Care of patients coinfected with HIV and hepatitis C virus: 2007 updated recommendations from the HCV–HIV International Panel

Abstract

Introduction Chronic hepatitis C (HCV) infection is currently one of the most clinically relevant comorbidities in the HIV population; overall, it affects one third of HIV-positive individuals [1]. Progression to end-stage liver disease occurs faster in coinfected patients [2–4] and decompensated cirrhosis is one of the main causes of hospitalization and death in this population [5–8]. However, the risk of hepatotoxicity using antiretroviral drugs is increased in subjects with underlying HCV infection [9,10]. Therefore, the optimal management of chronic HCV in HIV-positive patients is currently a priority. Several guidelines for caring for HCV infection in HIV-positive individuals have been released [11–15]. Because new and relevant information has recently appeared, it is convenient to update them. Eleven areas have been identified in which new recommendations are particularly needed: management of patients with persistently normal aminotransferases liver fibrosis assessment: when and how predictors of response to anti-HCV therapy in coinfected patients optimal dosages of pegylated interferon (pegIFN) and ribavirin (RBV) optimal duration of anti-HCV therapy treatment of non-responders and/or relapsers care of patients with end-stage liver disease treatment of acute HCV infection in HIV-infected individuals management of patients with multiple hepatitis viruses interactions between HCV medications and antiretroviral drugs hepatotoxicity of antiretroviral drugs. Patients with persistently normal aminotransferases The exact definition of persistently normal aminotransferases is not well established in patients with chronic HCV infection. Fluctuations in aspartate/alanine aminotransferases (AST/ALT) are frequent in HCV-related liver disease and differences in the prevalence of persistently normal ALT may reflect the length of follow-up and/or the number of biochemical determinations made [16–18]. We propose a definition requiring the demonstration of normal ALT in at least three consecutive tests made at least 2 months apart each, over a period of 12 months. One third of individuals who initially meet these criteria, however, may show ALT elevations as the period of observation extends [19–22]. Therefore, the characterization of patients with normal ALT should not be based on sporadic determinations of liver enzymes, and the term ‘asymptomatic’ or ‘healthy’ HCV carrier is inappropriate [22]. A further consideration is that the so-called ‘normal’ limit of aminotransferases has to be revisited, since recent studies have shown that aminotransferase levels in subjects without any liver injury [23] or in persons free of liver-related death on follow-up [24] are definitely lower than those accepted as normal in the past. The degree of aminotransferase elevation generally reflects the extent of liver inflammation. Around 25% of HCV-monoinfected patients show persistently normal ALT [19–22,25] and liver disease is generally less severe in this group [25–28]. Women tend to show more frequently persistently normal ALT than men [28], as well as subjects infected with HCV genotype 4 [29–31]. In contrast, patients with HCV genotype 3 show normal ALT less frequently [31]. As expected considering the immune-mediated nature of HCV-related liver disease, there is little correlation between serum HCV RNA and aminotransferases [32]. Few studies have been conducted so far in coinfected patients with normal ALT. Only 7–9% of this population show persistently normal liver enzymes [31,33]. Exposure to antiretroviral drugs, alcohol abuse and other conditions explain, on the one hand, the lower rate of normal ALT in HIV-positive patients with chronic HCV infection. On the other hand, significant liver fibrosis has been reported in up to 25–40% of coinfected patients with normal ALT [31,33], a prevalence higher than the 10–30% reported in HCV-monoinfected individuals [28,34]. In two recent studies, 12–14% of coinfected patients with normal ALT had cirrhosis on liver biopsy [33,35]. Since less than 15% of HCV-monoinfected individuals with minimal or absent liver fibrosis progress to cirrhosis within 15 years [36], and most patients with normal ALT have mild liver disease [37], these individuals have formerly not been considered for HCV therapy. Moreover, flares in ALT activity and lower treatment responses were reported in the past in patients with normal ALT exposed to IFN, which further discouraged their treatment. However, recent studies in HCV-monoinfected patients have alerted clinicians to the higher liver fibrosis progression in initially mild chronic HCV infection [38] and similar responses to pegIFN plus ribavirin RBV have been obtained in patients with normal than with elevated aminotransferases [39]. Recommendation Given that the prevalence of and progression to advanced liver fibrosis in patients with normal ALT is higher in HIV-positive patients [31,33], these patients should be considered for anti-HCV therapy. Treatment should be recommended based on patient's motivation, disease duration, fibrosis stage and virological profile regardless ALT levels [40]. Liver fibrosis assessment: when and how? The extent of hepatic fibrosis is the best prognostic factor of disease progression in patients with chronic HCV infection, and therefore it is worth considering this before initiating HCV therapy. Liver biopsy has been for many years the only tool to assess hepatic fibrosis. It has the advantage of providing additional information on other relevant histological findings, such as necroinflammation and steatosis. However, the development of non-invasive tools for staging hepatic fibrosis has been prompted by the several limitations of liver biopsy, such as its invasive nature, with occasional serious and even life-threatening complications [41]; sampling error owing to relatively small size and/or fragmentation of examined tissue [42] and/or to the inherent heterogeneity of hepatic fibrosis [43]; low acceptance by most patients; and relatively elevated cost [44]. Non-invasive procedures to assess liver fibrosis are currently divided into two major categories: imaging techniques, such as elastometry (FibroScan) [45–48] and serum biochemical markers (i.e., Fibrotest, APRI, SHASTA, FIB-4, Forn's index, etc.) [49–53]. These tools are generally accurate in discriminating between lack of fibrosis and advanced fibrosis but are less precise in distinguishing between intermediate fibrosis stages. Their predictive value is particularly good for advanced hepatic fibrosis and cirrhosis [54]. However, serum fibrosis markers are generally less reliable in coinfected patients, given the inflammatory nature of HIV disease and/or the frequent prescription of drugs in this population that may interfere with some fibrosis markers in the blood [55,56], as with bilirubin elevations in atazanavir therapy, gamma-glutamate transaminase abnormalities with non-nucleoside reverse transcriptase inhibitors, or cholesterol elevations associated with some protease inhibitors. In contrast, fibrosis staging using elastometry seems to be more reliable in this setting, avoiding such interference [48,57]. Elastometric measurements can be made in 10 min, be repeated periodically, are inexpensive and have more than 90% positive predictive value for advanced fibrosis [45–47]. When the diagnosis of a hepatic disease is clear by other means, as occurs with chronic HCV infection using virological markers (serum HCV RNA), the need for a liver biopsy to stage hepatic fibrosis and guide treatment decisions is currently no longer justified in most instances [58,59]. The higher response to pegIFN–RBV compared with that to standard IFN, the faster progression of HCV-related liver disease in the HIV setting and the chance to assess the virological response at earlier time-points to identify who will and who will not respond to therapy are all factors that allow the opportunity to prescribe HCV therapy to most patients while avoiding a liver biopsy [59]. The availability of easier means to assess liver fibrosis accurately has permitted this invasive procedure to be abandoned in most cases in routine clinical practice outside academic purposes. Moreover, these new tools have opened further opportunities to improve our knowledge of the natural history of HCV-related liver damage. Large cross-sectional and longitudinal studies have allowed recognition of (1) HCV genotype 3 as an independent predictor of accelerated liver fibrosis [60]; (2) different fibrosis thresholds in cirrhotic patients for developing distinct complications (e.g., esophageal varices, ascites or bleeding) [61]; and (3) of the possibility that severe liver fibrosis, including cirrhosis, can partially revert in at least a subset of patients who clear HCV after IFN therapy [62–64]. The information needed about hepatic fibrosis in chronic HCV infection is limited to that required to divide patients into those with and those without fibrosis (the latter group has not immediate need to be treated) and to recognize liver cirrhosis. Treatment is particularly needed for those with compensated cirrhotic disease; moreover, they should undergo periodic screening for esophageal varices and hepatocarcinoma, and overall are more prone to experience liver toxicity under antiretroviral therapy [65]. With this view, the distinction of histopathological stages of hepatic fibrosis based on a liver biopsy is currently unnecessary, avoiding the inherent problems derived from intra- and interobserver variations [66] and the other limitations mentioned above. Figure 1 summarizes the main variables that should be assessed before prescribing HCV therapy.Fig. 1: Main variables to assess in patients considered as candidates for hepatitis C (HCV) therapy. *Low viral load defined as HCV RNA < 500 000–800 000 IU/ml. Ab, antibody.Recommendation Information on liver fibrosis staging is important for therapeutic decisions in coinfected patients. However, a liver biopsy is not mandatory for considering the treatment of chronic HCV infection. A combination of non-invasive methods to assess liver fibrosis accurately predicts hepatic fibrosis in most cases. Predictors of response to hepatitis C therapy Baseline serum HCV RNA and HCV genotype are the main predictors of sustained virological response (SVR) to pegIFN–RBV in coinfected [11,14,67,68] as in HCV-monoinfected patients. Several other variables, however, may influence treatment responses, although generally to a lesser extent (Table 1). They can be grouped in three categories, determining a better outcome as follows: (1) host (younger age, non-black ethnicity, lower body mass index, lack of insulin resistance), (2) HCV status (elevated ALT, less advanced hepatic fibrosis), and (3) treatment schedule (optimal doses of pegIFN and/or RBV, enough length of therapy, good adherence). In addition, treatment outcomes could be better depending on some HIV variables, such as higher CD4 cell counts [69] or low HIV load, although it may just reflect a better tolerance of the anti-HCV medication in this subset of patients [70].Table 1: Factors associated with sustained virological response to HCV therapy.Particular attention has recently been paid to the negative impact of insulin resistance on HCV treatment response [71]. Insulin resistance is quite prevalent in coinfected patients at least in part because of the use of certain antiretroviral drugs [72,73]. Therefore, prevention of insulin resistance and/or its adequate management (even considering treatment with insulin-sensitizer agents when indicated) might improve HCV treatment outcomes in coinfected patients [74]. As in HCV-monoinfected patients, treatment adherence should be encouraged as much as possible. The ‘80/80/80’ rule is equally valid in coinfected patients, meaning that subjects who take more than 80% of pegIFN and of RBV doses during at least 80% of planned period of therapy respond significantly better than the rest [75]. Therefore, adequate selection of treatment candidates [76], psychological and/or psychiatric support [77] and use of growth factors to avoid dose reductions of either pegIFN and/or RBV [78,79] must all be encouraged in order to maintain adequate doses of anti-HCV medications in the majority of patients. The kinetics of HCV load in response to pegIFN–RBV is a reliable indicator of treatment efficacy. The availability of sensitive quantitative tools to closely monitor HCV decays under treatment has permitted the recognition of early time-points with high predictive value of SVR. Overall, the early virological response to HCV therapy divides patients into those sensitive and those refractory to therapy. Nearly 20% of HCV-monoinfected subjects do not show a significant reduction in HCV viremia (defined as a decline > 1 log IU/ml) during the first month of pegIFN–RBV [80], and this figure increases up to 30% in coinfected patients [81]. In virological responders, the best positive predictive value for SVR is achieved when a negative serum HCV RNA is attained at week 4 of therapy (rapid virological response, RVR), while the best negative predictive value for SVR is seen when HCV RNA falls < 2 log IU/ml at week 12 [67,68,82–86]. Higher baseline HCV RNA levels in coinfected patients compared with HCV-monoinfected individuals may explain why they achieve undetectable HCV viremia at week 4 less frequently and, therefore, achieve SVR less often [87]. Coinfected patients may show slower HCV decays on HCV therapy [88]. Interestingly, this could be overcome at least partially using higher RBV doses [81]. The so-called ‘2-log stopping rule’ refers to the strong predictive value of non-response at the week 12 assessment of virological response [80]. The failure to achieve HCV RNA declines > 2 log IU/ml (early virological response) at this time point permits the premature discontinuation of anti-HCV therapy, avoiding side effects and costs, when there is no chance of attaining the main goal of anti-HCV therapy, which is eradication of HCV infection. Fortunately, this rule works as well in coinfected as in HCV-monoinfected patients [67,68,82–86]. By comparison, a negative serum HCV RNA 6 months after completing anti-HCV therapy, which defines SVR, correlates with the long-term clearance of serum HCV as well as with histological and clinical improvements in most patients [89–91]. Therefore, ‘occult’ HCV infections with the potential worry of late HCV relapses are very rare. Recommendation The achievement of SVR can be predicted on the basis of negative serum HCV RNA at week 4 of therapy. On the other hand, a reduction < 2 log IU/ml in HCV RNA at week 12 and/or the presence of detectable viremia at week 24 both predict lack of SVR; accordingly these patients should be advised to stop prematurely anti-HCV therapy. Optimal dosages of pegylated interferon and ribavirin Adequate exposure to RBV is crucial to maximize responses to anti-HCV therapy [92–94]. Weight-based dosing seems well able to balance the highest efficacy and the lowest limiting toxicities of the drug, namely anemia. Pharmacokinetic studies have shown a good correlation between RBV plasma levels and HCV RNA responses [95,96]. Therefore, the use of fixed low doses of RBV (800 mg/day) in most trials conducted so far in coinfected patients could explain lower SVR [67,68,82–85,97–100]. The use of higher RBV doses (1000–1200 mg/day) in the PRESCO trial has confirmed this assumption, since the overall SVR in this trial (50%) is the highest reported so far in coinfected patients [101]. Figure 2 shows the proportion of patients achieving SVR in pivotal trials as a function of distinct doses of RBV and HIV status. Clearly, while HCV/HIV-coinfected patients may respond less, low RBV exposure may further impair treatment outcomes.Fig. 2: Proportion of patients with sustained virological response (SVR) in three different large trials in HIV-positive (pos) and HIV-negative (neg) patients using low or weight-based ribavirin (RBV) doses (intent-to-treat analysis).Optimal exposure to RBV could be particularly important in coinfected patients if the main mechanism of RBV action is hypermutagenesis [93,94,102,103]. Causing errors in the virus replication cycle, RBV activity should be maximized in HIV-positive individuals, in whom the immune-mediated effects of IFN are compromised. Moreover, the benefit of adequate RBV exposure might not be limited to patients infected with HCV genotypes 1–4 and may expand to genotype 3 [81]. In HCV-monoinfected individuals, a flat RBV dose of 800 mg/day is enough for genotype 3 [104], as long as therapy is provided for at least 24 weeks. However, shorter periods of therapy seem to require greater RBV doses in order to minimize relapses [105,106]. Anemia is the main drawback of increasing RBV dosing and may force a reduction in RBV dosage. When dose adjustments are made within the first weeks of therapy, reduced SVR may be expected [107], especially in patients with HCV genotypes 1–4. The use of zidovudine with pegIFN–RBV significantly increases the risk of developing severe anemia [108]. Therefore, when possible, zidovudine should be avoided and the use of erythropoetin should be encouraged in patients developing anemia under pegIFN–RBV in order to avoid the need for RBV dose reductions [78,79]. The efficacy of higher doses of pegIFN in coinfected patients has been explored in a few studies. In the CORAL-1 trial, the administration of 270 μg/week of pegIFN alpha-2a for the first 4 weeks did not improve the early virological response, whether measured as the proportion of patients with undetectable HCV load at week 4 or as reductions of > 2 log IU/ml HCV RNA at week 12, when compared with the administration of standard doses (180 μg/week) [109]. However, the size of the study population in that study was relatively small and nearly half the patients carried non-1 HCV genotypes. In contrast, data from studies conducted in HCV-monoinfected individuals suggest that there is a subset of patients who may benefit from exposure to higher doses of pegIFN [110] and this issue still warrants further investigation. Recommendation The current treatment of chronic HCV infection in HIV-positive persons should be pegIFN at standard doses plus weight-based RBV (1000 mg/day if < 75 kg and 1200 mg/day if > 5 kg). Optimal duration of therapy Studies conducted in HCV-monoinfected patients have shown that RVR, defined as undetectable HCV load at week 4, in patients treated with pegIFN–RBV may allow therapy to be shortened safely. Accordingly, treatment for only 12–16 weeks in patients with HCV genotype 3 [105,106] or for only 24 weeks in HCV genotype 1 [111,112] have been proposed for patients with RVR. The picture seems to be slightly different in coinfected patients. First, HCV load is generally higher in this population, which could explain why a smaller proportion reaches undetectable viremia at week 4 despite showing good early virological response [87]. Second, HCV clearance driven by IFN could be delayed in the HIV setting [86,88]. Third, the relapse rate upon completion of treatment might be increased in coinfected patients. This was shown to be the case for 24 weeks of therapy in HCV genotypes 2–3 in earlier trials [98,113]. For all these reasons, prior guidelines have recommended that duration of treatment in coinfected patients should be 48 weeks regardless HCV genotype [12,15]. It is important to note that the 2 log IU/ml rule at week 12 is also highly predictive of non-SVR in coinfected patients [12,15], which permits premature cessation of anti-HCV therapy when there is no chance of achieving a cure. Recent studies, however, have questioned these simple views to some extent. In a retrospective study conducted in coinfected patients with HCV genotypes 2–3, the subset who reached undetectable HCV RNA at week 4 could safely stop therapy at week 24, with minimal risk of relapse [114]. Similar findings have been reported in another recent Irish study [115]. However, a retrospective substudy of the APRICOT trial has shown that patients with HCV genotype 1 with low baseline HCV RNA and RVR obtained high rates of SVR (61%) and did not relapse [116], suggesting that shorten periods of therapy could have been enough in those patients. Overall, all these preliminary data encourage the provision of shorten periods of therapy on the basis of viral response at week 4, and clearly studies specifically designed to confirm this hypothesis in coinfected patients are needed. It might be the case that relapses could be limited to the subset of patients with high baseline HCV load and/or advanced fibrosis despite experiencing RVR, in whom 48 weeks of therapy would still be advisable [117]. In some patients with slow virological response, extended periods of treatment may permit SVR to be achieved [118]. Detectable viremia at week 4 seems to identify a subset of patients with genotypes 1–4 who may benefit from longer duration of therapy provided that it proves to be effective (> 2 log IU/ml fall in HCV RNA at week 12 by undetectable viremia at week However, the main with extended periods of therapy is This may be particularly in coinfected individuals, given that a tolerance of the medication has on outcomes in many trials trials designed to the efficacy of extended periods of therapy in coinfected patients without RVR therefore, this the information the that shorter periods of therapy could be advised in patients with HCV genotypes 2–3 with RVR, as long as HCV load is there is good there is not advanced hepatic fibrosis and weight-based RBV dosing is For the rest of the patients with HCV genotypes 2–3, 48 weeks of therapy could still be In patients with HCV genotypes of treatment 48 weeks could be recommended in the of RVR if the medication is well However, as high rates might limit the benefit of this optimal duration of hepatitis C (HCV) therapy in HCV/HIV-coinfected patients. patients with baseline low viral load and minimal liver fibrosis. The current treatment of chronic HCV infection in HIV-positive persons should be pegIFN plus weight-based RBV for 48 weeks. Patients infected with HCV genotype 2–3 and RVR could benefit from shorter of therapy. In contrast, of HCV genotypes 1 and 4 with early virological response but not RVR might benefit from extended of therapy. Treatment of non-responders and/or relapsers A number of coinfected patients have been exposed to without achieving SVR. These patients to be at risk for progression to end-stage liver disease, including the development of 2 summarizes the distinct these of which may require a distinct 2: of and for patients coinfected with hepatitis C and HIV who are to prior prior (i.e., shorter duration, low RBV with standard who prematurely therapy owing to side or who were to the medication are not treatment and could be better considered as patients. This group should be to current in The data that subjects who a prior of therapy may achieve but lower SVR rates than patients Overall, the for treatment response in patients are higher as the efficacy of the was lower the efficacy of pegIFN–RBV also on whether it was virological non-response or relapse to the prior and overall responses tend to be better in the should be defined as patients who an optimal of therapy with pegIFN–RBV at doses and to achieve early virological response (> 2 log IU/ml decline in HCV RNA at week or undetectable HCV load at week are patients who HCV after stopping a of therapy. of any of these patients permits SVR to be although relapsers might benefit more than virological of treatment and/or use of higher than recommended doses of HCV medications may slightly improve response rates although SVR will When the achievement of SVR is not the goal of therapy may be to or progression of liver Interestingly, hepatic fibrosis in of coinfected patients despite not attaining SVR Since liver in these studies were at baseline and after completion of it should be that this benefit most will as time as in studies conducted both in HCV-monoinfected and in coinfected patients for longer periods of However, these that histological improvements while on HCV therapy or reflect the effects of IFN and the for whether therapy with pegIFN could liver disease when HCV eradication is not Large trials in HCV-monoinfected and coinfected virological non-responders are currently under to this hypothesis Several should be in with to these trials in coinfected patients. First, it is not whether CD4 cell and effects of IFN may be in these patients when the is provided for long Second, side effects and of will limit the administration of pegIFN in a proportion of these patients. drugs HCV are particularly for the large and of coinfected patients who to clear HCV with the current progression of liver disease will shorten their in the of to those 3 summarizes the main anti-HCV drugs in the the efficacy and of these drugs in coinfected patients should be without for the of trials conducted in HCV-monoinfected With the of should encourage these anti-HCV in and relapsers to prior of HCV therapy are a population and therapeutic in should be of end-stage liver disease The management of coinfected persons with advanced liver cirrhosis is They should be for staging of liver disease and management of liver-related complications such as ascites and Because of an increased risk of life-threatening complications during pegIFN–RBV therapy, persons with hepatic are not candidates for therapy to liver is therapy may significantly improve and outcomes in HIV-positive patients with hepatic and, therefore, should not be However, the effective treatment of HIV in persons with advanced cirrhosis may be owing to in hepatic of antiretroviral drugs and risk of liver injury this liver is the treatment for coinfected patients with stage or C cirrhosis (Table In a recent study 24 HIV-positive was similar to that and HIV-negative 12, 24 and months after liver rates were and HIV-positive patients and and HIV-negative patients. However, when only patients were there was an significant in coinfected compared with HCV-monoinfected The rates at 2 and 3 years were and in HCV-monoinfected subjects and and in coinfected patients. Factors associated with were to CD4 cell counts < detectable plasma HIV RNA and HCV infection for liver in HIV-infected patients with end-stage liver HIV infection should no longer be considered a to liver However, coinfected patients and highly problems including HCV infection and interactions between agents and protease Accordingly, liver in