| |
Drug-Induced Liver Injury Associated with Antiretroviral Therapy that Includes HIV-1 Protease Inhibitors
|
| |
| |
A Review of Data From Studies on Hepatotoxicity (elevations in liver enzyme tests) For patients on Protease Inhibitor Regimens
Clinical Infectious Diseases 2004;38:S90-S97
Mark S. Sulkowski
Johns Hopkins University School of Medicine, Baltimore, Maryland
The present supplement in CID summarizes presentations made at a special conference sponsored by Boehringer Ingelheim International entitled "Hepatic Safety and HAART." It took place in Santa Barbara, California, on 79 June 2002.
The authors say: the data derived from clinical trials and cohort studies indicate that elevations in serum aminotransferase levels are relatively common in HIV-infected patients receiving PI-based ART. Among individual PIs, full-dose ritonavir is the only specific drug that has been identified as an independent risk factor for the development of grade 3 and/or 4 severe HAART-related liver injury. In addition, these studies also indicated that HBV or HCV coinfection substantially increases the risk of developing significant liver enzyme elevationss after the initiation of HAART. However, most studies have shown that the great majority of HCV-coinfected patients treated with HAART do not develop severe hepatotoxicity.
--additional research is needed to understand the complex interactions that lead to drug-induced liver injury in HIV-infected persons. Furthermore, new HIV-1 PIs should be specifically evaluated for potential hepatic effects, particularly in patients with chronic viral hepatitis
Key Summary points made by the authors of this article written by Jules Levin:
--Hepatic injury attributed specifically to PI regimens alone are difficult to assess because of the presence of many complicating factors, such as drug-drug interactions, the clinical condition of the patient, and the hepatic effects of various comorbid diseases
--Although clinicians caring for HIV-infected patients should be aware of these specific hepatic effects of PIs, there are no data to support withholding PI-based HAART in patients coinfected with HCV or HBV.
-- indinavir and atazanavir were associated with increases in unconjugated bilirubin in 6%-40% of patients…. that clinically resembles Gilbert syndrome.. this phenomenon was not associated with signs or symptoms of hepatocellular injury, such as increases in serum liver enzyme levels; in addition, levels of bilirubin returned to normal after the discontinuation of the drug in all subjects. In clinical trials of Reyataz, significant increases in the total bilirubin level (>2.5 times the upper limit of normal [ULN]) were observed in 22%-47% of patients treated with atazanavir; however, clinical jaundice or scleral icterus was observed in only 7%-8% of subjects, and discontinuation of therapy due unconjugated hyperbilirubinemia was rare (<1%). this phenomenon does not reflect hepatocellular injury and should not be considered in the spectrum of hepatotoxicity related to PIs.
--in one study (Swiss HIV Cohort) liver enzyme elevations occurred in 18% of patients on PI-HAART. Of the 29 patients who were HBsAg positive, 13 (45%) had liver enzyme elevations, and, of the 57 patients who were anti-HCV positive, 19 (33%) developed liver enzyme elevations. Among the patients who were neither HBsAg nor anti-HCV positive, 38 (12%) of 306 had liver enzyme elevations. Having HCV or HBV increased risk significantly (5x for HCV & 9x greater for HBV) for severe ALT elevations.
--in another study, severe hepatotoxicity occurred in 10 (10%) of 96 of patients receiving PIs, in 8 (9%) of 90 of patients receiving nonnucleoside reverse-transcriptase inhibitors (NNRTIs), and in 3 (9%) of 35 receiving both PIs and NNRTIs.
--In a study discussed below (Aceti et al), severe hepatotoxicity was reported in 44 patients (3.2%) and 50% of these patients were taking ritonavir as part of their therapy (the article doesn't say if this was full dose ritonavir, but I presume it was). Having HBV or HCV increased risk for severe liver toxicity by 5-7 times.
--In a different study the rate of severe hepatoxicity (>5 x the upper limit of normal of ALT which is 30-40 IU/L) was 10%, but 48% of these patients were taking full dose ritonavir (>400 mg twice daily). Today, most patients use 100 mg twice daily as a boost for other PI. Coinfection with HCV or HBV can increase risk for see hepatoxicity by 3 times but "most coinfected patients (88%) did not experience significant hepatotoxicity".
-- in the study by Wit et al., the use of low-dose ritonavir based ART (i.e., 200 mg/day) was not associated with any cases of grade 4 hepatotoxicity. Furthermore, in a randomized controlled trial that compared lopinavir therapy boosted with low-dose ritonavir and nelfinavir, only 4.5% of lopinavir/ritonavir recipients developed an AST or ALT level >5 times the ULN, which was similar to the incidence observed in nelfinavir recipients (5.2%). Similarly, Vora et al. reported that the addition of low-dose ritonavir to indinavir therapy for 19 patients coinfected with HBV or HCV was not associated with significant increases in serum ALT or AST levels…..emerging data indicates that the use of low-dose ritonavir to "boost" the levels of other PIs (e.g., lopinavir or indinavir) is not associated with significantly higher incidence of severe hepatotoxicity than is observed with other PIs, such as nelfinavir
--In one study alcohol abuse was associated with severe hepatoxicity while on HAART.
What is mechanism of action of hepatoxicity on HAART:
--There is some evidence to support the association of immune restoration (e.g., increasing CD4 T cell count) as a result of HAART and the development of acute liver injury related to viral hepatitis
-- The second potential mechanism of PI-associated liver injury is an alteration in metabolism, because all PIs are metabolized in liver by various isoforms of the cytochrome P450 system
-- Hsu et al. reported that, in patients with mild hepatic impairment, ritonavir exposure (measured as the area under the curve [AUC]) increased by 40%; 95% CI, 17%66%) and peak concentration increased by 27% (95%, CI 5%53%), compared with values in control patients.
--nelfinavir AUC increased by 58% in patients with moderate to severe liver disease, which indicates that the activity of the CYP2C19 isoenzyme was suppressed in patients with liver disease.
--In a similar manner, Veronese et al., studying the pharmacokinetics of amprenavir (APV) showed that the AUC was increased in patients with cirrhosis and suggested a reduction in the dose of the PI when given to patients with Child-Pugh scores >5
ABSTRACT: Since their introduction, hepatotoxicity has been associated with the use of human immunodeficiency virus (HIV)1 protease inhibitors (PIs). However, the complexity of the HIV-infected patient and the combinations of medications used to treat HIV complicate the understanding of the independent effects of PIs in the development of drug-induced liver injury (DILI). I discuss the current understanding of PI-associated hepatotoxicity. Of the PI regimens studied, the greatest risk of DILI has been observed among patients receiving full-dose ritonavir. Similarly, hepatitis B and/or C virus coinfection has been associated with a greater risk of DILI, compared with those withno hepatitis. Although the specific mechanism by which viral hepatitis increases this risk is not known, patients with cirrhosis may have decreased cytochrome P450 activity, leading to increased PI exposure. Clearly, further research is needed to define the interaction of PIs and chronic viral hepatitis in the development of DILI.
Since their introduction in the mid-1990s, protease inhibitors (PIs) have become one of the integral parts of the standard treatment of HIV infection as outlined in multiple national and international guidelines for the management of persons infected with HIV. However, the use of PI-based HAART has been associated with adverse effects such as diabetes mellitus and lipodystrophy (fat redistribution/accumulation syndrome) and other metabolic abnormalities. In addition, the prescribing information for all PIs approved by the US Food and Drug Administration (FDA) includes the following warning: (1) hepatitis, including cases resulting in hepatic failure and death, has been reported in patients taking PIs; and (2) there may be an increased risk for alanine aminotransferase and/or aspartate aminotransferase (ALT/AST) elevations in patients with preexisting liver disease or underlying hepatitis B virus (HBV) or hepatitis C virus (HCV) infection. An increased frequency of AST/ALT monitoring should be considered for these patients.
However, to date, hepatotoxicity has been associated with all currently used antiretroviral (ARV) drug regimens. Although the mechanisms of drug-induced liver injury are poorly defined, it is likely that many factors contribute to the hepatic effects of a specific drug. For example, patient-specific factors, such as age, stage of HIV disease, comorbid diseases (e.g., chronic viral hepatitis), and concurrent abuse or use of alcohol and/or other hepatotoxic substances are likely to affect the development of liver injury in individual patients. Similarly, characteristics of specific medicationssuch as hepatic metabolism, drug-drug interactions, and intrinsic hepatic effectsmay contribute to drug-related hepatotoxicity.
UNCONJUGATED HYPERBILIRUBINEMIA ASSOCIATED WITH SPECIFIC PIS: INDINAVIR AND ATAZANAVIR
In clinical trials, indinavir and atazanavir were associated with increases in unconjugated bilirubin in 6%-40% of patients. However, this phenomenon was not associated with signs or symptoms of hepatocellular injury, such as increases in serum liver enzyme levels; in addition, levels of bilirubin returned to normal after the discontinuation of the drug in all subjects. Research by Zucker et al. demonstrated that indinavir directly inhibits the activity of the hepatic enzyme UDP-glucuronosyltransferase (UGT), leading to the development of a reversible, asymptomatic, indirect hyperbilirubinemia that clinically resembles Gilbert syndrome. Not surprisingly, the effect of indinavir on serum levels of unconjugated bilirubin is most pronounced in persons with a polymorphism in the gene UGT1, which is associated with Gilbert syndrome. In indinavir-treated HIV-infected patients lacking the Gilbert polymorphism, serum bilirubin levels were increased by a mean of 0.34 mg/dL, whereas, in those patients who were either homozygous or heterozygous for the polymorphism, serum bilirubin levels were increased by a mean of 1.45 mg/dL. More recently, studies of atazanavir have demonstrated potent inhibition of the activity of UGT similar to that observed with indinavir. In clinical trials, significant increases in the total bilirubin level (>2.5 times the upper limit of normal [ULN]) were observed in 22%-47% of patients treated with atazanavir; however, clinical jaundice or scleral icterus was observed in only 7%-8% of subjects, and discontinuation of therapy due unconjugated hyperbilirubinemia was rare (<1%). Nonetheless, the manufacturer of atazanavir (Bristol-Meyers Squibb) recommends that alternative therapies be considered if jaundice or scleral icterus presents cosmetic concerns for patients. Taken together, although increases in unconjugated or indirect bilirubin are commonly seen in HIV-infected patients who receive atazanavir and indinavir, this phenomenon does not reflect hepatocellular injury and should not be considered in the spectrum of hepatotoxicity related to PIs.
HEPATOTOXICITY ASSOCIATED WITH PIs
Hepatotoxicity, which is defined as a significant increase in serum ALT or AST levels that occurs after the initiation of ART, has occurred at rates of 1%-9.5% in randomized controlled trials conducted for the purpose of obtaining FDA approval for the PI studied ("registration trials"). Overall, in these registration studies, liver enzyme elevations (LEEs) were rarely associated with symptoms or dose limiting, and permanent liver injury was a distinctly uncommon event. However, subjects enrolled in such registration trials were generally required to have serum liver enzyme levels within the normal range. Furthermore, the prevalence of coinfection with HBV or HCV was generally lower in such study populations, compared with the prevalence observed in clinical settings in the developed world. Thus, the ability to generalize these data to clinical practice settings may be limited.
Indeed, shortly after licensure of the first PIs in the United States, the potential for hepatotoxicity associated with the PIs was highlighted by the publication of several case reports detailing severe liver injury after the use of PIs in HIV-infected patients coinfected with HBV or HCV. Subsequently, a number of large cohorts have been evaluated to determine the incidence of and risk factors for significant LEEs in HIV-infected adults after the administration of HAART.
Early during the era of PI-based HAART, a prospective cohort study was conducted at a university-based, urban HIV clinic in Baltimore, Maryland, to determine the incidence of severe hepatotoxicity during ARV therapy among 298 patients prescribed a new ARV regimen and to define the role of chronic viral hepatitis in its development. The PI-treated group consisted of 212 patients (71%) who were taking PIs for the first time. Ritonavir, with or without saquinavir, was the regimen used by 50 patients (17%), indinavir by 117 patients (39%), nelfinavir by 51 patients (17%), and dual nucleoside reverse-transcriptase inhibitors (NRTIs) by 87 patients (29%). All patients had pretreatment liver enzyme levels measured within 6 months of initiating the new drug regimen. HCV antibody was present in 48% of the patients, and hepatitis B surface antigen (HBsAg) was present in 3.3%. Patients infected with HCV had higher pretreatment AST and ALT levels than did HCV-uninfected patients (P < .001 for both enzymes). In this study, severe hepatotoxicity was defined as grade 3 or 4 change in AST or ALT levels during antiretroviral treatment. Patients with pretreatment serum AST and ALT levels within the normal range (AST level of <35 U/L and ALT level of <31 U/L) were classified on the basis of changes relative to the ULN: grade 3, 5.1-10 times ULN and grade 4, >10 times the ULN. To avoid selection bias favoring the inclusion of persons with chronic viral hepatitis, patients with elevated pretreatment serum AST and ALT levels (higher than the ULN) were classified on the basis of changes relative to baseline value rather than ULN: grade 3 was defined as a level 3.65 times baseline and grade 4 as a level of >5 times baseline.
Among all subjects, the incidence of severe hepatotoxicity was 10.4% (95% CI, 7.2%-14.4%). However, the incidence of severe hepatotoxicity was higher in patients who received ritonavir than in those who received dual NRTIs, indinavir, nelfinavir, or saquinavir (without concurrent ritonavir). Ritonavir use was associated with 48% of all cases of severe hepatotoxicity and with the highest incidence of toxicity (incidence, 30%; 95% CI, 17.9%44.6%). In this study, the risk of severe hepatotoxicity with the use of NRTIs was similar to that with use of indinavir, nelfinavir, or saquinavir (without the concurrent use of ritonavir). Hepatotoxicity of any grade was observed in 83 (54%) of 154 patients infected with HCV, compared with 56 (39%) of 144 uninfected patients (P = .009). However, no severe hepatotoxicity developed in 139 (88%) of 158 patients with evidence of chronic HBV or HCV infection. In a multivariate analysis, only ritonavir use and a CD4 cell count increase of >0.5 x 109 cells/L were associated with severe hepatotoxicity. Among individuals who did not receive ritonavir-containing regimens, HCV and/or HBV coinfection was associated with an increased risk of severe hepatotoxicity (relative risk [RR], 3.7; 95% CI, 1.011.8). However, most coinfected patients (88%) did not experience significant hepatotoxicity.
In a similar study that involved patients observed at a single university clinic in the Swiss HIV Cohort Study, den Brinker et al. evaluated the risk of hepatotoxicity after the initiation of PI-containing HAART for HIV-infected patients. They followed-up 394 patients after the initiation of HAART, 7% of whom were HBsAg positive and 14% of whom were anti-HCV positive. As expected, compared with patients without coinfection, patients with chronic hepatitis had a higher incidence of LEEs. Overall, 70 (18%) patients developed significant LEEs. Of the 29 patients who were HBsAg positive, 13 (45%) had LEEs, and, of the 57 patients who were anti-HCV positive, 19 (33%) developed LEEs. Among the patients who were neither HBsAg nor anti-HCV positive, 38 (12%) of 306 had LEEs. In a multivariate analysis, the presence of HBsAg or anti-HCV was associated with a greater risk of having an LEE, with an RR of 2.78 (95% CI, 1.505.16) and 2.46 (95% CI, 1.434.24), respectively (both P = .001). However, 63% of patients with significant LEEs were able to continue the same ARV regimen with interruption. Of interest, among patients with chronic HBV infection, 38% lost hepatitis B e (Be) antigen or developed antihepatitis Be after the initiation of HAART, which suggests that LEEs in these patients were related to HBV-related immune reconstitution.
In a study conducted in Spain, Nuñez et al analyzed the risk factors associated with the use of PI-containing ART in a cohort of 222 patients starting their regimens. Coinfection with HCV and HBV was diagnosed in 84 patients (38%) and 11 patients (5%), respectively. Severe hepatotoxicity occurred in 10 (10%) of 96 of patients receiving PIs, in 8 (9%) of 90 of patients receiving nonnucleoside reverse-transcriptase inhibitors (NNRTIs), and in 3 (9%) of 35 receiving both PIs and NNRTIs. Treatment was discontinued in 6 individuals who experienced a severe hepatic injury. In a univariate analysis, the following factors were all associated with severe hepatic injury: injection drug use (RR, 3.5; 95% CI, 1.49.1), presence of anti-HCV antibodies (RR, 3.7; 95% CI, 1.49.6), and alcohol abuse (RR, 8.7; 95% CI, 2.530.4). Of importance, in this cohort, alcohol abuse contributed substantially to the development of severe hepatotoxicity. Given the relatively high prevalence of alcohol abuse among HIV-infected persons, these data highlight the need to counsel and treat alcoholism among persons starting ARV therapy, particularly those with chronic viral hepatitis.
To evaluate the evolution of liver enzyme changes over time in HIV-infected persons treated with ART, Aceti et al.performed a retrospective study of 1325 consecutive HIV patients who were treated with antiretroviral therapy that included at least 1 PI for at least 6 months. In this cohort, 616 patients (45%) patients were anti-HCV positive, 54 (4.1%) patients were HBsAg positive, and 115 (8.7%) tested positive for both HBV and HCV. Severe liver toxicity was defined as ALT levels >5 times ULN. Severe hepatotoxicity was reported in 44 patients (3.2%). Twenty-one (50%) of 44 patients were receiving ritonavir as part of their therapy. Univariate analysis showed that infection with HBV (OR, 4.9; 95% CI, 2.59.7) or HCV (OR, 6.8; 95% CI, 4.610.1) were both independent risk factors for an occurrence of hepatotoxicity after 6 months of therapy. Similar to prior cohort studies, this study compared the incidence of severe and mild (any grade) hepatotoxicity in coinfected patients after 6, 12, and 24 months of therapy. Ritonavir use was associated with significantly greater incidence of hepatotoxicity after 6 (P < .0001) and 12 (P = .029) months but not after 24 months of therapy.
More recently, the incidence of severe hepatotoxicity associated with HAART in the Swiss HIV Cohort Study cohort was reevaluated by Wit et al. In this study, grade 4 liver enzyme elevations were defined as AST and/or ALT elevations 10 times ULN and an absolute increase in aminotransferase level of >200 U/L. Between July 1996 and January 2000, 560 patients initiated HAART, of whom 8.8% and 10.7% were HBV coinfected and HCV coinfected, respectively. Overall, 35 (6.3%) of 560 patients developed grade 4 hepatotoxicity, of whom 6 (17.1%) were symptomatic and 12 (34%) discontinued ARV therapy. In multivariate analysis, independent risk factors for grade 4 liver injury were higher baseline ALT levels (hazard ratio [HR], 1.05 for each 10-U increase), chronic HBV infection (HR, 9.2), chronic HCV infection (HR, 5.0), the use of first-line potent ARV combination regimens in patients without prior NRTI treatment (HR, 2.8), recent start of nevirapine (HR, 9.6) or ritonavir (HR, 4.9) therapy, and female sex (HR, 2.8). Furthermore, among patients chronically coinfected with HBV, discontinuing the use of lamivudine (3TC) was associated with the development of grade 4 LEEs (HR, 6.8). Taken together, these studies indicate that the use of HAART that included full-dose ritonavir (i.e., >400 mg of ritonavir daily) was associated with an increased risk of liver injury, compared with other PI-based ARV regimens.
Of significance, in the study by Wit et al., the use of low-dose ritonavir based ART (i.e., 200 mg/day) was not associated with any cases of grade 4 hepatotoxicity. Furthermore, in a randomized controlled trial that compared lopinavir therapy boosted with low-dose ritonavir and nelfinavir, only 4.5% of lopinavir/ritonavir recipients developed an AST or ALT level >5 times the ULN, which was similar to the incidence observed in nelfinavir recipients (5.2%). Similarly, Vora et al. reported that the addition of low-dose ritonavir to indinavir therapy for 19 patients coinfected with HBV or HCV was not associated with significant increases in serum ALT or AST levels.
Thus, among currently available PIs, full-dose ritonavir has been associated with a high incidence of significant hepatotoxicity, and the use of full-dose ritonavir has been consistently identified as an independent risk factor for the development of severe hepatic injury after adjustment for other risk factors such as chronic viral hepatitis. However, there are emerging data to indicate that the use of low-dose ritonavir to "boost" the levels of other PIs (e.g., lopinavir or indinavir) is not associated with significantly higher incidence of severe hepatotoxicity than is observed with other PIs, such as nelfinavir.
Taken together, the data derived from clinical trials and cohort studies indicate that elevations in serum aminotransferase levels are relatively common in HIV-infected patients receiving PI-based ART. Among individual PIs, full-dose ritonavir is the only specific drug that has been identified as an independent risk factor for the development of grade 3 and/or 4 severe HAART-related liver injury. In addition, these studies also indicated that HBV or HCV coinfection substantially increases the risk of developing significant LEEs after the initiation of HAART. However, most studies have shown that the great majority of HCV-coinfected patients treated with HAART do not develop severe hepatotoxicity.
MECHANISMS OF PI-RELATED HEPATOTOXICITY
The pathogenesis of PI-associated liver injury is not known. The majority of published studies have described the occurrence of LEE temporally related to the initiation of HAART, but few studies have systematically investigated the nature of the liver injury. To date, clinical trials and observational cohort studies have not consistently identified other factors such as age, sex, or virological or immune response to HAART as being associated with the development of PI-related hepatotoxicity. Furthermore, given the lack of information regarding liver disease stage (i.e., hepatic fibrosis stage), such studies have been unable to better define which HCV-infected patients are at the greatest risk of HAART-induced liver injury.
Nonetheless, several potential mechanisms have been proposed. There is some evidence to support the association of immune restoration (e.g., increasing CD4 T cell count) as a result of HAART and the development of acute liver injury related to viral hepatitis. The evidence that restored immune recognition of a viral pathogen is responsible for increases in liver enzymes is strongest for hepatitis B, for which seroconversion has been reported in the setting of HAART. Overall, the evidence is less convincing for the occurrence of immune-related disease flares related to chronic hepatitis C infection. However, one recent study reported that HAART-associated hepatotoxicity was associated with increases in serum HCV-specific IgG and soluble CD26 dipeptidyl peptidase IV enzyme activity in 11 HCV/HIV-coinfected patients. Thus, although an enhanced immune response to hepatitis B or C may account for some episodes of liver enzyme elevation after the initiation of HAART, this proposed mechanism cannot explain the occurrence of drug-induced liver injury in HIV-infected patients without concurrent viral hepatitis.
The second potential mechanism of PI-associated liver injury is an alteration in metabolism, because all PIs are metabolized in liver by various isoforms of the cytochrome P450 system. Accordingly, some experts have speculated that liver injury may be caused by supertherapeutic serum levels of PIs in patients coinfected with viral hepatitis or with preexisting liver disease. Recently, Hsu et al. reported that, in patients with mild hepatic impairment, ritonavir exposure (measured as the area under the curve [AUC]) increased by 40%; 95% CI, 17%66%) and peak concentration increased by 27% (95%, CI 5%53%), compared with values in control patients.
In a retrospective analysis of data from a cohort of patients, a significant increase in the incidence of nephrolithiasis in patients coinfected with HIV and HCV or HBV was identified, which suggests that relatively high serum concentrations of indinavir occurred in patients with liver disease and contributed to the development of indinavir kidney stones. Of interest, the US prescribing information for indinavir contains the statement, "the dosage of indinavir should be reduced to 600 mg every 8 hours in patients with mild-to-moderate hepatic insufficiency due to cirrhosis".
Similarly, in a study of the pharmacokinetics of nelfinavir and its metabolites, Khaliq et al. found that the nelfinavir AUC increased by 58% in patients with moderate to severe liver disease, which indicates that the activity of the CYP2C19 isoenzyme was suppressed in patients with liver disease. In a similar manner, Veronese et al., studying the pharmacokinetics of amprenavir (APV) showed that the AUC was increased in patients with cirrhosis and suggested a reduction in the dose of the PI when given to patients with Child-Pugh scores >5. The dosage of 1200 mg b.i.d. for a patient without cirrhosis should be reduced to 450 mg b.i.d. for subjects with Child-Pugh scores of 58 and further lowered to 300 mg b.i.d. if the Child-Pugh score is 915.
However, although there is strong evidence that decompensated liver disease is associated with substantial decreases in the metabolism of some PIs, there are no published data linking elevated serum PI levels to the development of hepatotoxicity. In one study, Reijers et al. evaluated the relationship of hepatotoxicity and PI exposure in HIV-infected patients receiving a 4-drug regimen that included both nelfinavir and saquinavir. Mild to severe elevations of liver enzymes occurred in 12 of 65 patients in the study. In four of these patients, the LEEs led to a discontinuation of study medication. Measured over the course of the study, plasma nelfinavir and saquinavir levels of patients with LEEs were no different from those from patients who did not have LEEs. However, drug levels were not measured at the time of the toxicity, thus limiting the interpretation of these data. Nonetheless, although prospective studies are currently under way to determine the association of serum PI concentration and the development of drug-induced liver injury, there is insufficient evidence to recommend the routine use of therapeutic drug monitoring of PIs as a means to decrease the risk of HAART-associated hepatotoxicity.
CONCLUSIONS
Hepatic injury attributed specifically to PI regimens alone are difficult to assess because of the presence of many complicating factors, such as drug-drug interactions, the clinical condition of the patient, and the hepatic effects of various comorbid diseases. Nonetheless, across a number of studies, coinfection with chronic HBV and/or HCV has been consistently associated with a greater risk of severe liver injury compared with patients who have concurrent liver disease. Although the etiopathogenesis of drug-induced liver injury in persons with chronic viral hepatitis remains uncertain, several potential mechanisms have been proposed, such as pathogen-specific immune reconstitution syndrome and impaired drug metabolism caused by liver disease. Indeed, several studies have indicated that a person with significant liver disease may have a higher serum concentration of hepatically metabolized PIs; however, to date, there are few data that firmly link high concentrations of PIs and hepatotoxicity. Accordingly, without additional supporting data, the therapeutic drug monitoring of PIs in persons with chronic liver disease cannot be routinely recommended.
Among specific PIs, the use of full-dose ritonavir appears to be associated with the highest risk of developing LEEs. However, recent data suggest that the use of low dose ritonavir to pharmacologically boost other PIs, such as lopinavir, is not associated with greater risk of liver injury compared with other PIs, such as nelfinavir or indinavir. Other PIs (e.g., indinavir and atazanavir) are associated with a benign increase in unconjugated bilirubin in HIV-infected patients, because of their direct inhibition of hepatic UTG activity.
Clearly, additional research is needed to understand the complex interactions that lead to drug-induced liver injury in HIV-infected persons. Furthermore, new HIV-1 PIs should be specifically evaluated for potential hepatic effects, particularly in patients with chronic viral hepatitis. Finally, although clinicians caring for HIV-infected patients should be aware of these specific hepatic effects of PIs, there are no data to support withholding PI-based HAART in patients coinfected with HCV or HBV.
|
|
| |
| |
|
|
|
