Hepatic steatosis with stavudine in HIV/hepatitis C virus co-infection

Abstract

Listen

Translate article

With interest we read the study on hepatic steatosis in subjects co-infected with HIV and hepatitis C virus (HCV) by Sulkowski and colleagues [1]. They detected hepatic steatosis in 40% of patients who had received antiretroviral therapy (ART) for more than 2 years, and identified the historical use of stavudine as a modifiable risk factor for steatosis in this population. Interestingly, however, the use of stavudine at the time of biopsy and the cumulative stavudine exposure were not determinants of hepatic steatosis in the cohort of Sulkowski et al. [1], as was the use of didanosine, which like stavudine is a relatively strong polymerase gamma inhibitor. In order to strengthen their argument, Sulkowski et al. [1] discussed a similar analysis of liver biopsies and argued that the use of stavudine, but not the use of other nucleoside reverse transcriptase inhibitors (NRTI) was associated with hepatic mitochondrial DNA depletion [2]. This work, however, was not cited correctly because mtDNA depletion was not found exclusively with stavudine exposure, but also with the other D-drug NRTI (didanosine and zalcitabine). Similar data were obtained in cultured hepatocytes [3]. We have taken this opportunity and re-analysed the data of our cross-sectional study of 78 HIV/HCV-co-infected individuals according to the criteria of Sulkowski et al. [1]. In our patients, 47% of antiretrovirally treated patients had no steatosis, 23% had grade 1, 15% had grade 2, 13% had grade 3, and 2% had grade 4 steatosis. In order to assess the impact of D-drug and PI use on hepatic steatosis in HIV/HCV co-infection, patients were classified according to their antiretroviral drugs at the time of biopsy (Table 1): patients naive to any antiretroviral therapy (n = 9); individuals receiving neither D-drugs nor protease inhibitors (PI; n = 24); individuals treated without D-drugs but with PI (n = 11); patients receiving D-drugs but no PI (n = 18); and those receiving both ART components (n = 16). There were no statistical differences between all groups with respect to demographic, virological and immunological determinants, and groups did not differ with regard to the duration of current and cumulative ART. Importantly, the prevalence of hepatic steatosis did not vary significantly between these groups (P = 0.53). Similarly, no significant differences were identified when we focused on macro or microvesicular steatosis (P = 0.53, respectively P = 0.95) although mtDNA was depleted in HIV/HCV-co-infected subjects receiving D-drugs (P < 0.0001). Neither overall hepatic steatosis (P = 0.90) nor macro and microvesicular steatosis (P = 0.90, respectively P = 0.99) and mtDNA levels (P = 0.64) were influenced by the use of PI. However, when we compared individuals with fat in over 5% of hepatocytes we found a slight trend towards a higher prevalence of hepatic steatosis in patients treated with both D-drugs and PI, compared with their D-drug and PI-negative counterparts (P = 0.26).Table 1: Characteristics of patients with HIV/hepatitis C virus co-infection.We then focused on the use of stavudine in particular. Steatosis was observed in 53% of patients who had ever used stavudine (n = 51, mean grade 1.1, range 0–4), compared with the identical prevalence of steatosis (53%) in patients who had never used stavudine (mean steatosis grade 0.8, range 0–3). Analysis according to the current use of stavudine revealed no significant differences because 56% of patients (n = 30) who were treated with stavudine at the time of biopsy had hepatic steatosis (mean steatosis grade 1.1, range 0–4), compared with 52% of individuals whose antiretroviral regimen was devoid of stavudine at that time (mean steatosis grade 0.9, range 0–3). Whereas Sulkowski and colleagues [1] had not detected hepatic steatosis in any subject who had never received stavudine or a PI, we detected steatosis in four out of six such patients (two patients had grade 1, one patient had grade 2 and one patient had grade 3 steatosis). A direct comparison between the two datasets may be limited because in our study all patients were Caucasian and 55% were infected with HCV genotype 1, whereas in the study of Sulkowski et al. [1], the patients were mostly black and carried HCV genotype 1. We also acknowledge that we may not have had sufficient power to rule out a true relationship between hepatic steatosis and stavudine (and D-drugs in general), but conclude that the current data do not support the notion that stavudine is an important contributor to hepatic steatosis. Such a relationship, although mechanistically plausible, may be obscured by other confounders. Acknowledgement This work was supported by the BMBF, Kompetenznetz HIV/AIDS (grant number: 01KI0211).