Population Pharmacokinetic Models and Individualized Bayesian Dose Optimization in HIV-Infected Patients Michael Neely

1. Poster #71 Population Pharmacokinetic Models and Individualized Bayesian Dose Optimization in HIV-Infected Patients Michael Neely MD and Roger Jelliffe MD Los Angeles, California, USA Laboratory of Applied Pharmacokinetics www.lapk.org Introduction Standard antiretroviral dosing is size-based for children and fixed for adults. This may be problematic for an individual patient. If the patient is not “average” then the “average” dose may not achieve the efficacy goal, or may be toxic. Dose-dependent and dose-independent toxicity are difficult to distinguish. Sub-optimal adherence may be undetected prior to development of virologic failure and resistance; alternatively, sub-optimal dosing may be mistaken for poor adherence. There is little or no information on when to change from pediatric to adult dosing. There is no ability to adjust the recommended dose in an individual patient with any assurance of success. All of these problems can cause poor outcomes: viral resistance, toxicity, unnecessary regimen changes or combinations of these. There are now several clinical investigations [1-7] and numerous reviews or position papers, e.g. [8-13], that affirm the usefulness of incorporating measurement of antiretroviral drug concentrations into the clinical management of selected HIV-infected patients. Here, we report four case vignettes of HIV-infected patients whose therapy was optimized using an approach and software developed in our lab. These vignettes illustrate the benefit of dose individualization and optimization in common clinical scenarios. Methods Antiretroviral population pharmacokinetic models, each with oral absorption into a single compartment, were constructed using the MM-USC*PACK software collection [14] (available at www.lapk.org). PK parameter estimates obtained or derived from over 30 published studies were used to generate, by Monte Carlo simulation with noise, populations of n=50 for each drug. Final parameter values for the models in this report are shown in Table 1. Simulated populations were then analyzed using the Non-Parametric Adaptive Grid (NPAG) program in MM-USC*PACK [15] to generate a population PK model for each antiretroviral drug. The models were applied as part of comprehensive clinical care to outpatients in our HIV clinic using MM-USC*PACK’s multiple-model, Bayesian adaptive control to individualize therapy. Results/Case Histories Patient 1 was a 13-year old antiretroviral-naïve African boy (Tanner stage 2), started on an efavirenz-based regimen at the recommended dose for his age. After 2 weeks, his mother reported that he was too drowsy to attend school, more severe than the typical transient drowsiness after starting efavirenz. Suspecting that he was a genetic slow metabolizer, we empirically reduced his dose by half, and a week later measured a serum concentration of 1.37 mg/L 22 hours after his previous dose. His 24-h trough concentration was predicted to remain above a target of 1 mg/L [22]; therefore, he continued on this dose. A follow-up sample confirmed his therapeutic concentrations on 50% dose, and he has maintained an undetectable HIV viral load with no further somnolence for the past 2 years. (Figure 1A) Patient 2 was a 10 year-old girl (Tanner stage 2) who weighed 30.7 kg. Based on the standard pediatric dose of 55 mg/kg, given formulation limitations, she was prescribed 1875 mg. Since the recommended “maximum” is the adult dose of 1250 mg, we measured a random serum concentration of 4.9 mg/L 4 hours after her previous dose to ensure that she was not in a toxic range. Her predicted peak concentration was 5.5 mg/L and her 12-h trough was 2.1 mg/L, both within a suggested therapeutic range of 1 - 6 mg/L [22], and she never demonstrated toxicity despite the continued “supra-maximal” dose. (Figure 1B) Patient 3 was a 45 year-old woman (Tanner stage 5) with a long history of medication intolerance. She was started on a fos-amprenavir containing regimen (without ritonavir), 2 x 700 mg tablets twice daily. After starting the new regimen, she complained of daytime fatigue, which she attributed to the morning dose. She enquired about taking the entire dose at night. Prior to making changes, we measured a serum amprenavir concentration of 1.4 mg/L 4.5 hours after her previous dose. Modeling suggested that although 2800 mg once daily would not maintain her trough concentration above the minimum target of 0.23 mg/L [22], a regimen of 1 tablet at 8am followed by 3 tablets at 6pm (a 10-14 hour schedule) would achieve this goal. She was changed to the latter regimen, and has achieved an undetectable viral load without any further complaints of fatigue. A follow-up level of 0.9 mg/L 4 hours after the morning dose on the new regimen confirmed that her predicted troughs were likely to be therapeutic. (Figure 1C) Patient 4 was a 14-year old male (Tanner stage 4) with poor adherence and limited treatment options. To encourage better adherence, he was changed to a once-daily regimen that included atazanavir given in combination with low-dose ritonavir. There were no published pediatric PK data at the time. The usual adult dose of atazanavir is 300 mg when given with ritonavir, so he was started on 200 mg based on his small size. We obtained a random atazanavir concentration of 0.782 mg/L 18 hours after his previous dose. His predicted trough concentration was 0.380 mg/L, above the minimum target of 0.150 mg/L [23], so the dose was continued. He initially achieved an undetectable viral load but persistently poor adherence (self-reported) allowed his viral load to rebound partially to about 2000 copies/mL, despite a second confirmed therapeutic concentration of atazanavir. (Figure 1D) B A C D Figure 1 – MM-USC*PACK output for each patient. Black lines are weighted average Bayesian-posterior predicted concentrations, red dots are measured serum concentrations, and blue lines are dose events. Intervals between measured serum concentrations have been compressed for clarity. Target concentrations have been added as a reference. A) Patient 1, efavirenz; B) Patient 2, nelfinavir; C) Patient 3; amprenavir given as fos-amprenavir; D) Patient 4, atazanavir. Conclusions Our methods and software for converting reported PK data into population PK models can be used locally to optimize safety and efficacy in individual patients. Successful therapeutic drug management tailored to patients representing four scenarios was presented: 1) altered metabolism 2) supra- “maximal” doses related to pediatric vs. adult dosing 3) altered dosing schedules 4) dosing with limited relevant published PK data Individualized Bayesian adaptive control can move population PK/PD models beyond their current primary domain of drug development to the optimized care of patients. Table 1: Mean model parameter values. Models based on pediatric studies were used for patients with a Tanner Sexual Maturity Rating stage ≤3. References [1] Fletcher CV, Anderson PL, Kakuda TN et al. Concentration-controlled compared with conventional antiretroviral therapy for HIV infection. AIDS. 2002; 16(4):551-560. [2] Best BM, Goicoechea M, Witt MD et al. A randomized controlled trial of therapeutic drug monitoring in treatment-naive and -experienced HIV-1-infected patients. J Acquir Immune Defic Syndr 2007 Dec 1. 1946;433-442. [3] Boyd MA, Siangphoe U, Ruxrungtham K et al. The use of pharmacokinetically guided indinavir dose reductions in the management of indinavir-associated renal toxicity. J Antimicrob Chemother 2006 Jun. 1957;1161-1167. [4] Burger D, Hugen P, Reiss P et al. Therapeutic drug monitoring of nelfinavir and indinavir in treatment-naive HIV-1-infected individuals. AIDS 2003 May 23. 1917;1157-1165. [5] Cleijsen RM, van de Ende ME, Kroon FP et al. Therapeutic drug monitoring of the HIV protease inhibitor atazanavir in clinical practice. J Antimicrob Chemother 2007 Oct. 1960;897-900. [6] Durant J, Clevenbergh P, Garraffo R et al. Importance of protease inhibitor plasma levels in HIV-infected patients treated with genotypic-guided therapy: pharmacological data from the Viradapt Study. AIDS 2000 Jul 7. 1914;1333-1339. [7] Wasmuth JC, Lambertz I, Voigt E et al. Maintenance of indinavir by dose adjustment in HIV-1-infected patients with indinavir-related toxicity. Eur J Clin Pharmacol 2007 Oct. 1963;901-908. [8] Dahri K, Ensom MH. Efavirenz and nevirapine in HIV-1 infection : is there a role for clinical pharmacokinetic monitoring? Clin Pharmacokinet 2007. 1946;109-132. [9] Fraaij PL, Rakhmanina N, Burger DM et al. Therapeutic drug monitoring in children with HIV/AIDS. Ther Drug Monit 2004 Apr. 1926;122-126. [10] Wertheimer BZ, Freedberg KA, Walensky RP et al. Therapeutic drug monitoring in HIV treatment: a literature review. HIV Clin Trials 2006 Mar -Apr. 2007;59-69. [11] US Department of Health and Human Services: Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Available at: http://www.aidsinfo.nih.gov (Last Update: May 4, 2006) Accessed: Jul 7, 2006 [12] Acosta EP, Gerber JG. Position paper on therapeutic drug monitoring of antiretroviral agents. AIDS Res Hum Retroviruses. 2002; 18(12):825-834. [13] Acosta EP, King JR. Methods for integration of pharmacokinetic and phenotypic information in the treatment of infection with human immunodeficiency virus. Clin Infect Dis. 2003; 36(3):373-377. [14] Jelliffe RW. The USC*PACK PC programs for population pharmacokinetic modeling, modeling of large kinetic/dynamic systems, and adaptive control of drug dosage regimens. Proc Annu Symp Comput Appl Med Care. 1991;922-924. [15] Bustad A, Terziivanov D, Leary R et al. Parametric and nonparametric population methods: their comparative performance in analysing a clinical dataset and two Monte Carlo simulation studies. Clin Pharmacokinet. 2006; 45(4):365-383. [16] Hirt D, Urien S, Jullien V et al. Age-related effects on nelfinavir and M8 pharmacokinetics: a population study with 182 children. Antimicrob Agents Chemother. 2006; 50(3):910-916. [17] Csajka C, Marzolini C, Fattinger K et al. Population pharmacokinetics and effects of efavirenz in patients with human immunodeficiency virus infection. Clin Pharmacol Ther. 2003; 73(1):20-30. [18] Starr SE, Fletcher CV, Spector SA et al. Efavirenz liquid formulation in human immunodeficiency virus-infected children. Pediatr Infect Dis J. 2002; 21(7):659-663. [19] Wire MB, Shelton MJ, Studenberg S. Fosamprenavir : clinical pharmacokinetics and drug interactions of the amprenavir prodrug. Clin Pharmacokinet. 2006; 45(2):137-168. [20] Kashuba AD, Tierney C, Downey GF et al. Combining fosamprenavir with lopinavir/ritonavir substantially reduces amprenavir and lopinavir exposure: ACTG protocol A5143 results. AIDS. 2005; 19(2):145-152. [21] Colombo S, Buclin T, Cavassini M et al. Population Pharmacokinetics of Atazanavir in Patients with Human Immunodeficiency Virus Infection. Antimicrob Agents Chemother. 2006; 50(11):3801-3808. [22] Kappelhoff BS, Crommentuyn KM, de Maat MM et al. Practical guidelines to interpret plasma concentrations of antiretroviral drugs. Clin Pharmacokinet. 2004; 43(13):845-853. [23] Gonzalez de Requena, D. et al. Atazanavir Ctrough is associated with efficacy and safety: definition of therapeutic range. Poster Abstracts, 12th Conference on Retroviruses and Opportunistic Infections in Boston, MA. Abstract #645, 2005 Acknowledgements This work was supported by Department of Health and Human Services, NIH-NIAID, 1 K23 AI076106-01 and NIH-NBIB, R01 EB005803-01A1