Risk factors based vessel-specific prediction for stages of coronary artery disease using Bayesian quantile regression machine learning method: Results from the PARADIGM registry.

BACKGROUND AND HYPOTHESIS: The recently introduced Bayesian quantile regression (BQR) machine-learning method enables comprehensive analyzing the relationship among complex clinical variables. We analyzed the relationship between multiple cardiovascular (CV) risk factors and different stages of coronary artery disease (CAD) using the BQR model in a vessel-specific manner. METHODS: From the data of 1,463 patients obtained from the PARADIGM (NCT02803411) registry, we analyzed the lumen diameter stenosis (DS) of the three vessels: left anterior descending (LAD), left circumflex (LCx), and right coronary artery (RCA). Two models for predicting DS and DS changes were developed. Baseline CV risk factors, symptoms, and laboratory test results were used as the inputs. The conditional 10%, 25%, 50%, 75%, and 90% quantile functions of the maximum DS and DS change of the three vessels were estimated using the BQR model. RESULTS: The 90th percentiles of the DS of the three vessels and their maximum DS change were 41%-50% and 5.6%-7.3%, respectively. Typical anginal symptoms were associated with the highest quantile (90%) of DS in the LAD; diabetes with higher quantiles (75% and 90%) of DS in the LCx; dyslipidemia with the highest quantile (90%) of DS in the RCA; and shortness of breath showed some association with the LCx and RCA. Interestingly, High-density lipoprotein cholesterol showed a dynamic association along DS change in the per-patient analysis. CONCLUSIONS: This study demonstrates the clinical utility of the BQR model for evaluating the comprehensive relationship between risk factors and baseline-grade CAD and its progression.

Fractional centralities on networks: Consolidating the local and the global.

We propose a new centrality incorporating two classical node-level centralities, the degree centrality and the information centrality, which are considered as local and global centralities, respectively. These two centralities have expressions in terms of the graph Laplacian L, which motivates us to exploit its fractional analog L^{γ} with a fractional parameter γ. As γ varies from 0 to 1, the proposed fractional version of the information centrality makes intriguing changes in the node centrality rankings. These changes could not be generated by the fractional degree centrality since it is mostly influenced by the local aspect. We prove that these two fractional centralities behave similarly when γ is close to 0. This result provides its complete understanding of the boundary of the interval in which γ lies since the fractional information centrality with γ=1 is the usual information centrality. Moreover, our computation for the correlation coefficients between the fractional information centrality and the degree centrality reveals that the fractional information centrality is transformed from a local centrality into being a global one as γ changes from 0 to 1.

Network Analysis of Cardiac Remodeling by Primary Mitral Regurgitation Emphasizes the Role of Diastolic Function.

BACKGROUND: Topological data analysis (TDA) can generate patient-patient similarity networks by analyzing large, complex data and derive new insights that may not be possible with standard statistics. OBJECTIVES: The purpose of this paper was to discover novel phenotypes of chronic primary mitral regurgitation (MR) patients and to analyze their clinical implications using network analysis of echocardiographic data. METHODS: Patients with chronic moderate to severe primary MR were prospectively enrolled from 11 Asian tertiary hospitals (n = 850; mean age 56.9 ± 14.2 years, 57.9% men). We performed TDA to generate network models using 14 demographic and echocardiographic variables. The patients were grouped by phenotypes in the network, and the prognosis was compared by groups. RESULTS: The network model by TDA revealed 3 distinct phenogroups. Group A was the youngest with fewer comorbidities but increased left ventricular (LV) end-systolic volume, representing compensatory LV dilation commonly seen in chronic primary MR. Group B was the oldest with high blood pressure and a predominant diastolic dysfunction but relatively preserved LV size, an unnoticed phenotype in chronic primary MR. Group C showed advanced LV remodeling with impaired systolic, diastolic function, and LV dilation, indicating advanced chronic primary MR. During follow-up (median 3.5 years), 60 patients received surgery for symptomatic MR or died of cardiovascular causes. Kaplan-Meier curves demonstrated that although group C had the worst clinical outcome (P < 0.001), group B, characterized by diastolic dysfunction, had an event-free survival comparable to group A despite preserved LV chamber size. The grouping information by the network model was an independent predictor for the composite of MR surgery or cardiovascular death (adjusted HR: 1.918; 95% CI: 1.257-2.927; P = 0.003). CONCLUSIONS: The patient-patient similarity network by TDA visualized diverse remodeling patterns in chronic primary MR and revealed distinct phenotypes not emphasized currently. Importantly, diastolic dysfunction deserves equal attention when understanding the clinical presentation of chronic primary MR.

Effects of catastrophic financial loss on suicide risk: evidence from Korean stock market crash in October 2008.

PURPOSE: The negative effect of catastrophic financial loss on suicide risk is widely perceived but hardly studied in-depth because of various difficulties in designing studies. We empirically investigated the effect utilizing the stock market crash event in October 2008 in South Korea. METHODS: We extracted stock market investor data from Korea Exchanges, and mortality data from Microdata Integrated Service of individuals aged 30-60 years. We calculated age-standardized monthly suicide rate per 100,000 persons according to sex and age, and developed intervention analysis with multiplicative seasonal ARIMA model to isolate the effect of the stock market crash on suicide rate. RESULTS: More than 11% of people aged 30-60 years were directly investing in stocks during stock market crash. In October 2008, both KOSPI and KOSDAQ indexes dropped by 22.67% and 30.14%, respectively. In November 2008, the suicide rate in males 30-60 years increased by > 40% compared to the expected levels if there had been no market crash, and in females aged 30-40 and 40-50 years, it increased by 101.84% and 74.81%, respectively. The effect appeared to persist in males, whereas it degenerated with time in females during our sampling period. Suicide was more pronounced in younger age groups and females. CONCLUSION: In this first in-depth study, the effect of catastrophic financial loss negatively affects suicide risk for an extended period, indicating health and financial authorities should provide a long-term financial and psychological support for people with extreme financial loss.

Predicting survival in heart failure: a risk score based on machine-learning and change point algorithm.

OBJECTIVE: Machine learning (ML) algorithm can improve risk prediction because ML can select features and segment continuous variables effectively unbiased. We generated a risk score model for mortality with ML algorithms in East-Asian patients with heart failure (HF). METHODS: From the Korean Acute Heart Failure (KorAHF) registry, we used the data of 3683 patients with 27 continuous and 44 categorical variables. Grouped Lasso algorithm was used for the feature selection, and a novel continuous variable segmentation algorithm which is based on change-point analysis was developed for effectively segmenting the ranges of the continuous variables. Then, a risk score was assigned to each feature reflecting nonlinear relationship between features and survival times, and an integer score of maximum 100 was calculated for each patient. RESULTS: During 3-year follow-up time, 32.8% patients died. Using grouped Lasso, we identified 15 highly significant independent clinical features. The calculated risk score of each patient ranged between 1 and 71 points with a median of 36 (interquartile range: 27-45). The 3-year survival differed according to the quintiles of the risk score, being 80% and 17% in the 1st and 5th quintile, respectively. In addition, ML risk score had higher AUCs than MAGGIC-HF score to predict 1-year mortality (0.751 vs. 0.711, P < 0.001). CONCLUSIONS: In East-Asian patients with HF, a novel risk score model based on ML and the new continuous variable segmentation algorithm performs better for mortality prediction than conventional prediction models. CLINICAL TRIAL REGISTRATION: Unique identifier: INCT01389843 https://clinicaltrials.gov/ct2/show/NCT01389843 .

Topological Data Analysis of Coronary Plaques Demonstrates the Natural History of Coronary Atherosclerosis.

OBJECTIVES: This study sought to identify distinct patient groups and their association with outcome based on the patient similarity network using quantitative coronary plaque characteristics from coronary computed tomography angiography (CTA). BACKGROUND: Coronary CTA can noninvasively assess coronary plaques quantitatively. METHODS: Patients who underwent 2 coronary CTAs at a minimum of 24 months' interval were analyzed (n = 1,264). A similarity Mapper network of patients was built by topological data analysis (TDA) based on the whole-heart quantitative coronary plaque analysis on coronary CTA to identify distinct patient groups and their association with outcome. RESULTS: Three distinct patient groups were identified by TDA, and the patient similarity network by TDA showed a closed loop, demonstrating a continuous trend of coronary plaque progression. Group A had the least coronary plaque amount (median 12.4 mm3 [interquartile range (IQR): 0.0 to 39.6 mm3]) in the entire coronary tree. Group B had a moderate coronary plaque amount (31.7 mm3 [IQR: 0.0 to 127.4 mm3]) with relative enrichment of fibrofatty and necrotic core (32.6% [IQR: 16.7% to 46.2%] and 2.7% [IQR: 0.1% to 6.9%] of the total plaque, respectively) components. Group C had the largest coronary plaque amount (187.0 mm3 [IQR: 96.7 to 306.4 mm3]) and was enriched for dense calcium component (46.8% [IQR: 32.0% to 63.7%] of the total plaque). At follow-up, total plaque volume, fibrous, and dense calcium volumes increased in all groups, but the proportion of fibrofatty component decreased in groups B and C, whereas the necrotic core portion decreased in only group B (all p < 0.05). Group B showed a higher acute coronary syndrome incidence than other groups (0.3% vs. 2.6% vs. 0.6%; p = 0.009) but both group B and C had a higher revascularization incidence than group A (3.1% vs. 15.5% vs. 17.8%; p < 0.001). Incorporating group information from TDA demonstrated increase of model fitness for predicting acute coronary syndrome or revascularization compared with that incorporating clinical risk factors, percentage diameter stenosis, and high-risk plaque features. CONCLUSIONS: The TDA of quantitative whole-heart coronary plaque characteristics on coronary CTA identified distinct patient groups with different plaque dynamics and clinical outcomes. (Progression of AtheRosclerotic PlAque DetermIned by Computed TomoGraphic Angiography Imaging [PARADIGM]; NCT02803411).

Tumor spread through air spaces (STAS): prognostic significance of grading in non-small cell lung cancer.

Tumor spread through air spaces (STAS) is an invasive pattern of lung cancer that was recently described. In this study, we investigated the association between the extent of STAS and clinicopathological characteristics and patient outcomes in resected non-small cell lung cancers (NSCLCs). STAS has been prospectively described from 2008 and graded its extent with a two-tiered system (STAS I: <2500 µm [one field of ×10 objective lens] from the edge of tumor and STAS II: ≥2500 µm from the edge of tumor) from 2011 in Seoul National University Bundang Hospital. We retrospectively analyzed the correlations between the extent of STAS and clinicopathologic characteristics and prognostic significance in 1869 resected NSCLCs. STAS was observed in 765 cases (40.9%) with 456 STAS I (24.4%) and 309 STAS II (16.5%). STAS was more frequently found in patients with adenocarcinoma (ADC) (than squamous cell carcinoma), pleural invasion, lymphovascular invasion, and/or higher pathologic stage. In ADC, there were significant differences in recurrence free survival (RFS), overall survival (OS), and lung cancer specific survival (LCSS) according to the extent of STAS. In stage IA non-mucinous ADC, multivariate analysis revealed that STAS II was significantly associated with shorter RFS and LCSS (p < 0.001 and p = 0.006, respectively). In addition, STAS II was an independent poor prognostic factor for recurrence in both limited and radical resection groups (p = 0.001 and p = 0.023, respectively). In conclusion, presence of STAS II was an independent poor prognostic factor in stage IA non-mucinous ADC regardless of the extent of resection.