- Email: [email protected]
Pharmacokinetic and pharmacodynamic modeling of pegylated-interferon alfa
Editorial
Pharmacokinetic and pharmacodynamic modeling of pegylated-interferon alfa Rositsa B. Dimova, Andrew H. Talal* Center for the Study of Hepatitis C and Division of Gastroenterology and Hepatology, Weill Cornell Medical College, New York, NY, USA
See Article, pages 460–467
Hepatitis C virus (HCV) infection affects approximately 170 million individuals worldwide [1,2]. Among those exposed to HCV, between 50% and 80% develop chronic infection that may result in progressive liver fibrosis and consequently in cirrhosis and/ or hepatocellular carcinoma. Among HIV-infected people, approximately 15–30% are co-infected with HCV, and the hepatitis virus is a leading cause of death in these individuals [3,4]. Unfortunately, therapeutic efficacy is diminished in HIV/HCV co-infected compared to HCV mono-infected patients to current standard therapy, which consists of pegylated-interferon (PEGIFN) and ribavirin (RBV) [3]. Two separate formulations of PEGIFN have been approved, alfa-2a and alfa-2b, each with different pharmacokinetic profiles [5], and these medications are typically administered for 48 weeks. They can be difficult to tolerate; consequently, therapy in those unlikely to respond should be limited. According to current guidelines, patients should be treated for at least 12 weeks to determine whether a 2-log HCV RNA decline has occurred, thereby justifying treatment continuation [6]. However, the identification of parameters that could be used earlier for therapeutic intervention would be of tremendous importance. Successful treatment outcome for HCV in PEG-IFN/RBV treated patients has been defined as undetectable HCV RNA in serum 6 months after the end of treatment (sustained virological response [SVR]). Nonresponders (NR) are those in whom serum HCV RNA is detectable 6 months post treatment cessation. SVR occurs in 27–40% of HIV/HCV coinfected treated patients [3]. Mathematical modeling of HCV RNA decay has been used to assess the effectiveness of anti-HCV treatment [7,8]. Early viral dynamic models that used standard IFN alfa assumed constant effectiveness, which was appropriate when IFN was administered thrice weekly. In contrast, when patients were treated with PEGIFN alfa-2b, time-varying IFN concentration was observed [9]. To account for these changes, Powers et al. incorporated pharmacokinetic/pharmacodynamic parameters into a viral dynamic model
Received 21 April 2010; accepted 22 April 2010 * Corresponding author. Address: Weill Cornell Medical College 525 E. 68th Street, Box 319 New York, NY 10065, USA. Fax: +1 212 746 7977. E-mail address: [email protected] (A.H. Talal). Abbreviations: HCV, hepatitis C virus; HIV, human immunodeficiency virus; PEGIFN, pegylated-interferon; RBV, ribavirin; SVR, sustained virological response; NR, nonresponder; PK, pharmacokinetics; PD, pharmacodynamics.
[10]. Pharmacokinetics (PK) establishes the connection between drug inflow and concentration in blood and includes parameters such as drug absorption, elimination, and blood volume. Pharmacodynamics (PD) establishes the connection between drug concentration and clinical outcome. PK/PD of PEG-IFN/RBV has been modeled through partial differential equations constructed to explain the drug’s mechanism of action [11]. Using mechanistic models, one can estimate in vivo parameters that are not directly measurable, such as how effective a drug (i.e., IFN) is in blocking virus production, or to make inferences about other clinically-useful variables, such as the duration of treatment necessary to eradicate the virus [12,13]. Limitations of the mechanistic models include difficulty in model validation as well as computational issues that may arise due to the approximating and iterative nature of the algorithms used to estimate the parameters and their convergence. An important question is whether PK/PD properties of PEGIFN/RBV can identify HCV-infected patients early in treatment who are unlikely to respond. To address this issue, we constructed a dynamic model that incorporates PK/PD parameters and applied it to data obtained from 24 HIV/HCV co-infected patients treated with PEG-IFN alfa-2b/RBV [14]. We found that EC50, the median individual effective concentration of PEG-IFN, was significantly lower in SVR compared with NR patients. Additionally, the estimated PK parameters did not differ significantly between the two patient groups. Recent studies also evaluated the utility of PK/PD parameters for their ability to identify NR patients early in therapy [15–17]. Rozenberg et al. evaluated PK/PD and viral parameters and their association with treatment outcome in 23 HIV/HCV co-infected patients treated with PEG-IFN alfa-2b. These investigators additionally studied whether PK/PD parameters differ between African Americans and Caucasians and found that the EC50 was significantly lower in Caucasians. Moreover, patients who achieved an SVR maintained serum concentrations of IFN above the EC90, an improved estimator of antiviral sensitivity than EC50, for longer periods of time than NR patients. In comparison to the above-referenced studies that investigated PK/PD properties of PEG-IFN alfa-2b in co-infected patients [14,15], in the current issue of the Journal of Hepatology, Dahari et al. performed a comprehensive analysis of PK/PD of PEG-IFN alfa-2a in 26 co-infected patients [18]. This analysis is particu-
Journal of Hepatology 2010 vol. 53 j 418–420
JOURNAL OF HEPATOLOGY larly important because of the differences in pharmacokinetics between the two PEG-IFN preparations and, in addition, their decreased therapeutic efficacy in co-infected patients. Dahari et al. used mathematical models together with two-stage, individual-based, nonlinear least squares method to estimate the PK/PD parameters associated with PEG-IFN. Their main purpose was to discover PK/PD-related parameters that could be used to identify potential NR patients. PEG-IFN concentration and HCV RNA levels were sampled frequently until week 12. Five of the 26 patients were excluded from the modeling due to early treatment discontinuation. The authors estimated the PK parameters for the first week and for the twelfth week of treatment. They also estimated the maximum PEG-IFN effectiveness during the first dose, e7max, and the infected cell loss rate, d, for each patient. Then, they used nonparametric statistical methods to compare the estimated quantities between SVR and NR, between HCV genotypes 1 and 3, and between African Americans and Caucasians. The authors identified and described four viral kinetic patterns. In patients with a triphasic HCV RNA decay pattern, they fit the combined-PD model that includes additional terms for hepatocyte proliferation. In patients with a biphasic decline, they fit the combined biphasic-PD model. Based on their results, Dahari et al. conclude that none of the PK parameters can significantly differentiate between NR and SVR. However, they identified two PD parameters with 100% positive predictive value for the identification of SVR patients: (1) the maximum PEG-IFN effectiveness during the first dose, e7max, and (2) the infected cell loss rate, d. When comparing the PK/PD parameters between individuals infected with HCV genotypes 1 and 3; these authors concluded that EC50 is significantly lower, and that d is significantly increased in genotype 3-infected individuals. In addition, they did not identify significant differences in the viral kinetic and pharmacodynamic parameters between Caucasians and African Americans. An important issue is the statistical technique used in the analysis of PK/PD data. In general, two statistical approaches, the two-stage nonlinear least squares approach (individual PK), and the nonlinear mixed effects model (population PK) (Table 1) have been established for the analysis of PK/PD data. Briefly, the two-stage (individual PK) approach requires that a large number of observations are obtained for each patient at the same time points in order to obtain parameter estimates. The first stage of this procedure consists of using nonlinear regression (nonlinear least squares) to estimate the PK/PD parameters for each individual separately. In the second stage, these parameters are summarized through computation of descriptive statistics (means, variances), and a subsequent analysis of dependencies between the estimated parameters and patient-related covariates can be performed. The three studies discussed above [14,15,18] used
this statistical technique. An alternative approach for analysis of PK/PD data is based on the nonlinear mixed effects models. This method is especially appropriate in sparse or unbalanced data situations (i.e., patients may have different sampling times or may have missed some measurements) and when the investigator is interested in making generalizations about the population from which the study subjects were randomly selected. In contrast to the two-stage approach, in this procedure the population parameters are estimated directly, and inter and intra-subject variability is incorporated into the model. As the presence of missing values is a common problem in studies of HCVinfected persons because of poor compliance among some HCVinfected populations, the population PK approach may be a useful tool for analyzing such data. The US Food and Drug Administration has issued guidelines concerning the use of population PK in the process of drug development [19]. In conclusion, identifying PK/PD models that describe the relationship between the concentration of the drug in the blood, the dose of the drug, other PK parameters and covariates is a very important aspect in the area of HCV therapeutics. Besides IFN, exploring differences in RBV pharmacokinetics, as has been done in various studies [17,20], and the PK/PD of directly acting antivirals, are additional considerations that may differentiate SVR and NR patients. Explaining inter and intra-subject variability in the PK of a drug, or in viral kinetics, is a central goal of the analysis of PK/PD parameters. Studies investigating the mechanism of action of IFN performed to date have not identified PK parameters that significantly differ between SVR and NR patients. In contrast, some PD parameters have been identified with the ability to differentiate between these two patient groups. Dahari et al. investigated the important problem of modeling the pharmacokinetics of PEG-IFN alfa-2a and the HCV viral kinetics in HIV/HCV co-infected patients, which is motivated by the widespread use of PEGIFN alfa-2a in clinical practice. The investigators found that the maximum PEG-IFN effectiveness during the first dose, and the infected cell loss rate can differentiate SVR from NR patients. Further research is needed to confirm the findings concerning the use of viral kinetic parameters as prognostic tools of treatment outcome and to determine their role in patient management.
Financial disclosure Dr. Andrew Talal has received research support from Schering Plough (presently Merck) and Merck, and research support and is on the speaker’s bureau of Genentech, a member of the Roche group.
Table 1. Summary of the characteristics of the population (nonlinear mixed effects model) pharmacokinetics and the individual two-stage (nonlinear least squares method) pharmacokinetic approach. Individual Two-Stage Pharmacokinetics
Population Pharmacokinetics
Focus on individual Tools: nonlinear regression model based upon nonlinear least squares method Data from each individual are modeled separately Needs rich data for each subject
Focus on population Tools: nonlinear mixed effects model Data from all individuals are modeled simultaneously Handles dense, sparse, and unbalanced data (even one observation per subject) Accounts for the different sources of variability (between-subject, within-subject) Important factors such as demographic, environmental, pathophysiological, etc., which explain subject variability, can be identified Can be used also for individualized inference
Journal of Hepatology 2010 vol. 53 j 418–420
419
Editorial References [1] Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C virus infection. Lancet Infect Dis 2005;5:558–567. [2] Davis GL, Keeffe EB, Balart LA. Advances in liver disease: highlights from the 56th Annual Meeting of the American Association for the Study of Liver Disease. Rev Gastroenterol Disord 2006;6:48–61. [3] Kadam JS, Talal AH. Changing treatment paradigms: hepatitis C virus in HIVinfected patients. AIDS Patient Care STDS 2007;21:154–168. [4] Gonzalez SA, Talal AH. Hepatitis C virus in human immunodeficiency virusinfected individuals: an emerging comorbidity with significant implications. Semin Liver Dis 2003;23:149–166. [5] Foster GR. Pegylated interferons for the treatment of chronic hepatitis C: pharmacological and clinical differences between peginterferon-alpha-2a and peginterferon-alpha-2b. Drugs 2010;70:147–165. [6] Ghany MG, Strader DB, Thomas DL, Seeff LB. Diagnosis, management, and treatment of hepatitis C: an update. Hepatology 2009;49:1335–1374. [7] Neumann AU, Lam NP, Dahari H, Gretch DR, Wiley TE, Layden TJ, et al. Hepatitis C viral dynamics in vivo and the antiviral efficacy of interferon-a therapy. Science 1998;282:103–107. [8] Talal AH, Shata MT, Markatou M, Dorante G, Chadburn A, Koch R, et al. Virus dynamics and immune responses during treatment in patients coinfected with hepatitis C and HIV. J Acquir Immune Defic Syndr 2004;35:103–113. [9] Glue P, Fang JW, Rouzier-Panis R, Raffanel C, Sabo R, Gupta SK, et al. Pegylated interferon-alpha2b: pharmacokinetics, pharmacodynamics, safety, and preliminary efficacy data. Clin Pharmacol Ther 2000;68:556–567. [10] Powers KA, Dixit NM, Ribeiro RM, Golia P, Talal AH, Perelson AS. Modeling viral and drug kinetics: hepatitis C virus treatment with pegylated interferon alfa-2b. Semin Liver Dis 2003;23 (Suppl. 1):13–18. [11] Dahari H, Shudo E, Ribeiro RM, Perelson AS. Mathematical modeling of HCV infection and treatment. Methods Mol Biol 2009;510:439–453.
420
[12] Bonate PL. Pharmacokinetic–pharmacodynamic modeling and simulation. New York, NY: Springer; 2006. [13] Shudo E, Ribeiro RM, Perelson AS. Modeling HCV kinetics under therapy using PK and PD information. Expert Opin Drug Metab Toxicol 2009;5:321–332. [14] Talal AH, Ribeiro RM, Powers KA, Grace M, Cullen C, Hussain M, et al. Pharmacodynamics of PEG-IFN alpha differentiate HIV/HCV coinfected sustained virological responders from nonresponders. Hepatology 2006;43:943–953. [15] Rozenberg L, Haagmans BL, Neumann AU, Chen G, McLaughlin M, LevyDrummer RS, et al. Therapeutic response to peg-IFN-alpha-2b and ribavirin in HIV/HCV co-infected African-American and Caucasian patients as a function of HCV viral kinetics and interferon pharmacodynamics. AIDS 2009;23:2439–2450. [16] Asahina Y, Izumi N, Umeda N, Hosokawa T, Ueda K, Doi F, et al. Pharmacokinetics and enhanced PKR response in patients with chronic hepatitis C treated with pegylated interferon alpha-2b and ribavirin. J Viral Hepat 2007;14:396–403. [17] Nicot F, Legrand-Abravanel F, Lafont T, Dubois M, Saune K, Pasquier C, et al. Serum concentrations of ribavirin and pegylated interferon and viral responses in patients infected with HIV and HCV. J Med Virol 2008;80:1523–1529. [18] Dahari H, Affonso de Araujo ES, Haagmans BL, Layden TJ, Cotler SJ, Barone AA, et al. Pharmacodynamics of PEG-IFN alpha-2a in HIV/HCV co-infected patients: Implications for treatment outcomes. J Hepatol 2010;53:460–467. [19] Food and Drug Administration. Guidance for industry: population pharmacokinetics. Rockville, MD: US Department of Health and Human Services; 1999. [20] Dahari H, Markatou M, Zeremski M, Haller I, Ribeiro RM, Licholai T, et al. Early ribavirin pharmacokinetics, HCV RNA and alanine aminotransferase kinetics in HIV/HCV co-infected patients during treatment with pegylated interferon and ribavirin. J Hepatol 2007;47:23–30.
Journal of Hepatology 2010 vol. 53 j 418–420
Pharmacokinetic and pharmacodynamic modeling of pegylated-interferon alfa Rositsa B. Dimova, Andrew H. Talal* Center for the Study of Hepatitis C and Division of Gastroenterology and Hepatology, Weill Cornell Medical College, New York, NY, USA
See Article, pages 460–467
Hepatitis C virus (HCV) infection affects approximately 170 million individuals worldwide [1,2]. Among those exposed to HCV, between 50% and 80% develop chronic infection that may result in progressive liver fibrosis and consequently in cirrhosis and/ or hepatocellular carcinoma. Among HIV-infected people, approximately 15–30% are co-infected with HCV, and the hepatitis virus is a leading cause of death in these individuals [3,4]. Unfortunately, therapeutic efficacy is diminished in HIV/HCV co-infected compared to HCV mono-infected patients to current standard therapy, which consists of pegylated-interferon (PEGIFN) and ribavirin (RBV) [3]. Two separate formulations of PEGIFN have been approved, alfa-2a and alfa-2b, each with different pharmacokinetic profiles [5], and these medications are typically administered for 48 weeks. They can be difficult to tolerate; consequently, therapy in those unlikely to respond should be limited. According to current guidelines, patients should be treated for at least 12 weeks to determine whether a 2-log HCV RNA decline has occurred, thereby justifying treatment continuation [6]. However, the identification of parameters that could be used earlier for therapeutic intervention would be of tremendous importance. Successful treatment outcome for HCV in PEG-IFN/RBV treated patients has been defined as undetectable HCV RNA in serum 6 months after the end of treatment (sustained virological response [SVR]). Nonresponders (NR) are those in whom serum HCV RNA is detectable 6 months post treatment cessation. SVR occurs in 27–40% of HIV/HCV coinfected treated patients [3]. Mathematical modeling of HCV RNA decay has been used to assess the effectiveness of anti-HCV treatment [7,8]. Early viral dynamic models that used standard IFN alfa assumed constant effectiveness, which was appropriate when IFN was administered thrice weekly. In contrast, when patients were treated with PEGIFN alfa-2b, time-varying IFN concentration was observed [9]. To account for these changes, Powers et al. incorporated pharmacokinetic/pharmacodynamic parameters into a viral dynamic model
Received 21 April 2010; accepted 22 April 2010 * Corresponding author. Address: Weill Cornell Medical College 525 E. 68th Street, Box 319 New York, NY 10065, USA. Fax: +1 212 746 7977. E-mail address: [email protected] (A.H. Talal). Abbreviations: HCV, hepatitis C virus; HIV, human immunodeficiency virus; PEGIFN, pegylated-interferon; RBV, ribavirin; SVR, sustained virological response; NR, nonresponder; PK, pharmacokinetics; PD, pharmacodynamics.
[10]. Pharmacokinetics (PK) establishes the connection between drug inflow and concentration in blood and includes parameters such as drug absorption, elimination, and blood volume. Pharmacodynamics (PD) establishes the connection between drug concentration and clinical outcome. PK/PD of PEG-IFN/RBV has been modeled through partial differential equations constructed to explain the drug’s mechanism of action [11]. Using mechanistic models, one can estimate in vivo parameters that are not directly measurable, such as how effective a drug (i.e., IFN) is in blocking virus production, or to make inferences about other clinically-useful variables, such as the duration of treatment necessary to eradicate the virus [12,13]. Limitations of the mechanistic models include difficulty in model validation as well as computational issues that may arise due to the approximating and iterative nature of the algorithms used to estimate the parameters and their convergence. An important question is whether PK/PD properties of PEGIFN/RBV can identify HCV-infected patients early in treatment who are unlikely to respond. To address this issue, we constructed a dynamic model that incorporates PK/PD parameters and applied it to data obtained from 24 HIV/HCV co-infected patients treated with PEG-IFN alfa-2b/RBV [14]. We found that EC50, the median individual effective concentration of PEG-IFN, was significantly lower in SVR compared with NR patients. Additionally, the estimated PK parameters did not differ significantly between the two patient groups. Recent studies also evaluated the utility of PK/PD parameters for their ability to identify NR patients early in therapy [15–17]. Rozenberg et al. evaluated PK/PD and viral parameters and their association with treatment outcome in 23 HIV/HCV co-infected patients treated with PEG-IFN alfa-2b. These investigators additionally studied whether PK/PD parameters differ between African Americans and Caucasians and found that the EC50 was significantly lower in Caucasians. Moreover, patients who achieved an SVR maintained serum concentrations of IFN above the EC90, an improved estimator of antiviral sensitivity than EC50, for longer periods of time than NR patients. In comparison to the above-referenced studies that investigated PK/PD properties of PEG-IFN alfa-2b in co-infected patients [14,15], in the current issue of the Journal of Hepatology, Dahari et al. performed a comprehensive analysis of PK/PD of PEG-IFN alfa-2a in 26 co-infected patients [18]. This analysis is particu-
Journal of Hepatology 2010 vol. 53 j 418–420
JOURNAL OF HEPATOLOGY larly important because of the differences in pharmacokinetics between the two PEG-IFN preparations and, in addition, their decreased therapeutic efficacy in co-infected patients. Dahari et al. used mathematical models together with two-stage, individual-based, nonlinear least squares method to estimate the PK/PD parameters associated with PEG-IFN. Their main purpose was to discover PK/PD-related parameters that could be used to identify potential NR patients. PEG-IFN concentration and HCV RNA levels were sampled frequently until week 12. Five of the 26 patients were excluded from the modeling due to early treatment discontinuation. The authors estimated the PK parameters for the first week and for the twelfth week of treatment. They also estimated the maximum PEG-IFN effectiveness during the first dose, e7max, and the infected cell loss rate, d, for each patient. Then, they used nonparametric statistical methods to compare the estimated quantities between SVR and NR, between HCV genotypes 1 and 3, and between African Americans and Caucasians. The authors identified and described four viral kinetic patterns. In patients with a triphasic HCV RNA decay pattern, they fit the combined-PD model that includes additional terms for hepatocyte proliferation. In patients with a biphasic decline, they fit the combined biphasic-PD model. Based on their results, Dahari et al. conclude that none of the PK parameters can significantly differentiate between NR and SVR. However, they identified two PD parameters with 100% positive predictive value for the identification of SVR patients: (1) the maximum PEG-IFN effectiveness during the first dose, e7max, and (2) the infected cell loss rate, d. When comparing the PK/PD parameters between individuals infected with HCV genotypes 1 and 3; these authors concluded that EC50 is significantly lower, and that d is significantly increased in genotype 3-infected individuals. In addition, they did not identify significant differences in the viral kinetic and pharmacodynamic parameters between Caucasians and African Americans. An important issue is the statistical technique used in the analysis of PK/PD data. In general, two statistical approaches, the two-stage nonlinear least squares approach (individual PK), and the nonlinear mixed effects model (population PK) (Table 1) have been established for the analysis of PK/PD data. Briefly, the two-stage (individual PK) approach requires that a large number of observations are obtained for each patient at the same time points in order to obtain parameter estimates. The first stage of this procedure consists of using nonlinear regression (nonlinear least squares) to estimate the PK/PD parameters for each individual separately. In the second stage, these parameters are summarized through computation of descriptive statistics (means, variances), and a subsequent analysis of dependencies between the estimated parameters and patient-related covariates can be performed. The three studies discussed above [14,15,18] used
this statistical technique. An alternative approach for analysis of PK/PD data is based on the nonlinear mixed effects models. This method is especially appropriate in sparse or unbalanced data situations (i.e., patients may have different sampling times or may have missed some measurements) and when the investigator is interested in making generalizations about the population from which the study subjects were randomly selected. In contrast to the two-stage approach, in this procedure the population parameters are estimated directly, and inter and intra-subject variability is incorporated into the model. As the presence of missing values is a common problem in studies of HCVinfected persons because of poor compliance among some HCVinfected populations, the population PK approach may be a useful tool for analyzing such data. The US Food and Drug Administration has issued guidelines concerning the use of population PK in the process of drug development [19]. In conclusion, identifying PK/PD models that describe the relationship between the concentration of the drug in the blood, the dose of the drug, other PK parameters and covariates is a very important aspect in the area of HCV therapeutics. Besides IFN, exploring differences in RBV pharmacokinetics, as has been done in various studies [17,20], and the PK/PD of directly acting antivirals, are additional considerations that may differentiate SVR and NR patients. Explaining inter and intra-subject variability in the PK of a drug, or in viral kinetics, is a central goal of the analysis of PK/PD parameters. Studies investigating the mechanism of action of IFN performed to date have not identified PK parameters that significantly differ between SVR and NR patients. In contrast, some PD parameters have been identified with the ability to differentiate between these two patient groups. Dahari et al. investigated the important problem of modeling the pharmacokinetics of PEG-IFN alfa-2a and the HCV viral kinetics in HIV/HCV co-infected patients, which is motivated by the widespread use of PEGIFN alfa-2a in clinical practice. The investigators found that the maximum PEG-IFN effectiveness during the first dose, and the infected cell loss rate can differentiate SVR from NR patients. Further research is needed to confirm the findings concerning the use of viral kinetic parameters as prognostic tools of treatment outcome and to determine their role in patient management.
Financial disclosure Dr. Andrew Talal has received research support from Schering Plough (presently Merck) and Merck, and research support and is on the speaker’s bureau of Genentech, a member of the Roche group.
Table 1. Summary of the characteristics of the population (nonlinear mixed effects model) pharmacokinetics and the individual two-stage (nonlinear least squares method) pharmacokinetic approach. Individual Two-Stage Pharmacokinetics
Population Pharmacokinetics
Focus on individual Tools: nonlinear regression model based upon nonlinear least squares method Data from each individual are modeled separately Needs rich data for each subject
Focus on population Tools: nonlinear mixed effects model Data from all individuals are modeled simultaneously Handles dense, sparse, and unbalanced data (even one observation per subject) Accounts for the different sources of variability (between-subject, within-subject) Important factors such as demographic, environmental, pathophysiological, etc., which explain subject variability, can be identified Can be used also for individualized inference
Journal of Hepatology 2010 vol. 53 j 418–420
419
Editorial References [1] Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C virus infection. Lancet Infect Dis 2005;5:558–567. [2] Davis GL, Keeffe EB, Balart LA. Advances in liver disease: highlights from the 56th Annual Meeting of the American Association for the Study of Liver Disease. Rev Gastroenterol Disord 2006;6:48–61. [3] Kadam JS, Talal AH. Changing treatment paradigms: hepatitis C virus in HIVinfected patients. AIDS Patient Care STDS 2007;21:154–168. [4] Gonzalez SA, Talal AH. Hepatitis C virus in human immunodeficiency virusinfected individuals: an emerging comorbidity with significant implications. Semin Liver Dis 2003;23:149–166. [5] Foster GR. Pegylated interferons for the treatment of chronic hepatitis C: pharmacological and clinical differences between peginterferon-alpha-2a and peginterferon-alpha-2b. Drugs 2010;70:147–165. [6] Ghany MG, Strader DB, Thomas DL, Seeff LB. Diagnosis, management, and treatment of hepatitis C: an update. Hepatology 2009;49:1335–1374. [7] Neumann AU, Lam NP, Dahari H, Gretch DR, Wiley TE, Layden TJ, et al. Hepatitis C viral dynamics in vivo and the antiviral efficacy of interferon-a therapy. Science 1998;282:103–107. [8] Talal AH, Shata MT, Markatou M, Dorante G, Chadburn A, Koch R, et al. Virus dynamics and immune responses during treatment in patients coinfected with hepatitis C and HIV. J Acquir Immune Defic Syndr 2004;35:103–113. [9] Glue P, Fang JW, Rouzier-Panis R, Raffanel C, Sabo R, Gupta SK, et al. Pegylated interferon-alpha2b: pharmacokinetics, pharmacodynamics, safety, and preliminary efficacy data. Clin Pharmacol Ther 2000;68:556–567. [10] Powers KA, Dixit NM, Ribeiro RM, Golia P, Talal AH, Perelson AS. Modeling viral and drug kinetics: hepatitis C virus treatment with pegylated interferon alfa-2b. Semin Liver Dis 2003;23 (Suppl. 1):13–18. [11] Dahari H, Shudo E, Ribeiro RM, Perelson AS. Mathematical modeling of HCV infection and treatment. Methods Mol Biol 2009;510:439–453.
420
[12] Bonate PL. Pharmacokinetic–pharmacodynamic modeling and simulation. New York, NY: Springer; 2006. [13] Shudo E, Ribeiro RM, Perelson AS. Modeling HCV kinetics under therapy using PK and PD information. Expert Opin Drug Metab Toxicol 2009;5:321–332. [14] Talal AH, Ribeiro RM, Powers KA, Grace M, Cullen C, Hussain M, et al. Pharmacodynamics of PEG-IFN alpha differentiate HIV/HCV coinfected sustained virological responders from nonresponders. Hepatology 2006;43:943–953. [15] Rozenberg L, Haagmans BL, Neumann AU, Chen G, McLaughlin M, LevyDrummer RS, et al. Therapeutic response to peg-IFN-alpha-2b and ribavirin in HIV/HCV co-infected African-American and Caucasian patients as a function of HCV viral kinetics and interferon pharmacodynamics. AIDS 2009;23:2439–2450. [16] Asahina Y, Izumi N, Umeda N, Hosokawa T, Ueda K, Doi F, et al. Pharmacokinetics and enhanced PKR response in patients with chronic hepatitis C treated with pegylated interferon alpha-2b and ribavirin. J Viral Hepat 2007;14:396–403. [17] Nicot F, Legrand-Abravanel F, Lafont T, Dubois M, Saune K, Pasquier C, et al. Serum concentrations of ribavirin and pegylated interferon and viral responses in patients infected with HIV and HCV. J Med Virol 2008;80:1523–1529. [18] Dahari H, Affonso de Araujo ES, Haagmans BL, Layden TJ, Cotler SJ, Barone AA, et al. Pharmacodynamics of PEG-IFN alpha-2a in HIV/HCV co-infected patients: Implications for treatment outcomes. J Hepatol 2010;53:460–467. [19] Food and Drug Administration. Guidance for industry: population pharmacokinetics. Rockville, MD: US Department of Health and Human Services; 1999. [20] Dahari H, Markatou M, Zeremski M, Haller I, Ribeiro RM, Licholai T, et al. Early ribavirin pharmacokinetics, HCV RNA and alanine aminotransferase kinetics in HIV/HCV co-infected patients during treatment with pegylated interferon and ribavirin. J Hepatol 2007;47:23–30.
Journal of Hepatology 2010 vol. 53 j 418–420