RESUMO
AIM: To describe the magnetic resonance imaging (MRI) features of HIV-associated obliterative portopathy (HIV-OP) and determine the most indicative appearance of this condition on MRI by using a retrospective case-control study. METHODS: MRI examinations of 24 patients with HIV-OP (16 men, 8 women; mean age = 48 ± 6.6 [SD] years; age range, 35-71 years) were analyzed by two blinded observers and compared with those obtained in 18 HIV-infected patients with hepatic cirrhosis (14 men, 4 women; mean age = 51 ± 3.4 [SD] years; age range, 35-60 years). Images were qualitatively and quantitatively analyzed with respect to imaging presentation. Comparisons were performed using uni- and multivariate analyses. RESULTS: Regular liver contours had the highest accuracy for the diagnosis of HIV-OP (83%, 35 of 42; 95% confidence interval [CI], 69-93%) and was the most discriminating independent variable for the diagnosis of HIV-OP (odds ratio, 51; 95%CI, 4.96-1272%) (p < 0.0001). At multivariate analysis, the width of segment 4 in millimeters (OR = 1.23 [95%CI, 1.05-1.44%]; p = 0.011) and the presence of regular liver contours (OR = 7.69 [95%CI, 1.48-39.92%]; p = 0.015) were the variables independently associated with the diagnosis of HIV-OP. CONCLUSIONS: Regular liver contours are the most discriminating independent variable for the diagnosis of HIV-OP but have limited accuracy. Familiarity with this finding may help differentiate HIV-OP from cirrhosis in HIV-infected patients. KEY POINTS: ⢠Regular liver contour is the most discriminating independent variable for the diagnosis of HIV-OP (odds ratio = 51) with 83% accuracy. ⢠At multivariate analysis, the width of segment 4 in millimeters and the presence of regular liver contours are the variables independently associated with the diagnosis of HIV-OP. ⢠MRI helps diagnose HIV-OP in the presence of several categorical findings, which are more frequently observed in HIV-OP patients than in HIV patients with cirrhosis.
