| |
HCV Genotype 3 Associated with ART Hepatoxicity, Perhaps Due to Fatty Liver
|
| |
| |
Influence of Genotype 3 Hepatitis C Coinfection on Liver Enzyme Elevation in HIV-1-Positive Patients After Commencement of a New Highly Active Antiretroviral Regimen: Results From the EPOKA-MASTER Cohort
JAIDS Journal of Acquired Immune Deficiency Syndromes: Volume 41(2) 1 February 2006 pp 180-185
".....HIV-positive patients coinfected with HCV genotype 3 displayed a higher risk of relevant hepatotoxicity (than other HCV genotypes), independently from other clinical variables. The impact of HCV genotype outweighed the role of drugs in determining hepatotoxicity.....neither use of triple-NRTI regimens or PI-containing regimens (either boosted or not boosted with ritonavir) were shown to be associated with hepatotoxicity considering use of NNRTI regimens as reference category....Previous studies correlated HCV-3 infection with liver steatosis (fatty liver) in patients not infected with HIV...liver steatosis has been correlated with higher grade of fibrosis among HCV-monoinfected and among HIV/HCV-coinfected patients and with greater necroinflammatory activity. Because advanced liver histologic stage was correlated with HAART-associated hepatotoxicity in patients coinfected with HIV and HCV, it can be speculated that liver steatosis induced by HCV-3 could promote fibrogenesis, thus rendering liver more susceptible to drug injury. Therefore, liver damage correlated with steatosis may increase susceptibility to cumulative toxicity induced by drugs, either direct or mediated by proinflammatory cytokines..."
BACKGROUND
Highly active antiretroviral therapy (HAART) is associated with a number of serious and potentially life-threatening adverse events, including liver transaminase elevation or liver injury (ie, hepatotoxicity); this is one of the most frequent reasons for treatment withdrawal in HIV-infected patients with hepatitis C virus (HCV) coinfection.1 Numerous factors have been implicated as potentially dangerous in this respect. Previous studies have demonstrated that there is an association between coinfection with HCV and the occurrence of hepatotoxicity in HIV-infected patients initiating HAART.2-9
Moreover, among HCV genotypes, genotype 3 (HCV-3) has been suggested to play a role. Although this evidence is of potential interest for risk assessment, the role of HCV-3 in the elevation of transaminases following HAART is still debated. Only 2 reports exist so far, providing conflicting results. A first report by Nunez et al10 indicated that HCV-3 is an independent risk factor for liver enzyme elevation following HAART in a small cohort of HIV/HCV-coinfected subjects. By contrast, in another study,11 HCV-3 was not significantly associated with acute liver enzyme elevation; unfortunately the number of patients was too small, precluding multivariate analysis. The objective of our study was to assess the impact of HCV-3 on liver enzyme elevation in a large population of patients, using a multivariable model that could account for possible impact exerted by other factors, including the impact of modern antiretroviral (ARV) regimens.
Authors- Torti, Carlo MD*; Lapadula, Giuseppe MD*; Puoti, Massimo MD*; Casari, Salvatore MD*; Uccelli, Maria Cristina MD*; Cristini, Graziella MD*; Bella, Daniele MD*; Pastore, Giuseppe MD ; Ladisa, Nicoletta MD ; Minoli, Lorenzo MD ; Sotgiu, Giovanni MD ; Caputo, Sergio Lo MD ; Bonora, Stefano MD ; Carosi, Giampiero MD*
From the *Istituto di Malattie Infettive e Tropicali, Universita di Brescia, Brescia; Clinica di Malattie Infettive, Policlinico di Bari, Bari; Istituto di clinica delle Malattie Infettive, IRCCS S. Matteo, Pavia; Divisione di Malattie Infettive, Ospedale SM Annunziata, Firenze; and Istituto di Malattie Infettive, Ospedale Amedeo di Savoia, Torino, Italy.
Abstract
Background: The independent role of hepatitis C virus (HCV) genotype 3 in liver transaminase elevation following highly active antiretroviral regimens is still controversial.
Methods: Analysis of data from a cohort of 492 HIV/HCV-coinfected patients was conducted using an intention-to-treat approach. Incidence of grade ≥III liver transaminase elevation was estimated per 100 patient-years of follow-up. Univariate and multiple proportional hazards regression analysis of factors that may predict liver enzyme elevation was performed.
Results: The incidence of grade ≥III hepatotoxicity was 25 per 100 patient-years among patients coinfected with HCV genotype 3 and 11 per 100 patient-years among those with other genotypes. On multiple proportional hazard regression analysis, time-to-grade ≥III liver enzyme elevation was directly correlated with HCV genotype 3 (hazards ratio [HR]: 2.0, 95% CI: 1.3 to 2.9; P = 0.001), male gender (HR: 2.7; 95% CI: 1.3 to 5.7; P = 0.007), chronic hepatitis B virus infection (HR: 2.9, 95% CI: 1.5 to 5.9; P = 0.002), and alanine aminotransferase level at baseline (per 10 IU/L HR: 1.10; 95% CI: 1.06 to 1.15; P < 0.001). In the same model, higher CD4+ T-cell counts at baseline were inversely correlated with risk of hepatotoxicity (HR: 0.998; 95% CI: 0.997 to 0.999; P = 0.036). Moreover, among patients experienced to antiretroviral drugs, previous grade ≥III hepatotoxicity (HR: 2.8; 95% CI: 1.8 to 4.3; P < 0.001) was an adjunctive independent risk factor.
Conclusions: HIV-positive patients coinfected with HCV genotype 3 displayed a higher risk of relevant hepatotoxicity, independently from other clinical variables. The impact of HCV genotype outweighed the role of drugs in determining hepatotoxicity.
Incidence of Hepatotoxicity by HCV Genotype
In the ITT model, 100 patients showed grade ≥III liver enzyme elevation; thus the overall incidence was 14 per 100 patient-years of follow-up. The incidence of grade ≥III was 25 per 100 patient-years among patients coinfected with HCV-3 and 11 per 100 patient-years among those infected with other genotypes. Eighteen patients showed grade IV hepatotoxicity, with an incidence of 2 per 100 patient-years. Ratio of grade IV hepatotoxicity was 4 per 100 patient-years among patients coinfected with HCV-3 and 2 per 100 patient-years among those infected with other genotypes. Among 492 patients, 28 (6%) experienced grade ≥III liver enzyme elevation within the first 12 weeks, with an overall prevalence of 8% (11/132) among patients coinfected with HCV-3 and 5% (17/460) among those coinfected with an HCV genotype other than 3. Difference between groups was evident also at later timepoints. Among patients coinfected with HCV-3, 28 of 132 (21%) experienced grade ≥III liver enzyme elevation by 48 weeks of follow-up, in contrast to 40 of 360 (11%) of those coinfected with genotypes other than 3. Importantly, survival analysis confirmed the higher risk of liver enzyme elevation in patients coinfected with HCV-3 (Fig. 1; P < 0.001).
Patients infected with HCV-3 had mean values of ALT (P < 0.001) and AST (P < 0.001) enzymes significantly higher than those found in patients infected with different genotypes, both at baseline and during the follow-up, both in the ITT (Fig. 2) and in the OT analysis.
Predictors of Relevant Hepatotoxicity
Univariate model conducted by the ITT database showed that male gender, HCV-3, chronic HBV infection, previous exposure to ARV drugs (yes vs. no), higher baseline ALT, higher baseline AST, and lower baseline CD4+ T-cell count were associated with the risk of hepatotoxicity with a P value ≦0.20, such that they were imputed in the multiple proportional hazards regression analysis. In this model the following factors were associated with a higher risk of hepatotoxicity: male gender (P = 0.007), HCV-3 (P = 0.001), presence of chronic HBV infection (P = 0.002), higher baseline ALT (P < 0.001), and higher CD4 T-cell count at baseline (P = 0.036) (Table 2). Interestingly, when a separate Cox analysis was conducted in patients experienced to ARV, previous grade ≥III hepatotoxicity was shown to be an adjunctive independent risk factor (hazards ratio [HR]: 2.8; 95% CI: 1.8 to 4.3; P < 0.001).
OT analyses confirmed these results, providing the opportunity to test for the possible impact of types of prescribed ARV regimens on the risk of hepatotoxicity. Although HCV-3 continued to be an independent risk factor (HR: 1.9; 95% CI: 1.2 to 3.1; P = 0.006), neither use of triple-NRTI regimens or PI-containing regimens (either boosted or not boosted with ritonavir) were shown to be associated with hepatotoxicity considering use of NNRTI regimens as reference category (HR: 1.0; 95% CI: 0.5 to 2.0; P = 0.993 and HR: 0.9; 95% CI: 0.5 to 1.4; P = 0.60, respectively).
Association of Hepatotoxicity With Genotypes Other Than 3
To explore possible association of other HCV genotypes, in particular 2 and 4, with hepatotoxicity, a supplementary analysis was performed comparing the risk of hepatotoxicity among 480 patients infected by a single HCV genotype. In comparison with HCV-3, the presence of either HCV genotype 1 (HR: 0.6; 95% CI: 0.4 to 0.8; P = 0.006) or HCV genotype 4 (HR: 0.2; 95% CI: 0.1 to 0.4; P < 0.001) was shown to be significantly associated with a lower risk of hepatotoxicity. Even the presence of HCV genotype 2 tended to be protective, but the association was not statistically significant (HR: 0.2; 95% CI: 0.03 to 1.7; P = 0.148). The same analysis was performed with HCV genotype 1 as reference category, to reveal possible differences among HCV genotypes different from 3. In this model, HCV genotype 4 was shown to be protective (HR: 0.3; 95% CI: 0.1 to 0.8; P = 0.014) compared with genotype 1, whereas the difference between genotype 2 and 1 did not reach statistical significance (HR: 0.4; 95% CI: 0.1 to 3.0; P = 0.381). These results were also confirmed at multivariable hazard regression analysis.
DISCUSSION
The only 2 studies available10,11 conflict on the issue of how HCV genotype might influence the occurrence of liver transaminase elevation after HAART. Inconsistency between the 2 studies could be due to the small number of patients, which therefore reduces the statistical power so that reliable multivariate analysis could not be performed. Our study was conducted in a cohort comprising a higher number of patients infected by different HCV genotypes, confirming the association between HCV-3 coinfection and occurrence of transaminase elevation during HAART, as previously reported.10 Importantly, the apparent impact of HCV genotype outweighed the role of drugs in determining hepatotoxicity.
It is still unclear why HCV-3 may represent an additional risk for developing relevant transaminase elevation during HAART. Previous studies correlated HCV-3 infection with liver steatosis in patients not infected with HIV,13-15 probably due to a virus-specific cytopathic effect.14,16,17 In its turn, liver steatosis has been correlated with higher grade of fibrosis among HCV-monoinfected13,18-20 and among HIV/HCV-coinfected patients21 and with greater necroinflammatory activity.13,21 Because advanced liver histologic stage was correlated with HAART-associated hepatotoxicity in patients coinfected with HIV and HCV,22 it can be speculated that liver steatosis induced by HCV-3 could promote fibrogenesis, thus rendering liver more susceptible to drug injury. Therefore, liver damage correlated with steatosis may increase susceptibility to cumulative toxicity induced by drugs, either direct or mediated by proinflammatory cytokines.23 The demonstration of events of liver toxicity not restricted to the first weeks of therapy in the present study would support a hypothetical mechanism for this synergism.
Specific immune response was demonstrated to be more prominent against HCV-3 and was probably implicated in better virologic response rate to interferon-based antiviral therapy observed with this genotype, although a cause rather than an effect relationship has not been demonstrated.24-26 Therefore, our observation may be consistent with an immune restoration disease more prominent in HCV-3-coinfected patients. A higher CD4 T-cell count at baseline was associated with a lower risk of hepatotoxicity, possibly indicating that risk of hepatotoxicity is due to a restoration of immune response in those who are deficient. Altogether, these findings suggest that HCV genotype (especially HCV-3) may influence specific immune response, which could be restored by HAART and could be implicated in determining hepatotoxicity. This hypothesis requires further pathogenic investigations.
The information provided may be clinically relevant and helpful for the therapeutic management of disease in coinfected patients, because the impact of HCV genotype outweighed the role of drugs in determining hepatotoxicity. The absence of an effect of NNRTI, including nevirapine, is somehow unexpected. Analysis of the original EPOKA-A cohort demonstrated an NNRTI class effect but only in HAART-experienced patients, probably because treatment options were restricted by the emergence of HIV drug resistance, whereas, in naive patients, NNRTI regimens (including nevirapine) were preferentially used in those with a low risk for hepatotoxicity.27 The present analysis was conducted in a subset of mixed (drug-naive and experienced) patients, thus reducing statistical power so that the impact of nevirapine could not be detected. Conversely, the absence of an effect of PIs (either single or boosted) was not unexpected.27,28 To our knowledge, the only strong influence was exerted by use of full-dose ritonavir,5 which was never prescribed in our cohort.
The independent effect of other factors in determining hepatotoxicity, such as chronic infection with HBV, elevated ALT at baseline, and previous grade ≥III hepatotoxicity, is well recognized and is discussed elsewhere.29 By contrast, the possible role of male gender was demonstrated by some authors,30 but not by others,2,3,6 depending on heterogeneity of the cohorts studied.
Finally, our study suggested, for the first time, a lower risk of hepatotoxicity related to HCV genotype 4 compared with genotype 1, and this may require further pathogenic and clinical studies.
Possible limitations of this study are related to its observational nature. In particular, several variables (especially alcohol abuse) that were not reliably detected in the cohort overall should be further evaluated. Moreover, patient differences, in particular baseline liver enzyme levels and CD4 T-cell count at nadir, could have influenced the results, even after checking for the different impact of these variables through multivariate analysis.
In conclusion, a cautious approach and strict monitoring should be applied in patients infected by HCV-3, especially in men, coinfected by HBV, with high baseline ALT values, low baseline CD4 T-cell count, and suffering from previous hepatotoxicity. HCV genotyping before initiating HAART may therefore be a useful tool for assessing the risk of transaminase flare in the course of therapy. Given that HCV-3 has the best response rate after treatment with pegylated interferon plus ribavirin, HCV treatment may be particularly advisable in these patients before starting HAART, preventing most hepatotoxicity events. In light of our observation, we believe that studies on hepatotoxicity in HIV/HCV-coinfected patients should include HCV genotype as covariate. This may also explain different prevalence of hepatotoxicity detected among different cohorts.
|
|
| |
| |
|
|
|
