DIAGNOSTIC ROLE OF DIFFUSION-WEIGHTED IMAGE OF THE LIVER WITH MAGNETIC RESONANCE IMAGING IN PREDICTING ABSTINENCE DISORDERS IN PATIENTS WITH ALCOHOLIC LIVER DISEASE

Objective. To assess the diagnostic role of diffusion-weighted images of the liver with magnetic resonance imaging in predicting abstinence disorders in patients with alcoholic liver disease. Methods. A total of 122 patients with ALD aged 48±5.4 years were examined. The survey algorithm we used included: performing liver DWI with MRI (n=122) with b-value values of 100/600/1000, ultrasound of abdominal organs with clinical elastography – 97 (80%) patients. Trepan liver biopsy was chosen as a reference method (n=64). Results. The patients were monitored for 2.5 years. The terms of follow-up were selected individually, depending on the results of clinical and laboratory research methods. A high correlation was established (r=0.879), when comparing clinical elastography and quantitative indicators of DWI of the liver, at admission and during dynamic observation of patients, also at the middle level, the data obtained correlated with the results of trephine biopsy of the liver (r=0.721). After 3 months, 6 (15%) of 40 patients showed normalization of biochemical blood test parameters with no diffusion restriction according to the results of DWI of the liver. Based on the results obtained, a high correlation was noted between changes in the biochemical blood test and MRI data in the DWI mode. After 9 months of follow-up, according to DWI data, 34 patients showed persistence of cytolysis syndrome and limited diffusion on DWI of the liver. After collecting an additional history and clarifying the details of the lifestyle of the patients' relatives, it was found that these patients continued to consume alcoholic beverages against the background of the received treatment, which was manifested by the presence of diffusion restriction on MRI in the DWI mode, which was a magnetic resonance sign of the presence of inflammatory processes in the structure of the parenchyma liver. After 12 months, positive dynamics – the absence of diffusion restriction according to the results of DWI of the liver was noted in 34 patients, which indicates the effectiveness of using the qualitative characteristics of DWI of the liver to assess the violation of the abstinence regimen (AUROC=0.906 (95% CI 0.872-0.916)). But in 16 (13%) patients from this group, changes in the biochemical blood test were noted, but no diffusion limitation was noted according to the DWI of the liver. Patients (n=16) underwent a correction of the received treatment – after 1 month there was a positive trend. There was a correlation of quantitative parameters of DWI of the liver with clinical forms of ALD, regardless of the presence or absence of diffusion restriction (r=0.936). Next, we assessed the prognostic and diagnostic significance of the developed criteria for DWI of the liver for patients with ALD on admission. The results of the study indicated the effectiveness of using the diagnostic and prognostic model of MRI in the DWI mode for patients with ALD on admission and in dynamic observation. Conclusions. 1. A high correlation was found between the quantitative parameters of DWI of the liver and clinical elastography (r=0.879) at admission and follow-up. Average correlation relationship of DWI of the liver with the results of trephine biopsy of the liver in patients with ALD on admission and follow-up (r=0.721). There was a high correlation between the results of DWI of the liver on MRI with the data of clinical and laboratory parameters in dynamic observation of patients with ALD: no diffusion limitation – positive (r=0.887); yes – negative (r=0.887). The high prognostic and diagnostic value of DWI of the liver in assessing the violation of the abstinence regimen in patients with ALD was established (AUROC=0.906 (95% CI 0.872-0.916)). Prognostic and diagnostic criteria for liver DWI on MRI in patients with ALD at admission: qualitative characteristic – AUROC=0.846 (95% CI 0.811-0.862), quantitative characteristic – AUROC=0.909 (95% CI 0.879-0.912); with dynamic observation: qualitative characteristic – AUROC=0.949 (95% CI 0.907-0.965), quantitative characteristic – AUROC=0.917 (95% CI 0.876-0.932).