Novel Machine Learning Identifies 5 Asthma Phenotypes Using Cluster Analysis of Real-World Data.

BACKGROUND: Asthma classification into different subphenotypes is important to guide personalized therapy and improve outcomes. OBJECTIVES: To further explore asthma heterogeneity through determination of multiple patient groups by using novel machine learning (ML) approaches and large-scale real-world data. METHODS: We used electronic health records of patients with asthma followed at the Cleveland Clinic between 2010 and 2021. We used k-prototype unsupervised ML to develop a clustering model where predictors were age, sex, race, body mass index, prebronchodilator and postbronchodilator spirometry measurements, and the usage of inhaled/systemic steroids. We applied elbow and silhouette plots to select the optimal number of clusters. These clusters were then evaluated through LightGBM's supervised ML approach on their cross-validated F1 score to support their distinctiveness. RESULTS: Data from 13,498 patients with asthma with available postbronchodilator spirometry measurements were extracted to identify 5 stable clusters. Cluster 1 included a young nonsevere asthma population with normal lung function and higher frequency of acute exacerbation (0.8 /patient-year). Cluster 2 had the highest body mass index (mean ± SD, 44.44 ± 7.83 kg/m2), and the highest proportion of females (77.5%) and Blacks (28.9%). Cluster 3 comprised patients with normal lung function. Cluster 4 included patients with lower percent of predicted FEV1 of 77.03 (12.79) and poor response to bronchodilators. Cluster 5 had the lowest percent of predicted FEV1 of 68.08 (15.02), the highest postbronchodilator reversibility, and the highest proportion of severe asthma (44.9%) and blood eosinophilia (>300 cells/µL) (34.8%). CONCLUSIONS: Using real-world data and unsupervised ML, we classified asthma into 5 clinically important subphenotypes where group-specific asthma treatment and management strategies can be designed and deployed.

Blood Cholesterol and Triglycerides Associate with Right Ventricular Function in Pulmonary Hypertension.

Background: Blood lipids are dysregulated in pulmonary hypertension (PH). Lower high-density lipoproteins cholesterol (HDL-C) and low-density lipoproteins cholesterol (LDL-C) are associated with disease severity and death in PH. Right ventricle (RV) dysfunction and failure are the major determinants of morbidity and mortality in PH. This study aims to test the hypothesis that dyslipidemia is associated with RV dysfunction in PH. Methods: We enrolled healthy control subjects (n=12) and individuals with PH (n=30) (age: 18-65 years old). Clinical characteristics, echocardiogram, 2-[18F] fluoro-2-deoxy-D-glucose positron emission tomography (PET) scan, blood lipids, including total cholesterol (TC), triglycerides (TG), lipoproteins (LDL-C and HDL-C), and N-terminal pro-B type Natriuretic Peptide (NT-proBNP) were determined. Results: Individuals with PH had lower HDL-C [PH, 41±12; control, 56±16 mg/dL, p<0.01] and higher TG to HDL-C ratio [PH, 3.6±3.1; control, 2.2±2.2, p<0.01] as compared to controls. TC, TG, and LDL-C were similar between PH and controls. Lower TC and TG were associated with worse RV function measured by RV strain (R=-0.43, p=0.02 and R=-0.37, p=0.05 respectively), RV fractional area change (R=0.51, p<0.01 and R=0.48, p<0.01 respectively), RV end-systolic area (R=-0.63, p<0.001 and R=-0.48, p<0.01 respectively), RV end-diastolic area: R=-0.58, p<0.001 and R=-0.41, p=0.03 respectively), and RV glucose uptake by PET (R=-0.46, p=0.01 and R=-0.30, p=0.10 respectively). NT-proBNP was negatively correlated with TC (R=-0.61, p=0.01) and TG (R=-0.62, p<0.02) in PH. Conclusion: These findings confirm dyslipidemia is associated with worse right ventricular function in PH.

Knowledge, Attitude, and Practice Toward Cardiovascular Diseases in the Lebanese Population.

Background & Objective: Cardiovascular diseases (CVD) are the leading cause of death globally. Assessing CVD knowledge, attitude, and practice (KAP) is necessary to spread awareness about CVD in Lebanon, their corresponding risk factors, and behaviors in which individuals can avoid or minimize the possibility of developing a CVD. Subjects & Methods: This was a case-control analytical study that targeted 921 CVD and non-CVD subjects. A questionnaire form was used to collect data related to patients' demographics, socioeconomic status, habits, medical and family history, KAP towards CVD, and source of information. Data was analyzed using SPSS v.25. Results: Data from 921 participants were distributed over the CVD group (52.6% males aged 58.3 ± 13.7 years [n = 460]) and the non-CVD group (47.7% males aged 36.3 ± 15.4 years [n = 461]). CVD patients were significantly older than non-CVD subjects (p < 0.001). All three KAP scores of both groups were of poor to fair levels. Both CVD knowledge and attitude mean scores in CVD patients (26.6 ± 5.2 over 40 [66.50%] and 63.3 ± 10.2 over 85 [74.47%], respectively) were significantly higher than the ones of non-CVD subjects (23.5 ± 7.9 over 40 [58.75%] and 61.4 ± 12.4 over 85 [72.74%], respectively, p < 0.001). However, the CVD mean practice score was significantly lower in CVD patients (6.0 ± 1.7 over 9 [66.67%]) than the one of non-CVD subjects (6.3 ± 2.2 over 9 [70.00%] p < 0.001). Mostly, educational level (p < 0.001), governorate (p < 0.01), and smoking (p < 0.001) were predictors of KAP CVD in both groups. Conclusion: With an overall limited knowledge, attitude, and practice toward CVDs, the Lebanese population (with CVD or non-CVD) needs targeted national campaigns about CVD according to the identified predictors of CVD KAP to prevent and to alleviate the complications due to CVDs.