Improve Risk Stratification In Patients Research Articles

Integration of clinical, pathological, radiological, and transcriptomic data improves prediction for first-line immunotherapy outcome in metastatic non-small cell lung cancer

Immunotherapy is improving the survival of patients with metastatic non-small cell lung cancer (NSCLC), yet reliable biomarkers are needed to identify responders prospectively and optimize patient care. In this study, we explore the benefits of multimodal approaches to predict immunotherapy outcome using multiple machine learning algorithms and integration strategies. We analyze baseline multimodal data from a cohort of 317 metastatic NSCLC patients treated with first-line immunotherapy, including positron emission tomography images, digitized pathological slides, bulk transcriptomic profiles, and clinical information. Testing multiple integration strategies, most of them yield multimodal models surpassing both the best unimodal models and established univariate biomarkers, such as PD-L1 expression. Additionally, several multimodal combinations demonstrate improved patient risk stratification compared to models built with routine clinical features only. Our study thus provides evidence of the superiority of multimodal over unimodal approaches, advocating for the collection of large multimodal NSCLC datasets to develop and validate robust and powerful immunotherapy biomarkers.

Abstract 4137272: The Impact of Genotype in Dilated and Arrhythmogenic Cardiomyopathy: Insights from the SHaRE Registry

Background: The impact of genotype on the risk of arrhythmia and heart failure in dilated and arrhythmogenic cardiomyopathy (DCM/ACM) is not well defined. Better understanding of genetic risk factors is needed to improve patient risk stratification, management, and outcomes. Methods: In this observational study, we leveraged the international Sarcomeric Human Cardiomyopathy Registry (SHaRe) to analyze the effect of arrhythmic gene variants on clinical events in patients with DCM/ACM (excluding arrhythmogenic right ventricular cardiomyopathy). Patients were stratified by the presence of pathogenic variants in 8 genes most commonly associated with ACM (AR(+): LMNA , DSP , FLNC , RBM20 , PLN , SCN5A , DES , and TMEM43 ). We assessed for risk of clinical events including ventricular arrhythmia (VA), atrial fibrillation (AF), heart failure (HF), death, and the overall composite of any of these events. Analyses were adjusted for age at diagnosis, sex, and baseline left ventricular ejection fraction (EF). Patients were also stratified by EF ≥50% in subgroup analyses. Results: A total of 1373 genotyped adults with DCM/ACM from 6 centers were included with median follow-up of 5.5 (IQR 2.4-10.4) years. Mean diagnosis age was 48 (± 15) years, and 39.5% were female. The cohort included 185 (13.5%) AR(+) patients. AR(+) compared to genotype-negative patients (N=486, 35.4%) had increased risk of VA (HR 5.02, 95% CI 3.46-7.29) and AF (HR 3.02, 95% CI 2.20-4.10). Risk of HF (HR 2.32, 95% CI 1.55-3.49) and the overall composite (HR 2.12, 95% CI 1.68-2.67) were also increased ( Figure ). Risk of each event remained significant after adjusting for age, sex, and baseline EF ( Table ). AR(+) patients with EF ≥50% also had increased risk of VA (HR 4.56, 95% CI 2.16-9.63), AF (HR 2.94, 95% CI 1.72-5.03), and the overall composite (HR 2.15, 95% CI 1.41-3.28). Conclusions: Genetic testing enables powerful risk stratification for both arrhythmias and heart failure in DCM/ACM independent of age, sex, or EF.

Role of common variants and new genes associated with arrhythmogenic cardiomyopathy

Abstract Introduction Arrhythmogenic Cardiomyopathy (ACM) is an inherited cardiac disease, predominantly affecting males, characterized by fibro-fatty myocardial replacement mainly in the right ventricle. It leads to 5% sudden cardiac death in affected young adults (<25 yo) without an implantable cardioverter defibrillator. Early diagnosis is crucial for appropriate clinical management to reduce mortality and prevent disease progression. Rare variants in desmosomal genes (PKP2,DSC2,DSG2,DSP and JUP) are found in 50% of cases. However, challenges persist since we observe a low penetrance in relatives (28-58%) and a high variability in expressivity and severity annihilating any preventive strategy. Objective We aim to decipher the genetic architecture of ACM by investigating the role of common variants in disease susceptibility/severity and to idendify new genes associated to ACM. Method A Genome Wide Association Study (GWAS) was performed on 1079 ACM index patients and 7305 controls, followed by conditional analysis. New candidate genes were knocked out (KO) in Zebrafish and/or hiPSC using CRISPR/Cas9. Impedance and external field potential measurements were performed on hiPSC derived cardiomyocytes (CM-hiPSC). Results The GWAS identified 5 significant loci (Figure 1, P<5.10-8), including 2 desmosomal genes (PKP2 a n d DSG2) and 3 new candidates (KLF12, CTNNA3, and ARL17B/WNT3). Conditional analysis revealed 3 additional loci with suggestive p-values (P<1.10-6), highlighting 3 new susceptible loci. Among them, we uncovered key non-coding regulatory regions near desmosomal genes (DSC2 and DSG2). Initial findings in PKP2-KO zebrafish present hearts with structural defects, while CM-hiPS KLF12-KO(-/-) exhibited altered interbeat interval and relaxation time compared to isogenic controls. Conclusion Our findings reveal ACM's complex genetic architecture, implicating common variants in disease susceptibility. A polygenic risk score will be implemented to improve patient risk stratification. Three novel loci/genes for ACM as well as specific regulatory regions of desmosomal genes have been highlighted. Their functional studies in isogenic hiPSC and Zebrafish models will offer new insights in ACM underlying molecular mechanism and potential therapeutic targets.

Five steps in performing machine learning for binary outcomes

BackgroundThe use of machine learning (ML) in cardiovascular and thoracic surgery is evolving rapidly. Maximizing the capabilities of ML can help improve patient risk stratification and clinical decision making, improve accuracy of predictions, and improve resource utilization in cardiac surgery. The many nuances and intricacies of ML modeling need to be understood to appropriately implement these technologies in the clinical research setting. This primer provides an educational framework of ML for generating predicted probabilities in clinical research and illustrates it with a real-world clinical example. MethodsWe focus on modeling for binary classification and imbalanced classes, a common scenario in cardiothoracic surgery research. We present a 5-step strategy for successfully harnessing the power of ML and performing such analyses, and demonstrate our strategy using a real-world example based on data from the National Surgical Quality Improvement Program pediatric database. ConclusionsCollaboration among surgeons, care providers, statisticians, data scientists, and information technology professionals can help to maximize the impact of ML as a powerful tool in cardiac surgery.

An engineered human cardiac tissue model reveals contributions of systemic lupus erythematosus autoantibodies to myocardial injury

Systemic lupus erythematosus (SLE) is a heterogenous autoimmune disease that affects multiple organs, including the heart. The mechanisms of myocardial injury in SLE remain poorly understood. In this study, we engineered human cardiac tissues and cultured them with IgG from patients with SLE, with and without myocardial involvement. IgG from patients with elevated myocardial inflammation exhibited increased binding to apoptotic cells within cardiac tissues subjected to stress, whereas IgG from patients with systolic dysfunction exhibited enhanced binding to the surface of live cardiomyocytes. Functional assays and RNA sequencing revealed that, in the absence of immune cells, IgG from patients with systolic dysfunction altered cellular composition, respiration and calcium handling. Phage immunoprecipitation sequencing (PhIP-seq) confirmed distinctive IgG profiles between patient subgroups. Coupling IgG profiling with cell surfaceome analysis identified four potential pathogenic autoantibodies that may directly affect the myocardium. Overall, these insights may improve patient risk stratification and inform the development of new therapeutic strategies.

Cardiovascular magnetic resonance parametric mapping in the risk stratification of patients affected by chronic myocarditis.

Chronic myocardial inflammation is the substrate for arrhythmias and dilated cardiomyopathy onset, causing morbidity and mortality. Cardiovascular magnetic resonance (CMR) is the noninvasive gold standard for myocardial inflammation detection, due to the high sensitivity of the parametric mapping techniques. However, the potential prognostic capabilities of CMR mapping have not been studied in the setting of chronic myocarditis. This is a retrospective study on consecutive patients undergoing CMR with suspicion of chronic myocarditis from September 2017 to November 2021. CMR was acquired according to 2018 Lake Louise Criteria recommendations. The outcome (chronic heart failure, recurrent chronic myocarditic chest pain, ICD/PM implantation, arrhythmias [Lown class ≥ 2]) was collected at follow-up. The extent and degree of native T1, T2, and extracellular volume fraction alterations were used to create multivariate binary logistic regression models for outcome prediction, with or without left ventricle ejection fraction; their AUCs were compared with DeLong test. Differences between other parameters were assessed using Chi-square test, Fisher's exact test, or Mann-Whitney U-test. The population included 88 patients (age 43 [32-52] yo), mostly male (53/88, 60%). After a median follow-up of 21 (17-34) months, 31/88 (35%) patients experienced the outcome. The model based on the extension of mapping alterations and LV dysfunction reached a good predictability (AUC 0.71). The model based on the intensity of mapping alterations and LV dysfunction had a very good performance (AUC 0.80). The quantitative analysis of CMR mapping parameters indicative of myocardial damage severity may improve risk stratification in patients with chronic myocarditis. The intensity of myocardial damage, assessed as the degree of native T1, T2, and ECV alteration, together with left ventricle dysfunction, improved patient risk stratification. Further prospective studies will be necessary for validation before clinical application. Risk stratification of patients affected by chronic myocarditis is an unmet clinical need. Cardiovascular MRI (CMR) can role in risk stratification thanks to its multiparametric capabilities of tissue characterization. A model based on CMR parametric mapping and left ventricle ejection fraction can predict arrhythmia, heart failure, and recurrent symptoms.

Changes in Global Longitudinal Strain after TAVI: Additional Prognostic Value over Cardiac Damage in Patients with Severe Aortic Stenosis.

Background: Previous studies demonstrated the prognostic value of baseline cardiac damage staging as well as left ventricular global longitudinal strain (LVGLS) in patients undergoing transcatheter aortic valve implantation (TAVI). The aim of the present study was to evaluate the changes in cardiac damage stage and LVGLS after TAVI and to investigate their prognostic values when integrated into the follow-up assessment. Methods: Patients with severe aortic stenosis undergoing TAVI were hierarchically classified into cardiac damage stages based on echocardiographic criteria before TAVI and at a 6-month follow-up. At the same time, LVGLS was measured. The staging system included stage 0 = no signs of cardiac damage; stage 1 = LV damage; stage 2 = mitral or left atrial damage; stage 3 = pulmonary vasculature or tricuspid damage; and stage 4 = right ventricular damage. The primary endpoint was all-cause mortality. Results: A total of 620 patients were included. At follow-up, LVGLS significantly improved, and the improvement was similar among each baseline cardiac damage stage. Follow-up LVGLS values were divided into quintiles, and each quintile was integrated into the cardiac damage staging, leading to a reclassification of 308 (50%) patients. At the time of a median follow-up at 48 (IQR 31-71) months starting from the 6-month follow-up after TAVI, 262 (38%) patients had died. A multivariable Cox regression model showed that LVGLS-integrated cardiac damage staging at follow-up had an incremental prognostic value over the baseline assessment (HR per 1-stage increase 1.384; 95% CI 1.152-1.663; p < 0.001). Conclusions: The integration of LVGLS with conventional echocardiographic parameters of cardiac damage at a 6-month follow-up after TAVI can improve patient risk-stratification.

Noninvasive Atherosclerotic Phenotyping: The Next Frontier into Understanding the Pathobiology of Coronary Artery Disease.

Despite recent advances, coronary artery disease remains one of the leading causes of mortality worldwide. Noninvasive imaging allows atherosclerotic phenotyping by measurement of plaque burden, morphology, activity and inflammation, which has the potential to refine patient risk stratification and guide personalized therapy. This review describes the current and emerging roles of advanced noninvasive cardiovascular imaging methods for the assessment of coronary artery disease. Cardiac computed tomography enables comprehensive, noninvasive imaging of the coronary vasculature, and is used to assess luminal stenoses, coronary calcifications, and distinct adverse plaque characteristics, helping to identify patients prone to future events. Novel software tools, implementing artificial intelligence solutions, can automatically quantify and characterize atherosclerotic plaque from standard computed tomography datasets. These quantitative imaging biomarkers have been shown to improve patient risk stratification beyond clinical risk scores and current clinical interpretation of cardiac computed tomography. In addition, noninvasive molecular imaging in higher risk patients can be used to assess plaque activity and plaque thrombosis. Noninvasive imaging allows unique insight into the burden, morphology and activity of atherosclerotic coronary plaques. Such phenotyping of atherosclerosis can potentially improve individual patient risk prediction, and in the near future has the potential for clinical implementation.

Comprehensive single-cell transcriptomic profiling reveals molecular subtypes and prognostic biomarkers with implications for targeted therapy in esophageal squamous cell carcinoma

BackgroundEsophageal squamous cell carcinoma (ESCC) is a genetically heterogeneous disease with poor clinical outcomes. Identification of biomarkers linked to DNA replication stress may enable improved prognostic risk stratification and guide therapeutic decision making. We performed integrated single-cell RNA sequencing and computational analyses to define the molecular determinants and subtypes underlying ESCC heterogeneity. MethodsSingle-cell RNA sequencing was performed on ESCC samples and analyzed using Seurat. Differential gene expression analysis was used to identify esophageal cell phenotypes. DNA replication stress-related genes were intersected with single-cell differential expression data to identify potential prognostic genes, which were used to generate a DNA replication stress (DRS) score. This score and associated genes were evaluated in survival analysis. Putative prognostic biomarkers were evaluated by Cox regression and consensus clustering. Mendelian randomization analyses assessed the causal role of PRKCB. ResultsHigh DRS score associated with poor survival. Four genes (CDKN2A, NUP155, PPP2R2A, PRKCB) displayed prognostic utility. Three molecular subtypes were identified with discrete survival and immune properties. A 12-gene signature displayed robust prognostic performance. PRKCB was overexpressed in ESCC, while PRKCB knockdown reduced ESCC cell migration. ConclusionsThis integrated single-cell sequencing analysis provides new insights into the molecular heterogeneity and prognostic determinants underlying ESCC. The findings identify potential prognostic biomarkers and a gene expression signature that may enable improved patient risk stratification in ESCC. Experimental validation of the role of PRKCB substantiates the potential clinical utility of our results.

Abstract 2137: Racial disparities in triple-negative breast cancer (TNBC) mortality rates in Hampton Roads Virginia and the effects of SIAH expression on prognosis

Abstract Background: While high-resolution imaging and advancements in therapies have significantly improved breast cancer survival rates, 43,170 breast cancer patients will die in the United States in 2023 alone. Triple-negative breast cancer (TNBC) is the most aggressive breast cancer subtype and disproportionately affects BRCA1 mutation carriers and young black women. Black/African American (AA) patients have the highest mortality and the shortest survival of any racial/ethnic group in the US. Persistent cancer racial disparities remain due to a variety of risk factors. The SEER/CDC data show an 8% lower breast cancer incidence rate and a 41% higher mortality rate in Black/AA patients compared to their white counterparts. Methods: Chart review was conducted using Sentara MD Office/EPIC and VOA iKnowMedicine portals to update tumor relapse, metastasis, and survival in 577 TNBC patients. Supported by strong evidence in developmental, evolutionary, and cancer biology, we hypothesize that persistent EGFR/K-RAS/SIAH pathway activation is a major driving force of TNBC malignancy, racial disparity, early relapse, and high mortality. We propose to integrate SIAH expression to augment the existing clinicopathological parameters to improve patient risk stratification, therapy quantification, and relapse/survival prediction at the 1st-line neoadjuvant settings. Results: We report that cancer disparity and high mortality rates are even more pronounced in our racially-diverse communities in Hampton Roads Virginia. We discovered that SIAH is a tumor-specific, therapy-responsive, and prognostic biomarker in TNBC. High SIAH expression in residual tumors reflects tumor-driving EGFR/K-RAS/SIAH pathway activation (ON) that will predict cancer disparity, treatment resistance, early relapse, and poor survival. Low SIAH expression in residual tumors reflect EGFR/K-RAS/SIAH pathway inactivation (OFF) that will predict tumor remission and prolonged survival. Conclusion: We detect a major racial disparity of TNBC patients at Sentara-EVMS-VOA. Our local Black/AA TNBC patients have a 1.6-fold higher mortality rate than their White counterparts. Encouraged by our preliminary data, we aim to develop a SIAH-centered biomarker panel by measuring the EGFR/K-RAS/SIAH pathway activation (ON)/inactivation (OFF), and use SIAH as a new prognostic biomarker to risk stratify patients, detect cancer racial disparities, forecast tumor relapse, and predict patient survival at 1st-line neoadjuvant settings. By targeting SIAH, individualized treatments can be created to allow for increased survival rates in at risk individuals with more aggressive TNBCs. Citation Format: Claire B. Piatak, Taylor N. Drake, Ashleigh Hannah, Jonathan Baker, Caroline L. Dasom, Emily L. Breeding, Janet S. Winston, Billur Samli, Rick Jansen, Michael Danso, Richard A. Hoefer, Amy H. Tang. Racial disparities in triple-negative breast cancer (TNBC) mortality rates in Hampton Roads Virginia and the effects of SIAH expression on prognosis [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 2137.

Polygenic Risk Scores for Glaucoma Onset in the Ocular Hypertension Treatment Study

Primary open-angle glaucoma (POAG) is a highly heritable disease, with 127 identified risk loci to date. Polygenic risk score (PRS) may provide a clinically useful measure of aggregate genetic burden and improve patient risk stratification. To assess whether a PRS improves prediction of POAG onset in patients with ocular hypertension. This was a post hoc analysis of the Ocular Hypertension Treatment Study. Data were collected from 22 US sites with a mean (SD) follow-up of 14.0 (6.9) years. A total of 1636 participants were followed up from February 1994 to December 2008; 1077 participants were enrolled in an ancillary genetics study, of which 1009 met criteria for this analysis. PRS was calculated using summary statistics from the largest cross-ancestry POAG meta-analysis, with weights trained using 8 813 496 variants from 449 186 cross-ancestry participants in the UK Biobank. Data were analyzed from July 2022 to December 2023. From February 1994 to June 2002, participants were randomized to either topical intraocular pressure-lowering medication or close observation. After June 2002, both groups received medication. Outcome measures were hazard ratios for POAG onset. Concordance index and time-dependent areas under the receiver operating characteristic curve were used to compare the predictive performance of multivariable Cox proportional hazards models. Of 1009 included participants, 562 (55.7%) were female, and the mean (SD) age was 55.9 (9.3) years. The mean (SD) PRS was significantly higher for 350 POAG converters (0.24 [0.95]) compared with 659 nonconverters (-0.12 [1.00]) (P < .001). POAG risk increased 1.36% (95% CI, 1.08-1.64) with each higher PRS decile, with conversion ranging from 9.52% (95% CI, 7.09-11.95) in the lowest PRS decile to 21.81% (95% CI, 19.37-24.25) in the highest decile. Comparison of low-risk and high-risk PRS tertiles showed a 2.0-fold increase in 20-year POAG risk for participants of European and African ancestries. In the subgroup randomized to delayed treatment, each increase in PRS decile was associated with a 0.52-year (95% CI, 0.01-1.03) decrease in age at diagnosis (P = .047). No significant linear association between PRS and age at POAG diagnosis was present in the early treatment group. Prediction models significantly improved with the addition of PRS as a covariate (C index = 0.77) compared with the Ocular Hypertension Treatment Study baseline model (C index = 0.75) (P < .001). Each 1-SD higher PRS conferred a mean hazard ratio of 1.25 (95% CI, 1.13-1.44) for POAG onset. Higher PRS was associated with increased risk for POAG in patients with ocular hypertension. The inclusion of a PRS improved the prediction of POAG onset. ClinicalTrials.gov Identifier: NCT00000125.

Application of GPT-4 foundation model for risk prediction and stratification in colorectal cancer.

219 Background: Deep Learning Language Models (LLMs) show significant potential in analyzing complex data, particularly in oncology, where they can predict and stratify patient risks, detect non-linear interactions, and weigh multiple variables. This study applies LLMs to stage II/III localized colon cancer patients to assess their effectiveness in identifying risk subgroups. Methods: In this retrospective study, we examined a cohort of stage II and III localized colorectal cancer patients diagnosed from September 2016 to December 2022 at La Paz University Hospital. Utilizing the GPT-4 API beta version, patients were stratified into high and low-risk groups for relapse based on various risk parameters outlined in the table. Notably, adjuvant treatment data was excluded from the input. The model was prompted with the question, "Given the following risk parameters for localized colorectal cancer {param}, guess whether the patient is in low risk or high risk for relapse," and the model provided dichotomic responses categorized as either 1 (high risk) or 0 (low risk). Results: Patient characteristics are depicted in the table. During a median follow-up of 23.68 months, 63 recurrence events and 73 deaths were observed. The GPT-4 model stratified patients into low (37.9%) and high-risk (62.1%) groups, with the survival curves showing a significant difference in outcomes between the groups. Low-risk patients had a relapse-free rate of 93% at 36 months vs 63% in the high-risk group. In univariate analysis, the high-risk variable had a hazard ratio (HR) of 4.26 (95% CI: 2.03-8.95; p &lt; 0.005). When stratified by adjuvant treatment, no significant difference in relapse probability was observed for those receiving therapy. However, without adjuvant therapy, high-risk patients had an HR of 12.41 (95% CI: 3.72-41.97; p &lt; 0.005) for relapse. Conclusions: GPT-4 demonstrated its potential in predicting oncology outcomes by accurately understanding colorectal cancer risk factors. To fully harness the benefits of LLMs in the future of oncology, they must be fine-tuned on extensive oncological datasets to improve patient risk stratification and treatment selection, ultimately leading to better patient outcomes.[Table: see text]

Early real-world experience monitoring circulating tumor DNA in resected early-stage non–small cell lung cancer

ObjectiveThe study objective was to evaluate the impact of monitoring circulating tumor DNA on the detection and management of recurrence in patients with resected early-stage non–small cell lung cancer. MethodsBetween October 2021 and March 2023, postoperative circulating tumor DNA was monitored in patients with non–small cell lung cancer (N = 108). Longitudinal blood samples (n = 378 samples) were collected for prospective circulating tumor DNA analysis at 3-month intervals after curative-intent resection. A tumor-informed assay was used for the detection and quantification of circulating tumor DNA. The primary outcome measure was a circulating tumor DNA–positive result. The secondary outcome measure was changes in practice after a circulating tumor DNA–positive result. ResultsThe mean age of the patients in this cohort was 68.1 years. Of the 108 patients, 12 (11.1%) were circulating tumor DNA positive at least at 1 timepoint postsurgery, of whom 8 (66.7%) had a clinically evident recurrence and the remaining 4 had limited clinical follow-up. Of the 10 patients with recurrent disease, 8 demonstrated circulating tumor DNA positivity and the remaining 2 patients had brain-only metastases. Postoperative clinical care was altered in 100% (12/12) of circulating tumor DNA–positive patients, with 58.3% (7/12) receiving an early computed tomography scan and 100% (12/12) receiving an early positron emission tomography computed tomography scan as part of their surveillance strategy. Among the patients who received an early positron emission tomography scan, 66.6% (8/12) were positive for malignant features. ConclusionsRoutine monitoring of tumor-informed circulating tumor DNA after curative intent therapy improved patient risk stratification and prognostication.

The 31-Gene Expression Profile Test Outperforms AJCC in Stratifying Risk of Recurrence in Patients with Stage I Cutaneous Melanoma.

Patients with stage I cutaneous melanoma (CM) are considered at low risk for metastasis or melanoma specific death; however, because the majority of patients are diagnosed with stage I disease, they represent the largest number of melanoma deaths annually. The 31-gene expression profile (31-GEP) test has been prospectively validated to provide prognostic information independent of staging, classifying patients as low (Class 1A), intermediate (Class 1B/2A), or high (Class 2B) risk of poor outcomes. Patients enrolled in previous studies of the 31-GEP were combined and evaluated for recurrence-free (RFS) and melanoma-specific survival (MSS) (n = 1261, "combined"). A second large, unselected real-world cohort (n = 5651) comprising clinically tested patients diagnosed 2013-2018 who were linked to outcomes data from the NCI Surveillance, Epidemiology, and End Results (SEER) Program registries was evaluated for MSS. Combined cohort Class 1A patients had significantly higher RFS than Class 1B/2A or Class 2B patients (97.3%, 88.6%, 77.3%, p < 0.001)-better risk stratification than AJCC8 stage IA (97.5%) versus IB (89.3%). The SEER cohort showed better MSS stratification by the 31-GEP (Class 1A = 98.0%, Class 1B/2A = 97.5%, Class 2B = 92.3%; p < 0.001) than by AJCC8 staging (stage IA = 97.6%, stage IB = 97.9%; p < 0.001). The 31-GEP test significantly improved patient risk stratification, independent of AJCC8 staging in patients with stage I CM. The 31-GEP provided greater separation between high- (Class 2B) and low-risk (Class 1A) groups than seen between AJCC stage IA and IB. These data support integrating the 31-GEP into clinical decision making for more risk-aligned management plans.

Evolving perspectives on aortic stenosis: the increasing importance of evaluating the right ventricle before aortic valve intervention.

Aortic stenosis (AS) was historically considered a disease of the left side of the heart, with the main pathophysiological impact being predominantly on the left ventricle (LV). However, progressive pressure overload in AS can initiate a cascade of extra-valvular myocardial remodeling that could also precipitate maladaptive alterations in the structure and function of the right ventricle (RV). The haemodynamic and clinical importance of these changes in patients with AS have been largely underappreciated in the past. Contemporary data indicates that RV dilatation or impairment identifies the AS patients who are at increased risk of adverse clinical outcomes after aortic valve replacement (AVR). It is now increasingly recognised that effective quantitative assessment of the RV plays a key role in delineating the late clinical stage of AS, which could improve patient risk stratification. Despite the increasing emphasis on the pathological significance of RV changes in AS, it remains to be established if earlier detection of these changes can improve the timing for intervention. This review will summarise the features of normal RV physiology and the mechanisms responsible for RV impairment in AS. In addition, we will discuss the multimodality approach to the comprehensive assessment of RV size, function and mechanics in AS patients. Finally, we will review the emerging evidence reinforcing the negative impact of RV dysfunction on clinical outcomes in AS patients treated with AVR.

Multimarker Approach to Improve Risk Stratification of Patients Undergoing Transcatheter Aortic Valve Implantation

BackgroundA blood multimarker approach may be useful to enhance risk stratification in patients undergoing TAVI. ObjectivesThe objective of this study was to determine the prognostic value of multiple blood biomarkers in transcatheter aortic valve implantation (TAVI) patients. MethodsIn this prospective study, several blood biomarkers of cardiovascular function, inflammation, and renal function were measured in 362 patients who underwent TAVI. The cohort was divided into 3 groups according to the number of elevated blood biomarkers (ie, ≥ median value for the whole cohort) for each patient before the procedure. Survival analyses were conducted to evaluate the association between blood biomarkers and risk of adverse event following TAVI. ResultsDuring a median follow-up of 2.5 (IQR: 1.9-3.2) years, 34 (9.4%) patients were rehospitalized for heart failure, 99 (27%) patients died, and 113 (31.2%) met the composite endpoint of all-cause mortality or heart failure rehospitalization. Compared to patients with 0 to 3 elevated biomarkers (referent group), those with 4 to 7 and 8 to 9 elevated biomarkers had a higher risk of all-cause mortality (HR: 1.54 [95% CI: 0.84-2.80], P = 0.16, and HR: 2.81 [95% CI: 1.53-5.15], P < 0.001, respectively) and of the composite endpoint (HR: 1.65 [95% CI: 0.95-2.84], P = 0.07, and HR: 2.67 [95% CI: 1.52-4.70] P < 0.001, respectively). Moreover, adding the number of elevated blood biomarkers into the clinical multivariable model provided significant incremental predictive value for all-cause mortality (Net Reclassification Index = 0.71, P < 0.001). ConclusionsAn increasing number of elevated blood biomarkers is associated with higher risks of adverse clinical outcomes following TAVI. The blood multimarker approach may be helpful to enhance risk stratification in TAVI patients.

Machine learning: the what, why, and how

There is a growing interest in the ability of machine learning to predict outcomes in shoulder arthroplasty. Machine learning is the use of computer systems that can adapt and learn by using algorithms to draw statistical significance and predictions from data. Multiple studies have shown that machine learning algorithms can help predict postoperative outcomes, including patient-reported outcome measures (PROMs), range of motion, complications, and unplanned readmissions. Ultimately, the promise of machine learning in shoulder arthroplasty is to improve patient risk stratification, procedure selection, and long-term outcomes. Modern algorithms rely on numerous input variables, although recent efforts seek to reduce the required number of variables to improve efficiency and decrease the clinical burden. A review of recent machine learning studies allows for a deeper understanding of the potential use of predictive analytics in improving outcomes of shoulder arthroplasty.

Should Cystatin C eGFR Become Routine Clinical Practice?

Kidney function assessment is crucial for diagnosing and managing kidney diseases. Glomerular filtration rate (GFR) is widely used as an indicator of kidney function, but its direct measurement is challenging. Serum creatinine, a commonly used marker for estimating GFR (eGFR), has limitations in accuracy and sensitivity. Cystatin C, a protein freely filtered by the glomerulus, has emerged as a promising alternative marker for kidney function. It is unaffected by muscle mass and shows stronger associations with cardiovascular disease and mortality than creatinine. Various equations have been developed to estimate GFR using creatinine or cystatin C alone or in combination. The CKD-EPIcreat-cys equation combining both markers demonstrates improved accuracy in GFR estimation, especially for individuals with eGFR values of 45-59 mL/min/1.73 m2. Cystatin C-based estimates of GFR outperform creatinine-based estimates in predicting clinical outcomes and identifying patients at higher risk, particularly in elderly and non-white ethnic groups. Cystatin C offers advantages over creatinine as a marker of kidney function. It is not influenced by non-kidney factors and provides more accurate estimation of GFR, aiding in the early detection of kidney disease and predicting adverse outcomes. Incorporating cystatin C into routine kidney function assessment may improve patient risk stratification and guide clinical decision-making. However, widespread adoption of cystatin C testing requires increased availability and accessibility in clinical laboratories. Further research and implementation efforts are needed to fully realize the potential of cystatin C in kidney function assessment and improving patient outcomes.

Clinical value of perivascular fat attenuation index and computed tomography derived fractional flow reserve in identification of culprit lesion of subsequent acute coronary syndrome.

To explore the potential of perivascular fat attenuation index (FAI) and coronary computed tomography angiography (CCTA) derived fractional flow reserve (CT-FFR) in the identification of culprit lesion leading to subsequent acute coronary syndrome (ACS). Thirty patients with documented ACS event who underwent invasive coronary angiography (ICA) from February 2019 to February 2021 and had received CCTA in the previous 6 months were collected retrospectively. 40 patients with stable angina pectoris (SAP) were matched as control group according to sex, age and risk factors. The study population has a mean age of 59.3 ± 12.3 years, with a male prevalence of 81.4%. The plaque characteristics, perivascular fat attenuation index (FAI), and coronary computed tomography angiography-derived fractional flow reserve (CT-FFR) of 32 culprit lesions and 30 non-culprit lesions in ACS patients and 40 highest-grade stenosis lesions in SAP patients were statistically analyzed. FAI around culprit lesions was increased significantly (-72.4 ± 3.2 HU vs. -79.0 ± 7.7 HU, vs. -80.4 ± 7.0HU, all p < 0.001) and CT-FFR was decreased for culprit lesions of ACS patients [0.7(0.1) vs. 0.8(0.1), vs.0.8(0.1), p < 0.001] compared to other lesions. According to multivariate analysis, diameter stenosis (DS), FAI, and CT-FFR were significant predictors for identification of the culprit lesion. The integration model of DS, FAI, and CT-FFR showed the significantly highest area under the curve (AUC) of 0.917, compared with other single predictors (all p < 0.05). This study proposes a novel integrated prediction model of DS, FAI, and CT-FFR that enhances the diagnostic accuracy of traditional CCTA for identifying culprit lesions that trigger ACS. Furthermore, this model also provides improved risk stratification for patients and offers valuable insights for predicting future cardiovascular events.

Machine Learning Algorithms to Predict Breast Cancer Recurrence Using Structured and Unstructured Sources from Electronic Health Records.

Recurrence is a critical aspect of breast cancer (BC) that is inexorably tied to mortality. Reuse of healthcare data through Machine Learning (ML) algorithms offers great opportunities to improve the stratification of patients at risk of cancer recurrence. We hypothesized that combining features from structured and unstructured sources would provide better prediction results for 5-year cancer recurrence than either source alone. We collected and preprocessed clinical data from a cohort of BC patients, resulting in 823 valid subjects for analysis. We derived three sets of features: structured information, features from free text, and a combination of both. We evaluated the performance of five ML algorithms to predict 5-year cancer recurrence and selected the best-performing to test our hypothesis. The XGB (eXtreme Gradient Boosting) model yielded the best performance among the five evaluated algorithms, with precision = 0.900, recall = 0.907, F1-score = 0.897, and area under the receiver operating characteristic AUROC = 0.807. The best prediction results were achieved with the structured dataset, followed by the unstructured dataset, while the combined dataset achieved the poorest performance. ML algorithms for BC recurrence prediction are valuable tools to improve patient risk stratification, help with post-cancer monitoring, and plan more effective follow-up. Structured data provides the best results when fed to ML algorithms. However, an approach based on natural language processing offers comparable results while potentially requiring less mapping effort.