10th International Workshop on Clinical Pharmacology of HIV Therapy
Courtney V. Fletcher, Pharm.D.
Dean and Professor
College of Pharmacy
University of Nebraska Medical Center
%CV, percent coefficient of variation
ACTG, adult AIDS clinical trials group
ARV, antiretroviral drug
ART, antiretroviral drug therapy
AUC, area under the concentration-time curve
Cmin, minimum drug concentration
CNS, central nervous system
CSF, cerebrospinal fluid
CYP, cytochrome P450 drug metabolizing enzymes
IC50, concentration of drug required to inhibit viral replication in vitro by 50%
IQ, inhibitory quotient
NRTI, nucleoside reverse transcriptase inhibitor
NNRTI, non-nucleoside reverse transcriptase inhibitor
PACTG, pediatric AIDS clinical trials group
PBMCs, peripheral blood mononuclear cells
PI, inhibitor of HIV protease
r or RTV, ritonavir
TDM, therapeutic drug monitoring
The 10th International Workshop on Clinical Pharmacology of HIV Therapy was held April 15-17, 2008 in Amsterdam, the Netherlands. This location was in recognition of the first workshop held in Noordwijk, the Netherlands. This year's workshop featured 83 abstracts including 22 oral presentations, six invited lectures, two round table discussions and a clinical case discussion session. This 10th Workshop set an attendance record, with just over 200 participants.
John Gerber, M.D., gave the opening presentation on the evolution and contribution of clinical pharmacology to HIV therapy. John divided the more than 20 year history of ARV drug development into NRTI, PI and post-PI phases. In this NRTI phase, he noted a pivotal contribution was the demonstration that knowledge of NRTI-triphosphates (the intracellular, pharmacologically-active moieties) provided a basis for more informed, convenient and less toxic doses of these drugs. ZDV was originally dosed every 4 hours in a daily dose of 1500 mg/day. Knowledge of ZDV-triphosphate provided a basis for the current twice-daily dosing interval, because the half-life of ZDV-triphosphate is 7 hours whereas the plasma half-life of ZDV is only 1 hour. The PI phase was marked by the development of ARVs that were substrates, inducers and inhibitors of drug metabolizing enzymes and drug transporters, and agents that had clear concentration-effect relationships. The substantial challenges of understanding and managing drug-drug interactions because of these characteristics of PIs persists with us today. The most important pharmacologic contribution during the PI phase was the development of RTV as a PK booster. This ability of RTV to increase concentrations and extend the half-life of other PIs, allowed PIs that otherwise would have no therapeutic use in humans (e.g. lopinavir and darunavir) to become FDA-approved, decreased the frequency of dosing and improved patient adherence, and increased the likelihood of therapeutic success. The post-PI phase, is likely to be marked by the introduction of pharmacogenetics (PG) into therapeutics. The best example presently is the ability of HLA-B5701 screening to prevent the hypersensitivity reaction associated with abacavir. John's talk was a wonderful overview of the contributions that pharmacology has made to HIV therapeutics and a great opening for this 10th anniversary meeting.
Craig Hendrix, M.D. gave an invited lecture on HIV prevention with topical microbicides and using clinical pharmacology to form a rational basis for the development of these agents. Dave Margolis, M.D. discussed persistent HIV replication and potential therapeutic strategies to target latency and low-level viremia. Charles Flexner, M.D. reviewed the effect of aging on human pharmacology - a (good) problem we would not have discussed 10 years ago. Saye Khoo, M.D. discussed the pharmacology of malaria and the challenges of treating the HIV and malaria co-infected person.
There were two round table discussions at the 10th Workshop. The first covered four different new models or paradigms for ARV development. Ed Acosta discussed new approaches to understanding the pharmacology of the anti-HIV effect through dose-response slopes and the potential of the instantaneous inhibitory potential (IIP) as a measure of drug activity. Chantal Csajka discussed strategies for the incorporation of pharmacogenetics into population PK modeling. Rick Bertz reviewed the application of microdosing to clinical development. Albertina Arien with Johnson and Johnson discussed the development of nanoformulations of ARVs. This was an excellent overview of this topic and an update of the efforts of J&J on the development of a nanoformulation of rilpivirine (TMC-278). A nanoformulation of rilpivirine was first reported at the 2008 CROI, and has recently been published (Eur J Pharm Biopharm. 2009 Mar 27.) Despite numerous therapeutic hurdles that will need to be overcome, I think the development of nanoformulations of ARVs, which might only need to be administered once monthly, is very exciting.
The International Workshops on Clinical Pharmacology of HIV Therapy have always provided a venue for interactions of scientists from academia, industry and regulatory authorities (e.g. FDA). This is one of the clear strengths and almost unique characteristics of this Workshop. At the 10th Workshop, a round table focused on challenges of the development of pediatric formulations and dose selection including regulatory perspectives. John Morris with Abbott Laboratories discussed the development of pediatric formulations, including meltrex, of RTV and LPV/RTV. Elisabeth Rook of the EMEA (European Medicines Agency) reviewed positions of the EMEA on pediatric labeling. Kellie Reynolds of the FDA discussed perspectives on dose selection, formulation issues, and the challenges with combination produces in children. This meeting is one of the very few venues where this kind of interchange takes place, and I think this is one of the most important characteristics of this meeting.
The abstracts that I have chosen to review in this report are those I felt were well designed, provocative, and had important clinical implications or relevance. I will discuss these abstracts in two sections: Drug-Drug Interactions and Pharmacokinetic-Pharmacodynamic-Pharmacogenetic (PK-PD-PG) Characteristics of ARVs and Therapeutic Implications. Preceding each of these sections is a summary paragraph of what I think the important findings and implications are. This is followed by a more detailed description of the abstracts.
I. Drug-Drug Interactions
Summary: Drug-Drug Interactions
Many drug-drug interaction studies whether in healthy volunteers or HIV-infected persons are performed after the first dose of one or both of the drugs of interest. Abstract O-01 showed a difference in the magnitude of the interaction from first dose to steady state. In their study with LPV/r and fexofenadine, the increase in the fexofenadine AUC was 4-fold after the first dose and 2.94-fold at steady state. The mechanism for the increase in AUC was an inhibition of the drug transporter P-glycoprotein (P-gp). The reason for a smaller degree of inhibition at steady state was that LPV/r, over time, induced the expression of P-gp, thus offsetting some of the inhibitory effect. Management of drug-drug interactions in HIV-infected persons requires knowledge of the interaction at steady state. This work tells us that in some cases first dose studies won't give us that information.
The interaction between RAL and ATV is of current interest because of potential mutually beneficial effects when the ATV is given twice daily. Abstract O-14 investigated RAL and ATV PK at doses of 400 mg twice daily for RAL, and 200 mg twice daily for ATV. The mean RAL trough concentration was 132 ng/mL, which is considerably higher than what might be expected in patients not taking ATV. This is from the "boosting" effect of ATV. The ATV mean trough was 227 ng/mL. While this is higher than what would be seen with the unboosted dose of 400 mg/day, it is much lower than the trough seen with ATV/RTV, which averages approximately 630 ng/mL. I think twice daily dosing of ATV and RAL has the potential to be a useful combination. However, the ATV dose is not known and the safety and efficacy of this combination have not been established. I believe clinicians should be quite cautions about using this regimen in patients until such data are available.
Three abstracts were presented that I thought provided new information about concomitant therapy in HIV-infected persons and drugs for treatment of malaria and TB. Abstract O-19 evaluated the interactions between atovaquone/proguanil (used for the prevention/treatment of malaria) and EFV, LPV/r and ATV/RTV. Significant reductions were seen in both atovaquone and proquanil concentrations. These data suggest the dose adjustments of the atovaquone or proguanil or both components may be needed. The appropriate management strategies for these interactions need to be determined.
Rifabutin is the preferred rifamycin in HIV-infected patients with active TB disease because it has a lower risk of clinically significant interactions with ARVs than dose rifampin. The recommended dose of rifabutin when given with a RTV-boosted PI is 150 mg every other day or three times weekly. This dosing regimen has been recommended from interaction studies in healthy volunteers; there are little data to confirm the appropriateness in HIV- and TB-coinfected persons. Abstracts O-21 and O-22 evaluated the interaction between rifabutin and LPV/r and ATV/RTV, respectively. There is some suggestion that the rifabutin PK are higher in these healthy volunteers than in HIV-infected patients, although no direct comparison is possible. What are clear from both abstracts are the rates of adverse reactions with rifabutin and LPV/r or ATV/RTV were considerably higher than with only rifabutin. These data are suggestive of PK and PD differences between healthy volunteers and HIV- and TB-infected persons. The recent publication of three case reports of acquired rifamycin resistance in patients taking the recommended dose of rifabutin with PI/RTV, adds to the uncertainty about the most appropriate dose of rifabutin. PK and PD studies in HIV- and TB-infected patients are needed.
Charles laPorte presented a study of effects of LPV/r on multiple intestinal drug transporters and metabolizing enzymes (abstract # O-01). These effects were assessed by performing duodenal biopsies in healthy volunteers who received a single dose of fexofenadine as a probe drug for the transporter P-glycoprotein (P-gp) following a single dose of LPV/r and then again after 12 days (to achieve steady-state) LPV/r. The AUC of fexofenadine increased 4-fold after single dose LPV/r, whereas the increase in AUC was lower at 2.9-fold after 12 days of LPV/r. 12 days of LPV/r dosing was found to induce the expression of several transporters and enzymes including ABCB1, CYP3A4, and SLCO1B1. ABCB1 encodes for P-gp, and the induction of ABCB1 provides an explanation why there was a decrease in the fexofenadine AUC over 12 days of LPV/r administration: the inhibitory effect seen after a single dose of LPV/r was offset by an induction of P-gp that only manifested after multiple dose LPV/r administration. An induction was also seen in SLCO1B1, and interestingly, the fold change in expression was less in women than in men. SLCO1B1 is important in drug transport, for example, the influx (or entry) of statins into the liver. SLCO1B1 appears to influence the PK of LPV (abstract 39 at CROI 2009) and ATV (see my discussion below on abstract O-04 from this Workshop). The inhibition of SLCO1B1 will increase plasma concentrations of a drug that is transported by SLCO1B1. The increase in rosuvastatin plasma concentrations seen with LPV/r is a result of inhibition of SLCO1B1. Thus, a lesser degree of induction of SLCO1B1 in women compared with men could provide a mechanism for higher plasma concentrations of a drug that is transported by SLCO1B1. There are several important take home messages from this study. First, single dose drug-drug interaction studies may not reflect the magnitude of an interaction at steady state. Second, sex differences in transporter induction, and perhaps baseline expression may be different in women and men and this may provide a basis for some of the gender differences in plasma concentrations and thereby perhaps response or toxicity seen with certain drugs including ARVs.
One potentially beneficial drug-drug interaction of current interest is that between atazanavir (ATV) and raltegravir (RAL). Two abstracts (O-13 and O-14) were presented that evaluated this interaction with different doses in HIV-infected persons. The primary metabolic pathway of RAL is glucuronidation and ATV is an inhibitor of glucuronidation. AT CROI 2009, the interaction between ATV and RAL was evaluated in healthy volunteers (abstract 696) receiving ATV, 300 mg twice daily and RAL, 400 mg twice daily. The AUC and Cmin of RAL were increased by 54% and 48%, respectively, by ATV co-administration. The ATV AUC and Cmin were reduced by 17% and 29%, respectively, relative to twice daily ATV without RAL. Of note, twice daily ATV dosing produces a greater than proportional increase in ATV concentrations. At this 10th Workshop, the PK of ATV and RAL were evaluated when both were given once daily in doses of 400 mg once daily (for ATV) and 800 mg once daily (for RAL). The mean RAL AUC was 13989 ng*h/mL (29 µM*h) and the mean Cmin was 34 ng/mL (70 nM). Relative to historical controls receiving a 400 mg single dose without ATV, the RAL AUC was increased 18% but the Cmin was substantially reduced to a ratio of 0.15 (versus a 400 mg dose). Abstract O-14 evaluated ATV and RAL PK at steady state in HIV-infected persons who were receiving ATV, 200 mg twice daily, and RAL, 400 mg twice daily. The geometric mean ATV AUC was 6257 ng*h/mL and Cmin was 227 ng/mL. The geometric mean RAL AUC was 9085 ng*h/mL and Cmin was 132 ng/mL. Short-term tolerance in both studies was acceptable. Once daily administration of ATV and RAL is intriguing; however, the effectiveness of once daily RAL (in the absence of ATV) is unknown. While co-administration of ATV with RAL would be expected to increase RAL concentrations, abstract O-13 suffers from no control group of patients who received once daily RAL without ATV. At this time, I think it is very premature to use once daily administration of RAL with ATV in routine patient care. We first must have clinical confirmation that once daily RAL is as effective as twice daily - remember the lesson from indinavir where twice daily was not as effective as three times daily despite the same daily dose and equivalent AUCs. The data from abstract O-14 provide a bit more guidance for continuing investigation ATV and RAL in twice daily of each dosing regimens. The ATV geometric mean Cmin at 200 mg twice daily was 227 ng/mL. 5 of 21 subjects had ATV Cmin values less than 150 ng/mL, the suggested minimum trough concentration. The ATV Cmin at the usual un-boosted dose of 400 mg once daily is approximately 120 ng/mL. Thus, as expected, the same daily dose of 400 mg/day given as 200 mg twice daily produced higher trough concentrations. Thus, twice daily ATV dosing produces a "boosting effect" on ATV concentrations; but it is not equivalent to that of RTV boosting, as ATV trough concentrations from a dose of 300/100 of ATV/RTV would average approximately 630 ng/mL. This could form a basis to evaluate a higher dose of ATV, such as the 300 mg twice-daily regimen used in the healthy volunteer study reported at CROI 2009. However, in that study, QRS widening (a cardiac conduction adverse effect) was observed with twice daily ATV dosing, and the clinical relevance of that effect needs careful investigation. Twice daily dosing of ATV and RAL has the potential to have a mutually beneficial drug-drug interaction. However, the ATV dose in particular is not known and the safety and efficacy of this combination have not been established. I believe clinicians should be quite cautions about using this regimen in patients until such data are available.
In the category of drug-drug interactions, I would like to comment on three abstracts with important implications for concomitant therapy in HIV-infected patients and drugs for prevention/treatment of malaria and TB. Matthijs van Luin and colleagues evaluated drug interactions among atovaquone/proguanil and EFV, LPV/r, and ATV/r in healthy volunteers (abstract O-19). Atovaquone/proguanil (Malarone) is used for the treatment and prevention of malaria. They found the atovaquone AUC was significantly reduced in the presence of each of these drugs. Geometric mean ratios were EFV, 0.25; LPV/r, 0.26; and ATV/RTV, 0.54. Proguanil concentrations were also reduced, with the magnitude varying whether the patients had polymorphisms in CYP2C19, which is the key enzyme for proguanil metabolism. In subjects with wild-type CYP2C19, geometric mean ratios for proguanil AUC were: EFV, 0.55; LPV/r, 0.42; and ATV/RTV, 0.34. These data suggest in HIV-infected persons taking EFV, LPV/r or ATV/RTV and atovaquone/proguanil that dose adjustments of the atovaquone or proguanil components or both may be needed. Further investigations to clarify clinical significance and management strategies are necessary.
Rifabutin is the preferred rifamycin in HIV-infected patients with active TB disease because it has a lower risk of clinically significant interactions with ARVs than dose rifampin. The recommended dose of rifabutin when given with a RTV-boosted PI is 150 mg every other day or three times weekly. Once- or twice-weekly dosing has been associated with increased rates of development of rifamycin resistance in patients with advanced HIV, thus adherence to rifabutin and the RTV-boosted PI (to maintain the drug interaction) is an essential element of success for this regimen. Drug interactions between rifabutin and PIs/RTV have been studied largely in healthy volunteers; PK data in HIV- and TB-infected patients are extremely sparse. Two abstracts were presented at this Workshop that raise uncertainty about the extrapolation of these data in healthy volunteers to HIV-infected patients. Abstract O-21 evaluated the interaction between rifabutin and LPV/r at the recommended adjusted doses. Rifabutin PK were compared in subjects who received 150 mg three times weekly plus LPV/r, 400/100 mg twice daily, with data from subjects who only received rifabutin 150 mg once daily. This study was prematurely discontinued as 13 of the 14 healthy volunteers experienced at least one rifabutin-associated adverse event during rifabutin + LPV/r co-administration. Adverse events included fever, rash, leucopenia, neutropenia and lymphopenia. The rifabutin AUC (as a sum of rifabutin and its metabolite) when given with LPV/r was approximately 3-fold higher compared to that without LPV/r. Rifabutin concentrations in these healthy volunteers appear (but there is no control group) higher than in HIV- and TB-infected patients, and the incidence of rifabutin-associated adverse events suggest differences (PD) as well. The authors concluded, and I agree, that further studies in healthy volunteers are not recommended. Abstract O-22 evaluated the interaction between ATV/RTV and rifabutin in healthy volunteers. In this study subjects received rifabutin 150 mg once daily and another group received rifabutin 150 mg TWICE weekly co-administered with ATV/RTV, 300/100 once daily. The rifabutin AUC (as a sum of rifabutin and its metabolite) was approximately 10% higher with the TWICE weekly regimen than what would be expected with the usual dose of 300 mg once daily (given without RTV). One subject who received rifabutin only discontinued therapy whereas 9 of 18 (50%) who received rifabutin with ATV/RTV discontinued therapy because of adverse reactions, neutropenia being the most common. The PK data from this study, taken in isolation, would suggest that a TWICE weekly rifabutin regimen would maintain plasma concentrations equivalent to rifabutin 300 mg once daily in healthy volunteers. The authors concluded a 150 mg twice to three times weekly, or every other day, rifabutin regimen when given with ATV/RTV should provide adequate rifabutin concentrations. The high rate of adverse reactions in the rifabutin plus ATV/RTV regimen is consistent with the findings from the abstract discussed above where the PI was LPV/r- in these 2 healthy volunteers studies there was a higher rate of adverse reactions when rifabutin was given with a RTV-boosted PI despite what appears to be an appropriate adjustment of the rifabutin dose based on plasma concentrations.
These two abstracts (O-21 and O-22) raise uncertainties of whether rifabutin PK and PD are different in healthy volunteers vs. HIV- and TB-infected patients. If true then this allows questions about the appropriateness of the current dose adjustment strategies, which are based on data in healthy volunteers. These issues are now a bit more concerning given a just published report of acquired rifamycin-resistant TB in three HIV- and TB-infected patients who received alternate day rifabutin and RTV-boosted PI therapy (Jenny-Avital ER, Clin Infect Dis 2009;48:1471-4). All of the usual caveats of n=3 case reports and the lack of objective documentation of medication adherence apply to this report. But, as the authors of these case reports stated "it is plausible that the currently recommended dose of rifabutin may be inadequate for some HIV-infected patients." I agree with this conclusion and it calls for investigations of rifabutin PK and PD in HIV- and TB-infected patients.
II. PK-PD-PG Characteristics of ARVs and Therapeutic Implications
Summary: PK-PD-PG Characteristics of ARVs and Therapeutic Implications
Investigators reported that RAL concentrations in cervicovaginal fluid after 7 days of dosing with 400 mg twice daily in healthy female volunteers were essentially the same as those in plasma (abstract O-06). This high degree of penetration provides support for further investigations of the virologic effect and potential use for HIV prophylaxis of RAL.
Raltegravir concentrations in the genital tract of healthy female volunteers were evaluated after a first dose of 400 mg and after 7 days of twice daily dosing with 400 mg (abstract O-06). On day 1, RAL concentrations in cervicovaginal fluid were detectable 6 hours after the first dose. On day 7, the median RAL plasma AUC was 11911 ng*h/mL. In cervicovaginal fluid the median RAL AUC was 9769 ng*h/mL. The cervicovaginal fluid to plasma concentration ratio at day 7 was 0.93. These data demonstrate that RAL concentrations in the female genital tract are essentially the same as those in plasma. The high degree of RAL penetration into the genital tract compartment provides support for further investigations into the virologic effect and potential use for RAL in HIV prophylaxis.
Two very similar pilot studies evaluated the ability of PK and PD information, expressed as the inhibitory quotient (IQ) to guide dose reductions of TPV/RTV (abstract O-15) and DRV/RTV (abstract O-16) in treatment experienced patients who had HIV-RNA < 50 copies/mL. The dose reduction for TPV/RTV was from 500/200 twice daily to 500/100 twice daily. For DRV/RTV the reduction was from 600/100 twice daily to 900/100 twice daily. In both studies these reductions were accomplished successfully in that HIV-RNA remained < 50 copies/mL. I think these data collectively illustrate the ability of TDM strategies implemented with both PK and PD information to simplify dosing, maintain the desired response or minimize the likelihood of adverse reactions.
Investigators from Torino, Genova and Liverpool explored whether the ability of patients who maintained therapeutic ATV trough concentrations without RTV was associated with certain genetic characteristics (abstract O-04). These investigators found the presence of two or more variant genotypes was predictive of ATV concentrations greater than 150 ng/mL. These findings allow an intriguing possibility, analogous to screening for HLA-B5701 to prevent ABC hypersensitivity. Here prospective screening would be to identify those patients who might be ideal candidates to receive unboosted ATV therapy, because they have genetic characteristics that confer a beneficial influence on ATV concentrations.
A genetic study from the Swiss HIV Cohort Study (abstract O-03) found those individuals who had genotypes associated with drug-related adverse reactions discontinued ATV, EFV and TDF more often than those who did not. Similar to the study above, the promise of pharmacogenetics may be to screen patients for the presence of these genotypes. If present, clinicians could avoid using these drugs, if possible or have a heightened awareness for the possibility of drug-related adverse reactions.
Two very similar pilot studies evaluated the ability of PK and PD information to guide dose reductions. The first (abstract O-15), evaluated the ability to reduce the dose of TPV/RTV from 500/200 mg twice daily to 500/100 mg twice daily. The second (abstract O-16) investigated a dose reduction of DRV/RTV from 600/100 twice daily to 900/100 once daily. Dose adjustments were guided by the TPV and DRV inhibitory quotients (IQ), which integrate PK information as the trough concentration with PD information in the form of in vitro viral susceptibility. These investigations were conducted in treatment experienced patients, who were receiving TPV/RTV or DRV/RTV containing ARV regimens at the standard doses, and had HIV-RNA < 50 copies/mL. Viral suppression was maintained in the 11 patients in the TPV/RTV study who switched to 500/100 twice daily, and an improvement in lipid profiles was observed. Similarly, viral suppression was maintained in all 15 patients who had their DRV/RTV dose reduced to 900/100 once daily, and when compared with a control group who continued on 600/100 twice-daily DRV/RTV, triglyceride concentrations in the QD group decreased whereas they remained stable in the BID group. Collectively these data indicate that the goals of TDM, to improve the likelihood of the desired response while minimizing adverse reactions were successfully implemented by using PK and PD information in the form of IQ values. These pilot studies provide support for larger investigations of this strategy.
Marco Siccardi and colleagues in Torino, Genova and Liverpool investigated the association of whether patients maintained ATV trough concentrations when given without RTV that were above the suggested threshold of 150 ng/mL with certain genetic characteristics (abstract O-04). 55 patients who received unboosted ATV as part of their ARV regimen were included in this study. The genes of interest were: the pregnane X receptor (PXR, which regulates the expression of CYP3A4 and ABCB1), ABCB1 (which encodes for the transporter P-glycoprotein) and the drug transporter SLCO1B1. These investigators found that variants of each of these three genes were associated with higher ATV trough concentrations. When the effects of these polymorphisms were summed in a multivariate logistic regression analysis, patients who had two or more of these variant genotypes was the only independent predictor of ATV concentrations greater than 150 ng/mL, with an odds ratio of 9.2. These findings allow an intriguing possibility, analogous to screening for HLA-B5701 to prevent ABC hypersensitivity. Here prospective screening would be to identify those patients who might be ideal candidates to receive unboosted ATV therapy, because they have genetic characteristics that confer a beneficial influence on ATV concentrations.
A significant challenge to the pharmacotherapy of HIV infection is the ability to identify patients who may experience drug-limiting adverse reactions. Abstract O-03 evaluated the association of pharmacogenetic markers with ARV drug discontinuation in an analysis of 471 persons from the Swiss HIV Cohort Study who initiated therapy in 2004-2007. The genetic markers selected had previously been shown to be associated with such drug-related adverse events as renal tubular acidosis, neurotoxicity, and hyperbilirubinemia. 157 (33%) of the 471 patients discontinued at least one ARV within 1 year after initiation. In antiretroviral drug specific analyses, those individuals who had genotypes associated with drug-related adverse reactions discontinued ATV, EFV and TDF more often than those who did not.
The promise of pharmacogenetics is captured exactly in these last two abstracts - to identify those patients who should not receive certain drugs, or need to receive lower doses, or who have the optimal response to a certain drug.