Background:  Liver biopsy is frequently used for establishing the prognosis and the need of therapy against hepatitis C virus (HCV) in co-infected patients. Transient elastometry (Fibroscan), a new non-invasive method to measure liver stiffness, has been validated to evaluate liver fibrosis in HCV-mono-infected patients. However, while it has been reported to be highly accurate in detecting advanced liver fibrosis and cirrhosis (F = 4), there are contradictory data regarding the discrimination of significant (F ≥2) from mild liver fibrosis. In addition, liver stiffness has been insufficiently validated in HIV/HCV co-infection. Our aim was to assess the reliability of reported cutoff values of liver stiffness to discriminate F = 4 and F ≥2 in HIV/HCV-co-infected subjects.

Methods:  Liver stiffness was measured in HIV/HCV-co-infected subjects who met the following inclusion criteria:  liver biopsy within 12 months of liver stiffness measurement; and no therapy for HCV infection between liver biopsy and liver stiffness measurement. Patients diagnosed with clinical cirrhosis were included, and a F4 fibrosis stage was assumed. Accuracy of liver stiffness to assess liver fibrosis was tested by the area under the receiver operating characteristic curves (AUROC).

Results:  The inclusion criteria were met by 101 patients, 37 of whom (37%) had F = 4, and 20 of those (54%) with clinical diagnosis of cirrhosis. AUROC (95% confidence interval CI) for liver stiffness were 0.95 (0.91 to 0.99) for F = 4 and 0.88 (0.82 to 0.95) for F ≥2. For the diagnosis of F = 4, cutoff values = 14.6 kPa, positive predictive value 79%, negative predictive value 93%; cutoff values = 17.6 kPa, positive predictive value 81%, negative predictive value 89%. For the diagnosis of liver fibrosis ≥2, a cutoff value of 7.2 showed a positive predictive value of 86% and a negative predictive value of 73%; 19 (19%) patients were incorrectly classified according to a cutoff value of 7.2. Thus, 10 (27%) of 37 patients with a liver stiffness value <7.2 showed F ≥2 in liver biopsy (9; 24%) with F = 2 and 1 (3%) with F = 3; and 9 (14%) of 64 subjects with a liver stiffness value ≥7.2 showed F<2 in liver biopsy (6  [9 %] with F = 1).

Conclusions:  Fibroscan fails to discriminate mild from significant liver fibrosis in a substantial number of HIV/HCV-co-infected patients. If a decision to start HCV therapy were based on the presence of a liver stiffness ≥7.2 kPa, as many as 27% of individuals carrying F ≥2 would not receive therapy.