The use of noninvasive tools with a high negative predictive value for significant liver fibrosis may help reduce the number of liver biopsies in patients with HIV mono-infection with abnormal transaminases, according to study results published in the Journal of Acquired Immune Deficiency Syndromes.

With an estimated 36.7 million people living with HIV worldwide, HIV infection is a major global health issue. However, the availability of combination antiretroviral therapy (ART) has led to near-normalization of life expectancy in people with HIV with access to testing and early ART. But data from 2009 to 2011 has shown people with HIV dying from other causes including chronic liver disease, which accounted for 10% of deaths during this period. Irrespective of co-infection, liver disease and abnormal liver blood tests (LFTs) are more prevalent in patients who have HIV than patients who do not; abnormal LFTs are usually caused by nonalcoholic fatty liver disease (NAFLD).

Clinicians struggle to characterize the etiology of abnormal LFTs in people with HIV mono-infection. Further, there is a lack of effort to validate tools of noninvasive fibrosis assessment in people with HIV. Therefore, this study assessed the underlying cause of liver disease and characterized the extent of fibrosis in patients with HIV mono-infection with abnormal transaminases by retrospectively evaluating all liver biopsies performed at the Royal Free Hospital in London, England, from 2000 to 2017. In addition, this study evaluated the diagnostic accuracy of FIB4 and FibroScan as noninvasive tools for fibrosis assessment.

A total of 97 patients with HIV mono-infection with abnormal transaminases were included in the study. The baseline characteristics of the cohort population included  mean age 47 years, 81% men, 66% white, and mean body mass index 27 kg/m2. Most patients (74%) had an undetectable HIV viral load and 89% were on ART at the time of biopsy. The median time since HIV diagnosis was 126 months. Patients had experience with the following drug regimens: nucleoside/tide reverse transcriptase inhibitors (92%); non-nucleoside/tide reverse transcriptase inhibitors (72%); protease inhibitors (68%); and d-drugs, ie, stavudine, didanosine, and zalcitabine, (55%).

The most common liver biopsy findings in this mono-infected HIV population were NAFLD (28%) and nonspecific changes (26%). Further, results showed that a significant percentage of the patients with raised transaminases had a completely normal histologic picture and the majority who underwent a liver biopsy had mild or no fibrosis. Significant fibrosis (defined as ≥F2 stage) was present in 20% of patients, whereas advanced fibrosis (defined as ≥F3 stage) was present in 11% of patients. However, most patients (63%) had no fibrosis.

Both the FIB4 and FibroScan demonstrated excellent performance in ruling out advanced fibrosis. The prevalence of FIB4 scores was <1.3 (38%), 1.3-2.67 (46%), and >2.67 (16%). FIB4 showed an 82% specificity and 95% negative predictive value for ruling out advanced fibrosis. The FibroScan values within 1 year of biopsy were available for 27 patients (28%) and 9 patients (33%) had liver stiffness >7.5 kPa. The FibroScan showed a sensitivity of 80%, specificity of 77%, and negative predictive value (NPV) of 94% for significant fibrosis.

Related Articles

Overall, the study authors concluded that, “[HIV monoinfected] patients should be triaged using noninvasive tools of fibrosis assessment, such as FIB4 and FibroScan; those with low values on these tests should only undergo biopsy if there are specific clinical concerns because the probability of advanced fibrosis is very low.”


Prat LI, Roccarina D, Lever R, et al. Etiology and severity of liver disease in HIV-positive patients with suspected NAFLD: lessons from a cohort with available liver biopsies. J Acquir Immune Defic Syndr. 2019;80:474-480.