the Disulfidoptosis-Related lncRNA Index (DRLI), integrating the molecular signatures of disulfidoptosis and long non-coding RNAs (lncRNAs) with the cellular heterogeneity of the tumor microenvironment, to predict clinical outcomes in patients with clear cell renal cell carcinoma (ccRCC). MethodsWe analyzed 530 tumor and 72 normal samples from The Cancer Genome Atlas (TCGA), employing k-means clustering based on disulfidoptosis-associated gene expression to stratify ccRCC samples into prognostic groups. lncRNAs correlated with disulfidoptosis were identified and used to construct the DRLI, which was validated by Kaplan-Meier and receiver operating characteristic curves. We utilized single-cell deconvolution analysis to estimate the proportion of immune cell types within the tumor microenvironment, while the ESTIMATE and TIDE algorithms were employed to assess immune infiltration and potential response to immunotherapy. ResultsThe Disulfidoptosis-Related lncRNA Index (DRLI) effectively stratified ccRCC patients into high and low-risk groups, significantly impacting survival outcomes (P < 0.001). High-risk patients, marked by a unique lncRNA profile associated with disulfidoptosis, faced worse prognoses. Single-cell analysis revealed marked tumor microenvironment heterogeneity, especially in immune cell makeup, correlating with patient risk levels. In prognostic predictions, DRLI outperformed traditional clinical indicators, achieving AUC values of 0.779, 0.757, and 0.779 for 1-year, 3-year, and 5-year survival in the training set, and 0.746, 0.734, and 0.750 in the validation set. Notably, while the constructed nomogram showed exceptional predictive capability for short-term prognosis (AUC = 0.877), the DRLI displayed remarkable long-term predictive accuracy, with its AUC value reaching 0.823 for 10-year survival, closely approaching the nomogram's performance. ConclusionsThe study introduces the DRLI as a groundbreaking molecular stratification tool for ccRCC, enhancing prognostic precision and potentially guiding personalized treatment strategies. This advancement is particularly significant in the context of long-term survival predictions. Our findings also elucidate the complex interplay between disulfidoptosis, lncRNAs, and the immune microenvironment in ccRCC, offering a comprehensive perspective on its pathogenesis and progression. The DRLI and the nomogram together represent significant strides in ccRCC research, highlighting the importance of molecular-based assessments in predicting patient outcomes.

Risk Stratification Using Right Ventricular Longitudinal Strain Ratio Derived from 13N-Ammonia PET in Patients with Ischemic Heart Disease.

Purpose To investigate whether right ventricular (RV) myocardial strain ratio (RVMSR) assessed using nitrogen 13 ammonia (13N-NH3) PET can predict cardiovascular events in patients with ischemic heart disease (IHD). Materials and Methods This retrospective study included 480 consecutive patients (mean age, 66 years ± 12 [SD]; 334 males and 146 females) with IHD who underwent 13N-NH3 PET. RVMSR was defined as the ratio of RV strain during stress to that at rest. The primary end point was major adverse cardiac events (MACEs), defined as cardiac death or heart failure hospitalization. The ability of RVMSR to predict MACE was assessed using receiver operating characteristic (ROC) curve and Kaplan-Meier analyses. Cox proportional hazards regression analysis was used to calculate hazard ratios (HRs) with 95% CIs. Results ROC curve analysis identified a sensitivity and specificity of 84% and 82%, respectively, for predicting MACE from RVMSR. Patients with reduced RVMSR (<110.2) displayed a significantly higher rate of MACE than those with a preserved RVMSR (34 of 240 vs four of 240; P < .001). Cox proportional hazards regression analysis of imaging parameters, including myocardial flow reserve, indicated that RVMSR was an independent predictor of MACE (HR, 0.94 [95% CI: 0.92, 0.97]; P < .001). Conclusion RVMSR was an independent predictor of MACE and has potential to aid in the risk stratification of patients with IHD. Keywords: Right Ventricular Myocardial Strain Ratio, Myocardial Flow Reserve, Ischemic Heart Disease, 13N-Ammonia Positron Emission Tomography Supplemental material is available for this article. © RSNA, 2024.

Independent validation of 30-day emergency department visit after initial chemotherapy A-PROACCT model.

e23243 Background: This study assessed model performance and, more importantly, the clinical usefulness of the PROACCT model created by Grant et al. (JAMA, 2019), refined as A-PROACCT by Stein et al. (JOP, 2023) with data from Orlando Health. The predictive performance of models anticipating acute care events within thirty days (ACE30) of initial chemotherapy administration (ICA) were evaluated. Methods: Stein et al. models classified high-risk patients as those whose model output risk score was at or above the 90th percentile. They studied 5153 ICAs between July and November of 2021, of which 559 (10.8%) had an ACE30. Their first model used drug category, age group, ED visits and hospitalizations in the prior year, and cancer type, and had sensitivity (Sen), specificity, (Spec), C-Statistic (C-Stat), and positive predictive value (PPV) of 0.28, 0.93, 0.79 and 0.18 respectively. Their second model added insurance category, number of anti-cancer agents, race, and BMI category and reported Sen, Spec, C-Stat, and PPV of 0.30, 0.94, 0.80 and 0.23 respectively. Of the high-risk patients, 33.6% and 34.1% had an ACE30 respectively. Of the low-risk patients, 8.3% and 8.1% had an ACE30 respectively. Our data were propagated through the two models. Results: Between February of 2012 and April 2021, Orlando Health had 15,483 distinct patient visits with ICAs. Overall, 2046 (13.2%) had an ACE30. Model one had Sen, Spec, C-Stat, and PPV of 0.19, 0.92, 0.65 and 0.27 respectively. Model two had Sen, Spec, C-Stat, and PPV of 0.19, 0.91, 0.65 and 0.25 respectively. Both models had an accuracy of 0.82. This demonstrated our data fit the Stein model. Of the 1460 ICAs identified as high-risk by model one, 387 (26.5%) had an ACE30, and of the 14,023 ICAs identified as low-risk, 1659 (11.8%) had an ACE30. Of the 1549 ICAs identified as high-risk by model two, 393 (25.4%) had an ACE30, and of the 13,934 ICAs identified as low-risk, 1653 (11.9%) had an ACE30. The difference of ACE30 rates between high risk and low risk groups were statistically significant (p &lt; 0.0001) in both models. This demonstrated both models were clinically useful in identifying chemotherapy patients at high risk of acute care events. Conclusions: These models allowed for clinically relevant stratification of patients into meaningful high-risk and low-risk groups (26.5% vs 11.8% with our data) for risk of acute care events within 30 days of initial chemotherapy administration. Nevertheless, only 26.5% of these high-risk patients actually had an acute care event within thirty days of initial chemotherapy administration. Further research is needed to improve the precision of identifying high-risk patients, by potentially adding more clinical variables, using alternative sampling methods, and employing nonlinear and more complex machine learning methodologies.

Suboptimally Controlled Diabetes in Pregnancy: A Review to Guide Antepartum and Delivery Management.

Diabetes mellitus is one of the most common complications in pregnancy with adverse maternal and neonatal risks proportional to the degree of suboptimal glycemic control, which is not well defined. Literature guiding providers in identifying and managing patients at highest risk of complications from diabetes is lacking. This article reviews the definition, epidemiology, and pathophysiology of suboptimal control of diabetes in pregnancy, including "diabetic fetopathy"; explores proposed methods of risk stratification for patients with diabetes; outlines existing antepartum management and delivery timing guidelines; and guides direction for future research. Original research articles, review articles, and professional society guidelines on diabetes management in pregnancy were reviewed. The reviewed available studies demonstrate worsening maternal and neonatal outcomes associated with suboptimal control; however, the definition of suboptimal based on parameters followed in pregnancy such as blood glucose, hemoglobin A1c, and fetal growth varied from study to study. Studies demonstrating specific associations of adverse outcomes with defined suboptimal control were reviewed and synthesized. Professional society recommendations were also reviewed to summarize current guidelines on antepartum management and delivery planning with respect to diabetes in pregnancy. The literature heterogeneously characterizes suboptimal glucose control and complications related to this during pregnancy in individuals with diabetes. Further research into antepartum management and delivery timing for patients with varying levels of glycemic control and at highest risk for diabetic complications is still needed.

Preoperative Plasma Insulin-Like Growth Factor-I and Its Binding Proteins-Based Risk Stratification of Patients Treated With Radical Nephroureterectomy for Upper Tract Urothelial Carcinoma

IntroductionWe evaluate the predictive and prognostic value of insulin-like growth factor-I (IGF-1), IGF binding protein-2 (IGFBP-2) and -3 (IGFBP-3) in patients treated with radical nephroureterectomy (RNU) for upper tract urothelial carcinoma (UTUC). MethodsThis is a retrospective analysis of a multi-institutional database comprising 753 patients who underwent RNU for UTUC and had a preoperative plasma available. Logistic and Cox regression analyses were performed. The discriminative ability and clinical utility of the models was calculated using the lasso regression test, area under receiver operating characteristics curves, C-index, and decision curve analysis (DCA). ResultsLower preoperative plasma levels of IGFBP-2 and -3 independently correlated with increased risks of lymph node metastasis, pT3/4 disease, nonorgan confined disease, and worse recurrence-free survival (RFS), cancer-specific survival (CSS), and overall survival (OS) (all P ≤ .004). The addition of both IGFBP-2 and -3 to a postoperative multivariable model, that included standard clinicopathologic characteristics, improved the model's concordance index by 10%, 9%, and 8% for RFS, CSS, and OS, respectively. On DCA, addition of both IGFBP-2 and -3 to base models improved their performance for RFS, CSS, and OS by a statistically and clinically significant margin. Plasma IGF-1 was not associated with any of outcomes. ConclusionsWe confirmed that a lower plasma levels of IGFBP-2 and -3 both are independent and clinically significant predictors of adverse pathological features and survival outcomes in UTUC patients treated with RNU. These findings might help guide the clinical decision-making regarding perioperative systemic therapy and follow-up scheduling.

The prognostic impact of transforming growth factor beta 2 (TGFB2) mRNA levels, in conjunction with interferon-gamma receptor activation of interferon regulatory factor 5 (IRF5) and expression of CD276/B7-H3 in low-grade gliomas.

e14564 Background: Low-grade glioma (LGG) tumors are characterized by a low infiltration of immune cells, requiring therapeutic interventions to boost the immune response. Methods: We analyzed mRNA expression datasets from the UCSC Xena web platform to compare normal brain tissue with LGG tumor samples to screen for upregulated genes. We also used cBioportal to determine the relationship between mRNA expression levels of 513 LGG patients and their overall survival (OS) outcomes. Results: Tumor-associated macrophage (TAM) markers, MSR1/CD204, CD86, and CD68, exhibited a 6-fold (P&lt;0.0001), 8.9-fold (P&lt;0.0001), and 15.6-fold increases in mRNA expression levels, respectively in LGG tumors. Both TGFB1 (4.1-fold increase, P&lt;0.0001) and TGFB2 (2.2-fold increase, P&lt;0.0001) were upregulated in brain tumors, suggesting that TGFB ligands are pivotal in establishing an immunosuppressive TME. In addition, mRNA upregulation of interferon-gamma receptors, IFNGR1 and IFNGR2, and signaling molecules STAT1, IRF1, and IRF5, pointed to an essential role for IFN-γ mediated remodeling of the TME. A tumor-associated antigen, CD276/B7-H3, mRNA expression showed a significant (P&lt;0.0001) 4.03-fold increase in tumor tissue. Multivariate Cox proportional hazards models investigating the interaction of TGFB2 and activation of IFNGR2, STAT1, IRF1, or IRF5 showed that the prognostic impact of high mRNA levels (25 percentile cut-off) of TGFB2high was independent of IFNGR2, STAT1, IRF1, or IRF5 mRNA levels (TGFB2high HR (95% CI) = 4.07 (2.35-7.06), 6 (3.62-10.11), 4.38 (2.67-7.17), and 4.48 (2.82-7.12) for models with IFNGR2, STAT1, IRF1, or IRF5 respectively). Patients with high levels of TGFB2 were overrepresented by LGG patients with IDH wild-type mutation status. The prognostic impact of TGFB2high and IDH wild-type observed by the increases in hazard ratios for TGFB2 (HR (95% CI range) = 2.02 (1.05-3.89)) and IDH wild-type (HR (95% CI range) = 4.44 (1.9-10.4)) were independent predictors of survival requiring risk stratification of patients identify LGG patients with IDH wild-type and TGFB2high in the design of clinical trials. The ratio of TGFB1:TGFB2 expression levels was also a prognostic indicator for OS, whereby patients who exhibited a 2.5-fold greater level of TGFB1 compared to TGFB2 mRNA levels experienced improved survival outcomes. A 1.8-fold higher TGFB2 levels, the mRNA expression corresponded to worse OS. Conclusions: The specific abrogation of TGFB2 and not TGFB1 by OT101, a TGFB2-specific synthetic phosphorothioate antisense oligodeoxynucleotide, is a prerequisite to employing a strategy to target TGFB pathways. IRF5 and CD276/B7-H3 as prognostic markers are potential targets for combination therapies with TGFB2 inhibitors.

Efficacy of CHA2DS2-VASc scores in predicting chronic kidney disease risk in patients treated in cardiac intensive care units

The CHA2DS2 -VASc score is a vital clinical tool for evaluating thromboembolic risk in patients with atrial fibrillation (AF). This study investigated the efficacy of the CHA2DS2 -VASc score in a cohort of 737 heterogeneous patients (mean age: 63 years) receiving care in cardiac intensive care units (CICUs), with a creatinine-based estimated glomerular filtration rate (eGFR) of ≥60 mL/min/1.73 m2 upon admission and discharge. Incident chronic kidney disease (CKD) was defined as the emergence of a new-onset eGFR<60 mL/min/1.73 m2, accompanied by a decline of >5 mL/min/1.73 m2 compared to that at discharge. The primary endpoint was the incidence of CKD, and the secondary endpoints included all-cause mortality, cardiovascular events, and progression to end-stage kidney disease. In this cohort, 210 (28%) patients developed CKD. Multivariate analyses revealed that CHA2DS2 -VASc score was a significant independent predictor of incident CKD, regardless of the presence of AF. Integration of CHA2DS2 -VASc scores with eGFR enhanced the predictive accuracy of incident CKD, as evidenced by the improved C-index, net reclassification improvement, and integrated discrimination improvement values (all p<0.05). Over the 12-month follow-up period, a composite endpoint was observed in 61 patients (8.3%), with elevated CHA2DS2 -VASc scores being independently associated with this endpoint. In conclusion, CHA2DS2-VASc scores have emerged as robust predictors of both CKD incidence and adverse outcomes. Their inclusion substantially refined the 12-month risk stratification of patients with preserved renal function hospitalized in the CICUs.

Dual analysis of IL-8 in non-small cell lung cancer: A clinical perspective on protein and RNA evaluation via imaging mass cytometry.

e15068 Background: The cytokine Interleukin-8 (IL-8) plays a pivotal role in the tumorigenesis and progression of non-small cell lung cancer (NSCLC), influencing both the tumor microenvironment and immune escape mechanisms. Given its significance, accurately assessing IL-8 levels at the protein and RNA levels could provide valuable insights into its prognostic value and inform therapeutic targeting strategies. This study utilizes imaging mass cytometry (IMC) to conduct a comprehensive evaluation of IL-8 expression in NSCLC, integrating both protein presence and gene expression analysis within the spatial context of the tumor microenvironment. Methods: Our approach leverages the capabilities of IMC to simultaneously detect protein and RNA levels of IL-8 in NSCLC tissue samples. For protein analysis, antibodies tagged with heavy metal isotopes specifically target IL-8, allowing for precise quantification and spatial localization within the tumor. Concurrently, RNA detection is achieved through hybridization probes designed for IL-8 mRNA, enabling the direct observation of gene expression patterns. This method provides a unique opportunity to correlate IL-8's protein abundance with its mRNA expression, offering insights into the regulatory mechanisms at play and their impact on the tumor's behavior and patient outcomes. Results: Preliminary data demonstrate a marked heterogeneity in IL-8 distribution at both the protein and RNA levels, with distinct patterns emerging that relate to tumor aggressiveness and immune infiltration. High IL-8 expression, identified through our integrated IMC approach, correlates with poor prognosis and reduced response to standard therapies, underscoring its potential as a biomarker for NSCLC management. Furthermore, the spatial analysis afforded by IMC highlights the microenvironmental niches where IL-8-mediated interactions could drive tumor progression and immune evasion. Conclusions: By employing IMC for the dual analysis of IL-8 protein and RNA in NSCLC, our study provides a comprehensive view of this cytokine's role in lung cancer pathology. The ability to visualize and quantify IL-8 expression within the spatial context of the tumor microenvironment opens new avenues for understanding its biological impact and for developing targeted therapies. Our findings advocate for the inclusion of IL-8 in the biomarker panel for NSCLC, with potential implications for patient stratification and personalized treatment planning. Future research will focus on expanding these observations to a larger cohort and exploring the therapeutic modulation of IL-8 as a strategy for improving NSCLC outcomes.