Pharmacologic Enhancement in Protease Inhibitor-based HAART: The Role of Ritonavir

By Stephen L. Becker, MD, and Lorna Thornton, MD
Dr. Becker is associate clinical professor of medicine, University of California School of Medicine, and Director of Pacific Horizon Medical Group, San Francisco, CA. Dr Thornton is in medical practice with the Pacific Horizon Medical Group, San Francisco, CA.

Abstract
Introduction
Why Do PIs Need to Be Boosted in the First Place?
    - Absorption
    - Distribution
    - Metabolism
    - Excretion
          • Cellular-Efflux Pumps and Drug Transporters
          • P-gp
          • MRP
          • Host Xenobiotic Response Elements
          • Summary of Factors Influencing PI Exposures
Benefits of Ritonavir Boosting
Liabilities of Ritonavir Boosting
    - Toxicity: Ritonavir-Related Toxicities
    - Toxicity: Primary PI-related Toxicities
    - Metabolic Alterations
    - Drug-Drug Interactions
Clinical Application of Selected Ritonavir-boosted PI Regimens
    - Lopinavir/Ritonavir
    - Saquinavir/Ritonavir
    - Amprenavir and Fos-Amprenavir/Ritonavir
    - Indinavir/Ritonavir
    - Atazanavir/Ritonavir
Can PIs Be Developed That Do Not Require Pharmacokinetic Enhancement?
Summary

Abstract

Having changed the landscape in the treatment of HIV infection, the clinical effectiveness of current protease inhibitors (PIs) remains limited. Foremost among the limitations of this class of agents are pharmacologic factors including complex metabolism, action of membrane drug transporters, and activation of the natural host defense mechanisms. These factors, as well as inadequate adherence, frequently engender the emergence of PI drug resistance and lead to regimen failure.

Co-administration of ritonavir can enhance exposures of a primary or parent PI, by inhibiting  metabolism and transporter-induced removal of drug from target cells.,  It is now appreciated that the enhanced levels of PI are associated with decreased rates of drug resistance, the most compelling reason to justify their use.  Adding ritonavir, however, is not without cost. Abnormalities of cholesterol and triglyceride metabolism (possibly increasing the risk of cardiovascular disease), gastrointestinal intolerance, and numerous drug-drug interactions can result and must be balanced against the pharmacokinetic improvement rendered by the addition of ritonavir. Understanding the pharmacologic origins for the variations in exposure levels of PIs, both between and within patients, is important for the successful use of these agents.


Introduction

The goal of highly active antiretroviral therapy (HAART) is to achieve maximal and durable suppression of HIV replication, reconstitution and preservation of immunologic function, and reduction of HIV-related morbidity and mortality. Without maximal suppression of HIV over the long term, drug-resistant variants will emerge, resulting in therapy failure and disease progression.

Maximal, durable HIV suppression is dependent upon achieving consistent, pharmacologically active drug levels in the intracellular compartment, the site of HIV replication. Clinical studies have repeatedly demonstrated the value of including a protease inhibitor (PI) in a HAART regimen.  PI-based therapy has resulted in decreased mortality, a lowered incidence of serious opportunistic infections and malignancy, and a lowered incidence of any AIDS-defining diagnosis among patients with HIV infection. 

The pharmacology of the PIs is complex.  The specific nature of their absorption, distribution, metabolism, and excretion can produce substantial variability in drug exposures among and between individuals. In some instances, plasma and intracellular levels fall below the necessary viral inhibitory concentrations.  Exposures less than those capable of inhibiting viral growth permit ongoing replication and the emergence of drug resistant variants. 

Thus, the fundamental success of PIs is jeopardized if pharmacologic exposures are not maintained.  Poor absorption (bioavailability) of these drugs, high degrees of protein binding in the serum, rapid metabolic degradation, and elimination by cellular efflux pumps have resulted in the need for frequent administration and imposed high pill burdens and strict meal requirements in the use of most of the approved PIs. Even with complete adherence to these complex regimens, substantial inter-subject variability exists. 

The reasons why levels of pharmacologically active PIs vary at the sites of their action and can become inadequate to suppress viral replication are the subject of this review.  This article discusses the rationale for pharmacologic enhancement of PI therapy (“boosting”) with ritonavir, itself an inhibitor of HIV protease, as a means of overcoming suboptimal drug exposures. This practice, now nearly routine involves clear benefits and risks. 

On the one hand, ritonavir-based pharmacologic enhancement of current-PI therapy improves overall exposures and reduces the likelihood of emerging drug resistance.  This strategy also improves the ability of therapy to overcome existing resistant viral variants and offset the enzyme-induction effect of certain co-administered drugs. In addition, it lessens pill burdens, lengthens dosing intervals, and, in some cases obviates food requirements. 

These gains explain why the use of unboosted PIs in the clinical arena has substantially declined and why co-formulations of PIs with ritonavir have been pursued commercially (eg, Kaletra^™, lopinavir and ritonavir). These benefits, however, come at a potential cost.  The increased PI drug exposures have led to higher rates of adverse drug effects, as well as a host of complex metabolic alterations and drug interactions.  This article describes the current understanding of these issues and their application to clinical practice. 


Why Do PIs Need to Be Boosted in the First Place?

Current PIs exhibit substantial variability in exposure, with minimum, or trough, concentrations (C_min) approaching or in some cases falling below the virus’ inhibitory concentration.  Enhancing or boosting the exposure of a primary PI with ritonavir or a secondary PI slows the metabolism of the primary agent and alters overall pharmacokinetic exposures in such a fashion as to raise C_min and area under the time-concentration curve (AUC).  Studies of virtually all available PIs show a correlation between trough concentrations and successful virologic outcomes. 

One factor that has been identified as critical to treatment success and the maintenance of PI exposure is patient adherence to the medication regimen. In HIV disease, and among PI-based regimens, adherence rates in excess of 95%, far higher than in other chronic illnesses, are needed to achieve these successful outcomes. Any discussion of pharmacokinetic enhancement must include an understanding of the unique issues of patient medication adherence.  Though no studies have demonstrated that pharmacokinetic enhancement has lessened the considerable requirement of strict medication adherence, boosted regimens, by virtue of the higher plasma concentration and extended half-life, may lessen the extent of needed  adherence.  Studies to test this hypothesis are underway.

Fundamental to any discussion of enhancing drug exposures is an understanding of the complex aspects of PI disposition.  Aspects of drug absorption, distribution, metabolism, and excretion define drug disposition.  Many factors including host genetics, natural defense mechanisms and HIV itself affect the pharmacokinetics and pharmacodynamics of the HIV PI class. 

Pharmacokinetics can be defined as those factors which determine drug concentration, while pharmacodynamics describe the clinical effect(s) achieved as a result of the given drug concentration.  Table 1 lists the essential determinants of PI exposure. The following discussion examines these factors in more detail.

Absorption

Drug absorption or bioavailability is affected by numerous factors including the presence of food in the stomach and levels of gastric acidity.  It is now clear that the PI drugs, upon entering the gastrointestinal tract, are acted upon by the drug metabolizing enzymes of the cytochome P450 system, specifically the isoenzyme 3A4 (CYP 3A4).  In addition the cellular efflux pump P-glycoprotein (P-gp) actively removes drug from the cells comprising the mucosal surface of the small intestine (enterocyte). 

The so called “first-pass” drug metabolism occurs at the level of the gut,, prior to enterohepatic and systemic circulation.  The expression of both CYP 3A4 and P-gp is variable (see below) and may determine overall pre-systemic and systemic exposure.  Agents that inhibit or induce CYP 3A4 and P-gp may have an effect at the gut as well as at the hepatic level.  Certain substances  (grapefruit juice, Seville oranges and certain teas) are examples and may affect pre-systemic exposures by preferentially acting on gut-expressed CYP 3A4. 

Distribution

The distribution of a drug throughout the body is dependent largely upon protein binding.  All HIV PIs are avidly bound to serum proteins.    Since only the non-protein–bound, free, drug is pharmacologically active, the extent and variability of protein binding is important to PI exposure.  PIs are bound by both serum albumin and α₁-acid glycoprotein (AAG).  Levels of AAG, an acute-phase reactant, vary in association with HIV disease activity.  Because AAG levels are increased in HIV-positive individuals, measured levels of free (unbound) drug are significantly reduced.  AAG further varies according to race, age, weight, and time of day. 

Most of the currently available PIs show considerable degrees of protein binding, 90% to 99% or greater. The exceptions are indinavir, atazanavir, and amprenavir, for which the mean protein binding levels are 60%, 86%, and 90%, respectively. Greater or lesser degrees of protein binding do not inherently confer benefit to one PI versus another. 

Metabolism

The complexity of the metabolism of the HIV PIs precludes a full discussion of this subject.  Drug metabolism occurs primarily in the liver, though increasingly it is appreciated that metabolism for some drugs also occurs at the level of the enterocyte.  All of the currently available PIs, or their active metabolites, act as substrates and inhibitors of the CYP450 3A4 isoenzyme.  Nelfinavir is the exception, being metabolized primarily through 2C19, although its active metabolite, M8, is metabolized by CYP 3A4. 

In addition to acting as substrates and inhibitors of CYP 3A4, many of the PIs are also CYP 3A4 inducers.  Among the PIs with induction properties are lopinavir, ritonavir, amprenavir, nelfinavir and the investigational agent tipranavir.  Predictably, use of PI combinations whereby both CYP 3A4 enzyme inhibition and induction occurs has been clinically difficult.  Formal pharmacokinetic studies are often lacking, and clinical results have varied.  The metabolism of PIs through the 3A4 isoenzyme system contributes to frequent drug-drug interactions, a hallmark of the protease inhibitor class.  Variability in the expression of CYP450 isoenzymes, including 3A4 is known to exist, and likely explains some of the observed inter-patient variability in PI exposures. Aside from these host factors, preliminary evidence suggests that HIV itself may affect the expression of 3A4 and thus alter drug exposures at various times during treatment.

Excretion

Excretion and elimination of drugs requires metabolism involving both phase I (oxidative) and phase II (conjugative) reactions.  These reactions convert drugs from non-polar and lipophilic compounds to more readily excreted forms that are polar and hydrophilic. Removal of a drug from target cells is another determinant of drug exposure and may increase substrate for subsequent metabolism and excretion. 

P-glycoprotein (P-gp) is the best studied of the energy-dependent, ABC, (ATP-binding cassette) cellular efflux pumps.  P-gp functions as a major transporter for the HIV PIs, and for this reason, the expression and function of this protein are determinants of intracellular concentrations of these agents.  The relationship between P-gp expression and pharmacodynamics is the subject of ongoing research efforts.  The role of cellular-efflux pumps and drug transporters is examined in more detail in the following discussion.

Cellular-Efflux Pumps and Drug Transporters

Drug transport and efflux systems such as P-gp and the multi-drug resistance proteins (MRPs) may play an important role in both the oral absorption and efficacy of PIs. The regulatory importance of these membrane transporters has been demonstrated extensively in oncology, specifically in relationship to drug treatment failures.  P-pg, in particular, has been well documented as a mediator of drug resistance in the treatment of cancer. The role of P-gp in the disposition of drugs more generally has only recently been described. 

With regard to HIV disease, one hypothesis is that the cellular membrane concentration (expression) of these drug-transporter elements is increased in protected, sanctuary, sites of HIV replication, often preventing the adequate accumulation of PIs in pharmacologically active concentrations. In HIV-infected patients, differences in P‑gp or MRP transport functions affecting PI activity could have several origins: genetic differences in expression and function, changes caused by the HIV infection itself, or activity of the drugs themselves, including ritonavir.

P-gp.  P-gp is the product of the MDR1 gene and is thus variably expressed. Several cell types intimately involved in HIV pathobiology express P-gp, especially the gut, central nervous system, testis, bone marrow, lymphocytes, and placenta.  P-gp itself also appears to be involved in certain aspects of the pathogenesis of HIV infection, by affecting membrane binding and cellular entry of the virus.

Analogous to their interaction with CYP 3A4, all of the PIs are substrates for P-gp, and all, to some degree, are also inhibitors. Amprenavir, indinavir, and ritonavir appear to be actively transported mainly by P-gp; indinavir and ritonavir also appear to be transported by mechanisms involving MRPs. Ritonavir, among the PIs, is the most potent inhibitor of P-gp efflux, as it is of the CYP450 3A4 isoform.

The first demonstration of the relationship between P-gp expression and HIV pharmacokinetic and pharmacodynamic outcomes was conducted among a selected cohort of patients attaining virologic control from treatment with either nelfinavir or efavirenz.  P-gp is encoded by the MDR-1 gene, for which at least 3 genotypes have been determined. The TT genotype results in under-expression of P-gp, whereas the CC genotype causes over-expression.  CT genotype is the heterozygous, intermediate, form.

In the study,  patients with the TT genotype had the lowest plasma drug levels, and perhaps reflecting lessened efflux and greater intracellular concentrations, had the greatest gains in CD4 cell count.  Patients expressing the CC genotype had higher plasma drug levels than the TT genotype patients and the lowest increase in CD4 cell count.  Results for the patients with the CT genotype were intermediate. 

This work was important because it was the first to correlate host genetic factors with both drug exposure and immunologic outcome in HIV infection.  Current estimates indicate that 80% of West Africans and African Americans and 20% of whites are of the CC genotype.  The implications of the findings from this study require confirmations as the results suggest the likelihood of blunted response to therapy in patients with the CC genotype, particularly black Africans, and those of African ancestry. 

This suggests the potential role for pre-treatment host genotyping in order to maximize treatment benefit.  It should be understood that a patient’ appearance, a phenotype, may not adequately reflect host genetics or genotype.  This is true for most populations, but especially those such as American blacks where high degrees of racial inter-mixing occur.

MRP.  MRP-1 is another drug transporter involved in PI efflux. MRP-1 has been identified in HIV sanctuary sites. Recently, ritonavir has been shown to inhibit the MRP-1 efflux pump, thus providing another mechanism of enhancing intracellular drug exposures.  Other MRPs have also been identified as determinants of PI exposure.

Host Xenobiotic Response Elements

Potential determinants of drug exposure are the orphan nuclear receptors responsible for host defense against xenobiotic compounds, including drugs.  The steroid xenobiotic receptor (SXR, also known as the pregnane X receptor, or PXR) and the constitutive adrostane receptor (CAR) are members of this family of orphan nuclear receptors that when activated by xenobiotic compounds induce the processes of metabolism, transport, and excretion. The orphan nuclear receptors activate specific CYP enzyme systems and cellular efflux pumps. For SXR, this primarily involves induction of the CYP 3A4 isoenzyme and activation of both P-gp and MRP-2. 

Among the numerous activators (ligands) for SXR are steroids, phenobarbital, taxol, calcium channel blockers, and the PIs, most prominently, ritonavir.  The importance of SXR in clinical pharmacology has only recently been appreciated; it may have a special role in HIV infection. Ritonavir provides an intriguing example. When ritonavir is used to improve the pharmacokinetic exposure of a PI, it does so by directly inhibiting CYP 3A4, P-gp, and MRP-1.  Mediated through SXR are indirect effects that result in CYP 3A4, P-gp and MRP-1 induction or activation.  The overall exposure of the boosted PI is thus the net result of this dynamic balance between opposing ritonavir-induced effects.  Work is currently underway to define any variability in expression of SXR, or other nuclear receptors involved in drug disposition, which, if present, would be expected to further influence this dynamic balance.

Summary of Factors Influencing PI Exposures

As the preceding discussion has demonstrated, the origins of inadequate exposure to PIs are diverse (Table 2).  Serum protein binding may occur to differing extents depending on HIV disease activity.  Protein binding also varies by age, race, and weight and at different times within the same individual.  Metabolism of the PI by the CYP450 isoenzyme system can be variable, resulting in fluctuations in drug levels over time. Alterations in the expression of  transporters such as P-gp and the MRPs could result in decreased availability at the site of drug action. Finally, the binding of PIs to the SXR or other nuclear receptors that are responsible for clearance of xenobiotic substances could be an additional factor lowering effective serum drug concentrations.

Benefits of Ritonavir Boosting

Ritonavir pharmacokinetic enhancement, or boosting, can increase the primary PI exposure, but it does not realize the same therapeutic benefit of the 2 PIs. Low doses of ritonavir, usually 100 mg or 200 mg per day, are insufficient to contribute to antiviral activity but are sufficient to inhibit CYP 3A4 metabolism and P-gp–mediated efflux.  For all boosted PIs, ritonavir increases the C_min, the minimum drug concentration, the pharmacokinetic parameter most closely associated with successful virologic outcome. The area under the concentration-time curve (AUC) of the primary PI is also increased by the addition of ritonavir.  Ritonavir boosting, depending on the primary PI, may or may not increase the C_max , the maximum drug concentration.  The C_max is dependent on the exact pharmacokinetic profile of the primary PI and the extent of metabolism through CYP 3A4.  

Several studies involving lopinavir/ritonavir (Kaletra) and fos-amprenavir/ritonavir tested against the unboosted PI nelfinavir have clearly demonstrated the benefit of boosted versus unboosted PIs. Primary among these benefits are the reduced rates of drug resistance seen in those treated with boosted PIs.  This benefit extended not only to variants resistant to the PI but to resistance mutations in the NRTI class of co-administered drugs as well. 

The importance of this observation cannot be overstated.  Emergence of drug resistance is the greatest limitation to current therapy.  Strategies that can diminish drug resistance therefore assume great importance.  Current Federal HIV guidelines favor the use of boosted PIs primarily for this reason. 

Boosting can also have important dosing and drug administration benefits.  Boosting of PIs can lengthen the dosing intervals, from 3 times daily to twice daily, and, for some, from twice daily to once daily.  Pill burden generally decreases. In the case of indinavir, the requirement for dosing on an empty stomach is obviated.  These changes may result in improved efficacy, increased potency and regimen durability.  Anecdotal reports suggest that some patients find the boosted regimens easier to adhere to.  In one study, an evaluation comparing a single-PI regimen (indinavir) with a twice‑daily ritonavir-boosted regimen, 87% of 429 patients preferred the boosted regimen for its ease and convenience. The ritonavir-enhanced regimen was described as easier to take as prescribed than was the regimen involving indinavir alone.  Given the critical importance of medication adherence to successful treatment outcomes, these benefits should not be underestimated. 

Liabilities of Ritonavir Boosting

The liabilities of boosting PIs with ritonavir can be broadly considered to involve issues of drug toxicity, induced metabolic alterations, and drug-drug interactions.

Toxicity: Ritonavir-Related Toxicities

Direct toxic effects of ritonavir include gastrointestinal effects, increased risks of hepatotoxicity, and changes in serum lipid concentrations. The latter is of concern because of the potential for adverse effects on the vascular endothelium and thus increased risk of coronary artery and cerebrovascular disease.

Toxicity: Primary PI-related Toxicities

Another study of indinavir 800 mg/ritonavir 100 mg twice daily versus indinavir alone also reported more frequent toxicities in the ritonavir-boosted group (P < 0.05), with no increased virologic benefit over traditional indinavir dosing. Moderate-to-severe diarrhea and nausea were the most common side effects, but no dose reductions occurred as a result of these events

The examples cited above have focused on indinavir-based therapies, demonstrating the association between indinavir C_max and drug toxicity.  Ritonavir enhancement of indinavir has a consistent effect of elevating the C_max. With other boosted PIs, lesser degrees of elevation of the C_max  results, and the primary PI associated toxicities may not be increased.  An interesting study to be launched next year will compare boosted versus unboosted atazanavir.  It will be of value to determine the extent to which ritonavir drives the well-described atazanavir related toxicities.

Metabolic Alterations

The most worrisome metabolic alteration associated with ritonavir boosting of PIs is the effect on serum lipids. Clinically significant elevations in serum cholesterol and triglycerides, to levels that could have long-term clinical relevance, are reported with nearly all ritonavir-enhanced regimens (atazanavir appears to be the sole exception). All of the currently available PIs, except atazanavir, whether boosted or not, elevate total cholesterol, low-density cholesterol, and triglyceride concentrations above the levels at which the National Cholesterol Education Program recommends treatment in a significant number of patients.

Ritonavir boosting appears to raise lipid levels further beyond the basal rate of the primary PI.  Given the varying effects of the primary PIs on serum lipids, the additive effect of ritonavir enhancement may be different depending on the primary PI.  As noted above the compelling, and as yet unanswered question, is whether these lipid elevations will be associated with excess cardiovascular and cerebrovascular events.  

Several studies have confirmed the lipid effects seen with ritonavir-enhanced regimens. In a comparison of lopinavir/ritonavir with nelfinavir, the incidence of hypertriglyceridemia ( ≥750 mg/dL) was 11% in the lopinavir/ritonavir group but only 2% in the nelfinavir group. Similarly, 10% of the patients who received lopinavir/ritonavir had total cholesterol levels ³300 mg/dL, versus 6% of those who received nelfinavir. In another study, conducted among 429 patients randomly assigned to remain on indinavir therapy or switch to indinavir 400 mg/ritonavir 400 mg, total cholesterol and triglyceride levels were significantly elevated over baseline values at 24 weeks in the patients who received indinavir/ritonavir. Total cholesterol levels >300 mg/dL were observed in 16% of the indinavir/ritonavir group, as compared with 4% of the indinavir group. Triglyceride levels were >750 mg/dL in 24% of the patients who received indinavir/ritonavir, versus only 12% of those receiving indinavir alone. 

Data from two atazanavir salvage studies show that the favorable lipid profile seen with atazanavir alone is not lost with ritonavir enhancement.  In the 043 study where unboosted atazanavir was compared to lopinavir/ritonavir, atazanavir was associated with a 2% drop in triglycerides from baseline, while the lopinavir/ritonavir arm showed an increase of 19%.   In the 045 study, where atazanavir 300mg was boosted with ritonavir 100mg, a 2% increase in triglycerides was observed, versus a 34% increase in the lopinavir/ritonavir group.

The ACTG Cardiovascular Focus Group recommends that PI-associated dyslipidemia be treated according to the National Cholesterol Education Program’s recommendations, adding the important caveat that drug interactions with lipid-lowering drugs be avoided.  While no studies have adequately addressed the issue, there is a sense among clinicians that the dosage of ritonavir used for PI enhancement is proportional to the lipid elevation.  This is an important point, as there are often several possible doses recommended for ritonavir and the primary PI, and an optimal dosage combination has not been established.

Drug-Drug Interactions

Numerous drug-drug interactions exist for the HIV PI class.  This results from the CYP3A4 inhibition that is characteristic of the class.  Ritonavir is the most potent inhibitor of the 3A4 isoform, and is thus associated with a significant number of clinically relevant drug interactions.  It is clear from the preceding discussion that the co-administration of ritonavir with other PIs is an attempt to take clinical advantage of an intentionally produced drug-drug interaction. 

However, clinically deleterious interactions can be expected when ritonavir is used with any drug that functions as a substrate for CYP 3A4. The need to treat dyslipidemia in PI-treated patients provides an example.  Several of the potent 3^‑hydroxy-3-methylglutaryl coenzyme A (HMG Co-A) reductase inhibitors (statins), used to treat dyslipidemia, are metabolized by the 3A4 isoenzyme system.  Studies of simvastatin and atorvastatin showed increases in the AUC of both drugs of 31.6-fold and 4.5-fold, respectively, when administered together with saquinavir/ritonavir. When these drugs were used in combination with nelfinavir 1250 mg twice daily, the AUC of simvastatin increased more than 400%, and that of atorvastatin increased by 74%. Consequently, the use of simvastatin needs to be avoided in patients taking nelfinavir, and atorvastatin should be used with caution.

In contrast, when pravastatin (which is primarily renally eliminated) was administered in combination with saquinavir and ritonavir, its pharmacokinetics were essentially unchanged. Thus, pravastatin is an appropriate choice within the class for the treatment of PI-associated dyslipidemia.  Other drugs used to treat hyperlipidemia, including the fibrates and niacin are not substrates of CYP 3A4 and are thus not affected by the co-administration of ritonavir.

The DHSS maintains an updated web-based resources for protease inhibitor related drug-drug interactions and should be consulted by interested clinicians for the most recently identified drug interactions.

Clinical Application of Selected Ritonavir-boosted PI Regimens


Lopinavir/Ritonavir

Lopinavir/ritonavir was developed according to an understanding of the potential benefits of pharmacokinetic enhancement.  The only ritonavir co-formulated preparation, this agent has demonstrated impressive effectiveness in both treatment-naïve and treatment-experienced patients.

Lopinavir/ritonavir is the prototype boosted PI, and is arguably the gold standard in the class.  Pharmacokinetic enhancement demonstrated by lopinavir/ritonavir, and later confirmed by another ritonavir enhanced PI, attenuates the emergence and development of PI resistance.  This was first demonstrated in the trial that compared lopinavir/ritonavir to nelfinavir in treatment naïve patients.   In this study, 25 of 76 patients failing the nelfinavir arm had emergent PI resistance mutations versus 0 of 37 on the lopinavir/ritonavir arm.   As expected the use of lopinavir/ritonavir in treatment experienced patients is associated with viral resistance evolution. 

Despite the impressive effectiveness of the lopinavir/ritonavir regimens, they have been limited by problems of tolerability and toxicity in a moderate number of patients. To what extent these toxicities are the result of ritonavir or the parent PI, lopinavir, is difficult to determine. 

Saquinavir/Ritonavir

The combination of saquinavir/ritonavir has been used since the mid-1990s, both with nucleoside reverse transcriptase inhibitors (NRTIs) and as dual PI therapy without NRTIs.  The combination was initially designed with higher-dose ritonavir (400 mg twice daily). However, a series of pharmacokinetic and pharmacodynamic studies have demonstrated that other dosage combinations, when paired with NRTIs, are effective and better tolerated. 

A twice-daily regimen of saquinavir 1000 mg/ritonavir 100 mg has received wide clinical use, and a once-daily combination of saquinavir 1600/ritonavir 100 mg is currently under study.  Saquinavir was originally introduced in 1995 as a hard gel capsule formulation (Invirase), which exhibited poor bioavailability and was subsequently replaced by a soft gel formulation (Fortovase).  This preparation improves bioavailability but is associated with a higher overall incidence of gastrointestinal adverse events. Recent studies have shown that when the hard gel preparation of saquinavir is co-administered with ritonavir, exposures are slightly greater than those obtained with the soft gel formulation.  It is likely that all future use of saquinavir/ritonavir combinations will be with the hard gel formulation (Invirase).

Saquinavir has several potential advantages over several of the other approved PIs.  Cholesterol and triglyceride elevations are often less marked in saquinavir-containing regimens, and the agent lacks enzyme induction effects on CYP 3A4, which may result in fewer drug-drug interactions. 

The role of dual PI saquinavir-enhanced regimens is under study.  Preliminary data on the use of saquinavir with atazanavir  and lopinavir/ritonavir  have prompted larger ongoing studies, including needed dosage optimization studies.

Amprenavir and Fos-Amprenavir/Ritonavir

With the recent approval of the pro-drug formulation of amprenavir (Agenerase), fos-amprenavir (Lexiva) will be the form of this agent in widespread use.  It is important to understand that all clinically relevant pharmacologic parameters of fos-amprenavir are the same as those of amprenavir.  The pro-drug formulation allows a significantly reduce pill burden and removes the bothersome vitamin E from the excipient of amprenavir. 

The well designed clinical development program of fos-amprenavir/ritonavir included both treatment-naïve (versus nelfinavir) and experienced (versus lopinavir/ritonavir) patients.  Data from these studies provides for a more detailed understanding of the drug’s utility than the more meager studies that lead to the approval of amprenavir. 

The understanding of the pharmacokinetics of boosted amprenavir substantially outpaced the clinical demonstration of its effectiveness.  The Food and Drug Administration approved 2 amprenavir/ritonavir combinations, a 600 mg/100 mg twice-daily regimen and a once-daily 1200 mg/200 mg regimen.  Similar doses of fos-amprenavir, 700/100 bid, 1400/200 and unboosted fos-amprenavir are approved.  Data from the treatment-experienced patients where fos-amprenavir was tested against lopinavir/ritonavir demonstrated that the bid regimen was comparable to the comparator arm, while the once daily regimen performed significantly less well.  For all patients with prior PI experience fos-amprenavir should only be used bid.

Fos-amprenavir was studied among treatment-experienced patients when used alone or in combination with lopinavir/ritonavir versus lopinavir/ritonavir alone.  This study sponsored by the ACTG was halted after completion of a planned pharmacokinetic study.  This study revealed that the exposures of both fos-amprenavir and lopinavir were lowered significantly when the agents were used together.  The degree of reduction of all meaningful pharmacokinetic parameters were far greater than those seen with the original amprenavir formulation. 

It is clear that fos-amprenavir and lopinavir/ritonavir cannot be used together.  As the current marketing plan for fos-amprenavir calls for the removal of amprenavir from the market, this creates a dilemma for both patients and clinicians in the management of patients currently successfully maintained on amprenavir-lopinavir/ritonavir combinations

Indinavir/Ritonavir

The optimal dose of indinavir boosted with ritonavir has not yet been established. There is clear evidence of increased indinavir-related toxicity at all commonly used doses (400/400, 800/100 and 800/200). This combination has, however, shown some ability to overcome unboosted indinavir-resistant variants. Indinavir is normally taken 3 times daily without food or with a light meal. The pharmacokinetic profile of indinavir is improved when indinavir is combined with ritonavir. Co-administration of indinavir and ritonavir produces a 4.5-fold increase in C_max and a 1.8-fold increase in AUC, both substantially in excess of the in vitro IC₅₀, thus permitting twice-daily dosing without regard to meal requirements.

Indinavir is the only PI used in a boosted regimen for which failure of therapy with the primary PI, followed by indinavir with ritonavir enhancement, has been the subject of formal study.   Whether a similar pattern of response would be seen with other PIs is unknown, but might reasonably be expected. 

Indinavir/ritonavir, 400/100 has been studied in several French cohorts.  Used primarily in treatment-naïve patients, this dosage combination appears well tolerated, and corresponding pharmacokinetic data confirm exposures greater than that achieved by indinavir alone, and in excess of anticipated IC₅₀ for wild type virus.  If indinavir is to regain a role in modern HIV therapeutics, further study of this dose will be required.

Atazanavir/Ritonavir

Atazanavir has been studied in both treatment-naïve and experienced patients.  It appears to have comparable activity to both nelfinavir and efavirenz when given unboosted in naïve patients.  In treatment experienced patients atazanavir was tested against lopinavir/ritonavir in two studies in boosted (300mg atazanavir/100mg ritonavir) and unboosted form. 

Unboosted atazanavir was inferior to lopinavir/ritonavir, while in its boosted form it appears to have comparable activity.  It may be premature to state that atazanavir could or should be used for all patients in preference to lopinavir/ritonavir.  In subanalyses of the study cited above it appears that boosted atazanavir did not perform as well as lopinavir/ritonavir in patients with a greater number of baseline PI or NRTI mutations. 

Further study of boosted atazanavir will be forthcoming, including a study of boosted versus unboosted atazanavir in treatment-naïve patients.  These studies and clinical experience should delineate whether ritonavir boosting increases the well described toxicities associated with atazanavir. 

Can PIs Be Developed That Do Not Require Pharmacokinetic Enhancement?

Summary

PIs have revolutionized the treatment of HIV infection and AIDS through durable suppression of HIV replication, restoration and preservation of immune function, and marked reduction in clinical events. Nonetheless, the currently available PIs have pharmacokinetic profiles that can limit their efficacy, and most are now paired with ritonavir to improve exposures and enhance clinical efficacy.

Ritonavir enhancement is, however, not without risk. Drug toxicity can result, secondary to elevated plasma levels of the primary PI or of ritonavir itself, and triglyceride and cholesterol levels can become markedly elevated in some. Numerous drug-drug interactions and activation of xenobiotic response elements may prove to limit the benefits of boosted therapy. Finally, the long-term effect of chronic inhibition of the CYP enzyme systems and cellular efflux pumps by ritonavir is unknown.

The primary challenge in drug development of the HIV PIs today is to fashion drugs that are active against viral strains exhibiting multiple-drug resistance. However, the next generation of PIs must also provide for better inherent pharmacologic exposures than do the currently available agents. These new agents must increase effectiveness with only limited additional toxicity, though, if they are be useful in the long-term treatment strategies required for HIV infection and AIDS.

02/01/04

References

MP Dube and others. Guidelines for the Evaluation and Management of Dyslipidemia in Human Immunodeficiency Virus (HIV)-Infected Adults Receiving Antiretroviral Therapy: Recommendations of the HIV Medicine Association of the Infectious Disease Society of America and the Adult AIDS Clinical Trials Group (IDSA Guidelines). Clinical Infectious Diseases 37: 613-627. September 3, 2003.

Becker SL. Expert Review of Anti-infective Therapy 1(13): 403-413. 2003.

Dresser GK, Spence JD, Bailey DG. Pharmacokinetic-pharmacodynamic consequences and clinical relevance of cytochrome P450 3A4 inhibition. Clin Pharmacokinet 2000, 38:41–57.

Dussault I, Lin M, Hollister K, Wang EH, Synold TW, Forman BM. Peptide mimetic HIV protease inhibitors are ligands for the orphan receptor SXR. J Biol Chem 2001, 276:33309–33312.  

Fellay J, Marzolini C, Meaden ER et al for the Swiss HIV Cohort Study. Response to antiretroviral treatment in HIV-1-infected individuals with allelic variants of the multidrug resistance transporter 1: a pharmacogenetics study. Lancet 2002, 359:30–36.     

Fichtenbaum CJ, Gerber JG, Rosenkranz SL et al the NIAID AIDS Clinical Trials Group. Pharmacokinetic interactions between protease inhibitors and statins in HIV seronegative volunteers: ACTG Study A5047. AIDS 2002, 16:569–577.

Huisman MT, Smit JW, Schinkel AH. Significance of P-glycoprotein for the pharmacology and clinical use of HIV protease inhibitors. AIDS 2000, 14:237–242.   

Piscitelli SC, Gallicano KD. Interactions among drugs for HIV and opportunistic infections. N Engl J Med 2001, 344:984–996.  

Park-Wyllie LY, Scalera A, Tseng A, Rourke S. High rate of discontinuations of highly active antiretroviral therapy as a result of antiretroviral intolerance in clinical practice: missed opportunities for adherence support? AIDS 2002, 16:1084–1086.         

Table 1. Determinants of PI Exposure

Table 2. Key Pharmacokinetic Characteristics of Selected PIs