Prevalencia y factores de riesgo de hígado grado en pacientes con infección por VIH

Abstract

Introduction: Liver involvement is a major cause of morbidity and consequently mortality among people living with HIV. Increasingly, HIV infection has been linked to fatty liver disease. Objective: To estimate the prevalence and risk factors of fatty liver disease in patients with HIV infection under antiretroviral treatment (ART) at the ‘Hospital General Universitario de Albacete’. To compare ultrasonography with transhepatic elastography (TE) in these patients, evaluating non-invasive markers. Material and methods: An observational, cross-sectional and descriptive study is performed, collecting somatometric variables, cardiovascular risk factors and HIV-specific data. Non-invasive steatosis and fibrosis scores are also carried out. In addition, in order to identify groups of patients in risk of hepatic steatosis and fibrosis, TE and ultrasonography are performed. Descriptive, bivariable and multivariable analyses are done, evaluating the validity of diagnostic tests and ROC curves. HIV infected patients enrolled in the study between August 2018 and February 2020. Results and conclusions: A total of 201 patients were included, with 81.3% of them presenting no symptoms and 94% with undetectable viral load. Of all the patients, 74.1% are men with greater weight (p < 0.001), height (p < 0.001) and abdominal girth (p < 0.001) than women. At the same time, the study shows that men present more metabolic syndrome (p = 0.06) and consume greater amounts of alcohol (p < 0.05). However, men have been followed up for 13.95 years, compared to 22 years for women (p < 0.01). In contrast, women present higher CD4/CD8 fraction values: 0.84 for men and 1.08 for women (p = 0.05). 14 drugs are prescribed in different combinations, with Dolutegravir (53.2%), Emtriciatabine (51.74 %) and Tenofovir alafenamide (50.74 %), being the most used. A high hepatic involvement prevalence is shown from ultrasonography (15.5%) and transhepatic elastography (19%) in the HIV population of Albacete, with these values being greater than those from general population. The internal validity of transhepatic elastography with respect to liver ultrasound has a sensitivity of 61.29 % and a specificity of 88.82 %, presenting a negative predictive value of 92.63 % in the context of the study population. Correlation between the values of transhepatic elastography and non-invasive markers is low or very low. The high hepatic involvement prevalence shown from ultrasonography and transhepatic elastography in HIV population of Albacete surpasses the general population. The great majority of this HIV population has more than five years of evolution since the HIV diagnosis. When comparing the internal validity of transhepatic elastography with liver ultrasound, a high negative predictive value of 92.63% has been obtained. This value indicates that TE should be considered as a test capable of considering patients as healthy. However, given that the correlation between the values of transhepatic elastography and non-invasive markers is low or very low, this has scarce clinical utility. A logistic regression model capable of predicting the liver involvement from the ultrasound has been established. This model considers the patient’s age, body mass index, the fact of being diabetic or not and the TE value. The model can be useful, as it requires few variables commonly managed in a typical consultation without needing regulated ultrasound. In-depth TE studies, preferably with long-term follow-up, are necessary