ABSTRACT
Familial combined hyperlipidemia (FCH) is a very common inherited lipid disorder, characterized by a high risk of developing cardiovascular (CV) disease and metabolic complications, including insulin resistance (IR) and type 2 diabetes mellitus (T2DM). The prevalence of non-alcoholic fatty liver disease (NAFLD) is increased in FCH patients, especially in those with IR or T2DM. However, it is unknown how precociously metabolic and cardiovascular complications appear in FCH patients. We aimed to evaluate the prevalence of NAFLD and to assess CV risk in newly diagnosed insulin-sensitive FCH patients. From a database including 16,504 patients, 110 insulin-sensitive FCH patients were selected by general practitioners and referred to the Lipid Center. Lipid profile, fasting plasma glucose and insulin were determined by standard methods. Based on the results of the hospital screening, 96 patients were finally included (mean age 52.2 ± 9.8 years; 44 males, 52 females). All participants underwent carotid ultrasound to assess carotid intima media thickness (cIMT), presence or absence of plaque, and pulse wave velocity (PWV). Liver steatosis was assessed by both hepatic steatosis index (HSI) and abdomen ultrasound (US). Liver fibrosis was non-invasively assessed by transient elastography (TE) and by fibrosis 4 score (FIB-4) index. Carotid plaque was found in 44 out of 96 (45.8%) patients, liver steatosis was found in 68 out of 96 (70.8%) and in 41 out of 96 (42.7%) patients by US examination and HSI, respectively. Overall, 72 subjects (75%) were diagnosed with steatosis by either ultrasound or HSI, while 24 (25%) had steatosis excluded (steatosis excluded by both US and HSI). Patients with liver steatosis had a significantly higher body mass index (BMI) compared to those without (p < 0.05). Steatosis correlated with fasting insulin (p < 0.05), liver stiffness (p < 0.05), BMI (p < 0.001), and inversely with high-density lipoprotein cholesterol (p < 0.05). Fibrosis assessed by TE was significantly associated with BMI (p < 0.001) and cIMT (p < 0.05); fibrosis assessed by FIB-4 was significantly associated with sex (p < 0.05), cIMT (p < 0.05), and atherosclerotic plaque (p < 0.05). The presence of any grade of liver fibrosis was significantly associated with atherosclerotic plaque in the multivariable model, independent of alcohol habit, sex, HSI score, and liver stiffness by TE (OR 6.863, p < 0.001). In our cohort of newly diagnosed, untreated, insulin-sensitive FCH patients we found a high prevalence of liver steatosis. Indeed, the risk of atherosclerotic plaque was significantly increased in patients with liver fibrosis, suggesting a possible connection between liver disease and CV damage in dyslipidemic patients beyond the insulin resistance hypothesis.
ABSTRACT
To realize a machine learning (ML) model to estimate the dose of low molecular weight heparin to be administered, preventing thromboembolism events in COVID-19 patients with active cancer. Methods: We used a dataset comprising 131 patients with active cancer and COVID-19. We considered five ML models: logistic regression, decision tree, random forest, support vector machine and Gaussian naive Bayes. We decided to implement the logistic regression model for our study. A model with 19 variables was analyzed. Data were randomly split into training (70%) and testing (30%) sets. Model performance was assessed by confusion matrix metrics on the testing data for each model as positive predictive value, sensitivity and F1-score. Results: We showed that the five selected models outperformed classical statistical methods of predictive validity and logistic regression was the most effective, being able to classify with an accuracy of 81%. The most relevant result was finding a patient-proof where python function was able to obtain the exact dose of low weight molecular heparin to be administered and thereby to prevent the occurrence of VTE. Conclusions: The world of machine learning and artificial intelligence is constantly developing. The identification of a specific LMWH dose for preventing VTE in very high-risk populations, such as the COVID-19 and active cancer population, might improve with the use of new training ML-based algorithms. Larger studies are needed to confirm our exploratory results.