Peculiarities of the Ukrainian model of Fracture Risk Assessment (FRAX®) among patients with liver cirrhosis accompanied by impaired bone mineral density: its diagnostic and prognostic value

Abstract

Introduction. The problem of osteoporotic fractures and the evaluation thresholds for intervention in patients with liver cirrhosis (LC) remains obscure so far. Ukrainian model of fracture risk assessment (FRAX®) has never been implemented among patients with LC in Ukraine. The aim of the study. To find out the peculiarities of the Ukrainian model of Fracture Risk Assessment, its diagnostic and prognostic value for implementation among patients with liver cirrhosis accompanied by impaired bone mineral density. Materials and methods. 90 patients with LC (27 women and 63 men aged 18 to 66 years) were randomly assigned into the study. Stratification into groups was based on information about bone condition. 72 patients were included into an experimental group (EG, patients with impaired bone mineral density (IBMD), which was divided into two subgroups – EG A (patients with osteopenia, 46) and EG B (patients with osteoporosis, 26). Control group (CG) included 18 patients without IBMD. The peculiarities of the fracture risk factors and evaluation thresholds according to the Ukrainian FRAX® model (2019) amoung patients with LC with bone disorders were established (significant differences between frequency of features in groups and substantial stochastic associations of features with IBMD or its manifestations were investigated). The diagnostic characteristcs (diagnostic value, predictive value, likelihood ratio) of the detected features for IBMD in general, osteopenia and osteoporosis in particular, were revealed, and after that the post-test probability of certain bone disorders was determined among all patients with LC in the case of applying the identified features. The results. It was found that although most of the risk factors occurred more often in patients with bone disorders, significant differences were detected only between the frequency of previous fractures in EG and CG, including EG B and CG, and EG A and EG B; between the frequency of cases of normal body weight, as well as overweight in EG and CG, including EG B and CG. The evaluation thresholds according to the Ukrainian FRAX® model also differed significantly: the values above the upper evaluation threshold – in EG B and CG and in EG A and EG B; the intermediate values of fracture risk – in EG A and CG; the values below the lower evaluation threshold – in EG and CG, as well as in EG A and CG and in EG B and CG, including. Bone disorders had a substantial direct stochastic association in the following cases: IBMD in general – with the previous fractures, normal body weight and values above the upper evaluation threshold; osteopenia – with the previous fractures, normal body weight and intermediate values of fracture risk; osteoporosis – with the previous fractures, normal body weight and values above the upper evaluation threshold. All manifestations of bone disorders had substantial negative stochastic association with overweight and values below the lower evaluation threshold, as well as osteoporosis with short height (indicates that features are inherent for normal bone mineral density). It was found out that fracture risk factors and evaluation thresholds according to the Ukrainian FRAX® model are mainly single-vector markers, since they can confirm the disease being detected, or deny it in the case they are absent. The previous fractures are highly specific for IBMD, especially for osteoporosis, and can be useful for confirming these disorders being present in patient with LC. The normal body weight is medium-specific for IBMD and for osteoporosis, but can be more useful for indicating IBMD if it is present, and excluding osteoporosis being absent. The values above the upper evaluation threshold according to the Ukrainian FRAX® model are highly specific for osteoporosis and can confirm osteoporisis being present. The intermediate values of fracture risk according to the Ukrainian FRAX® model are medium-specific for osteopenia, but can be more useful for excluding osteopenia if they are absent. The overweight, especially the values below the lower evaluation threshold, will most likely indicate normal bone mineral density. Conclusions. The use of the Ukrainian model of Fracture Risk Assessment (FRAX®) has certain peculiarities and can be valuable tool for detecting or excluding impaired bone mineral density in patients with liver cirrhosis.