JAIDS Journal of Acquired Immune Deficiency Syndromes:
A Randomized Controlled Trial of Therapeutic Drug Monitoring in Treatment-Naive and -Experienced HIV-1-Infected Patients
Best, Brookie M PharmD*; Goicoechea, Miguel MD*; Witt, Mallory D MD†; Miller, Loren MD†; Daar, Eric S MD†; Diamond, Catherine MD‡; Tilles, Jeremiah G MD‡; Kemper, Carol A MD§; Larsen, Robert MD∥; Holland, Diane T MPhil*; Sun, Shelly MS*; Jain, Sonia PhD*; Wagner, Glenn PhD¶; Capparelli, Edmund V PharmD*; McCutchan, J Allen MD*; Haubrich, Richard H MD*; the California Collaborative Treatment Group 578 Study Team
From the *University of California, San Diego, San Diego, CA; †Harbor-University of California, Los Angeles (UCLA) Medical Center and the Los Angeles Biomedical Research Institute, David Geffen School of Medicine at University of California, Los Angeles, Los Angeles, CA; ‡University of California, Irvine, Irvine, CA; §Santa Clara Valley Medical Center, CA; ∥University of Southern California, Los Angeles, CA; and ¶RAND Corporation, Santa Monica, CA.
Received for publication March 27, 2007; accepted July 30, 2007.
Supported by funds from the Universitywide AIDS Research Program of the University of California (center grants CC99-SD003, CC02-SD-003, and CH05-SD-607-005), the National Institute of Mental Health (RO1MH61695), the National Institute of Allergy and Infectious Disease (AI064086 and K23 AI066901), the National Institute for Child Health and Human Development (5U10 HD031318), and the University of California, San Diego Center for AIDS Research, National Institute for Allergy and Infectious Diseases (AI36214). Quest Diagnostics Laboratories (San Clemente, CA) and Monogram Biosciences (San Francisco, CA) provided all routine laboratory assays.
Presented in part at the 13th Conference on Retroviruses and Opportunistic Infections, Denver, CO, February 5-8, 2006 (abstract 589), and 12th Conference on Retroviruses and Opportunistic Infections, Boston, MA, February 22-25, 2005 (abstract 640).
Correspondence to: Brookie M. Best, PharmD, University of California, San Diego, Skaggs School of Pharmacy and Pharmaceutical Sciences, School of Medicine, Department of Pediatrics, 200 West Arbor Drive, MC 8214, San Diego, CA 92013-8214 (e-mail: email@example.com).
Objective: To improve the utility of therapeutic drug monitoring (TDM) by defining the proportion of patients with and predictors of above or below target protease inhibitor (PI) or nonnucleoside reverse transcriptase inhibitor (NNRTI) concentrations.
Methods: This 48-week, multicenter, open-label clinical trial randomized patients to TDM versus standard of care (SOC). Serial pharmacokinetics, including a week-2 3-sample sparse collection, and expert committee TDM recommendations were given to TDM-arm patients' providers.
Results: Seventy-four (39%) of 190 patients had week-2 concentrations outside of targets and 122 (64%) of 190 had nontarget exposure at least once over 48 weeks. Providers accepted 75% of TDM recommendations. Among patients with below-target concentrations, more TDM-arm than SOC-arm patients achieved targets (65% vs. 45%; P = 0.09). Increased body weight and efavirenz or lopinavir/ritonavir use were significant predictors of nontarget concentrations. Patients at target and patients who achieved targets after TDM-directed dose modifications trended toward greater viral load reductions at week 48 than patients with below-target exposures (HIV RNA reductions: 2.4, 2.3, and 1.9 log10 copies/mL, respectively; P = 0.09).
Conclusions: Most patients had nontarget PI and/or NNRTI concentrations over 48 weeks. TDM recommendations were well accepted and improved exposure. Patients below TDM targets trended toward worse virologic response.
Although highly active antiretroviral therapy (HAART) for HIV infection reduces morbidity and mortality,1 regimens fail to suppress HIV replication in many HIV-infected patients. Low concentrations of protease inhibitors (PIs) and nonnucleoside reverse transcriptase inhibitors (NNRTIs) are associated with increased risk of treatment failure.2-8 Likewise, excessive antiretroviral (ARV) dose or drug concentrations increase the risk of treatment-related toxicity.9-12 Therapeutic drug monitoring (TDM) may identify and correct excessively high or low PI and/or NNRTI concentrations, thereby improving clinical response and/or minimizing toxicity. Treatment guidelines in the United States and Europe recommend using TDM to help optimize ARV therapy only in selected patients, such as those with renal or hepatic impairment, during pregnancy, with potentially concentration-related toxicities, or with suspected significant drug-drug or drug-food interactions.13,14
Studies describing the use of TDM in large unselected groups of patients to guide PI and/or NNRTI dosing have yielded conflicting results.15-20 Issues that may have an impact on the utility of TDM in routine clinical practice in the United States include the following: (1) how to incorporate clinical (CD4, HIV RNA, adherence) and genetic (P-glycoprotein and cytochrome P450 alleles) factors that may predict suboptimal plasma concentrations, (2) defining the best methodology for measuring drug concentrations (ie, single trough measure, sparse sample, intensive sampling), (3) standardizing the laboratory methodology for measuring ARV concentrations, (4) selecting the appropriate pharmacologic parameter to evaluate (eg, minimum concentration [Cmin], maximum concentration [Cmax], oral clearance [Cl/F]), (5) providing expert interpretation of the ARV concentrations, (6) recommending practical pharmacologic changes, and (7) determining how best to use viral susceptibility measures.21-28
Investigating the benefit of TDM has challenges. In clinical trials of unselected patients on ARV therapy, only a subset of patients would be expected to have suboptimal ARV concentrations and require dose adjustment. Thus, the virologic differences between a TDM and control arm would be small, and to detect significant differences in virologic efficacy between groups, a large sample size would be required. For example, as detailed by Khoo and colleagues,19 approximately 1000 to 2000 subjects would need to be randomized assuming that 30% of subjects would need dose adjustment, all dose adjustments are 100% successful in achieving target concentrations, virologic response would improve in half of the subjects with dose adjustments, and physician acceptance of dose change recommendations ranges from 35% to 75%. Such a resource-intensive trial would be unfeasible to perform for most clinical research groups. Therefore, to address questions surrounding TDM of ARVs in a focused and feasible manner, we conducted a clinical trial whose primary objectives were (1) to define the proportion of patients for whom an expert committee recommended a pharmacologic intervention to change PI and/or NNRTI exposure and (2) to define patient factors that predicted the need for a TDM recommendation.
MATERIALS AND METHODS
This was a randomized controlled, open-label, factorial TDM study in patients initiating or changing a PI- or NNRTI-based ARV regimen. The subject's medical provider selected ARV therapy based on CD4 cell count, HIV RNA level, prior treatment history, and HIV phenotype resistance assay. Central factorial randomization was computer generated, blocked (block size of 9), and stratified by clinic site (5 sites) and prior treatment (naive vs. experienced). The computer randomized patients at a ratio of 2:1 to TDM versus standard of care (SOC) and at a ratio of 1:1:1 to 1 of 2 adherence interventions or usual adherence care.29 Because no consensus exists about what represents an optimal ARV concentration, a committee of ARV pharmacologists and HIV treatment specialists (hereafter referred to as the “expert committee”), blinded to subjects' TDM randomization assignments, reviewed clinical, laboratory, and modeled concentration data at designated visits throughout the study and made recommendations on whether and how to alter ARV dosing. Committee recommendations were forwarded to the California Collaborative Treatment Group (CCTG) Data Center, which distributed them to TDM-arm patient providers only.
All patients gave informed consent, and research review committees at each institution approved the study. Inclusion criteria were plasma HIV RNA level >3000 copies/mL; intention to start a new PI- or NNRTI-based ARV regimen; age ≥18 years; life expectancy >12 months; and, in female patients, willingness to use birth control. Patients could be ARV naive, treatment experienced and on a stable regimen or treatment experienced and off ARV therapy for ≥2 months. Experienced patients acknowledged prior adherence lapses or a belief that they could benefit from adherence training. Exclusion criteria included active opportunistic infection or cancer, chronic diarrhea, malabsorption, treatment-limiting toxicity to the current ARV regimen, severe cognitive impairment, active drug or alcohol abuse, and currently pregnant or breast-feeding women. Any concomitant medications were allowed; those that had documented or suspected significant interactions with licensed PIs and/or NNRTIs as identified in the PI and/or NNRTI prescribing information were recorded.30-40
The CCTG Data Center coordinated real-time data distribution among the laboratories, the expert committee, and the sites. Adverse events were graded according to AIDS Clinical Trials Group (ACTG) toxicity tables.41 Phenotype resistance testing (Monogram Biosciences, San Francisco, CA) was done at screening and at regimen failures. Laboratory assessments at study weeks 2, 4, 6, 12, 18, 24, 32, 40, and 48 were performed centrally (Quest Diagnostics Laboratories, San Clemente, CA) and included CD4 lymphocyte count, HIV RNA level, chemistry panel, complete blood cell count (CBC) with differential and platelets, and fasting lipid profile (total cholesterol, triglycerides, high-density lipoprotein [HDL], low-density lipoprotein [LDL; calculated and actual), and very-low-density lipoprotein [VLDL]). Electronic bottle caps (medication event monitoring system [MEMS]) measured adherence to the subject's PI or NNRTI in all subjects.29
Two weeks after initiating or changing their regimen, plasma samples for PI and NNRTI concentrations in all subjects were drawn before dosing and 2 and 4 hours after witnessed dosing (sparse sample). Single random samples were drawn at all other visits. Validated reverse-phase high-performance liquid chromatography methods assayed currently licensed PIs and NNRTIs, including efavirenz, nevirapine, amprenavir, atazanavir, indinavir, lopinavir, nelfinavir, ritonavir, and saquinavir, at the University of California, San Diego (UCSD) Pediatric Pharmacology Laboratory. Control validation precision was <13% coefficient of variation and within 13% deviation from expected for accuracy. The laboratory successfully completed 7 rounds of the Pediatric AIDS Clinical Trials Group Proficiency Testing program over the course of the study.42
The posthoc subroutine in the computer program NONMEM (nonlinear mixed effects modeling), version 5 (ICON Development Solutions, Ellicott City, MD),43 estimated individual pharmacokinetic parameters with a Bayesian nonlinear curve-fitting approach. One-compartment models with pharmacokinetic parameters set to values derived from published studies were used. To ensure that each model could adequately describe the pharmacokinetics of that ARV, data sets of 200 subjects were simulated for each model, and these simulations for all PIs and NNRTIs produced expected values for the median Cmax and Cmin. To account for a lag time in PI and/or NNRTI absorption (defined as observed concentrations decreasing from before dosing to 2 hours after witnessed dosing), each subject was assigned a “yes/no” variable according to whether or not an absorption lag was observed during the sparse sampling. Each pharmacokinetic model had a lag time pharmacokinetic parameter in NONMEM, which would be incorporated into the individual pharmacokinetic parameter predictions only for those subjects with a documented absorption lag. Individual subjects' predicted steady-state average [Dose/(Cl/F·τ)], 2-hour, 4-hour, and 12- or 24-hour (trough concentration [Ctrough]) postdose concentrations were determined for all monitored PIs and NNRTIs. Pharmacokinetic parameters were estimated from the 2-week evaluation (3 samples), and were re-estimated from samples collected between 2- and 12-week, 12- and 24-week, and 24- and 48-week visits and after supplemental evaluations attributable to regimen changes. Each patient's Ctrough and average concentration were compared with that individual's viral baseline 50% inhibitory concentration (IC50) measured in 10% fetal calf serum (PhenoSense; Monogram Biosciences44), generating average concentration and Cmin inhibitory quotients (IQs). In subjects with additional viral phenotype testing performed during the course of the study, the PI and/or NNRTI concentrations were compared with the most recent IC50. To correct for protein binding, the patient's predicted PI and/or NNRTI concentrations were multiplied by the corresponding ARV free fraction30-40 for comparison with the IC50.
The expert committee recommended an increase, a decrease, or no change in exposure for each PI and NNRTI, along with suggested means to implement the recommendation (eg, improve adherence, take with food, increase dose, add ritonavir; (Fig. 1). The data center forwarded TDM-arm subject recommendations 4 weeks after study weeks 2, 12, and 24. Providers elected to follow recommendations as appropriate, recording changes made and their agreement with recommendations. The pharmacologic target for all study drugs was an average concentration greater than the average concentrations calculated from published oral clearance values (Table 1).30-40,45-56 In patients with drug-resistant virus, the pharmacologic target may have been higher so as to achieve average concentrations at least 4 times higher than viral IC50 concentrations (average concentration: IQ ≥4 to achieve concentrations that would suppress approximately 95% of viral replication). Additionally, in subjects with toxicity or slow virologic response, the Cmin and Cmax were considered when deciding on dose modification recommendations. All dose change recommendations conformed to commercially available dosage forms. No maximum or minimum doses were specified by this study, and a single subject could potentially have multiple dose changes.
The a priori definition of evaluable patients included those who were randomized and started a new regimen at baseline, hereafter referred to as the modified intent-to-treat (ITT) group. The primary hypothesis was that the true proportion of patients with nontarget PI and/or NNRTI concentrations triggering a TDM recommendation to change exposure was 40%.
The primary endpoints were (1) the proportion of subjects who had a TDM recommendation to change PI and/or NNRTI dosing and (2) the proportion of recommended changes that were carried out (TDM arm only). Secondary endpoints included (1) the proportion of recommended interventions in the TDM arm that achieved the target concentrations, (2) the proportion of patients with an HIV RNA level <400 copies/mL at 48 weeks compared between arms, and (3) the proportion of patients experiencing grade III/IV and/or treatment-limiting toxicity compared between arms.
The sample size was based on adequate power for the adherence intervention comparisons.29 With the calculated sample size of 270, the 95% confidence interval (CI) around an estimated TDM recommendation rate of 40% was 34% to 46%. Univariate analyses using χ2 (categoric) or ANOVA (continuous variables) compared factors for those with or without a TDM recommendation to change. Multivariate logistic regression models, with a TDM recommendation to change as the bivariate outcome variable, assessed the independent odds ratios of prognostic factors at weeks 2, 12, and 24. Variables included age, HIV RNA, race, prior regimen history, gender, current regimen, weight, albumin, body mass index (BMI), CD4 cell count, prior AIDS, adherence intervention group, and MEMS coverage. Virologic outcomes at weeks 24 and 48 and occurrence of treatment-limiting or grade III/IV toxicities were analyzed with t tests, Fisher exact tests, and χ2 tests, as appropriate.
Patients were enrolled between September 2001 and December 2003, with the final study visit in January 2005. Of 241 patients identified and screened during the study period, 230 were randomized (Fig. 2). Twenty-seven subjects missed the baseline visit, and 4 did not start a new ARV regimen; thus, the modified ITT population was 199 subjects. Twenty-five subjects discontinued within 24 weeks, and 27 discontinued between weeks 24 and 48. No differences were noted in time to dropout or absolute numbers of patients prematurely discontinuing between the TDM and SOC arms.
Demographics were similar between the TDM and SOC arms (Table 2) and also across the adherence intervention arms (data not shown). This group had advanced disease, with a median HIV RNA level >5.0 log10 copies/mL and CD4 cell counts <200 cells/mm3. More than two thirds of these subjects were treatment experienced. Most subjects were male and nonwhite.
Baseline ARV regimens included a ritonavir-boosted PI in more than half of the subjects, with similar regimens used in each arm (see Table 2). Common ARVs included tenofovir (66% of subjects), lopinavir/ritonavir (46%), and efavirenz (31%). Interpatient variability in concentrations was large. The median (interquartile range [IQR]) week 2 modeled Ctrough and peak and average concentrations expressed as a percentage of population-predicted values were 135% (82% to 201%), 91% (68% to 120%), and 111% (82% to 141%), with ranges from 0% to 743%, 7% to 255% and 18% to 308%, respectively. Intrapatient variability was also substantial, with median (IQR) dose-corrected percent differences in concentrations across visits of 86% (42% to 144%), 39% (18% to 68%), and 54% (27% to 97%) for Ctrough and peak and average concentrations, respectively. Median (range) intrapatient oral clearance from baseline to last study visit decreased for indinavir −21% (−45% to 6%); did not change by more than 5% for atazanavir, nevirapine, lopinavir and saquinavir; and increased for amprenavir 37% (−36% to 89%) and efavirenz 14% (−68% to 282%). A lag time in PI and/or NNRTI absorption was noted in 43 (22%) subjects taking a PI, with a mean concentration decrease of 30% ± 16% (range: 4% to 60%). Thirty-four of these 43 patients were taking lopinavir/ritonavir (ie, one third of subjects taking lopinavir/ritonavir had absorption lags).
Nine subjects never had a PI and/or NNRTI concentration drawn while on study. One hundred ninety subjects had at least 1 visit with PI and/or NNRTI concentrations determined, and 122 (64%) of 190 of these patients (89 in the TDM arm and 33 in the SOC arm) had a TDM recommendation to change PI and/or NNRTI exposure at least once during the study. At the week 2 study visit, 74 (39%) of 190 subjects (42% in the TDM arm [n = 54] and 32% in the SOC arm [n = 20]) had a recommendation to change PI and/or NNRTI exposure. The expert panel performed 502 separate TDM evaluations during the study, with 170 recommendations to change PI and/or NNRTI exposure in 122 subjects (34% of TDM evaluations, 95% CI: 30% to 38%). At week 48, an additional 145 TDM evaluations were performed, with 55 recommendations to change PI and/or NNRTI exposure. These were done to determine whether or not the patients met the pharmacologic targets but were not provided to sites, because patients had completed the study.
Among the 170 recommendations to change drug exposure, 9 were to decrease drug exposure, whereas 166 were to increase drug exposure. Several patients had recommendations to change more than 1 ARV at a single visit and were only counted once per visit in the 170 recommendations. The most common suggested change was to increase the dose of the NNRTI or primary PI (n = 111). Other suggestions included adding or increasing ritonavir (n = 7), modifying concomitant medications (n = 2), modifying diet (n = 1), decreasing dose (n = 4), discontinuing medication (n = 4), and counseling to improve adherence (n = 41). For lopinavir, the most common specific recommendation was to increase dose (n = 44 [53%]), followed closely by improving adherence (n = 35 [42%]). Doses were increased in 1 capsule (133/33-mg lopinavir/ritonavir) increments to 533/133 mg of lopinavir/ritonavir twice daily in 30 subjects, and to 666/166 mg of lopinavir/ritonavir twice daily in 2 subjects. Two subjects added supplemental ritonavir (100 mg) instead: the first to doses of 400/200 of lopinavir/ritonavir mg twice daily and the second to 800/300 mg of lopinavir/ritonavir once daily. Two subjects took decreased doses of 266/66 mg of lopinavir/ritonavir twice daily. In contrast, for NNRTIs, 52 (81%) subjects were to increase dose, with only 5 (8%) to improve adherence. For efavirenz specifically, 16 subjects took 800 mg once daily and 2 subjects further increased their doses to 1000 mg once daily. For other boosted PIs (excluding lopinavir), 12 (75%) subjects were to increase their dose, whereas only 1 (6%) was to improve adherence.
Clinicians implemented the TDM recommendations to change drug exposure 76% (95% CI: 68% to 84%) of the time. In this TDM group, 53 (60%) of 88 recommendations yielded target exposure at the next follow-up visit. In contrast, in the SOC arm, target concentrations were met at subsequent visits in subjects with nontarget exposure for only 16 (36%) of 45 recommendations to change drug exposure, in the absence of providers receiving the TDM reports. Patients with nontarget concentrations in the TDM arm subsequently achieved the pharmacologic target exposure more often, on average, than SOC-arm patients (65% vs. 45%; P = 0.09).
Predictors of Above- or Below-Target ARV Exposure
For the week 2 TDM evaluations, independent factors predicting nontarget PI and/or NNRTI concentrations included greater body weight, efavirenz use, and lopinavir use (Table 3). Patients who ever had a change recommended had similar predictors. Predictors at weeks 12 and 24 included higher BMI, lower baseline hemoglobin, male gender, prior HIV treatment experience, prior NNRTI use, and a previous recommendation to change. In the subset who achieved targeted exposure at week 12 after an initial recommendation to change exposure, adherence (therapeutic coverage at week 12 as measured by MEMS devices) was a significant predictor of success.
Safety and Tolerability
The most commonly reported adverse effect was diarrhea/loose stools. Other reported adverse effects included nausea, fatigue/malaise/asthenia, vomiting, sleeping problems/insomnia/dreams, flatulence/gas/distension, and dizziness/lightheadedness/fainting. Rates of provider-reported treatment-limiting toxicities or grade III/IV toxicities did not differ between the TDM and SOC groups at weeks 24 and 48, or within the subset of patients for whom a dose increase was recommended at the initial TDM evaluation. The time to treatment-limiting or grade III/IV toxicities also did not differ. When patients were grouped according to presence or absence of treatment-limiting toxicities, median trough percent of population-predicted concentrations were similar (161% vs. 155%), but maximum PI and/or NNRTI concentrations were higher in those with treatment-limiting toxicities (126% vs. 100%; P = 0.04). No differences were seen in pharmacokinetic parameters between the groups of patients with or without grade III/IV toxicity.
This study enrolled a heterogeneous population and was not designed to detect a virologic difference between TDM and SOC arms, and no significant differences in virologic outcomes were observed. The proportions of patients with viral loads <400 copies/mL and <50 copies/mL were similar between the groups at the study midpoint and end. At week 24, 34% of subjects in the TDM arm and 39% of subjects in the SOC arm had an HIV RNA level <400 copies/mL. At week 48, the proportions were 31% in the TDM arm and 28% in the SOC arm. An HIV RNA level <50 copies/mL was achieved in the TDM arm in 13% of subjects and in 11% of SOC subjects at week 24; at week 48, 11% and 14% of TDM and SOC patients achieved this viral outcome, respectively. Patients who dropped out of the study prematurely were considered failures.
To explore the impact of TDM-directed augmented exposure on virologic response better, a posthoc analysis stratified all subjects from both arms into 3 groups: (1) at or above the target concentrations at all TDM evaluations (n = 68), (2) nontarget concentrations triggered a recommendation to change PI and/or NNRTI exposure and achieved target at next follow-up visit (n = 53), and (3) nontarget concentrations that never achieved target (n = 32). Subjects who had a recommendation to change drug exposure but did not have a follow-up study visit to assess achievement of target concentrations were excluded (n = 37). In a linear regression model adjusting for baseline RNA and treatment experience, groups 1, 2, and 3 had mean changes from baseline to week 48 HIV RNA log10 of −2.4, −2.3 and −1.9, respectively (P = 0.09).
This study describes the use of TDM for a heterogeneous group of treatment-naive and experienced HIV-infected subjects starting a new PI- or NNRTI-based regimen. One third of subjects had nontarget concentrations 2 weeks after starting therapy, similar to that observed in prior TDM studies.16,19,20 Notably, nearly two thirds of subjects had nontarget PI and/or NNRTI exposure at least once during the 48-week study, indicating that plasma ARV concentrations are dynamic and may benefit from repeated and initial pharmacologic monitoring.
Greater baseline body weight and use of efavirenz or lopinavir/ritonavir were statistically significant independent predictors of nontarget PI and/or NNRTI concentrations. Our finding that increasing body weight is related to nontarget concentrations is consistent with the finding by van der Leur and colleagues57 that body weight was significantly inversely related to lopinavir concentration ratios (individual concentrations divided by expected concentrations). Whereas specific weight cutoffs were not assessed, a BMI >30 would be 1 SD above the mean body weight observed in our study population and suggests that TDM of PIs or NNRTIs be considered. Prediction of need for TDM by efavirenz and lopinavir use may be attributable simply to the pronounced pharmacokinetic variation observed in this group. Another potential explanation is that although most enzyme induction/inhibition by efavirenz and ritonavir occurs within a few weeks, additional minor induction/inhibition of metabolic activity may occur over longer times, eventually significantly affecting ARV concentrations. For efavirenz, this is supported by our observation that oral clearance within a patient increased over time. Our observation that amprenavir oral clearance increased over time as well is less straightforward. Oral clearances of ritonavir-boosted PIs increased, decreased, or did not change over time in this study. The effects of the primary PIs may influence this complex interplay of enzyme modification by ritonavir, or differential adherence changes between boosted regimens over time could explain these oral clearance change differences.
Our subjects and medical providers were receptive to the expert panel's recommendations, with three quarters of the recommendations implemented, similar to several other TDM studies.15-18 In contrast, 2 previous TDM trials had 30% to 35% acceptance rates.19,20 One explanation for our high provider acceptance rate was that we incorporated extensive clinical information in the dose recommendation process. This resulted in optimizing overall therapy for each patient rather than optimizing a drug concentration in isolation. Given the variability of concentrations with time, some patients in the SOC arm achieved target exposure without changing their ARV regimen. Therapy optimization in TDM-arm patients achieved target concentrations nearly twice as often as in SOC-arm patients, however.
Similar to our study, the 2 randomized controlled trials of TDM that showed virologic benefit followed patients for more than 48 weeks and used more robust pharmacologic metrics than trough thresholds to adjust therapy.16,18 In addition, Fletcher and colleagues18 reported the only TDM study to monitor all active ARVs and ARV metabolites that contribute to viral suppression in the regimen. The study regimens were unboosted indinavir and nelfinavir given to treatment-naive patients. These regimens are not widely utilized, and the impact of TDM for treatment-experienced patients was not evaluated. Four randomized controlled TDM studies did not show virologic benefit (although other benefits were discussed). Unlike our study, 2 followed subjects for only 12 weeks and 2 had low acceptance rates of the pharmacologic recommendations,15,17,19,20 probably contributing to lack of power to detect an intervention difference in those studies.
One limitation of TDM is that target ARV concentrations are poorly delineated. Our targets were to exceed the predicted median population concentrations (potentially higher if viral resistance was present) rather than minimal trough values. With this more aggressive target, assessment for and absence of increased toxicity in the TDM arm were important safety outcomes. The results of this study must be interpreted in light of our somewhat aggressive targets. Current consensus recommendations for minimum target Ctrough are lower than the average concentration targets used in this study, and fewer subjects would be deemed “below target” with these less aggressive guidelines.58 For our most common study medications, lopinavir and efavirenz, recommended guideline troughs are 1 μg/mL, as compared with our study targets of 3.8 and 1.4 μg/mL. Considering all study PIs and NNRTIs, our targets were a median (range) of 4.2 (1.4 to 11) times higher than the consensus trough targets. These consensus guideline targets are for treatment-naive patients, however, and do not apply to the two thirds of our subjects who were treatment experienced. No current validated target concentrations exist for treatment-experienced patients beyond a consensus expert opinion that they need more aggressive concentration targets.
Another limitation is that TDM is not applied to all active components of an ARV regimen that contribute to viral suppression. Also, the ideal time to perform ARV TDM is unknown. Changes should be implemented quickly to avoid prolonged therapy with suboptimal doses; however, enzyme induction/inhibition takes time, and early concentrations may not reflect ARV exposure later in therapy.59 The time delay in implementing TDM in research settings may have diminished the potential benefit in this current study and prior studies,15-17 and faster clinical care implementation (within weeks rather than a month) may contribute to benefits shown in observational and clinic-based cohorts.60,61 Finally, the optimal times to sample are unknown and may be different for each ARV. Benefits of the sparse strategy (3 samples around witnessed dosing) included the relative ease of incorporating it into clinical practice, although providing more reliable information than a single random sample. For example, we observed absorption lags that could mask subtherapeutic exposure if predose concentrations are within 30% of the minimal threshold.
One study limitation is that drug-food interactions may have contributed to the observed variability in nelfinavir and lopinavir/ritonavir capsule pharmacokinetics. Such interactions may not be as important with widespread use of the new lopinavir/ritonavir tablet formulation. Second, implementation of suggested dose changes in the TDM arm was optional, which decreases the power to show a difference between arms. This design was to allow for potential instances of patient refusal to change dose and/or cases in which additional patient information available to the caregiver but not to the centralized expert committee would have altered the recommendations. Even though dose changes were not mandatory, provider acceptance of dose change recommendations was high. The expertise of various HIV subspecialists and the rich patient information, including viral susceptibility phenotypes available for interpretation by the expert committee, may have increased provider acceptance of recommendations in this research setting over what would be seen with TDM in routine clinical practice. Finally, patient adherence is inextricably linked to pharmacologic assessments, and plasma concentrations have been used to assess adherence, although they only provide insight for the several doses before the sample.62,63 This study used adherence interventions,29 electronic monitoring, and a witnessed dose evaluation to address this confounder. Additionally, “improve adherence” was a specific recommendation to increase exposure in those with widely variable or inconsistent concentrations. Potential contributors to the large number of adherence recommendations were that this group self-identified as needing improved adherence as part of the inclusion criteria and that subjects on lopinavir had a high incidence of diarrhea, which may have triggered nonadherence or more variable absorption.
In our treatment-naive and experienced HIV-infected patients taking PIs and/or NNRTIs from 5 sites, we found that patients receiving potent HAART regimens had widely variable ARV concentrations and that nearly two thirds had nontarget concentrations at least once over 48 weeks. TDM allows detection of these at-risk patients and provides a tool for the clinician to intervene before regimen failure, approximately doubling the likelihood of achieving target concentrations. Although this study was not designed to assess virologic outcomes, our exploratory findings in this study population were that patients who achieved target concentrations of PIs and NNRTIs trended toward a better virologic response over 48 weeks than those with nontarget concentrations. In addition to factors identified in the US treatment guidelines,13 patients with increased weight are candidates for TDM as well as patients on efavirenz or lopinavir. TDM should be performed after regimen initiation and periodically thereafter, ideally basing dose change recommendations on more than a single sample result, particularly for PIs. This study described a potential TDM approach for PIs and NNRTIs that could be used in future studies of ARV TDM to explore efficacy and toxicity endpoints.
The authors thank the patients, families, and study teams at the 5 CCTG participating sites: Santa Clara Valley Medical Center (SCVMC), University of California at Irvine (UCI), Harbor-UCLA Medical Center, UCSD, and University of Southern California (USC).
1. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet. 2002;360:119-129.
2. Masquelier B, Breilh D, Neau D, et al. Human immunodeficiency virus type 1 genotypic and pharmacokinetic determinants of the virological response to lopinavir-ritonavir-containing therapy in protease inhibitor-experienced patients. Antimicrob Agents Chemother. 2002;46:2926-2932.
3. Acosta EP, Henry K, Baken L, et al. Indinavir concentrations and antiviral effect. Pharmacotherapy. 1999;19:708-712.
4. Burger DM, Hugen PW, Aarnoutse RE, et al. Treatment failure of nelfinavir-containing triple therapy can largely be explained by low nelfinavir plasma concentrations. Ther Drug Monit. 2003;25:73-80.
5. Descamps D, Flandre P, Calvez V, et al. Mechanisms of virologic failure in previously untreated HIV-infected patients from a trial of induction-maintenance therapy. Trilege (Agence Nationale de Recherches sur le SIDA 072) Study Team. JAMA. 2000;283:205-211.
6. Burger DM, Hoetelmans RM, Hugen PW, et al. Low plasma concentrations of indinavir are related to virological treatment failure in HIV-1-infected patients on indinavir-containing triple therapy. Antivir Ther. 1998;3:215-220.
7. Veldkamp AI, Weverling GJ, Lange JM, et al. High exposure to nevirapine in plasma is associated with an improved virological response in HIV-1-infected individuals. AIDS. 2001;15:1089-1095.
8. Duval X, Mentre F, Lamotte C, et al. Indinavir plasma concentration and adherence score are codeterminant of early virologic response in HIV-infected patients of the APROCO cohort. Ther Drug Monit. 2005;27:63-70.
9. Dieleman JP, Gyssens IC, van der Ende ME, et al. Urological complaints in relation to indinavir plasma concentrations in HIV-infected patients. AIDS. 1999;13:473-478.
10. Marzolini C, Telenti A, Decosterd LA, et al. Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS. 2001;15:71-75.
11. Gutierrez F, Navarro A, Padilla S, et al. Prediction of neuropsychiatric adverse events associated with long-term efavirenz therapy, using plasma drug level monitoring. Clin Infect Dis. 2005;41:1648-1653.
12. Ray JE, Marriott D, Bloch MT, et al. Therapeutic drug monitoring of atazanavir: surveillance of pharmacotherapy in the clinic. Br J Clin Pharmacol. 2005;60:291-299.
14. Gazzard B, BHIVA Writing Committee. British HIV Association (BHIVA) Guidelines for the Treatment of HIV-infected Adults with Antiretroviral Therapy, 2005. HIV Med. 2005;6(Suppl 2):1-61.
15. Bossi P, Peytavin G, Ait-Mohand H, et al. GENOPHAR: a randomized study of plasma drug measurements in association with genotypic resistance testing and expert advice to optimize therapy in patients failing antiretroviral therapy. HIV Med. 2004;5:352-359.
16. Burger D, Hugen P, Reiss P, et al. Therapeutic drug monitoring of nelfinavir and indinavir in treatment-naive HIV-1-infected individuals. AIDS. 2003;17:1157-1165.
17. Clevenbergh P, Garraffo R, Durant J, et al. PharmAdapt: a randomized prospective study to evaluate the benefit of therapeutic monitoring of protease inhibitors: 12 week results. AIDS. 2002;16:2311-2315.
18. Fletcher CV, Anderson PL, Kakuda TN, et al. Concentration-controlled compared with conventional antiretroviral therapy for HIV infection. AIDS. 2002;16:551-560.
19. Khoo SH, Lloyd J, Dalton M, et al. Pharmacologic optimization of protease inhibitors and nonnucleoside reverse transcriptase inhibitors (POPIN)-a randomized controlled trial of therapeutic drug monitoring and adherence support. J Acquir Immune Defic Syndr. 2006;41:461-467.
20. Torti C, Quiros-Roldan E, Regazzi M, et al. A randomized controlled trial to evaluate antiretroviral salvage therapy guided by rules-based or phenotype-driven HIV-1 genotypic drug-resistance interpretation with or without concentration-controlled intervention: the Resistance and Dosage Adapted Regimens (RADAR) study. Clin Infect Dis. 2005;40:1828-1836.
21. Aarnoutse RE, Schapiro JM, Boucher CA, et al. Therapeutic drug monitoring: an aid to optimising response to antiretroviral drugs? Drugs. 2003;63:741-753.
22. Acosta EP, Gerber JG. Position paper on therapeutic drug monitoring of antiretroviral agents. AIDS Res Hum Retroviruses. 2002;18:825-834.
23. Back D, Gatti G, Fletcher C, et al. Therapeutic drug monitoring in HIV infection: current status and future directions. AIDS. 2002;16(Suppl 1):S5-37.
24. Boffito M, Acosta E, Burger D, et al. Therapeutic drug monitoring and drug-drug interactions involving antiretroviral drugs. Antivir Ther. 2005;10:469-477.
25. Boffito M, Acosta E, Burger D, et al. Current status and future prospects of therapeutic drug monitoring and applied clinical pharmacology in antiretroviral therapy. Antivir Ther. 2005;10:375-392.
26. Burger DM, Aarnoutse RE, Hugen PW. Pros and cons of therapeutic drug monitoring of antiretroviral agents. Curr Opin Infect Dis. 2002;15:17-22.
27. Kappelhoff BS, Crommentuyn KM, de Maat MM, et al. Practical guidelines to interpret plasma concentrations of antiretroviral drugs. Clin Pharmacokinet. 2004;43:845-853.
28. Van Heeswijk RP. Critical issues in therapeutic drug monitoring of antiretroviral drugs. Ther Drug Monit. 2002;24:323-331.
29. Wagner GJ, Kanouse DE, Golinelli D, et al. Cognitive-behavioral intervention to enhance adherence to antiretroviral therapy: a randomized controlled trial (CCTG 578). AIDS. 2006;20:1295-1302.
30. Fortovase prescribing information. Nutley, NJ: Roche Laboratories; 2003.
31. Kaletra prescribing information. North Chicago, IL: Abbott Laboratories; 2005.
32. Agenerase prescribing information. Research Triangle Park, NC: GlaxoSmithKline; 2005.
33. Invirase prescribing information. Nutley, NJ: Roche Laboratories; 2005.
34. Viramune prescribing information. Ridgefield, CT: Boehringer Ingelheim Pharmaceuticals; 2005.
35. Crixivan prescribing information. Whitehouse Station, NJ: Merck & Co.; 2006.
36. Lexiva prescribing information. Research Triangle Park, NC: GlaxoSmithKline; 2006.
37. Reyataz prescribing information. Princeton, NJ: Bristol-Myers Squibb Company; 2006.
38. Viracept prescribing information. New York, NY: Pfizer; 2006.
39. Rescriptor prescribing information. New York, NY: Pfizer; 2006.
40. Sustiva prescribing information. Princeton, NJ: Bristol-Myers Squibb Company; 2006.
42. Holland DT, DiFrancesco R, Connor JD, et al. Quality assurance program for pharmacokinetic assay of antiretrovirals: ACTG proficiency testing for pediatric and adult pharmacology support laboratories, 2003 to 2004: a requirement for therapeutic drug monitoring. Ther Drug Monit. 2006;28:367-374.
43. Beal SL, Sheiner LB, Boeckmann AJ, eds. NONMEM User's Guides, (1989-2006). ICON Development Solutions, Ellicott City, MD.
44. PhenoSense™ HIV, Monogram Biosciences, Inc., South San Francisco, CA.
45. Falloon J, Piscitelli S, Vogel S, et al. Combination therapy with amprenavir, abacavir, and efavirenz in human immunodeficiency virus (HIV)-infected patients failing a protease-inhibitor regimen: pharmacokinetic drug interactions and antiviral activity. Clin Infect Dis. 2000;30:313-318.
46. Sadler BM, Gillotin C, Lou Y, et al. Pharmacokinetic and pharmacodynamic study of the human immunodeficiency virus protease inhibitor amprenavir after multiple oral dosing. Antimicrob Agents Chemother. 2001;45:30-37.
47. Boffito M, Kurowski M, Kruse G, et al. Atazanavir enhances saquinavir hard-gel concentrations in a ritonavir-boosted once-daily regimen. AIDS. 2004;18:1291-1297.
48. Villani P, Regazzi MB, Castelli F, et al. Pharmacokinetics of efavirenz (EFV) alone and in combination therapy with nelfinavir (NFV) in HIV-1 infected patients. Br J Clin Pharmacol. 1999;48:712-715.
49. Hsu A, Granneman GR, Cao G, et al. Pharmacokinetic interaction between ritonavir and indinavir in healthy volunteers. Antimicrob Agents Chemother. 1998;42:2784-2791.
50. Stein DS, Fish DG, Bilello JA, et al. A 24-week open-label phase I/II evaluation of the HIV protease inhibitor MK-639 (indinavir). AIDS. 1996;10:485-492.
51. Burger DM, van Rossum AM, Hugen PW, et al. Pharmacokinetics of the protease inhibitor indinavir in human immunodeficiency virus type 1-infected children. Antimicrob Agents Chemother. 2001;45:701-705.
52. Murphy RL, Brun S, Hicks C, et al. ABT-378/ritonavir plus stavudine and lamivudine for the treatment of antiretroviral-naive adults with HIV-1 infection: 48-week results. AIDS. 2001;15:F1-F9.
53. Riska P, Lamson M, MacGregor T, et al. Disposition and biotransformation of the antiretroviral drug nevirapine in humans. Drug Metab Dispos. 1999;27:895-901.
54. van Heeswijk RP, Veldkamp AI, Mulder JW, et al. Once-daily dosing of saquinavir and low-dose ritonavir in HIV-1-infected individuals: a pharmacokinetic pilot study. AIDS. 2000;14:F103-F110.
55. Veldkamp AI, van Heeswijk RP, Mulder JW, et al. Steady-state pharmacokinetics of twice-daily dosing of saquinavir plus ritonavir in HIV-1-infected individuals. J Acquir Immune Defic Syndr. 2001;27:344-349.
56. Pellegrin I, Breilh D, Birac V, et al. Pharmacokinetics and resistance mutations affect virologic response to ritonavir/saquinavir-containing regimens. Ther Drug Monit. 2001;23:332-340.
57. van der Leur MR, Burger DM, la Porte CJ, et al. Retrospective TDM database analysis of interpatient variability in the pharmacokinetics of lopinavir in HIV-infected adults. Ther Drug Monit. 2006;28:650-653.
58. la Porte CJL, Back DJ, Blaschke T, et al. Updated guidelines to perform therapeutic drug monitoring for antiretroviral agents. Reviews in Antiviral Therapy. 2006;3:3-14.
59. Hsu A, Granneman GR, Witt G, et al. Multiple-dose pharmacokinetics of ritonavir in human immunodeficiency virus-infected subjects. Antimicrob Agents Chemother. 1997;41:898-905.
60. Forsyth SF, French PD, Macfarlane E, et al. The use of therapeutic drug monitoring in the management of protease inhibitor-related toxicity. Int J STD AIDS. 2005;16:139-141.
61. Boffito M, Bonora S, Raiteri R, et al. Pharmacokinetic evaluation of indinavir and indinavir/ritonavir-containing antiretroviral regimens in a clinical setting. Ther Drug Monit. 2002;24:574-576.
62. Hugen PW, Burger DM, Aarnoutse RE, et al. Therapeutic drug monitoring of HIV-protease inhibitors to assess noncompliance. Ther Drug Monit. 2002;24:579-587.
63. Podsadecki TJVB, Tousset EP, Rode RA, et al. White coat compliance patterns make therapeutic drug monitoring (TDM) a potentially unreliable tool for assessing long-term drug exposure. Presented at: 13th Conference on Retroviruses and Opportunistic Infections; Denver; 2006.
This article has been cited 15 time(s).
Advanced Drug Delivery ReviewsDynamical models of biomarkers and clinical progression for personalized medicine: The HIV contextAdvanced Drug Delivery Reviews
Status epilepticus resulting from severe efavirenz toxicity in an HIV-infected patient
AIDS Reader, 18(7):
Current Drug Metabolism
CYP2B6: New insights into a historically overlooked cytochrome P450 isozyme
Current Drug Metabolism, 9(7):
Journal of Antimicrobial ChemotherapyPharmacokinetic variability of antiretroviral drugs and correlation with virological outcome: 2 years of experience in routine clinical practiceJournal of Antimicrobial Chemotherapy
Cochrane Database of Systematic ReviewsTherapeutic drug monitoring of antiretrovirals for people with HIVCochrane Database of Systematic Reviews
Fundamental & Clinical PharmacologyBenefit of therapeutic drug monitoring of protease inhibitors in HIV-infected patients depends on PI used in HAART regimen - ANRS 111 trialFundamental & Clinical Pharmacology
Journal of Clinical PharmacologyPractical therapeutic drug management in HIV-infected patients: Use of population pharmacokinetic models supplemented by individualized Bayesian dose optimizationJournal of Clinical Pharmacology
Simultaneous population pharmacokinetic model for lopinavir and ritonavir in HIV-infected adults
Clinical Pharmacokinetics, 47():
Therapeutic Drug MonitoringPopulation Pharmacokinetics of Atazanavir in Human Immunodeficiency Virus-Infected PatientsTherapeutic Drug Monitoring
JAIDS Journal of Acquired Immune Deficiency SyndromesIncreased Baseline Body Weight Is a Risk Factor Associated With Virological Failure While on Antiretroviral TreatmentJAIDS Journal of Acquired Immune Deficiency Syndromes
JAIDS Journal of Acquired Immune Deficiency SyndromesEfavirenz Dose Reduction Is Safe in Patients With High Plasma Concentrations and May Prevent Efavirenz DiscontinuationsJAIDS Journal of Acquired Immune Deficiency Syndromes
JAIDS Journal of Acquired Immune Deficiency SyndromesIncreasing Age at HIV Seroconversion From 18 to 40 Years Is Associated With Favorable Virologic and Immunologic Responses to HAARTJAIDS Journal of Acquired Immune Deficiency Syndromes
anti-HIV agents; antiretroviral therapy; clinical trials; HIV; drug monitoring; pharmacokinetics
© 2007 Lippincott Williams & Wilkins, Inc.
Highlight selected keywords in the article text.