Prevention and management of antiretroviral therapy-related toxicity has emerged as a major issue for HIV/AIDS treatment and care [1,2]. Hepatotoxicity is now well described as a component of the broad spectrum of antriretroviral therapy toxicity [3–12]. Elevations in serum hepatic enzymes have been described in association with all major classes of antiretroviral therapy [3–12], with several underlying mechanisms proposed: mitochondrial toxicity in association with several nucleoside reverse transcriptase inhibitors (NRTI); hypersensistivity reaction in association with non-nucleoside reverse transcriptase inhibitors (NNRTI) and immune reconstitution disease in association with underyling chronic viral hepatitis.
Previously identified risk factors for severe hepatotoxicity in people receiving highly active antiretroviral therapy (HAART) include commencement of HAART in antiretroviral therapy-naive patients, recent commencement of nevirapine or high-dose ritonavir, chronic hepatitis B virus (HBV) or chronic hepatitis C virus (HCV) coinfection, elevated baseline alanine aminotransferase (ALT) level and female gender [3–12].
Both access to antiretroviral therapy and clinical research infrastructures are limited in developing countries; consequently, most information in relation to therapeutic toxicity comes from developed country settings. The growing impetus for access to therapy for people with HIV infection in all settings will see increasing numbers of people in developing countries receiving antiretroviral therapy . The limited infrastructure for therapeutic monitoring in many settings makes an understanding of rates and predictors of antiretroviral therapy toxicity even more crucial.
The present study assessed hepatotoxicity incidence and risk factors among people commenced on antiretroviral therapy in studies conducted by the HIV-Netherlands Australia Thailand Research Collaboration (HIV-NAT), in Bangkok, Thailand.
The study population included all HIV-infected adults participating in randomized controlled trials of antiretroviral therapy at HIV-NAT, Bangkok, Thailand who initiated therapy between December 1996 and March 2001. Patients were recruited from the HIV outpatient clinic of Chulalongkorn Hospital. In accordance with trial protocols, clinical and laboratory data were collected at screening (within 4 weeks of antiretroviral therapy commencement), baseline (just prior to antiretroviral therapy commencement), and at weeks 4, 8, 12, 24, 36 and 48.
Clinical and laboratory data
Data collected from patients at baseline were age, gender, risk group for HIV infection, past or present occurrence of HIV-related illnesses as classified according to the Centers for Disease Control and Prevention (CDC) 1993 guidelines , weight, CD4 and CD8 cell counts, HIV viral load, prior antiretroviral therapy use, liver function profile and past or present occurrence of symptoms and signs related to hepatitis.
In addition, at each subsequent scheduled visit, clinical data on CDC and non-CDC adverse events, and laboratory data were collected. Standard laboratory testing included full blood count, serum chemistries, ALT and aspartate aminotransferase (AST) levels, CD4 and CD8 cell counts, and plasma HIV viral load. Furthermore, HBV (HBsAg) and HCV (anti-HCV antibody) testing was performed retrospectively on baseline blood samples stored at −70°C by Cobas Core enzyme-linked immunosorbent assays (Roche Diagnostic Systems, Branchburg, New Jersey, USA). Coinfection with HBV was defined as the detection of HBsAg at baseline. Since HCV infection has a very high rate of persistence , patients were considered to have HCV coinfection when anti-HCV antibody was detected at baseline.
Definitions of hepatotoxicity grades
Hepatotoxicity grades were based on ALT level and defined in accordance with AIDS Clinical Trials Group criteria , in the following manner: grade 1, 1.25–2.5 times the upper limit of normal (× ULN); grade 2, 2.6–5.0 × ULN; grade 3, 5.1–10 × ULN; grade 4, > 10 × ULN. In order to avoid selection bias favouring inclusion of patients with very high baseline ALT levels, severe hepatotoxicity was defined as grade 3 or grade 4 increases in ALT level and an increase in ALT level of greater than 100 U/l from baseline. Furthermore, medical files of all patients experiencing severe hepatotoxicity were reviewed with the aim to identify and exclude other potential causes of hepatic enzyme elevations (such as acute hepatitis A virus infection, heavy alcohol use, and non-antiretroviral drug-related hepatotoxicity).
Hepatotoxicity was examined for all HIV-infected patients receiving antiretroviral therapy during the study period. Incidence of severe hepatotoxicity following commencement of trial combination antiretroviral therapy was calculated by hepatitis status and antiretroviral therapy. Three hepatitis groupings were used: (i) no hepatitis (HBsAg−/anti-HCV−), (ii) HBV coinfection (HBsAg+/anti-HCV−) and (iii) HCV coinfection (anti-HCV+/HBsAg−). Antiretroviral therapy was grouped as (i) dual NRTI, without a protease inhibitor or NNRTI, (ii) dual NRTI plus protease inhibitor, (iii) dual NRTI plus protease inhibitor and low-dose ritonavir, and (iv) dual NRTI plus an NNRTI. All patients received at least two NRTI within their antiretroviral regimen.
Associations between severe hepatotoxicity and a range of demographic and clinical characteristics were assessed in univariate and multivariate Cox regression analyses. Variables included in these analyses were age, gender, hepatitis status, HIV risk group, prior AIDS, baseline CD4 and CD8 counts, baseline HIV viral load, prior antiretroviral therapy use, trial antiretroviral therapy regimen, baseline ALT and follow-up CD4 and CD8 counts.
Differences between groups were analysed using Student's t test or ANOVA for normally distributed continuous data, and Wilcoxon or Kruskal–Wallis test for data that were not normally distributed. Chi-square tests, where appropriate, were used for categorical data. Differences between groups were considered to be significant at a P value < 0.05. All reported P values are two-sided. Survival analyses were performed and Kaplan–Meier curves were obtained for time to development of severe hepatotoxicity. Poisson regression was used to determine 95% confidence intervals (CI) for hepatotoxicity incidence. SPSS version 9.0 for Windows (SPSS, Inc., Chicago, Illinois, USA) was used for all statistical analyses.
All protocols were approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand. All patients gave written, informed consent.
A total of 692 patients enrolled in eight randomized controlled trials between December 1996 and March 2001 were included in the study (Table 1). All patients received at least two NRTI, while 135 received a protease inhibitor-containing regimen and 215 received an NNRTI-containing regimen. Of those who received a protease inhibitor-containing regimen, 50 patients received low-dose ritonavir in addition to the main protease inhibitor component of the antiretroviral regimen. Prevalence of HBV, HCV and HBV/HCV coinfection was 8.7% (60/692), 7.2% (50/692), and 0.4% (3/692), respectively. Baseline demographic and clinical characteristics by hepatitis status are shown in Table 2. Patients with HCV coinfection were most likely to be male (72%) and to report injecting drug use as their HIV risk factor (10%). HBV-coinfected patients were most likely to have a prior AIDS diagnosis (25%). Patients with HBV and HCV coinfection were more likely to have a past history of acute hepatitis (12% and 10%, respectively). Other demographic and clinical parameters were similar across hepatitis groups, including mean age, weight, HIV viral load, and trial antiretroviral therapy regimen.
Following commencement of trial antiretroviral therapy regimens, 40 of 692 patients (5.8%; 95% CI, 4.1–7.9) developed severe hepatotoxicity, as defined by an ALT > 5 × ULN and increase of greater than 100 U/l from baseline. The incidence of severe hepatotoxicity was 6.1/100 person-years (95% CI, 4.3–8.3), with a median time to detection of 56 days [interquartile range (IQR), 28–168]. Only three patients (0.4%) developed grade 4 (ALT > 10 × ULN) hepatotoxicity, an incidence rate of 0.5/100 person-years (95% CI, 0.1–1.4).
In univariate analysis, factors associated with an increased risk of severe hepatotoxicity were HBV coinfection [relative risk (RR), 3.91; 95% CI, 1.73–8.83], HCV coinfection (RR, 3.02; 95% CI, 1.18–7.75), no prior antiretroviral therapy (RR, 3.65; 95% CI, 1.11–11.99) and commencement of an NNRTI-containing regimen (RR, 6.78; 95% CI, 3.04–15.07) (Table 3). There was a borderline decreased risk of severe hepatotoxicity among patients with a baseline CD4 cell count greater than or equal to 250 × 106 cells/l (RR, 0.48; 95% CI, 0.23–0.95). No significant differences in risk were found for age, sex, risk group for HIV infection, CDC class, CD8 cell count, HIV load, ALT level at baseline and changes in CD4 and CD8 cell count during follow-up.
In multivariate analysis, three factors were associated with risk of severe hepatotoxicity: HBV coinfection (RR, 3.20; 95% CI, 1.14–9.03), HCV coinfection (RR, 3.00; 95% CI, 1.17–7.80) and NNRTI-containing regimen (RR, 9.75; 95% CI, 3.02–31.47). Baseline CD4 cell count and prior antiretroviral therapy status were no longer associated with severe hepatotoxicity (Table 3).
Incidence of severe hepatotoxicity was considerably higher for HBV (15.3/100 person-years; 95% CI, 7.0–29.0) and HCV (13.3/100 person-years; 95% CI, 4.9–29.1) coinfected patients compared with HIV- monoinfected patients (4.5/100 person-years; 95% CI, 2.9–6.7) (Table 4; Fig. 1). However, median time to severe hepatotoxicity was similar for the three groups. Incidence of severe hepatotoxicity was also considerably higher in patients commenced on NNRTI- containing regimens (14.1/100 person-years; 95% CI, 9.5–20.1) than in those commencing dual NRTI only (2.2/100 person-years; 95% CI, 0.9–4.5), protease inhibitor-containing regimens (2.1/100 person-years; 95% CI, 0.3–7.7) and protease inhibitor/low-dose ritonavir-containing regimens (0.0/100 person-years; 95% CI, 0.0–7.4) (Table 5; Fig. 2). Among patients commenced on NNRTI-containing regimens, 14.0% (30/215) developed severe hepatotoxicity, compared with 2.1% (10/477) of patients commenced on other regimens. Higher incidence of severe hepatotoxicity among patients on NNRTI-containing regimens was largely associated with use of nevirapine (18.6/100 person-years; 95% CI, 10.6–29.8) and nevirapine/efavirenz combination (32.4/100 person-years; 95% CI, 16.5–57.7), with a significantly lower incidence for efavirenz (2.4/100 person-years; 95% CI, 0.3–7.8) (P = 0.01). The combination of underlying chronic viral hepatitis and use of nevaripine-containing regimens provided the highest risk of severe hepatotoxicity, with a similar incidence for HBV coinfection and nevirapine (57.4/100 person-years; 95% CI, 26.3–109.1), and HCV coinfection and nevirapine (72.2/100 person-years, 19.6–184.9).
Antiretroviral therapy was not permanently modified or stopped in any of the 40 patients who experienced severe hepatotoxicity. Therapy was temporarily modified according to specific clinical trial protocols. At week 48 following commencement of trial antiretroviral therapy, all 40 patients were receiving their original antiretroviral regimen assigned at baseline. Despite development of severe hepatotoxicity, and continuation of original antiretroviral therapy regimens, median ALT level in these patients had returned towards baseline level by week 48 (Fig. 3). No episodes of clinical acute hepatitis or deaths related to severe hepatotoxicity were seen in these patients during the study period.
Following commencement of combination antiretroviral therapy, incidence of severe hepatotoxicity and factors associated with its development among Thai patients with HIV infection are similar to those seen in studies in developed country settings. Coinfection with either HBV or HCV and commencement of nevirapine were the major factors associated with severe hepatotoxicity. Clinical symptoms of hepatitis were uncommon among patients with severe hepatotoxicity, and antiretroviral therapy regimens could be maintained. These findings are of considerable importance for development of therapeutic and monitoring strategies for people living with HIV in resource-constrained settings.
Approximately 6% of patients developed severe hepatotoxicity within the initial 48 weeks of commencement of trial combination antiretroviral therapy. Although rates of hepatotoxicity depend on definitions employed – studies based on ALT elevations have used 3 × ULN to 10 × ULN – the incidence of severe hepatotoxicity in our study (6.1/100 person-years) is consistent with other studies [3–12]. The most common predictors of severe hepatotoxicity in previous studies have been HBV and HCV coinfection, and use of either nevirapine or full-dose ritonavir therapy [3–12], with three of these four factors identified in our study. The incidence of both grade 4 (0.5/100 person-years) and symptomatic hepatotoxicity was extremely low in our study. In contrast, a recent Dutch cohort study reported a grade 4 hepatotoxicity incidence of 6.3/100 person-years . The higher proportion of patients in their study receiving nevirapine (37% versus 18%) or full-dose ritonavir (66% versus 7%), and the more standardized liver enzyme assessment with antiretroviral therapy dose modification in our study, may explain the contrasting rates of grade 4 hepatotoxicity.
Hepatoxicity is now well described in people with HIV receiving NNRTI-containing regimens, in particular nevirapine [9,11,12]. Furthermore, acute fulminant hepatitis has been documented in association with nevirapine use within postexposure prophylaxis regimens . The incidence of nevirapine-related severe hepatotoxicity in our study was higher than in others [9,11,12,18]; however, this may partly reflect the frequent and standardized measurement of liver enzymes within the trial protocols. Frequent liver enzyme monitoring made it possible to demonstrate the early nature of NNRTI-related hepatotoxicity, with the majority of events occurring in the initial 12 weeks of therapy. Alternative explanations for the relatively high incidence of NNRTI-related hepatotoxicity are a higher prevalence of viral hepatitis coinfection than in some other studies, and the relatively low body weight of our Thai patients. Despite the relatively high incidence of severe hepatotoxicity, there were no adverse clinical outcomes related to the events. However, in settings where frequent monitoring (with subsequent regimen modification) is not possible, nevirapine-related severe hepatotoxicity may produce adverse clinical outcomes. Development of standardized antiretroviral therapy regimens for such settings need to take into account potential toxicity, and its relationship to monitoring capabilities.
Protease inhibitor-containing therapy is also associated with severe hepatotoxicity, in particular with ritonavir [7,10–12,19–20]. Underlying chronic HCV infection also appears to increase this risk [10–12]. A minority (20%) of patients in our study received protease inhibitors (in combination with two NRTI); however, the incidence of severe hepatotoxicity was particularly low in this group and comparable to that in individuals receiving two NRTI only. The probable explanation for this finding was the absence of high-dose ritonavir, with only low boosting doses of ritonavir employed. A recent study that examined liver enzyme changes on switching from indinavir to indinavir-ritonavir in HIV/HCV-coinfected patients also demonstrated no significant hepatotoxicity .
Coinfection with HBV and HCV are recognized risk factors for hepatotoxicity [4–5,9,11,12]. Our findings in a Thai HIV-infected population support these other studies. Immune reconstitution disease has been hypothesized as one of the underlying mechanisms for this increased risk . There was, however, no association in our study between CD4 cell change following commencement of trial antiretroviral therapy and risk of severe hepatotoxicity. An alternative explanation is that established liver inflammation increases the risk of direct toxicity from antiretroviral therapy agents such as nevirapine. Incidence of severe hepatotoxicity was relatively low among patients coinfected with HBV or HCV who were receiving regimens that did not contain NNRTI. The extremely high risk of severe hepatotoxicity among HIV-infected patients who were HBV or HCV coinfected and commenced on nevirapine-containing regimens suggests that alternative regimens should be considered when coinfection is documented, or in settings where testing is not routinely performed but background prevalence is greater than 10%. This is particularly important for Thailand, where a national government-funded antiretroviral therapy programme is employing a generic triple combination antiretroviral therapy regimen that includes nevirapine. In countries where antiretroviral therapy options are limited, monitoring and surveillance of NNRTI-related hepatotoxicity is crucial.
Any grade of hepatotoxicity was seen in approximately half of our cohort (data not shown). Despite this rate of hepatotoxocity, by week 48 liver enzyme levels had returned to baseline levels with minimal modifications to antiretroviral therapy regimens. This pattern of liver enzyme changes following commencement of antiretroviral therapy is consistent with other studies [5,6]. Even those with severe hepatotoxicity in our study had a favourable biochemical and clinical outcome. The standardized and regular evaluation of liver enzymes in our study provides further reassurance that this is the general course of hepatotoxicity. However, the systematic monitoring employed by the trial protocols also ensured dose modifications in patients with severe hepatotoxicity. Our study does not explore the natural history of hepatotoxicity in a setting with limited monitoring capabilities nor the longer-term impact of hepatotoxicity on disease outcomes. A recent study demonstrated that patients with hepatic enzyme elevations 4 months following commencement of HAART had increased mortality over the subsequent 2 years .
Our study among a large Thai HIV-infected population showed similar patterns of antiretroviral therapy-related hepatotoxicity to other settings and demonstrated limited short-term adverse outcomes associated with severe hepatotoxicity. Despite these encouraging aspects, development of standardized antiretroviral therapy protocols should consider potential hepatotoxicity, and agents with higher rates may be best avoided in settings where monitoring capabilities are limited.
The authors gratefully acknowledge the efforts of clinical trial nurses and, in particular, the participation of all patients involved in randomized controlled trials at HIV-NAT.
Sponsorship: The clinical trials were sponsored by Boehringer Ingelheim, Merck Sharp & Dohme, and Bristol-Myers Squibb. The National Centre in HIV Epidemiology and Clinical Research is funded by the Commonwealth Department of Health and Ageing.
1. Powderly WG, Carr A. AIDS 2001. Clinical treatment. Overview. AIDS
2. Carr A, Workman C, Smith DE, Hoy J, Hudson J, Doong N, for the Mitochondrial Toxicity (MITOX) Study Group. Abacavir substitution for nucleoside analogs in patients with HIV lipoatrophy: a randomized trial. JAMA
3. Rodriguez-Rosado R, Garcia-Samaniego J, Soriano V. Hepatotoxicity after introduction of highly active antiretroviral therapy. AIDS
4. Savès M, Vandentorren S, Daucourt V, Marimoutou C, Dupon M, Couzigou P, et al. Severe hepatic cytolysis: incidence and risk factors in patients treated by antiretroviral combinations. Aquitaine Cohort, France, 1996–1998. AIDS
5. den Brinker M, Wit WNM, Wertheim-van Dillen PME, Jurriaans S, Weel J, van Leeuwen R, et al. Hepatitis B and C virus co-infection and the risk for hepatotoxicity of highly active antiretroviral therapy in HIV-1 infection. AIDS
6. Gisolf EH, Dreezen C, Danner SA, for the Prometheus Study Group. Risk factors for hepatotoxicity in HIV-1–infected patients receiving ritonavir and saquinavir with or without stavudine. Clin Infect Dis
7. Sulkowski MS, Thomas DL, Chaisson RE, Moore RD. Hepatotoxicity associated with antiretroviral therapy in adults infected with human immunodeficiency virus and the role of hepatitis C or B virus infection. JAMA
8. Monforte Ade A, Bugarini R, Pezzotti P, for the ICONA (Italian Cohort of Naive for Antiretrovirals) Study Group. Low frequency of severe hepatotoxicity and association with HCV coinfection in HIV-positive patients treated with HAART. J AIDS
9. Martinez E, Blanco JL, Arnaiz JA, Perez-Cuevas JB, Mocroft A, Cruceta A, et al
. Hepatotoxicity in HIV-1-infected patients receiving nevirapine-containing antiretroviral therapy. AIDS
10. Aceti A, Pasquazzi C, Zechini B, for the LIVERHAART Group. Hepatotoxicity development during antiretroviral therapy containing protease inhibitors in patients with HIV: the role of hepatitis B and C virus infection. J AIDS
11. Sulkowski MS, Thomas DL, Mehta SH, Chaisson RE, Moore RD. Hepatotoxicity associated with nevirapine- or efavirenz-containing antiretroviral therapy: role of hepatitis C and B infections. Hepatology
12. Wit FW, Weverling GJ, Weel J, Jurriaans S, Lange JM. Incidence of and risk factors for severe hepatotoxicity associated with antiretroviral combination therapy. J Infect Dis
13. Dore GJ, Cooper DA. Bridging the divide: global inequities in access to HIV/AIDS therapy. Med J Aust
14. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. Morb Mortal Wkly Rep
15. Lauer GM, Walker BD. Hepatitis C virus infection. N Engl J Med
16. AIDS Clinical Trials Group. Table for Grading Severity of Adult Adverse Experiences.
Rockville, MD: Division of AIDS, National Institute of Allergy and Infectious Diseases, 1992.
17. Centers for Disease Control and Prevention. Serious adverse events attributed to nevirapine regimens for postexposure prophylaxis after HIV exposures: worldwide, 1997–2000. JAMA
18. Palmon R, Koo BC, Shoultz DA, Dieterich DT. Lack of hepatotoxicity associated with nonnucleoside reverse transcriptase inhibitors. J AIDS
19. Rublein JC, Eron JJ Jr, Butts JD, Raasch RH. Discontinuation rates for protease inhibitor regimens containing ritonavir 600 mg versus ritonavir 400 mg plus saquinavir 400 mg. Ann Pharmacother
20. Arribas JR, Ibanez C, Ruiz-Antoran B, Pena JM, Esteban-Calvo C
, Frias J, et al
. Acute hepatitis in HIV-infected patients during ritonavir treatment. AIDS
21. Vora S, Michon C, Junet C, Balavoine JF, Renold-Moynier C, Yerly S, et al
. Switch from indinavir to ritonavir-indinavir regimen in patients treated with highly active antiretroviral therapy co-infected with hepatitis C is not associated with alteration of liver function tests. AIDS
22. John M, Flexman J, French MA. Hepatitis C virus-associated hepatitis following treatment of HIV-infected patients with HIV protease inhibitors: an immune restoration disease? AIDS
23. Lewden C, Raffi F, Cuzin L, Cailleton V, Vilde JL, Chene G, et al
. Factors associated with mortality in human immunodeficiency virus type 1–infected adults initiating protease inhibitor-containing therapy: role of education level and of early transaminase level elevation (APROCO-ANRS EP11 Study). J Infect Dis