点击显示 收起
Clinical HIV Research Unit, University of the Witwatersrand, Johannesburg
Department of Medicine, University of Cape Town, Cape Town, South Africa
Duke University Medical Center
Gilead Sciences, Inc., Durham, North Carolina
Case Western Reserve University, University Hospital of Cleveland, Cleveland, Ohio
California Pacific Medical Center, San Francisco
Human immunodeficiency virus (HIV)infected South African patients (n = 468) received blinded lamivudine or emtricitabine, stavudine, and either nevirapine or efavirenz (based on screening viral load). Baseline characteristics were analyzed in univariate and multivariate regression, to identify risk factors for hepatotoxicity (grade 3 or greater increase in serum aminotransferase levels). The occurrence of early hepatotoxicity was 17% in the nevirapine group and 0% in the efavirenz group and was balanced between the lamivudine and emtricitabine arms. Two subjects died of hepatic failure. Independent risk factors were body-mass index (BMI) <18.5, female sex, serum albumin level <35 g/L, mean corpuscular volume >85 fL, plasma HIV-1 RNA load <20,000 copies/mL, aspartate aminotransferase level <75 IU/L, and lactate dehydrogenase level <164 IU/L. The use of nevirapine in female patients with a low BMI should be discouraged.
The implementation of combination antiretroviral therapies has transformed the prognosis of HIV infection during the past decade [1]. Nonnucleoside reverse transcriptase inhibitors (NNRTIs) have been shown to be sometimes superior alternatives to protease inhibitors and have become popular as first-line therapies [2, 3]. However, the drugs in this class have been associated with early adverse effects [2, 4].
Nevirapine (Viramune; Boehringer Ingelheim) is an NNRTI used in first-line treatment regimens, and it has also shown efficacy in decreasing the vertical transmission of HIV-1 [2, 3]. The acute hepatic toxicity of nevirapine has led to a change in the product label in both the United States and the European Union [5]. In a large randomized, prospective controlled trial that included nevirapine as part of the antiretroviral regimen, we observed a high frequency of hepatotoxicity [6].
Objectives.
The objectives of these analyses were to identify risk factors and symptoms associated with the emergence of early hepatotoxicity (hereafter, "hepatotoxicity") and to characterize the effects of treatment interruption on hepatotoxicity. Hepatotoxicity was defined as treatment-emergent grade 3 (510 times the upper limit of normal ) or grade 4 (>10 times the ULN) increases in serum aminotransferase (alanine aminotransferase or aspartate aminotransferase ) levels. "Early" was defined as within the first 12 weeks of therapy.
Study population.
HIV-1infected men and nonpregnant women were recruited in South Africa and were enrolled if they had plasma HIV-1 RNA levels >5000 copies/mL and CD4+ cell counts 200 cells/mm3 and were antiretroviral therapy naive [6]. Subjects were stratified to receive either nevirapine or efavirenz on the basis of screening plasma HIV-1 RNA levels below or above 105 copies/mL, respectively. All subjects provided written, informed consent before being screened for study eligibility. The protocol was reviewed and approved by independent ethics committees and health authorities. The randomized, double-blind, double-dummy component of the multicenter trial compared the safety and efficacy of emtricitabine (200 mg once daily, Emtriva; Gilead Sciences) versus lamivudine (150 mg twice daily, Epivir; GlaxoSmithKline) in combination with stavudine (30 or 40 mg twice daily, based on body weight, Zerit; Bristol-Myers Squibb) and either nevirapine (200 mg once daily for 2 weeks followed by 200 mg twice daily, Viramune) or efavirenz (600 mg once daily, Sustiva; Bristol-Myers Squibb). Subjects were evaluated clinically at weeks 1, 2, 4, 8, and 12 and by laboratory evaluations at weeks 4, 8, and 12 and every 4 weeks thereafter. Hepatitis C antibody testing was performed retrospectively, but only in subjects with hepatotoxicity. Hepatitis B screening was performed prospectively. Because of a high occurrence of hepatotoxicity and the development of liver failure in 2 subjects, health authorities recommended the termination of enrollment in the nevirapine stratum and placed the clinical trial on hold.
Statistical methods.
Demographics and disease characteristics were compared by use of Fisher's exact and Wilcoxon rank sum tests. Univariate and multivariate logistic-regression models were retrospectively used to assess possible prognostic factors associated with hepatotoxicity occurring within 12 weeks of the introduction of nevirapine. Stepwise regression analyses were performed to assess which factors, when combined, were significantly associated with the occurrence of hepatotoxicity. In addition, continuous baseline factors were evaluated and dichotomized on the basis of their associations with hepatotoxicity; these were used as factors in the stepwise regression procedure. Factors tested in the stepwise regression evaluation included those that were considered to be potentially clinically relevant and those found to be significantly associated with hepatotoxicity according to the results of the univariate analysis. Time to event was compared between risk groups by use of Kaplan-Meier estimates and log-rank test.
Patient characteristics.
A total of 468 subjects were enrolled in the study between August 1999 and February 2000; 234 subjects were randomized to receive emtricitabine and 234 were randomized to receive lamivudine. Eighty-two percent of subjects (385/468) were enrolled in the nevirapine group. The baseline demographic and clinical characteristics of the 385 subjects who received nevirapine are described in table 1.
Characteristics of hepatotoxicity.
Overall, hepatotoxicity occurred in 66 of 468 subjects, or 14% of the subjects who received at least 1 dose of blinded study medication. The frequency of hepatotoxicity was 17% (66/385) in the nevirapine group and 0% (0/83) in the efavirenz group (P < .001). The frequency in men who received nevirapine was 12.8% (20 subjects), and that in women was 20.1% (46 subjects). In the nevirapine group, the occurrence of hepatotoxicity was comparable between subjects treated with emtricitabine (15%) and those treated with lamivudine (19%) (P = .28). The onset of hepatotoxicity (median, day 29; range, days 19426) was within the first 12 weeks of therapy in 80% (53/66) of subjects.
In an effort to further characterize this early hepatotoxicity, we restricted subsequent analyses to the subjects enrolled in the nevirapine group and who were diagnosed with hepatotoxicity within 12 weeks of treatment initiation. There were 23 cases among 194 emtricitabine-treated subjects (11.9%), compared with 30 among 191 lamivudine-treated subjects (15.7%) (P = .30).
Univariate analysis.
Table 1 describes the baseline demographic and clinical characteristics of subjects with or without hepatotoxicity who were enrolled in the nevirapine group. No correlation with chronic hepatitis status was observed.
Multiple regression analysis.
In the stepwise regression analysis, female sex, body-mass index (BMI) <18.5, serum albumin level <35 g/L, mean corpuscular volume (MCV) >85 fL, plasma HIV-1 RNA load <20,000 copies/mL, lactate dehydrogenase (LDH) level <164 IU/L, and AST level <75 IU/L were independently associated with the subsequent development of hepatotoxicity (table 1). After adjustment for these independent factors, the proportions of subjects with or without hepatotoxicity between the lamivudine and the emtricitabine arms remained similar (P = .3). Women with a BMI <18.5 had a 50% probability of developing hepatotoxicity, and those with a BMI 18.5 had a 17% probability; men with a BMI <18.5 had a 15% probability, and those with a BMI 18.5 had a 7% probability.
Outcomes.
Of the 53 subjects who developed hepatoxicity, 15 (28%) permanently discontinued therapy within a median time of 3.5 days. Fulminant hepatitis leading to hepatic failure and death occurred in 2 women (at weeks 4 and 5) who had received nevirapine. Both women had normal ALT levels at baseline and no hepatitis coinfection. In 18 subjects (34%), hepatotoxicity resolved while they were still receiving nevirapine. The median time to return to a grade 2 value for ALT/AST levels was 36.5 days (range, 1093 days) for subjects who continued to receive nevirapine and 19 days for those who discontinued therapy (range, 4163 days) (P = .03, Wilcoxon rank sum test).
Discussion.
Although the mechanisms of antiretroviral-induced hepatotoxicity are not well understood and are probably multifactorial, our data demonstrate a high risk (17%) of early hepatotoxicity associated with the use of nevirapine [4, 7]. The specific study population, which consisted mainly of African women with relatively high baseline CD4+ cell counts (mean, 398 cells/mm3), may have increased the likelihood of hepatotoxicity. These observations are important, because nevirapine is commonly used as a component of first-line therapy in resource-limited settings, especially in women [8]. Nevirapine-associated hepatotoxicity was observed with chronic dosing and is not associated, to our knowledge, with single-dose mother-to-childtransmission prophylaxis. The lack of hepatotoxicity in the efavirenz group and the balanced incidence between the emtricitabine and lamivudine groups strongly suggests that the hepatotoxic agent was nevirapine. Previous studies have identified an association of nevirapine with more-frequent hepatotoxic events, compared with other approved antiretroviral agents [4, 9]. An observed lower incidence of hepatotoxicity in previously reported trials of nevirapine may be explained by the limited number of women studied, a 3-month delay in the first ALT or AST measurement, the discontinuation of treatment in subjects presenting with a grade 2 rash (diffuse maculopapular rash), or variations in the definitions of hepatotoxicity. Interestingly, the frequency of hepatotoxicity after week 12 in our study was low (3%).
During the course of our study, 2 black, nonpregnant women who had a rash and severe hepatitis died 45 weeks after drug initiation, despite prompt drug withdrawal. Therefore, weekly monitoring of hepatic enzymes after the initiation of nevirapine therapy is needed to detect potentially reversible toxicities earlier than were identified by our monitoring scheme. Indeed, the first measurement of enzymes levels in our study was at week 4, which possibly delayed the recognition of hepatotoxicity. Rash, nausea, and fever were observed significantly more frequently in subjects with hepatotoxicity than in those without. Physicians should instruct subjects starting nevirapine to promptly report such symptoms, and consideration should be given to closer monitoring of such subjects. Drug discontinuation in our study was associated with a faster recovery from hepatotoxicity.
The higher MCV observed in subjects with hepatotoxicity in our study may have been associated with alcohol consumption and/or nutritional deficiency. Nevirapine-induced liver toxicity may be mediated by an immunoallergic reaction to the parent drug or a reactive metabolite favored by a relative overexposure of selected subjects, as has been observed with halothane, phenytoin, and sulfamethoxazole [7]. In favor of an immunoallergic mechanism is the association of hepatotoxicity, rash, fever, and an increase in eosinophil counts. Other researchers have reported an increased occurrence of nevirapine-associated rash among women, possibly through a comparable mechanism [10].
In our study, having a BMI <18.5 was an independent risk factor for the emergence of hepatotoxicity. This factor could, in turn, be responsible for a relative overexposure to nevirapine or one of its metabolites, thereby triggering toxicity. This is plausible, given that previous reports have shown a link between the concentration of nevirapine and the risk of toxicity [11, 12]. Another report demonstrated decreased clearance of nevirapine in subjects with lower weight or who were of black race [13]. It is also possible that a lower BMI reflects an overall poor nutritional status. Because nevirapine is extensively used in resource-limited settings, with high proportions of women and the potential for low BMI, prescriber awareness of these issues is essential. Indeed, in our study, women who had a BMI <18.5 had a 50% risk of developing hepatotoxicity. The clinical significance of lower levels of AST and LDH as predictors of toxicity remains unclear.
In a recently published analysis of nevirapine-associated hepatotoxicity, sex-dependent CD4+ cell count was found to be predictive of hepatotoxicity [14]. From this analysis, men who had a CD4+ cell count >400 cells/mm3 and women who had a CD4+ cell count >250 cells/mm3 are considered to be at greater risk for hepatotoxicity [15]. The correlation between the occurrence of hepatotoxicity and high CD4+ cell counts is consistent with an immunoallergic mechanism. Within our data set, these indices were not identified as independent predictors, possibly because of the exclusion of subjects with CD4+ cell counts <200 cells/mm3. However, it is possible that the viral load, in our study, stands as a surrogate indicator of the immune status of the subjects.
In the present study, subjects who initiated antiretroviral therapy with nevirapine had a high occurrence of hepatotoxicity within the first 12 weeks of therapy. The risk was higher in women who had a BMI <18.5. Clinicians should consider avoiding the administration of nevirapine in such subjects. Weekly monitoring of ALT/AST levels during the first months of therapy may identify early, and potentially reversible, drug-induced hepatotoxicity. The appearance of a rash, nausea, or fever during the first 12 weeks of therapy should trigger closer monitoring.
Acknowledgments
We acknowledge the contributions of study subjects and the staff who participated in the emtricitabine-302 study and thank Boehringer Ingelheim, Bristol-Myers Squibb, and DuPont-Pharma for providing the study drugs.
References
1. Centers for Disease Control and Prevention. New data show continued decline in AIDS deaths. 31 August 1999. Available at: http://www.cdc.gov/od/oc/media/pressrel/r99083/htm (Accessed 7 February 2005). First citation in article
2. Podzamcer D, Fumero E. The role of nevirapine in the treatment of HIV-1 disease. Expert Opin Pharmacother 2001; 2:206578. First citation in article
3. Harris M, Montaner JSG. Clinical uses of non-nucleoside reverse transcriptase inhibitors. Rev Med Virol 2000; 10:21729. First citation in article
4. Wit FWNM, Weverling GJ, Weel J, Jurriaans S, Lange JMA. Incidence and risk factors for severe hepatotoxicity associated with antiretroviral therapy. J Infect Dis 2002; 186:2331. First citation in article
5. Serious adverse events attributed to nevirapine regimens for post exposure prophylaxis after HIV exposuresworldwide, 19972000. MMWR Morb Mortal Wkly Rep 2001; 49:11536. First citation in article
6. Sanne I. Severe liver toxicity in patients receiving two nucleoside analogues and a non-nucleoside reverse transcriptase inhibitor [abstract PL 9.3]. AIDS 2000; 14(Suppl 4):S12. First citation in article
7. Lee WM. Drug-induced hepatotoxicity. N Engl J Med 2003; 349:47485. First citation in article
8. World Health Organization (WHO), HIV/AIDS Department. Scaling up antiretroviral therapy in resource-limited settings: guidelines for a public health approach: executive summary. Geneva: WHO, 2002. First citation in article
9. 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 2002; 35:1829. First citation in article
10. Bersoff-Matcha SJ, Miller WC, Aberg JA, et al. Sex differences in nevirapine rash. Clin Infect Dis 2001; 32:1249. First citation in article
11. de Requena DG, Nunez M, Jimenez-Nacher I, Soriano V. Liver toxicity associated with nevirapine. AIDS 2002; 16:2901. First citation in article
12. De Maat MM, Ter Heine R, Mulder JW, et al. Incidence and risk factors for nevirapine-associated rash. Eur J Clin Pharmacol 2003; 59:45762. First citation in article
13. De Maat MM, Huitema AD, Mulder JW, Meenhorst PL, Van Gorp EC, Beijen JH. Population pharmacokinetics of nevirapine in an unselected cohort of HIV-1 infected individuals, Br J Clin Pharmacol 2002; 54:37885. First citation in article
14. Stern JO, Robinson PA, Love J, Lanes S, Imperiale MS, Mayers DL. A comprehensive hepatic safety analysis of nevirapine in different populations of HIV infected patients. J Acquir Immune Defic Syndr 2003; 34(Suppl 1):S2133. First citation in article
15. Viramune tablets and oral suspension, US package insert. Ingelheim, German: Boehringer Ingelheim. First citation in article