Hepatic steatosis and steatohepatitis in human immunodeficiency virus/hepatitis C virus–coinfected patients


  • Potential conflict of interest: Nothing to report.


Hepatic steatosis (HS) is frequent in human immunodeficiency virus (HIV)- and hepatitis C virus (HCV)-coinfected patients. Antiretroviral therapy (ART) and metabolic alterations could induce HS. However, a protective effect of ART has been reported in a paired biopsy study. Thus, our aim was to examine the changes and predictors of HS progression among HIV/HCV-coinfected patients with sequential biopsies. We also evaluated the rates of steatohepatitis and factors associated thereof. HIV-infected patients with detectable serum HCV RNA, who underwent two biopsies, separated at least by 1 year, were included in this retrospective study. HS progression was defined as increase in one or more HS grades. The median (interquartile range) time between biopsies was 3.3 (2.0-5.2) years. Among 146 individuals, HS at baseline was observed in 86 (60%) patients and in 113 (77%) in the follow-up biopsy (P < 0.001). Progression of HS was observed in 60 (40%) patients. HS regressed in 11 (8%) patients. Factors associated with HS progression were changes in fasting plasma glucose (FPG) between biopsies (per 10 mg/dL increase; odds ratio [OR] [95% confidence interval; CI] = 1.4 [1.04-1.8]; P = 0.024) and cumulative use of dideoxynucleoside analogs (per year; OR [95% CI] = 1.5 [1.2-1.8]; P = 0.001). Persistent steatohepatitis or progression to steatohepatitis between biopsies was observed in 27 (18%) patients. Persistence of or progression to steatohepatitis was associated with progression ≥1 fibrosis stages between biopsies (OR [95% CI] = 2.4 [1.01-5.7]; P = 0.047). Conclusions: HS progresses frequently and regression is rarely observed in HIV/HCV-coinfected patients, including in those on ART. Cumulative exposure to dideoxynucleoside analogs and increases in FPG are related with HS progression. Stetatohepatitis is frequently observed in these patients and is linked to fibrosis progression. (HEPATOLOGY 2012)

Hepatic steatosis (HS) is a common condition in hepatitis C virus (HCV)-infected patients with human immunodeficiency virus (HIV) coinfection. Previous cross-sectional studies have reported on frequencies of HS between 30% and 70%.1-12 Besides its prevalence, the main clinical implication of HS is that it has been associated with liver fibrosis progression in several studies.2-7 Particularly, among patients without HIV infection with nonalcoholic fatty liver disease (NAFLD), those with steatohepatitis are at increased risk of fibrosis progression.13 Thus, a better understanding of modifiable risk factors that may contribute to HS development is critical. In this sense, previous studies have implicated several metabolic factors, such as obesity, hyperglycemia, hyperlipidemia, or lipodystrophy, in the development of HS.2-7 In addition, the use of certain reverse-transcriptase inhibitors, such as stavudine, didanosine, or efavirenz, has been associated with HS in some studies.4, 6, 8, 14 However, other studies failed to find this association.1-3, 5, 7 Thus, the relationship between HS and antiretroviral therapy (ART) remains to be elucidated.

The main limitation to the current knowledge of risk factors for HS is that the available information derives mainly from cross-sectional studies. Moreover, only one of these reports, to our knowledge, evaluated the frequency of steatohepatitis in HIV/HCV-coinfected patients.5 A recent longitudinal analysis of HIV/HCV-coinfected patients, who had undergone at least two liver biopsies, examined the rates of steatosis progression.15 The prevalence of HS at baseline was lower than that found in previous studies.1-11, 14 At the follow-up biopsy, HS did not progress in the majority of patients. Among progressors, ART was associated with a lower risk of HS progression. The reasons for these findings are unclear. The racial background of the study cohort, overwhelmingly composed of HCV genotype 1–infected African Americans, may partly explain these striking results. Thus, there is a need for additional studies assessing the rates of HS progression and the risk factors for progression, including the role of antiretroviral drugs, in HIV/HCV-coinfected subjects, as it has been claimed by some experts.16 Furthermore, there are no data on the changes in steatohepatitis over time in HIV/HCV coinfection.

In this study, we aimed at evaluating the changes in HS between liver biopsies and the predictors of HS progression among HIV/HCV-coinfected patients with sequential liver biopsies. We also assessed the rates of steatohepatitis and factors associated with the persistence and progression thereof in these patients.


ART, antiretroviral therapy; BMI, body mass index; CDC, Centers for Disease Control and Prevention; CI, confidence interval; DM, diabetes mellitus; ETR, end-of-treatment response; FPG, fasting plasma glucose; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HS, hepatic steatosis; IQR, interquartile range; IR, insulin resistance; NAFLD, nonalcoholic fatty liver disease; NAS, NAFLD activity score; OR, odds ratio; SVR, sustained virological response; TGs, triglycerides.

Patients and Methods

Design and Study Population.

This was a retrospective study carried out in paired liver biopsies performed in HIV/HCV-coinfected patients who attended nine Spanish hospitals from January 1989 to January 2008. An analysis of liver fibrosis progression in these sequential biopsies has been previously reported on.17 HIV-infected patients were included in the present study if they met the following: (1) active HCV infection, as determined by detectable serum HCV RNA; (2) underwent two liver biopsies, separated by at least 1 year; (3) liver biopsies have been performed as part of the assessment of HCV infection to establish the prognosis and/or to indicate treatment; and (4) no evidence of vascular, tumoral, biliary, or autimmune liver disease. Individuals with cirrhosis detected at the first liver biopsy were included for the present analysis. Biopsy samples with length lower than 15 mm or fragmented specimens were deemed as inadequate, and the corresponding patient was excluded.

Antiviral Therapy.

Patients received ART according to the availability of drugs and the recommendations of international guidelines and panels of experts in force during the study period. Therapy against HCV infection was prescribed according to the caring physician criteria, based on consensus recommendations in effect along the study period, usually guided by HCV genotype and liver fibrosis stage. End-of-treatment response (ETR) was defined as undetectable serum HCV RNA at the planned date of treatment cessation. Sustained virological response (SVR) was defined as undetectable serum HCV RNA 24 weeks after the end of treatment.

Liver Histology.

Liver steatosis and liver fibrosis were scored blindly by a central pathologist (M.A.J.). HS classification was based on the proportion of hepatocytes containing fat droplets using Brunt's criteria18 and was classified as follows: 0, absent steatosis; 1, less than 33% (i.e., mild HS); 2, 33%-66% (i.e., moderate HS); and 3, more than 66% (i.e., severe HS). Lobular inflammation was scored as follows: 0 = no foci; 1 = <2 foci (excludes 2 foci ×200 field); 2 = 2-4 (includes 2 and 4 foci ×200 field) 3 = >4 foci (excludes 4 foci ×200 field). Cytologic ballooning was classified as follows: 0, none; 1, few balloon cells; and 2, many cells/prominent ballooning. The NAFLD activity score (NAS) was calculated as the unweighted sum of steatosis, lobular inflammation, and hepatocellular ballooning scores.19 Scheuer's score20 was applied to stage fibrosis as follows: 0, absent fibrosis; 1, portal fibrotic expansion; 2, extension of fibrosis to the lobule, but with few septa; 3, bridging fibrosis with numerous septa, with architectural distortion without cirrhosis; and 4, cirrhosis. The length of biopsies was recorded to assess their adequacy. The median (interquartile range; IQR) length of the initial biopsy was 17 (15-25) mm, whereas the respective figure of the second biopsy was 18 (16-25) mm.

Statistical Analysis.

The primary outcome variable of the study was the progression in one or more grades in Brunt's score. The associations of the following baseline factors with HS progression were analyzed: gender, age, body mass index (BMI), self-reported daily alcohol intake, diagnosis of diabetes mellitus (DM) following the American Diabetes Association criteria, fasting plasma glucose (FPG), cholesterol and triglycerides (TGs), HCV genotype, and Centers for Disease Control and Prevention (CDC) stage C. In addition, the following variables between biopsies and their relationship with HS progression were assessed: BMI, self-reported daily alcohol intake, FPG, cholesterol and TGs, response to treatment against HCV, CD4 cell counts at liver biopsies, plasma HIV RNA viral load at liver biopsies, exposure to ART, and changes in fibrosis stage. Cumulative exposure to individual antiretroviral drugs was calculated as the period in years receiving each antiretroviral drug between biopsies. Blood tests were drawn within 1 month before the liver biopsies.

The secondary outcome variable was persistence of steatohepatitis between biopsies or progression to steatohepatitis in the final biopsy. Steatohepatitis was defined as NAS score ≥5.19 The associations between the above-stated baseline and follow-up variables with persistence of or progression to steatohepatitis were assessed.

Continuous variables are expressed as median (IQR), and categorical variables are presented as numbers (percentage). Mann-Whitney's U test was applied for comparisons of continuous variables between groups. Comparisons between categorical variables were made by the chi-square test or Fisher's exact test, when appropriate. For comparisons between related groups, Wilcoxon's signed-rank test for continuous variables and McNemar's test for the categorical ones were used. Odds ratios (ORs) (95% confidence intervals; CIs) of variables potentially related with the outcome variables were calculated by univariate logistic regression. Variables associated with the outcome variables with a P value ≤0.2 in univariate analyses were entered in multivariate logistic regression models. Multivariate models were adjusted by the difference in sample length between the first and second biopsies, regardless of its association with the outcome variables. Associations with a P value <0.05 after the multivariate analysis were considered significant. Statistical analysis was carried out using the SPSS 19 statistical software package (SPSS, Inc., Chicago, IL).

Ethical Aspects.

The study was designed and conducted following the Helsinki declaration. The Ethics Committee of the Hospital Universitario de Valme (Seville, Spain) approved the study.


Baseline Characteristics.

One hundred and forty-six patients were included in the study. The characteristics of these patients at the initial biopsy are summarized in Table 1. All patients were Caucasians of European ancestry. DM was diagnosed in 9 (6.5%) individuals. One hundred and twenty-five (86%) patients showed a BMI between 18.5 and 24.99 kg/m2, 6 (4%) individuals showed a BMI lower than 18.5 kg/m2, 11 (7.5%) had a BMI equal to or greater than 25 kg/m2 and lower than 30 kg/m2, and 4 (2.7%) showed a BMI greater than 30 kg/m2. The majority of patients received ART at baseline (Table 1). The distribution of individual antiretroviral drugs prescribed during the follow-up and the cumulative exposure to them is listed in Table 2.

Table 1. Baseline Characteristics of the Study Patients (n = 146)
  • Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma-glutamyl transferase; BMI, body mass index; CDC, Centers for Disease Control and Prevention; DM, diabetes mellitus; FPG, fasting plasma glucose; TGs, triglycerides.

  • *

    Median (IQR).

  • Available in 122 patients.

  • Available in 135 patients (10 patients with nontypable genotype and 1 not available).

  • §

    Available in 127 patients.

  • Available in 132 patients.

Age, years*37.8 (33-41)
Male gender, n (%)107 (73)
Injecting drug users, n (%)125 (86)
DM, n (%)9 (6.5)
Alcohol intake >50 g/day,§ n (%)31 (25)
BMI,* kg/m222.5 (21.9-23.4)
FPG,* mg/dL90 (80-94)
TGs,* mg/dL118 (98-180)
Cholesterol,* mg/dL172 (150-192)
CDC stage C, n (%)35 (24)
Nadir CD4 cell counts, cells/μL*215 (100-340)
CD4 cell counts,* cells/mcL495 (338-660)
Undetectable HIV viremia, n (%)88 (60)
Antiretroviral therapy, n (%)117 (80)
Estimated age at HCV infection, years*20.4 (18-26)
HCV genotype, n (%) 
 185 (58)
 22 (1.4)
 329 (20)
 419 (13)
Serum HCV RNA, log IU/mL*5.96 (5.59-6.37)
ALT,* IU/mL70 (52-121)
AST,* IU/mL63 (41-92)
GGT,* IU/mL105 (51-203)
Hepatic steatosis, n (%) 
 059 (40)
 157 (39)
 222 (15)
 38 (5.5)
Lobular inflammation, n (%) 
 08 (5.5)
 154 (37)
 254 (37)
 330 (21)
Cytologic ballooning, n (%) 
 196 (66)
 250 (34)
Liver fibrosis stage, n (%) 
 029 (20)
 149 (34)
 227 (19)
 330 (21)
 411 (7.5)
Table 2. Antiretroviral Therapy During the Follow-up (n = 146)
DrugsN (%)Median (IQR) Time, Years
Zidovudine65 (45)1 (0.8-2.0)
Abacavir28 (19)1.8 (0.9-2.6)
Tenofovir37 (25)2.7 (1.8-3.3)
Lamivudine117 (80)2.3 (1.3-3.4)
Dideoxynucleosides89 (61)1.7 (0.9-4.5)
Saquinavir26 (18)1.5 (1.1-2.8)
Indinavir25 (17)1.3 (1.1-2.2)
Nelfinavir24 (16)1.7 (1.3-2.4)
Lopinavir/r10 (6.8)2.2 (1.7-3.0)
Nevirapine53 (36)1 (0.5-3.0)
Efavirenz51 (35)1.3 (1-3)

Rate of Steatosis Progression.

The median (IQR) time between biopsies was 3.3 (2.0-5.2) years. HS at baseline was observed in 87 (60%) patients. Most patients with HS at the initial biopsy presented grade 1 HS (Fig. 1; refer to Supporting Table 1 for associations with baseline HS). In the second biopsy, HS was detected in 113 (77%) patients, 49 (34%) of whom bore grade 2 HS. The frequency of HS grades at the first and second biopsy is detailed in Fig. 1. The prevalence of moderate and severe HS was higher in the follow-up biopsies, compared with the baseline biopsies (Fig. 1). Progression of at least one grade of HS was observed in 60 (40%) patients at the second liver biopsy, and 8 (5%) patients progressed two or more grades of HS. Progression to grade 2 or 3 HS was observed in 34 (23%) patients. Only 11 (8%) patients experienced regression of one or more HS grades (refer to Supporting Table 2 for detailed changes in HS between biopsies).

Figure 1.

Frequency of grades of hepatic steatosis at the initial and follow-up biopsies (n = 146).

Figure 2.

Frequency of patients with progression of hepatic steatosis by the length of exposure to dideoxynucleosides, lamivudine, efavirenz, and nevirapine.

Changes in Different Factors Between Liver Biopsies.

Median CD4 cell counts were 495 (338-660) cells/μL at the initial biopsy and 540 (427-700) cells/μL at the follow-up biopsy (P = 0.021). At baseline, 88 (60%) patients showed undetectable plasma HIV RNA and 107 (73%) reached plasma HIV viremia below the detection level at the second biopsy (P = 0.018). Twenty-nine (20%) patients at the first biopsy and 28 (19%) subjects at the second biopsy were not under ART (P = 0.882).

The stage of liver fibrosis at the initial and follow-up biopsies were as follows: stage 0, 29 (20%) versus18 (12%); stage 1, 49 (34%) versus 39 (27%); stage 2, 27 (19%) versus 42 (29%); stage 3, 30 (21%) versus 24 (16%); and stage 4, 11 (7.5%) versus 23 (16%) (P = 0.019). Among 69 patients who received therapy against HCV, 4 (5.8%) achieved SVR and 26 (38%) ETR.

Factors Associated With Progression of Liver Steatosis.

Table 3 compares those with and without steatosis progression. CD4 cell counts, plasma HIV RNA, and use of ART were not related with HS progression. Cumulative exposure to dideoxynucleoside analogs (i.e., didanosine, stavudine, or zalcitabine) and efavirenz was associated with HS progression (Table 3; Fig. 2). Progression of fibrosis one or more stages was not associated with HS progression (Table 3). Increases in median FPG were significantly higher among patients with HS progression (Table 3). Changes in BMI, TGs, and cholesterol were not associated with HS progression (Table 3).

Table 3. Factors Associated With Progression of One or More Grades of Steatosis Between Liver Biopsies
CharacteristicSteatosis Progression (N = 60)No Steatosis Progression (N = 86)Unadjusted OR (95% CI)P UnivariateAdjusted OR (95% CI)P Multivariate
  • BMI, body mass index; CDC, Centers for Disease Control and Prevention; DM, diabetes mellitus; FPG, fasting plasma glucose; TGs, triglycerides; ETR: End of treatment response; SVR: Sustained virological response.

  • *

    Median (IQR).

  • Per 10-mg/mL increase.

  • Available in 135 patients.

  • Undetectable viremia at baseline and follow-up biopsies.

  • §

    Cumulative time on each drug for patients exposed to them.

  • Per year of exposure.

  • #

    Among 69 patients who received anti-HCV therapy.

Baseline age,* years38.1 (33.9-40.6)37.5 (32.7-41.0)1.0 (0.9-1.1)0.977  
Male gender, n (%)43 (72)56 (65)1.4 (0.7-2.8)0.405  
Alcohol intake >50 g/day during the follow-up15 (28)20 (25)1.2 (0.5-2.5)0.703  
Baseline BMI,* kg/m221.9 (22.6-23.5)21.9 (22.4-23.1)1.03 (0.89-1.2)0.544  
Change in BMI,* kg/m20.36 (−0.36, 1.72)0.36 (−0.07, 1.17)0.9 (0.8-1.2)0.773  
Baseline PFG,* mg/dL82 (89-93)80 (90-97)0.98 (0.95-1.0)0.213  
Change in PFG,* mg/dL4.5 (−4.0, 11.5)1 (−9, 9)1.04 (1.01-1.1)0.0081.4 (1.04-1.9)0.024
Baseline plasma TGs,* mg/dL85 (121-156)100 (133-200)0.997 (0.99-1.001)0.243  
Change in plasma TGs,* mg/dL6.5 (−10, 32)2.5 (−9, 44)1 (0.99-1.003)0.426  
Baseline plasma cholesterol,* mg/dL144 (170-192)151 (170-190)1.001 (0.99-1.01)0.786  
Change in plasma cholesterol,* mg/dL−1.5 (−32, 25)−0.5 (−31, −33)1 (0.99-1.01)0.355  
HCV genotype 3, n (%)11 (18)18 (21)1.2 (0.5-2.7)0.699  
Change in CD4 cell counts,* cells/μL42 (−67, 113)60 (−40, 156)1 (0.99-1.001)0.404  
Nadir CD4 cell counts,* cells/μL189 (91-310)258 (137-354)1.001 (0.99-1.002)0.248  
Undetectable HIV viremia,§ n (%)27 (45)31 (36)1.5 (0.7-2.8)0.278  
Antiretroviral therapy, n (%)51 (85)70 (81)1.3 (0.5-3.2)0.569  
Cumulative time on antiretroviral therapy between biopsies, years3.3 (1.6-5.2)2.7 (1.8-4.1)1.1 (0.9-1.3)0.1811.0 (0.9-1.2)0.884
Cumulative time on individual antiretroviral drugs between biopsies, years      
 Zidovudine*1 (0.5-2)1 (0.9-2.6)0.7 (0.4-1.2)0.1650.8 (0.5-1.3)0.432
 Abacavir*1.8 (0.9-2.6)1.8 (0.9-2.6)1 (0.5-2.1)0.986  
 Tenofovir*3 (2.0-3.8)2.4 (1.8-3.0)1.1 (0.8-1.7)0.567  
 Lamivudine*2.5 (1.3-3.6)2.2 (1.2-3.3)1.2 (0.9-1.4)0.1710.8 (0.6-1.02)0.077
 Dideoxinucleosides*5 (3.0-8.7)1.1 (0.7-1.9)2.2 (1.5-3.0)<0.0011.5 (1.2-1.8)<0.001
 Saquinavir*1.9 (1.2-3.5)1.3 (1-2.4)1.5 (0.7-2.9)0.280  
 Indinavir*1.3 (1.0-2.2)1.3 (1.1-2.1)1 (0.4-2.7)0.984  
 Nelfinavir*2.1 (1.3-2.7)1.6 (1.3-2.1)1.7 (0.8-3.5)0.1711 (0.5-1.8)0.994
 Lopinavir*2.2 (1.7-3.0)2.5 (1.8-3.2)0.5 (0.1-3.2)0.497  
 Efavirenz*3 (1.0-4.7)1 (0.9-2.4)1.6 (1.1-2.3)0.0181.1 (0.9-1.5)0.405
 Nevirapine*0.7 (0.5-0.8)2 (1-3)0.9 (0.7-1.2)0.448  
Fibrosis progression ≥1 stages between biopsies, n (%)26 (43)34 (40)1.2 (0.6-2.3)0.646  
Treatment against HCV, n (%)26 (43)43 (50)0.8 (0.4-1.5)0.427  
ETR or SVR,# n (%)12 (46)18 (42)0.9 (0.4-2.1)0.727  
Time between biopsies,* years3.4 (2.03-5.4)2.9 (1.8-4.7)1.04 (0.9-1.2)0.511  
Length difference between last and first biopsy, mm0 (0-2)0 (0-3)0.96 (0.9-1.06)0.5390.96 (0.9-1.09)0.548

After the multivariate analysis, cumulative exposure to dideoxynucleoside analogs and increases in FPG were independently related with progression of HS (Table 3). An analysis excluding patients with baseline cirrhosis yielded similar results (data not shown).

Steatohepatitis Persistence or Progression.

The median (IQR) NAS score was 3 (3-4) for the first biopsy and 4 (3-4) for the follow-up biopsy (P = 0.002; refer to Supporting Table 3 for associations with baseline steatohepatitis). The NAS score increased in 65 (45%) and decreased in 35 (24%) individuals between the initial and final biopsy. Steatohepatitis was detected in 24 (16%) patients in the first biopsy and in 27 (18%) subjects in the final biopsy (P = 0.602). Steatohepatitis persisted in 9 (38%) of 24 patients. Among 122 individuals without steatohepatitis initially, 18 (15%) showed progression (refer to Supporting Table 4 for detailed changes in NAS scores between biopsies).

Persistence of or progression to steatohepatitis was related with fibrosis progression (Table 4). There was a nonsignificantly longer exposure to dideoxynucleoside analogs and to ART among patients with persistent or progressive steatohepatitis (Table 4). Eleven (14%) patients with ART for <4 years and 10 (25%) subjects with ART for ≥4 years showed steatohepatitis persistence or progression (P = 0.119). After the multivariate analysis, the only variable independently associated with persistent or progressive steatohepatitis was liver fibrosis progression (Table 4). An analysis excluding patients with baseline cirrhosis yielded similar results (data not shown).

Table 4. Factors Associated With Persistence of or Progression to Steatohepatitis Between Liver Biopsies
CharacteristicPersistence/Progression of Steatohepatitis (N = 27)Absent/Regression of Steatohepatitis (N = 119)Unadjusted OR (95% CI)P UnivariateAdjusted OR (95% CI)P Multivariate
  • BMI, body mass index; CDC, Centers for Disease Control and Prevention; DM, diabetes mellitus; FPG, fasting plasma glucose; TGs, triglycerides; ETR: End of treatment response; SVR: Sustained virological response.

  • *

    Median (IQR).

  • Available in 135 patients.

  • Undetectable viremia at baseline and follow-up biopsies.

  • §

    Cumulative time on each drug for patients exposed to them.

  • Per year of exposure.

  • Among 69 patients who received anti-HCV therapy.

Baseline age,* years35.2 (32.5-38.5)38.2 (33.3-41.2)0.93 (0.86-1.01)0.0740.97 (0.87-1.09)0.628
Male gender, n (%)21 (78)78 (66)1.8 (0.69-4.9)0.219  
Alcohol intake >50 g/day during the follow-up9 (36)26 (24)1.8 (0.69-4.4)0.236  
Change in BMI,* kg/m20.36 (−0.35, 1.91)0.37 (0.0-1.4)0.98 (0.78-1.2)0.863  
Change in PFG,* mg/dL0 (−6.5, 13.5)2 (−5, 10)1.01 (0.98-1.04)0.498  
Change in plasma TGs,* mg/dL19.5 (0-57)2 (−12, 26)1 (0.99-1.01)0.235  
Change in plasma cholesterol,* mg/dL−6 (−19, 42)−0 (−33, 22)1 (0.99-1.01)0.628  
HCV genotype 3, n (%)4 (15)25 (21)0.65 (0.21-2.06)0.469  
Change in CD4 cell counts,* cells/μL57 (−62, 152)35 (−18, 119)1.0 (0.99-1.01)0.820  
Nadir CD4 cell counts,* cells/μL259 (76-389)200 (103-328)1.0 (0.99-1.003)0.185  
Undetectable HIV viremia, n (%)11 (41)47 (40)1.05 (0.45-2.5)0.905  
Antiretroviral therapy, n (%)21 (78)100 (84)0.67 (0.24-1.9)0.436  
Cumulative time on antiretroviral therapy between biopsies,§ years3.8 (1.7-5.0)2.9 (1.7-4.6)1.1 (0.9-1.3)0.553  
Cumulative time on invidual antiretroviral drugs between biopsies,§ years      
 Zidovudine*1.5 (0.8-2.5)1 (0.9-2.0)1.3 (0.77-2.0)0.363  
 Abacavir*1.8 (1.0-2.6)1.9 (0.9-2.6)0.9 (0.35-2.3)0.830  
 Tenofovir*1.9 (0.8-4.2)2.7 (1.9-3.3)0.79 (0.38-1.67)0.542  
 Lamivudine*2.5 (1.3-3.4)2.2 (1.2-3.4)0.95 (0.73-1.2)0.711  
 Dideoxinucleosides*5 (4-10)3 (1-5)1.4 (0.87-2.1)0.1751.1 (0.8-1.4)0.639
 Saquinavir*1.5 (0.8-2.2)1.5 (1.1-2.8)0.58 (0.1-3.5)0.553  
 Indinavir*1.9 (1.3-2.3)1.3 (0.91-2.1)1.3 (0.46-3.8)0.603  
 Nelfinavir*1.3 (0.8-1.6)1.7 (1.3-2.7)0.37 (0.08-1.7)0.204  
 Lopinavir*1.9 (0.8-2.1)2.96 (1.7-3.03)0.07 (0.002-2.8)0.1570.97 (0.39-2.4)0.950
 Efavirenz*1 (1-1)1.8 (1-3)0.95 (0.59-1.5)0.820  
 Nevirapine*0.6 (0.5-0.8)1 (0.5-3.0)0.5 (0.2-1.3)0.1730.61 (0.2-1.9)0.390
Fibrosis progression ≥1 stages between biopsies, n (%)16 (59)44 (36)1.2 (0.6-2.3)0.0342.4 (1.01-5.7)0.047
Treatment against HCV, n (%)16 (59)53 (44)1.8 (0.78-4.2)0.202  
ETR or SVR, n (%)6 (38)24 (45)0.73 (0.23-2.3)0.775  
Time between biopsies,* years3.8 (1.8-5.3)3.0 (1.8-5.1)1.04 (0.9-1.2)0.475  
Length difference between last and first biopsy, mm0 (0-5)0 (0-3)1.06 (0.9-1.2)0.5691.04 (0.9-1.2)0.507


HS severity progresses with time frequently in HIV/HCV-coinfected patients, both in those who receive ART and in those who do not. HS regression is rarely observed in this setting. Cumulative exposure to dideoxynucleoside analogs and increases in FPG are associated with HS progression. In addition, steatohepatitis is frequently observed in HIV/HCV-coinfected patients, and NAS score increases over time in these individuals. Steatohepatitis tends to be associated with more-prolonged exposures to ART and dideoxynucleoside analogs. Importantly, persistence of or progression to steatohepatitis is linked to fibrosis progression in HIV/HCV coinfection.

The results of the herein reported study are in contrast with the study by Woreta et al. that assessed HS progression in paired liver biopsies from HIV/HCV-coinfected patients.15 In that study, fewer patients presented HS at baseline and HS did not progress in approximately 90% of patients in the follow-up biopsies.15 On the contrary, in our study, 60% of patients showed some degree of HS in the initial biopsy, increases of 1 stage in HS was observed in 40% of patients, and progression to moderate or severe HS was observed in 23% of individuals. The reasons for such conflicting data are unclear. The participants in the study by Woreta et al. were overwhelmingly HCV genotype 1–infected African Americans,15 whereas patients in the present study were Caucasians with infection by more-diverse HCV genotypes. This may partly explain the lower prevalence and progression of HS in the study by Woreta et al., given that individuals with African ancestry might have a lower propensity to develop NAFLD.21 However, a recent meta-analysis did not find a significantly different prevalence of HS among HIV/HCV-coinfected African Americans.12 The high prevalence of HCV genotype 3 may partially account for the higher rates of HS in our study, given the association between this genotype and HS.2-4, 11 Nevertheless, HCV genotype 3 was not associated with HS progression in our study.

The role of ART in the development of HS is controversial. We found that HS progression between liver biopsies was associated with cumulative exposure to dideoxynucleoside analogs. This finding is in agreement with previous cross-sectional studies.4, 6, 14 Dideoxynucleoside analogs, susc as didanosine, stavudine, and zalcitabine, are potent inhibitors of mitochondrial DNA (mtDNA) polymerase-gamma, the enzyme responsible for mtDNA replication. mtDNA depletion impairs respiratory chain activity and thus inhibits mitochondrial β-oxidation, finally causing abnormal deposition of fatty acids in hepatocytes.22 However, most reported cross-sectional studies failed to find an association with ART or individual antiretroviral drugs.1-3, 5, 7 One possible explanation might be different exposures to dideoxynucleoside analogs across studies. Thus, the use of these drugs was highly prevalent in studies finding an association between them and HS.4, 6, 14

We found an association between progression of HS and cumulative exposure to efavirenz in the univariate analysis. Similar to our findings on dideoxynucleosides, the larger the time on efavirenz, the higher the frequency of patients with HS progression. There are some data that support the mitochondrial toxicity of efavirenz. In vitro, efavirenz induces bioenergetic stress in hepatic cells by inhibiting mitochondrial function through an acute mechanism that is independent of mtDNA replication.8 This leads to the accumulation of lipids in the cytoplasm through a mechanism mediated by the activation of adenosine monophosphate&activated protein kinase.8 In vivo, efavirenz is associated with lipoatrophy,23 a mitochondrial toxicity initially described among recipients of dideoxynucleosides. In the present study, the lack of an independent statistical association between efavirenz and HS progression might have been the result of the overwhelming effect of dideoxynucleosides and the relatively small sample size of the efavirenz treatment group. Importantly, efavirenz is currently recommended as a first-option drug to combine in initial ART regimens. Thus, the risk of HS progression among patients exposed to efavirenz needs further evaluation.

Cumulative ART exposure was associated with a lower risk of HS progression in a previous study.15 In addition, higher CD4 cell counts were also protective of HS progression.15 In our study, we found that markers of response to ART, such as CD4 cell counts and undetectable HIV viremia, improved between liver biopsies, confirming that most patients were receiving effective ART. In spite of this fact, HS increased in frequency and severity in the follow-up biopsy, and this observation was not related to CD4 cell counts or HIV viremia changes. Moreover, we found that cumulative dideoxynucleoside analog exposure was a predictor of HS progression, and that time on efavirenz between biopsies was associated, in the univariate analysis, with HS progression. Both dideoxynucleoside analogs and efavirenz display mitochondrial toxicity. On the contrary, a drug with a very low risk of mitochondrial toxicity, such as lamivudine, showed a statistical trend to less HS progression. Conflicting results between the present study and a previous report15 are difficult to explain on the sole basis of racial and HCV genotype influences. Our study data are consistent with many previous findings. Thus, ART is associated with increasing insulin resistance (IR), a mechanism involved in the pathogenesis of HS. Drugs typically related with mitochondrial toxicity, such as dideoxynucleosides and efavirenz, were associated with HS progression, whereas drugs without this side effect (i.e., lamivudine and nevirapine) were not. Prospective studies with uniform and predefined criteria for sequential evaluation of HS are warranted to better define the relationship between ART and specific drugs or families and HS progression. However, biopsy-based studies with this design will not be easily feasible, given the invasive nature of the technique and the necessity of repeated measures over short periods. Only sequential evaluations of HS with an accurate noninvasive technique can allow these studies.

Metabolic factors are associated with HS. Obesity is a well-characterized risk factor for HS. However, its role in HIV/HCV-coinfected patients seems to be weaker. BMI has been related with HS in cross-sectional studies2-4, 10 and in the only longitudinal study reported so far,15 but the magnitude of this association was small in cross-sectional studies.12 In our study, BMI was not related with progression of HS. One possible reason for the lack of association in our study is that very few patients were overweight or obese. Indeed, the majority of patients were within the limits of normal BMI. IR plays a central role in metabolic syndrome and it is another factor implicated in HS. In this regard, we found that FPG was associated with HS progression. Several previous cross-sectional studies found an association between fasting glucose and HS.4, 6, 11 However, hypertriglyceridemia, another component of metabolic syndrome, was not associated with HS progression in the present study. In this regard, a meta-analysis carried out in cross-sectional studies on HIV/HCV-coinfected patients failed to show a relationship of hypertriglyceridemia with HS.12

Persistent steatohepatitis or progression to steatohepatitis was observed in 18% of patients. This is a high rate of steatohepatitis, which is within the range observed in morbid obesity or DM.13 Our results are in agreement with a previous cross-sectional study on HIV/HCV coinfection that reported on a prevalence of steatohepatitis similar to the frequencies observed in the herein reported study.5 Notably, we found that persistent steatohepatitis or progression to steatohepatitis was associated with fibrosis progression in sequential biopsies. This result is in agreement with studies in paired liver biopsies among HIV-uninfected patients with NAFLD, where fibrosis progression was observed in the subset of individuals with steatohepatitis.24, 25 Moreover, cohort studies on NAFLD show that patients with steatohepatitis progress to more-serious liver disease and have higher liver-related mortality than those only with HS.26, 27

HS is very frequent in HIV/HCV-coinfected patients.1-12 HS has been linked with fibrosis progression in previous cross-sectional studies on HIV/HCV coinfection2-7 and in the only previous study on paired biopsies.15 Thus, HS itself is relevant in the setting of HIV and HCV dual infection. Because of this, steatohepatitis was a secondary endpoint in the present study. We did not find a correlation between HS and fibrosis progression, in agreement with some cross-sectional studies.1, 9-14 Rather, we observed that patients with features of steatohepatitis were at an increased risk of fibrosis progression. Different rates of underlying steatohepatitis in previous studies may account for the contradictory results on the relationship between HS and fibrosis in coinfected patients.

This study has a few limitations. First, HS may change fast in response to the modification of some factors, such as alcohol intake, overweight, or drugs. These changes could be missed by a paired biopsy study, particularly if the time between biopsies is very long. Importantly, under- or overdiagnosing the incidence of HS in liver biopsies resulting from changes in such modifiable factors may be a relevant reason for discrepant results among studies or for lack of detection of some associations. This limitation is inherent in every liver biopsy-based study. Additionally, the use of scores to classify HS precludes the detection of smaller changes in HS, within each HS category. The only theoretical solution to these drawbacks would be a prospective study with a frequent schedule of noninvasive assessments of HS using a reliable procedure. Second, alcohol intake was self-referred by patients. Probably because of this, we did not find any association between alcohol and HS. In this regard, previous cross-sectional studies also failed to find this relationship.1, 4-6, 9 In fact, history of alcohol abuse was associated with HS progression in a study on sequential liver biopsies, but present reporting was not.15 Third, some antiretroviral drugs used during the period of study have become obsolete, particularly dideoxynucleoside analogs and unboosted protease inhibitors. A number of newer antiretroviral drugs are available to combine that may have no mitochondrial toxicity and a better metabolic profile and thus potentially lead to lesser HS. Thus, data on the HS associated with those newer drugs are needed. Fourth, patients who accept to undergo repeated liver biopsies are highly selected. Usually, these patients are those more compliant with follow-up. Clinicians may indicate a follow-up biopsy if liver disease progression is suspected. Thus, rates of HS and steatohepatitis might be overestimated by paired liver biopsy studies as the study herein reported on. However, the frequency of HS in the present study is in agreement with previous cross-sectional studies.1-11 Moreover, rates of steatohepatitis were similar to those from a previous cross-sectional report.5 Finally, this study's results may not be applicable to other groups of patients, because very few overweight or obese individuals were included and the racial background of the study subjects was only Caucasian.

In summary, HS is frequently detected in HIV/HCV-coinfected patients with and without ART and high rates of progression to severe HS are observed in them. This is a major concern, given that among individuals with HS, those with features of steatohepatitis are at increased risk of fibrosis progression. Steatohepatitis was not influenced by the suppression of HIV replication, and this may be one of the reasons why fibrosis progresses fast in HIV/HCV-coinfected patients despite controlling HIV replication. In addition, HS may contribute to poorer response to interferon-based therapy against HCV. Cumulative exposure to dideoxynucleoside analogs and, perhaps, to efavirenz are factors associated with HS progression. Because of this, antiretroviral drugs with less potential impact on mitochondrial toxicity should be prioritized in HIV/HCV-coinfected patients. Finally, the natural history of HS and steatohepatitis in HIV/HCV coinfection needs further investigation, particularly in patients receiving the newer antiretroviral drugs.