Fatty liver prevalence high with HIV/HCV coinfection: 'use ddI or d4T cautiously'

Individuals coinfected with HIV and hepatitis C virus (HCV) should use ddI (didanosine, Videx/Videx EC) and d4T (stavudine, Zerit) with caution, according to an editorial commenting on the results of the largest study so far to report on factors associated with hepatic steatosis, or fatty liver - the accumulation of fatty acids in the liver - in coinfected patients. The study, to be published in the August 1^st issue of Clinical Infectious Diseases, found that 69% had steatosis and that the use of ddI or d4T increased the risk of steatosis almost fivefold.

Hepatic steatosis is frequently a consequence of obesity, high alcohol consumption and metabolic syndrome, affecting almost one-third of the United States population - although it is usually a benign condition. However, previous studies in HCV monoinfected individuals have shown that hepatic steatosis can result in accelerated progression of liver disease. Due to the body-shape changes and metabolic abnormalities that occur in HIV lipodystrophy syndrome, individuals coinfected with HIV/HCV may be at increased risk of developing steatosis.

Last year, researchers from Johns Hopkins School of Medicine found that hepatic steatosis was present in 40% of a cohort of 112 coinfected individuals. They also found that the risk of hepatic steatosis was greatest in those receiving treatment with d4T.

To further evaluate the prevalence of hepatic steatosis, the factors associated with it, and its relationship to liver fibrosis, investigators from four New England HIV clinics undertook a retrospective chart review of all HIV/HCV coinfected patients who had undergone liver biopsy between January 2000 through December 2003.

A total of 260 patients were evaluated, of whom 193 were eligible. Exclusion criteria included hepatitis B coinfection, history of steatosis-associated drug use, and cirrhosis. A further ten patients had inadequate biopsy specimens, resulting in a study population of 183 (median age 43 years; 79% male; 50% white, 27% Hispanic, 24% African American; 83% injection drug use history, with a median 23 years HCV infection). The majority (60%) were infected with HCV genotype 1: 8% had genotype 2, 16% had genotype 3, and 4% had genotype 4. At the time of the liver biopsy, 55 patients (30%) were not taking antiretroviral therapy.

The investigators found steatosis in 69% of the biopsy specimens (minimal 31%, mild 27%, moderate 18%, severe 1%). They explain a higher prevalence of steatosis in their cohort than in the Johns Hopkins cohort as due to a higher proportion of non-African Americans in their cohort. Previous studies have found that "African American patients with HCV monoinfection are at lower risk of steatosis than are white patients," they write.

In univariate analysis, presence of steatosis was associated with greater weight (p=0.011) and an undetectable viral load (p=0.019). There was a trend towards significance with a higher body mass index; and HCV genotype 3 infection, glucose and triglyceride levels were found to be of borderline significance.

The greatest odds of steatosis, however, were associated with the use of any antiretroviral agent (OR, 1.99; p=0.043) and any nucleoside analogue (OR, 2.14; p=0.024). When types of nucleoside analogues were stratified into dideoxynucleoside analogues (the D drugs, ddI and d4T) versus non-dideoxynucleoside analogues, the risk of steatosis increased with the D drugs (OR, 2.63 vs. 1.78; p=0.05). Use of ddI and d4T together increased the odds even further (OR, 3.38; 95% CI, 0.67-17.07).

Adding to the weight of the association with the nucleoside analogue class was the finding that although only 21 patients were taking triple nucleoside analogue therapy, 19 had evidence of hepatic steatosis (OR, 7.13; 95% CI, 1.51-33.57).

In multivariate analysis, only nucleoside analogue use was significantly associated with hepatic steatosis. Although the use of nucleoside analogues that were not D drugs was still associated with an increased risk of steatosis compared with no nucleoside analogue use, there was a wide confidence interval, making it of borderline significance. (OR, 2.65; 95% CI, 0.95-7.41; p=0.062).

However, when the investigators compared D drug use with no nucleoside analogue use, the increased risk was almost fivefold (OR, 4.63; 95% CI, 1.55-13.8; p=0.006). "This clinical observation is supported by in vitro and in vivo data that suggest that didanosine and stavudine have significant mitochondrial toxicity that exceeds that of other drugs," the investigators write.

Nevertheless, they note that the limitations of their cross-section study design should be taken into consideration, and add that "any association between steatosis and dideoxynucleoside use will need to be confirmed by longitudinal prospective studies." In addition, "the impact of prior or cumulative NRTI exposure could not be evaluated in our study and was likely important."

The investigators also found that steatosis was associated with stage of liver fibrosis (univariate OR, 1.37; 95% CI, 1.03-1.81; p=0.029). "These data are of vital clinical importance," they write, "because fibrosis progression occurs faster in HIV-HCV-coinfected patients than in patients with HCV infection alone."

They conclude by suggesting that "administration of antiretroviral agents with little or no mitochondrial toxicity (e.g., tenofovir, lamivudine, emtricitabine, or abacavir) may be preferred over other NRTIs whenever possible," a recommendation that is echoed and amplified in an accompanying editorial by Marija Zeremski and Andrew Talal of Weill Medical College at Cornell University.

They write that although "an association between antiretroviral therapy and steatosis in HIV-HCV-coinfected patients has not been a universal finding among the studies that have been performed to date, a connection between antiretroviral therapy and steatosis is becoming increasingly apparent."

Although the mechanism of NRTI-induced steatosis "remains obscure," they add, "mitochondrial toxicity is likely responsible for their harmful effects" and D drugs are "the antiretroviral therapy agents with the strongest capacity to deplete mtDNA through the interaction with DNA polymerase-y". They suggest that in addition to the direct effects of NRTIs on the development of steatosis, the indirect effects of the thymidine analogue, d4T, may also play a role, due to its association with lipoatrophy, insulin resistance and hyperlipidaemia.

They point out that McGovern and her colleagues observed microvesicular steatosis (small fat droplets of fatty infiltration) in the vast majority of cases (19% pure microvesicular; 50% mixed microvesicular and macrovesicular). "Because both microvesicular steatosis and use of NRTIs are implicated in mitochondrial toxicity, these observations support the role of NRTIs in the development of hepatic steatosis," they argue.

They conclude by saying that "in the clinical treatment of HCV-HIV-coinfected patients, and especially for those with steatosis, the D drugs should be used cautiously."