Ritonavir-boosted protease inhibitors in HIV therapy

Abstract

The advent of combination antiretroviral therapy has led to significant improvement in the care of HIV-infected patients. Originally designed as a protease inhibitor (PI), ritonavir is currently exclusively used as a pharmacokinetic enhancer of other protease inhibitors, predominantly due to ritonavir's potent inhibition of the cytochrome P450 3A4 isoenzyme. Ritonavir-boosting of PIs decrease pill burden and frequency of dosing. Boosted PIs are recommended for first-line therapy in treatment and play a key role in the management of treatment-experienced patients. Potential problems associated with PIs include metabolic abnormalities (e.g. dyslipidemia), increased cardiovascular risk, and drug interactions.

Key words::

• Antiretroviral therapy
• HIV
• protease inhibitors
• ritonavir

Key messages

• Modern protease inhibitors require the use of low-dose ritonavir to boost pharmacokinetic exposure through inhibition of metabolism via the cytochrome P450 3A4 enzyme pathway.

• Boosted protease inhibitors are recommended components of first-line therapy in treatment-naive patients, and new agents such as ritonavir-boosted darunavir play a major role in the management of treatment-experienced patients.

• Common adverse events include metabolic abnormalities including dyslipidemia and potentially cardiovascular disease.

• Ritonavir-boosting is frequently associated with drug interactions which require evaluation prior to initiation of protease inhibitor-based therapy, or prior to initiating new medications in those already receiving a boosted protease inhibitor.

Introduction

The advent of combination antiretroviral therapy revolutionized the care of HIV-infected patients, with profound impact on morbidity and mortality (1,2). Advances in the understanding of the HIV replication cycle and elucidation of unique viral proteins, such as the HIV protease, have allowed targeted drug development (3–8). Since the early randomized clinical trials of combination antiretroviral therapy just over one decade ago, over 20 agents have been approved for use in treating HIV-infected patients (9,10). Although treatment guidelines have evolved significantly over this time-period, the use of protease inhibitors (PIs), and in particular the use of ritonavir-boosted PIs, has remained a mainstay of therapy (Table I) (11). An understanding of the pharmacology of boosted PIs and the current clinical trial data for the use of the agents is important in the management of HIV-infected patients in the modern antiretroviral era.

Table I. Recommended first-line therapy for HIV-infected treatment-naive patients in 2010.

	European AIDS Clinical Society Guidelines (130)	International AIDS Society-USA Panel Guidelines (11)	US DHHS Guidelines (131)
Recommended dual NRTI component	Tenofovir-emtricitabine or Abacavir-lamivudine	Tenofovir-emtricitabine	Tenofovir-emtricitabine
Alternative NRTI component	Zidovudine-emtricitabine or Didanosine + lamivudine	Abacavir-lamivudine	Abacavir-lamivudine or Zidovudine-lamivudine
Recommended third agent	NNRTI: Efavirenz, Nevirapine Protease Inhibitors: Atazanavir/ritonavir, Darunavir/ritonavir, Lopinavir-ritonavir Saquinavir/ritonavir	Efavirenz or Atazanavir/ritonavir, Darunavir/ritonavir or Raltegravir	Efavirenz or Atazanavir/ritonavir, Darunavir/ritonavir, or Raltegravir
Alternative third agent	Fosamprenavir/ritonavir or Raltegravir	Lopinavir-ritonavir, Fosamprenavir/ritonavir	Nevirapine or Lopinavir-ritonavir, Fosamprenavir/ritonavir, Saquinavir/ritonavir

HIV protease

HIV replication includes the production of proteins from the three major HIV genes: gag, pol, and env. Together these genes encode the structural and replicative proteins of the virus, with gag encoding proteins involved in the core and nucleocapsid, env those of the surrounding envelope, and pol encoding the HIV reverse transcriptase, ribonuclease H, integrase, and protease (12). These pol-encoded replicative proteins are translated initially in the form of a large precursor polypeptide, gag-pol. The HIV protease is able to excise itself from this fusion protein and subsequently cleaves the remaining gag-pol chain into the other component proteins. The protease also is required for cleavage of separately translated gag polyprotein, although the env polyprotein is processed by host cell enzymes.

The protease itself is a homodimer of two 99-amino acid monomers. The HIV protease is classified as an aspartyl protease due to the presence of a conserved Asp-Thr-Gly sequence that each monomer contributes to the active catalytic site of the enzyme (4,13). There are five non-contiguous conserved regions within the protease including both the catalytic site and the substrate-binding domain (11,12). There is homology of approximately 50% between the proteases of HIV-1 and HIV-2 viruses, and the PIs thus have some activity against HIV-2 (13). At present in-vitro phenotypic assays of PI activity against HIV-2 have demonstrated decreased susceptibility to agents such as amprenavir and nelfinavir due to naturally occurring polymorphisms, while agents such as saquinavir, lopinavir, and darunavir exhibit the highest potency (14–16). Although aspartyl proteases are found within mammalian systems (e.g. renin), these proteins are asymmetric in comparison with the symmetric viral protein, and thus compounds with selective inhibition of viral protease have been developed.

Genotypic mutations may lead to altered protease structure and development of resistance. Protease inhibitors, in contrast to first-line non-nucleoside reverse transcriptase inhibitors (NNRTIs) and certain nucleoside reverse transcriptase inhibitors (NRTIs), require the development of multiple resistance mutations before drug effect is compromised. Guidelines for interpreting resistance mutations for protease inhibitors are updated on a regular basis (17).

Early protease inhibitors

Four PIs, saquinavir, indinavir, nelfinavir, and ritonavir, were licensed for use during the period 1995–1997. These are currently no longer recommended for first-line therapy in their original formulations (11). Use of these agents was limited by poor bioavailability and hence the need for more frequent dosing and higher pill burden with concomitant toxicities.

Saquinavir

Originally developed in a hard-gel formulation, saquinavir had limited bioavailability of only 4% (18). Saquinavir monotherapy studies showed only short-term virologic suppression, although combination therapy with nucleosides did have longer-term benefit (19,20). Long-term outcomes with the original hard-gel capsule were inferior to comparator PIs at the time (21). Tolerability was limited by gastrointestinal side-effects, particularly at high doses, and the pill burden required in some studies could be prohibitive (19). A soft-gel capsule formulation of saquinavir was developed and evaluated in conjunction with ritonavir-boosting (see below) (22,23), but this was discontinued in 2006 and replaced by a new 500 mg hard-gel formulation boosted by ritonavir (24,25).

Indinavir

Administered at a dose of 800 mg three times daily, indinavir could not be taken with food due to significant decrease in drug absorption (26). Although combination therapy including indinavir with two nucleoside analogs was shown to achieve durable virologic suppression, metabolic toxicities and nephrolithiasis have limited the use of indinavir in the modern antiretroviral era (9,27–29).

Nelfinavir

Nelfinavir had better bioavailability, which was further maximized when administered with a meal (particularly a high fat content meal) (30). Nelfinavir has a serum half-life of 3.5–5 hours and is metabolized via the cytochrome P450 CYP2C19 isoenzyme system; as such it does not benefit significantly from ritonavir-boosting which occurs primarily through ritonavir inhibition of the P450 3A4 isoenzyme (31). Nelfinavir was found to have antiviral activity in initial phase I monotherapy studies, with sustained virologic suppression of 1.6 log₁₀ at 12 months in some patients (32). Initial combination studies showed that the addition of nelfinavir (dosed 750 mg t.i.d.) to the contemporary standard of two nucleoside analogs resulted in 67% of patients achieving a viral load <400 copies/mL at 24 weeks compared to 7% in the placebo arm (33). At 48 weeks, 61% of those receiving nelfinavir 750 mg t.i.d. were able to achieve virologic suppression with viral load <50 copies/mL (33). Nelfinavir has been found to be inferior to non-nucleoside reverse transcriptase inhibitors (NNRTIs) when combined with two NRTIs (34,35). In an evaluation of initial therapy comprised either of four-drug regimens or standard three-drug regimens, the combination of nelfinavir and two NRTIs was inferior to virologic outcomes obtained with starting with zidovudine, lamivudine, and the NNRTI agent efavirenz (34). In addition, nelfinavir is inferior to boosted PIs such as lopinavir-ritonavir (see below) (36). Diarrhea is a common side-effect of nelfinavir, seen in the initial dose escalation studies and in clinical trials (33,37).

Ritonavir

The third protease inhibitor to be licensed, ritonavir was initially assessed for use as an independent agent before subsequent pharmacologic analysis demonstrated significant advantage to combination therapy with low-dose ritonavir and other PIs: the principle of ritonavir-boosting.

Ritonavir is available both in liquid formulation (80 mg/mL) and as capsules (100 mg). The capsules require refrigeration for long-term storage, while the liquid formulation does not. Ritonavir bioavailability is high (38,39) and may be increased if taken with a meal. Recently ritonavir tablets using melt-extrusion technology (Meltrex, SOLIQS, Abbott GmbH, Ludwigshaven, Germany) have been developed (40). Ritonavir is highly protein-bound with a half-life of 3–5 hours (41) and will decrease its own metabolism due to auto-induction of the CYP3A4 isoenzyme system (39). The ritonavir tablet formulation is stable at room temperature and does not require refrigeration. The tablet formulation has been compared to the standard soft-gel capsule in 24 subjects in a phase I open-label cross-over study. The tablet formulation of ritonavir was found to have higher area under the curve (AUC) (12% increase) and maximum concentration (C_max) (21% higher) than the capsule formulation (40). The tablet formulation of ritonavir was introduced in Europe in 2010 and will likely enter clinical use in North America in 2011 as a component of ritonavir-boosted protease inhibitor regimens (see below).

In monotherapy phase I/II studies, administration of ritonavir 600 mg b.i.d. resulted in drops in HIV RNA of at least 1.0 log₁₀ in the first 4 weeks, and a drop of 0.8 log₁₀ at week 32 in a small subset of patients (41,42). Ritonavir added to two nucleosides in treatment of patients with advanced HIV (CD4 cell counts <100 cells/mm³) was found to decrease the risk of opportunistic infections and prolong survival (43). Due to poor tolerance and toxicities, use of full-dose ritonavir is no longer recommended.

Ritonavir-boosted combination protease inhibitor therapy

Concomitant administration of ritonavir at sub-therapeutic levels (100–400 mg daily) has profound effects on the absorption and metabolism of other PIs (44). The metabolism of PIs is complex, with systemic drug concentrations reflecting the combined effects of first-pass metabolism due to the action of CYP3A4 isoenzyme and efflux pump P-glycoprotein within the gut enterocyte, and hepatic metabolism via the cytochrome P450 CYP3A4 isoenzyme (45). Systemic drug concentrations may also be affected by excretory mechanisms, in particular the effects of cellular transport systems such as multidrug resistance proteins (MRP-1) in addition to P-glycoprotein (46–48).

Ritonavir is a potent inhibitor of cytochrome P450 CYP3A4 isoenzyme present both in the intestinal tract and liver. Ritonavir may also play a role in limiting cellular transport of other PIs via the P-glycoprotein and MRP efflux channels (48,49). These effects have significant impact on the primary PI pharmacokinetic parameters, including the minimum concentration (C_min), AUC, and half-life, and thus on therapeutic efficacy. The use of low-dose ritonavir as a pharmacokinetic enhancer allows extended dosing intervals for the primary PI (once or twice daily dosing), with associated improvements in pill burden and adherence (50), and improves potency with decreased likelihood of resistance compared to unboosted PIs (51).

Ritonavir-boosted PIs are now one of the first-line regimens recommended for treatment-naive patients and play a primary role in second-line regimens and therapy for treatment-experienced patients (11).

Current ritonavir-boosted protease inhibitors

There are currently six protease inhibitors recommended for the treatment of HIV infection. These include lopinavir, atazanavir, fosamprenavir, darunavir, tipranavir, and, saquinavir. In all instances these are recommended in combination with ritonavir-boosting. Atazanavir is the only protease inhibitor which under special circumstances may be used without ritonavir-boosting. These agents (with the exception of tipranavir) have undergone extensive comparative evaluation in treatment-naive patients. Darunavir and tipranavir were originally developed in the setting of treatment-experienced patients and remain a key component of therapy in these patients.

Lopinavir-ritonavir

Lopinavir is structurally related to ritonavir; however, its poor oral bioavailability and extensive first-pass metabolism mandate it be used with ritonavir as a booster (52,53). After oral administration of a single 400 mg dose in healthy volunteers, peak lopinavir concentrations only briefly exceeded the mean 50% effective concentration (EC₅₀) (0.1 mg/L) of wild-type virus; however, addition of a single 50 mg ritonavir dose increased lopinavir C_max to 5.5 mg/L due to inhibition of CYP3A4 metabolism of lopinavir (52,54). Lopinavir-ritonavir is currently the only co-formulated boosted PI combination. Originally available as a soft-gelatin capsule or a liquid, a tablet formulation is now available. The tablet (Meltrex) formulation is bioequivalent to the capsule and liquid formulations but does not require co-administration with a meal (55,56). In addition, the tablet formulation is heat-stable, does not require refrigeration, and decreases pill burden from three soft-gel capsules to two tablets twice daily (56). The lopinavir C_max for the capsule and tablet formulations are 6.2 μg/mL and 8.0 μg/mL, respectively, with a half-life of 4–6 hours with multiple-dose administration (55).

Clinical studies. Initial dose ranging studies identified a recommended dose of lopinavir-ritonavir of 400 mg/100 mg twice daily in treatment-naive patients when combined with a dual nucleoside backbone (57). Four-year follow-up of patients receiving open-label lopinavir-ritonavir 400 mg/100 mg-based therapy demonstrated 70% suppression of viral load <50 copies/mL (intent-to-treat analysis) (58). Lopinavir-ritonavir 400 mg/100 mg twice daily was virologically superior to nelfinavir 750 mg three times daily in a randomized double-blind trial of 653 treatment-naive patients (36) (Table II). At 48 weeks 75% of those receiving lopinavir-ritonavir achieved a viral load <400 copies/mL compared to 63% of those receiving nelfinavir (P < 0.001). Similarly a greater proportion of patients achieved viral load suppression with viral load <50 copies/mL (67% versus 52%; P < 0.001). Patients receiving lopinavir- ritonavir demonstrated a greater proportion of sustained virologic response during the 48 week trial (84% versus 66%). When patients with viral loads >400 copies/mL during weeks 24–48 were assessed for genotypic resistance, it was found that no patient receiving lopinavir-ritonavir had evidence of genotypic PI resistance mutations compared to 33% of those treated with nelfinavir (36). Lopinavir-ritonavir was associated with higher increase in mean triglycerides (1.4 mmol/L versus 0.5 mmol/L; P < 0.001), but rates of diarrhea were similar, and overall discontinuations due to adverse events were comparable (3.4% versus 3.7%).

Table II. Outcomes of boosted protease inhibitor comparator trials in HIV-infected treatment-naive patients at 48 weeks.

		GEMINI	KLEAN	CASTLE	ARTEMIS
	Nelfinavir versus Lopinavir-ritonavir (36)	Saquinavir/ritonavir versus Lopinavir-ritonavir (25)	Fosamprenavir/ritonavir versus Lopinavir-ritonavir (74)	Atazanavir/ritonavir versus Lopinavir-ritonavir (85)	Darunavir/ritonavir versus Lopinavir-ritonavir (103)
n	653	337	878	883	689
Entry viral load criteria	Age >12 y; VL >400 copies/mL	Age >18 y; VL >10,000 copies/mL	Age >18 y; VL >1000 copies/mL	Age >18 y; VL >5000 copies/mL	Age >18 y; VL >5000 copies/mL
NRTI backbone	Stavudine + lamivudine	Tenofovir + emtricitabine	Abacavir + lamivudine	Tenofovir + emtricitabine	Tenofovir + emtricitabine
VL <400 copies/mL (%)	63% versus 75%	72.5% versus 74.7%	73% versus 71%	NA	NA
VL <50 copies/mL	52% versus 67%	64.7% versus 63.5%	66% versus 65%	78% versus 76%	84% versus 78%
CD4 cell count gain (mean or median) cells/mm^3	195 versus 207	178 versus 204	176 versus 191	203 versus 219	137 versus 141
Total cholesterol mean or median increase	1.24 mmol/L versus 1.37 mmol/L	26 mg/dL versus 31 mg/dL	61 mg/dL versus 52.5 mg/dL	17 mg/dL versus 38 mg/dL	0.5 mmol/L versus 0.8 mmol/L
Triglycerides mean or median increase	0.5 mmol/L versus 1.4 mmol/L	14 mg/dL versus 55 mg/dL	67 mg/dL versus 77.5 mg/dL	14 mg/dL versus 58 mg/dL	0.1 mmol/L versus 0.8 mmol/L

VL = plasma viral load.

Lopinavir-ritonavir has been compared to an NNRTI-based regimen with a 48 week comparison to efavirenz (59). In an open-label study of 757 patients randomized to lopinavir-ritonavir (400 mg/100 mg soft-gel capsules) or efavirenz with two NRTIs (42% received zidovudine and lamivudine, 34% tenofovir and lamivudine), 89% of those receiving efavirenz had a viral load <50 copies/mL versus 77% in those receiving lopinavir-ritonavir at 96 weeks (P = 0.03) (59). Of note, virological failure on the efavirenz-based regimen was associated with increased proportion of virologic resistance (9% versus 6%). Failure on efavirenz was associated with class resistance to the NNRTIs, but failure on lopinavir-ritonavir was not associated with major PI mutations (59). Similar results were obtained when this issue was evaluated among patients starting therapy with CD4 cell counts below 200/mm³ (60).

Originally approved as twice daily therapy, lopinavir-ritonavir is now licensed for once daily use in treatment-naive patients. In an open-label trial 38 patients were randomized to receive lopinavir- ritonavir 800 mg/200 mg once daily or 400 mg/100 mg twice daily in combination with stavudine and lamivudine (61). Outcomes were similar, with 74% of those receiving once daily, and 79% twice daily, achieving viral load <50 copies/mL at 48 weeks (61). The median pre-dose (trough) concentrations of once daily and twice daily dosing exceeded the in-vitro concentrations needed to inhibit replication of wild-type virus by 50% by 40 and 84-fold respectively (61). Similar outcomes were seen in two large randomized trials of 190 and 320 patients receiving once or twice daily lopinavir-ritonavir capsules (62,63). In an analysis of 190 patients randomized to once or twice daily lopinavir-ritonavir capsules, 70% (once daily) and 64% (twice daily) achieved virologic suppression at <50 copies/mL at 48 weeks (62). Diarrhea was more common in the once daily group (16% versus 5%; P = 0.036) (62). In an analysis of 320 patients randomized to once or twice daily lopinavir-ritonavir, the probability of achieving a sustained virologic response to therapy was similar overall, but the probability of sustained virologic response was better in those receiving twice daily therapy in the stratum with a base-line HIV RNA >100,000 copies/mL (63). In an assessment of the lopinavir- ritonavir tablet formulation, 664 patients were randomized to once or twice daily dosing of lopinavir-ritonavir tablets (64). At 48 weeks, 77% (once daily) and 76% (twice daily) achieved virologic suppression with viral load <50 copies/mL. In this analysis, rates of diarrhea were similar between those receiving once daily or twice daily lopinavir-ritonavir tablets in an initial 8 week lead-in phase consisting of either tablet or capsule formulation (64).

Saquinavir/ritonavir

Clinical studies. In early comparison trials saquinavir/ritonavir was found to be superior to indinavir/ ritonavir (22). In the MaxCMin1 study, 306 patients were randomized to boosted indinavir (800 mg/ 100 mg twice daily) or saquinavir soft-gel capsules (1000 mg/100 mg twice daily). Although virologic outcomes were similar between arms (27% virologic failure for those receiving indinavir/ritonavir versus 25% receiving saquinavir/ritonavir), 41% of those randomized to indinavir switched regimens compared to 27% of those receiving saquinavir, primarily due to adverse events (22).

Saquinavir/ritonavir (100mg/100mg twice daily) was compared to lopinavir-ritonavir in 324 patients enrolled in the MaxCMin2 study (23). At 48 weeks in an intent-to-treat analysis, 33% of those receiving saquinavir/ritonavir experienced virologic failure compared to only 18% in the lopinavir-ritonavir arm (P = 0.002 log-rank test) (23). In addition, 14% of those receiving lopinavir-ritonavir discontinued therapy compared to 30% of those receiving saquinavir/ritonavir. Discontinuations, as in the MaxCMin1 trial, were primarily due to adverse events (23).

Recently, the new formulation of saquinavir 500 mg hard-gel capsules boosted by ritonavir (1000 mg/100 mg twice daily) was compared to lopinavir-ritonavir both in conjunction with tenofovir/emtricitabine in 337 treatment-naive patients randomized in the open-label non-inferiority GEMINI trial (25) (Table II). At 48 weeks 64.7% versus 63.5% of those in the saquinavir/ritonavir arm and lopinavir-ritonavir arm, respectively, demonstrated virologic suppression with viral load <50 copies/mL. Saquinavir/ritonavir was non-inferior to lopinavir-ritonavir, with an estimated difference in proportion for non-inferiority of 1.14 (95% confidence interval -9.6 to 11.9) (25). Adverse events profiles were similar, although lopinavir-ritonavir was associated with a higher increase in triglycerides over 48 weeks (Table II).

Fosamprenavir/ritonavir

Amprenavir has been evaluated in clinical trials in combination with two nucleosides with and without ritonavir-boosting (65–68). Amprenavir was compared to indinavir in PI-naive patients and was found to be inferior, with 30% achieving viral load <400 copies/mL compared to those receiving indinavir (69). Amprenavir has now been withdrawn from the market and replaced by its prodrug fosamprenavir.

Fosamprenavir is converted to amprenavir within the epithelial cells of the intestinal tract (70). Fosamprenavir has a similar AUC pharmacokinetic profile to amprenavir but higher trough concentration (71), with median C_min of 0.35 μg/mL, t_max of 1.3 hours, and half-life of approximately 7 hours (69).

Clinical studies. Unboosted fosamprenavir 1400 mg twice daily has been compared to nelfinavir 1250 mg twice daily in combination with abacavir and lamivudine in 249 treatment-naive patients in the NEAT study (72). Patients randomized to receive fosamprenavir were more likely to demonstrate virologic suppression to viral load <400 copies/mL compared to nelfinavir (66% versus 51%, respectively) after 48 weeks. Fosamprenavir was better tolerated, with less diarrhea (5% versus 18%) than nelfinavir (72).

In the SOLO trial of 649 treatment-naive patients, boosted fosamprenavir (1400 mg/200 mg once daily) was compared to twice daily nelfinavir 1250 mg in combination with abacavir and lamivudine (73). Patients receiving boosted fosamprenavir had similar virologic outcomes, with 69% achieving a viral load <400 copies/mL compared to 68% of those receiving nelfinavir (73).

Fosamprenavir/ritonavir 700 mg/100 mg twice daily was compared to lopinavir-ritonavir 400 mg/ 100 mg twice daily in combination with abacavir-lamivudine among 878 treatment-naive patients in the KLEAN study (74). At 48 weeks 73% of those receiving fosamprenavir and 71% of those receiving lopinavir achieved viral load <400 copies/mL (Table II) (74). Non-inferiority was demonstrated, with the lower bound of the confidence interval for the treatment difference falling within the pre-specified lower bound of -12%. Treatment adverse events were similar between the two arms at 12% and 10%, respectively.

The use of fosamprenavir combined with a lower dose of ritonavir at 100 mg daily has also been evaluated (75). In the COL100758 study, 115 treatment-naive patients were randomized to receive either fosamprenavir/ritonavir 1400 mg/200 mg daily or fosamprenavir/ritonavir 1400 mg/100 mg daily in conjunction with abacavir-lamivudine (76). At 96 weeks in the intention-to-treat analysis, 66% of those receiving 100 mg ritonavir achieved a viral load <50 copies/mL versus 53% in those on 200 mg daily. The 100 mg ritonavir dosing was also associated with lower triglyceride elevations (27 mg/dL versus 48 mg/dL). Once daily fosamprenavir in combination with 100 mg ritonavir has also been compared to once daily atazanavir in the ALERT study (see below).

Atazanavir/ritonavir

Licensed in 2003, boosted atazanavir is currently a preferred first-line PI (11). Atazanavir 300 mg is given once daily when boosted with 100 mg of ritonavir. Atazanavir is rapidly absorbed, but absorption is dependent upon gastric pH, and therefore concomitant medications that block gastric pH, particularly proton pump inhibitors, should be avoided. Food exposure boosts atazanavir exposure by as much as 70%, with a half-life of 6–7 hours if unboosted and approximately 11 hours if boosted with ritonavir (77). Atazanavir C_min is increased 11-fold in the presence of ritonavir (78). Atazanavir exposure is lowered if combined with tenofovir, and as such ritonavir-boosted atazanavir is recommended for those patients receiving this combination (79,80) The proposed minimum effective concentration of atazanavir is 0.15 mg/mL, and the median trough concentration obtained using once daily atazanavir 300 mg/d in combination with ritonavir 100 mg/d was 0.71 mg/mL (interquartile range 0.39–0.91 mg/mL) (81). Higher atazanavir trough concentrations have been associated with increased indirect bilirubin levels, a commonly reported side-effect of this agent (82).

Clinical studies. Phase II dose-ranging studies compared unboosted atazanavir to nelfinavir in treatment-naive patients (83,84). After 48 weeks in combination with stavudine and lamivudine, unboosted atazanavir 400 mg/d had similar efficacy to nelfinavir, with 35% versus 34% achieving viral load <50 copies/mL (84).

Atazanavir/ritonavir has been compared to lopinavir-ritonavir in a large randomized trial in combination with tenofovir-emtricitabine (85). In the CASTLE study, 883 patients were randomized to atazanavir/ritonavir 300 mg/100 mg or lopinavir-ritonavir 400 mg/ 100 mg soft-gel capsules (Table II). At 48 weeks virologic suppression to <50 copies/mL was achieved in 78% and 76% of the atazanavir/ritonavir and lopinavir-ritonavir arms, respectively (difference 1.7%; 95% CI -3.8 to 7.1). Fewer patients receiving atazanavir experienced diarrhea (2% versus 11%), but 34% had documented grade 3–4 increases in total bilirubin versus <1% in those receiving lopinavir-ritonavir (85).

Atazanavir/ritonavir has also been compared to once daily fosamprenavir/ritonavir 1400 mg/100 mg in combination with tenofovir-emtricitabine in the ALERT study (86). In a 48 week comparison in 106 subjects, 83% of those receiving atazanavir/ritonavir 300 mg/100 mg achieved viral load <50 copies/mL compared to 75% receiving fosamprenavir (P = 0.43). Patients receiving fosamprenavir had higher triglyceride values at week 48 but fewer bilirubin-related adverse events (86).

Atazanavir has been compared to NNRTI therapy in treatment-naive patients. Unboosted atazanavir was compared to efavirenz in combination with zidovudine-lamivudine over 48 weeks in 810 patients (87). Virologic outcomes were similar, with 70% versus 64% achieving viral load <400 copies/mL for atazanavir and efavirenz, respectively, and no significant differences in cholesterol profiles were observed (87). Boosted atazanavir has been compared to efavirenz in the ACTG 5202 study (88). This was a four-arm study involving 1858 patients, comparing nucleoside backbones in a blinded fashion (abacavir-lamivudine versus tenofovir-emtricitabine) in combination with either boosted atazanavir or efavirenz (88). Interim analysis of this study found higher rates of virologic failure in those with base-line viral load >100,000 copies/mL who were prescribed abacavir-based combinations, regardless of third agent (88). Final 96 week outcomes demonstrated similar virologic outcomes for the comparison of boosted atazanavir and efavirenz, with HR 1.13 (95% CI 0.82–1.56) with an abacavir-based backbone and HR 1.01 (95% CI 0.70–1.46) for tenofovir-based regimens (89). Results of a randomized comparison of atazanavir/ritonavir with nevirapine (ARTEN study) have been presented in abstract form. Nevirapine (in combination with tenofovir-emtricitabine) demonstrated non-inferiority to atazanavir/ritonavir in 569 treatment-naive patients after 48 weeks, with 66.8% versus 65.3% achieving treatment response, respectively (90).

The use of atazanavir for treatment maintenance following sustained suppression with boosted PI-based regimens was evaluated in the SWAN and ARIES studies (91,92). In the SWAN study 419 patients were randomized to remain on a stable PI regimen (54% boosted, 68% of which was lopinavir-ritonavir) or changed to atazanavir (either unboosted or boosted if the background regimen also included tenofovir). Those switched to atazanavir (91% received unboosted atazanavir) had better virologic outcomes when compared to those who received older unboosted PI regimens but overall had fewer adverse events or cholesterol abnormalities (91). In the ARIES study, patients with virologic suppression after 36 weeks of atazanavir/ritonavir in conjunction with abacavir-lamivudine were randomized to continue on a boosted regimen or switch to unboosted atazanavir. Those randomized to unboosted atazanavir (210 patients) had similar rates of continued virologic suppression (86% versus 81%) to those receiving boosted atazanavir (209 patients), meeting the pre-specified non-inferiority criteria. Discontinuation of ritonavir resulted in improved bilirubin and lipid profiles (92).

New generation boosted protease inhibitors: darunavir and tipranavir

The next-generation protease inhibitors tipranavir and darunavir retain activity in the setting of multiple protease mutations and have clinical activity against multi-PI-resistant strains. These new ritonavir-boosted PIs, particularly darunavir, have contributed substantially to improve the virological rates of suppression among highly treatment-experienced patients, particularly those with multiple drug resistant HIV.

Darunavir

Darunavir is highly active against HIV protease, achieving higher concentrations above the IC₅₀ than the majority of other PIs (93). Darunavir binds with very high affinity and fits closely within the substrate pocket, giving it activity even against multidrug-resistant proteases (94,95). Darunavir exposure increases by approximately 30% if taken with food, reaching peak trough concentrations within 2.5–4 hours. Darunavir has been assessed at doses of 600 mg/100 mg twice daily or 800 mg/100 mg daily with C_trough values of 3.5 and 2.0 μg/mL, respectively (96).

Clinical studies. Initially assessed in treatment-experienced patients within the POWER 1 (318 subjects) and 2 (278 subjects) studies, darunavir/ritonavir at differing doses was compared to an investigator-selected control ritonavir-boosted PI, given in addition to an optimized background regimen over 24 weeks (97,98). Patients were treatment-experienced, with at least one PI mutation and exposure to a PI for at least 8 weeks before enrolment. Patients (n = 250) continued to be followed in an open-label fashion to 48 weeks (99). Those receiving darunavir/ritonavir 600 mg/100 mg twice daily achieved significantly higher levels of virologic suppression (45% versus 10%) than the comparator PIs (36% lopinavir-ritonavir), with comparatively fewer side-effects (99). Similar outcomes were seen in the open-label, non-randomized POWER 3 study (n = 246) (100). Darunavir/ritonavir was compared to lopinavir-ritonavir in treatment-experienced patients in the TITAN study (101). In this study, treatment experience was defined as prior exposure to highly active antiretroviral therapy (HAART) for at least 3 months, with no specific requirements for evidence of protease mutations. Overall 595 patients were randomized, with 47% overall being triple-class experienced, but with a median of 0 (range 0–6) primary PI mutations (101). At week 48, significantly more patients receiving darunavir/ ritonavir achieved viral load <50 copies/mL than lopinavir-ritonavir (71% versus 60%; P = 0.005), and it was considered non-inferior. In addition, selection of PI mutations was lower in the darunavir/ritonavir arm. The majority of those receiving lopinavir-ritonavir received soft-gel capsule formulations, and triglyceride elevations were higher in this group.

Simplification of dosing with the use of once daily darunavir in patients with limited treatment experience has been assessed (ODIN trial) (102). Patients with no underlying darunavir-associated resistance mutations and detectable viral load (>1000 copies/mL) on a stable HAART regimen were randomized to once daily (800 mg/100 mg) darunavir/ritonavir or standard twice daily dosing. Overall 590 patients were randomized, and at 48 weeks 72.1% of those receiving once daily dosing achieved virologic suppression compared to 70.9% in the twice daily arm, meeting study criteria for non-inferiority (102). However, patients in this study had high rates of active agents within the background regimen, compared to earlier darunavir studies in treatment-experienced patients, and also had low rates of primary protease mutations. Caution may be required before these results are applied to more experienced patient populations.

Once daily darunavir/ritonavir 800 mg/100 mg was compared to once or twice daily lopinavir-ritonavir (total daily dose 800 mg/200 mg), in combination with tenofovir-emtricitabine in treatment-naive patients in the ARTEMIS study (Table II) (103). In a phase III open-label study 689 patients received darunavir/ritonavir or lopinavir-ritonavir tablet formulation. At 48 weeks, 84% of darunavir/ritonavir and 78% of lopinavir/ritonavir patients had viral load <50 copies/mL, demonstrating non-inferiority. Response rates were higher in the darunavir/ritonavir arm for patients with viral load >100,000 copies/mL (103). The 96-week results of the ARTEMIS trial have recently been presented (104). These results confirm both statistical non-inferiority of darunavir/ ritonavir once daily in terms of virologic response, with 79% achieving viral load <50 copies/mL compared to 71% for lopinavir-ritonavir, with an estimated difference of 8.4%; 95% CI 1.9–14.8 (P < 0.001; per-protocol) and superiority (P = 0.012; intent-to-treat) (104).

Tipranavir/ritonavir

Tipranavir is a novel non-peptidic agent with activity against multidrug-resistant isolates (105,106). Tipranavir requires ritonavir-boosting for pharmacokinetic enhancement; dosing studies have identified a standard dose of 500 mg/200 mg twice daily, and absorption is increased with a meal (107).

Clinical studies. Tipranavir has been evaluated among treatment-experienced patients within the RESIST-1 and 2 studies, where tipranavir/ritonavir was compared to investigator-selected comparator ritonavir-boosted PI in conjunction with optimized background (108). Patients were triple-class experienced, with two or more prior PI-based regimens and genotypic evidence of >1 primary protease mutations. In pooled 48-week analysis 1,483 patients were randomized, and 678 remained on assigned treatment. At week 48 those receiving tipranavir were more likely to have viral load <50 copies/mL (22.8% versus 10.2%) than comparator PIs (108). In addition, increased treatment response rates were observed when additional active agents such as enfuvirtide were combined in the background regimen. Overall adverse events including gastrointestinal tolerability, cholesterol, and hepatic toxicity were more frequent in those receiving tipranavir (108). Tipranavir has been linked to intracranial hemorrhage, particularly in patients with predisposing factors or medications. Tipranavir may play a role in decreasing thromboxane B2 formation and subsequent inhibition of platelet aggregation (109).

Protease inhibitors as monotherapy

Clinical trials have been conducted evaluating the use of ritonavir-boosted atazanavir, and lopinavir and darunavir as single active agents, usually among virologically suppressed patients following an induction period with standard combination therapy. Potential theoretical benefits to the use of monotherapy regimens could include decreased pill burden, toxicities, and cost, as well as the ability to preserve drugs and drug classes for future use. However, these benefits may be overwhelmed by virologic breakthrough and development of resistance mutations if PI-only therapy was less effective.

Lopinavir-ritonavir monotherapy has been assessed in a number of randomized trials and observational studies (110–112). A systematic review of lopinavir-ritonavir monotherapy trials found a higher risk of therapy failure on monotherapy (pooled odds ratio 1.48; 95% CI 1.02–2.13; P = 0.037) (113). If patients who successfully resuppressed after reintroduction of NRTI backbone were counted as non-failures, then risk of short-term virological failure was no longer statistically significant (OR 1.05; 95% CI 0.72–1.53; P = 0.81) (113).

Darunavir/ritonavir monotherapy has been evaluated in a randomized clinical trial (MONET study) in which patients with virologic suppression for at least 24 weeks were randomized to darunavir/ ritonavir 800 mg/100 mg daily either as monotherapy or in conjunction with two NRTI agents (114). Treatment failure was defined as two consecutive viral load values >50 copies/mL or treatment switch. A total of 256 patients were randomized, and outcomes at 48 weeks were similar between those on monotherapy compared to standard triple-drug combination (84.3% versus 85.3%), demonstrating non-inferiority. The MONOI-ANRS 136 study evaluated 225 patients randomized to darunavir/ritonavir monotherapy over 96 weeks, with a primary end-point of patients with viral load <400 copies/mL at 48 weeks (115). In the intention-to-treat analysis 92% of patients on monotherapy achieved viral load <400 copies/mL compared to 87.5% (δ -4.9%; 90% CI -11.2 to 2.1). Although the difference was favorable, the use of monotherapy did not meet pre-specified criteria for non-inferiority based on the lower bounds of the confidence interval (115).

Atazanavir/ritonavir monotherapy has been evaluated in a small pilot study involving 34 patients with at least 12 months of virologic suppression on a PI-based regimen (116). Three patients (9%) experienced virologic failure by 24 weeks; two of three patients had low atazanavir trough concentrations. Another pilot study assessing the use of atazanavir monotherapy was terminated early after only 15 patients were enrolled due to evidence of virologic failure in 5 patients (117). Viral rebound occurred between weeks 12 to 16 and was not associated with atazanavir trough concentration.

A primary concern regarding the use of protease inhibitor monotherapy is the potential for loss of control of compartmentalized virus within cerebrospinal fluid (CSF) or other sanctuary sites. This was evaluated in an atazanavir monotherapy trial (ATARITMO) involving 30 patients (9 previously enrolled in an indinavir/ritonavir monotherapy trial) (118). Two patients failed atazanavir/ritonavir monotherapy, although one patient was considered to be in violation of the study protocol has he had had prior failure on indinavir. Evaluation of viral rebound within CSF and seminal fluid identified evidence of detectable viral load (>100 copies/mL) in 3/20 (CSF) and 2/15 (seminal fluid) patients evaluated in this manner, despite suppressed plasma viral load values (118). Further evidence of differential suppression of compartmentalized virus in patients receiving PI monotherapy has been documented in an open-label study of lopinavir-ritonavir monotherapy where 60 patients were randomized to monotherapy or continued combination therapy (119). The study was terminated prematurely due to viral rebound in six patients in the monotherapy arm. All patients had had previous CD4 nadir of <200 cells/mm³. Five patients failing monotherapy underwent CSF examination, and all had detectable CSF viral load (119). The MONOI darunavir monotherapy trial also documented two patients with neurologic symptoms and detectable CSF viral load despite suppressed plasma viral load, and both patients were receiving darunavir/ritonavir monotherapy (115).

At present, given the overall effectiveness and relatively low risk of adverse events with current triple-drug combinations, the use of PI monotherapy is not routinely recommended.

Drug interactions and potential toxicities associated with current protease inhibitors

Boosted PIs have the potential for significant drug interactions due to the potent inhibition of CYP3A4 isoenzyme by ritonavir and differing effects of the other PIs on this metabolic pathway (44). As such, metabolism of other medications may be inhibited, leading to increased drug exposure and associated toxicity. Medications which are also active at the CYP 450 isoenzyme system may lead to bidirectional interactions requiring dose alterations of the boosted PI. Common interactions include interactions with other HIV-related agents such as non-nucleoside reverse transcriptase inhibitors (NNRTI) or the chemokine receptor antagonist maraviroc. Lopinavir-ritonavir experiences significant decrease in exposure when combined with NNRTI agents, and routine dose increase to four soft-gel capsules or three tablets twice daily is recommended (54). Use of certain antiarrhythmics, antihistamines and benzodiazepines are contraindicated. Use of certain other medications such as anti-seizure medications, statins (simvastatin and lovastatin are contraindicated), nitric oxide inhibitors (sildenafil), and even inhaled corticosteroids (which have been associated with syndromes of steroid excess and associated adrenal suppression (120,121)) require close follow-up and dose adjustments. Drug interactions with rifampin pose a significant problem in the management of patients co-infected with tuberculosis, and dose-reduced rifabutin is recommended (122).

Although each boosted PI is associated with adverse events, overall class-wide adverse events must also be considered. Gastrointestinal upset was common with the older PIs, but it is diminished with the use of atazanavir or darunavir and is lower with the use of the tablet formulation of lopinavir-ritonavir compared to the soft-gel capsule formulation (123).

Metabolic abnormalities have been described with PI use. Early PIs were associated with abnormal fat accumulations (lipodystrophy), often with an abdominal or dorsocervical distribution (124,125). Insulin resistance is relatively common with the original PIs and is also seen with lopinavir-ritonavir use (54,126). The use of newer agents such as atazanavir at this stage appears to have relatively limited association with insulin resistance. Similarly, dyslipidemia is relatively common with older PIs, and lopinavir-ritonavir is generally associated with increased triglyceride levels compared to other PIs. PIs have also been associated with increased risk of cardiovascular disease in some observational cohort studies. An early association between PI use and cardiovascular events was reported shortly after the first PIs became available, but recently the large D:A:D cohort (Data collection on Adverse events of anti-HIV Drugs) with 23,437 patients has identified an association between PI use and the risk of myocardial infarction compared to NNRTI use, even after adjustment of other cardiovascular risks and lipid levels (127). Updated analysis has linked the use of indinavir, and lopinavir-ritonavir, to cardiovascular outcomes, although unrecognized confounders may yet play a role in the association (128). No links have yet been identified to newer protease inhibitors, but the follow-up period to date has been comparatively shorter. The mechanisms of this association are unclear but could involve abnormalities in inflammatory cascades or endothelial dysfunction. Saquinavir/ritonavir combinations have been found to be associated with conduction abnormalities in post-marketing surveillance, with the US Food and Drug Administration warning of potential PR or QT interval prolongation leading to increased risk of torsades-de-pointes, and as such base-line and follow-up electrocardiograms are required in all patients initiating these medications (129).

Role of current boosted protease inhibitors—conclusions

The combination of low-dose ritonavir to enhance the pharmacokinetic properties of a primary PI has contributed significantly to the evolution of highly active antiretroviral therapy. Ritonavir-boosted PIs are a key component of antiretroviral therapy in both treatment-naive patients and treatment-experienced patients. Boosted PIs such as atazanavir/ritonavir have overcome the concerns regarding pill burden, multiple daily dosing, and toxicity, yet still retain potent antiviral activity with high genetic barrier to resistance. Newer PIs such as darunavir/ritonavir are active against strains resistant to other PIs and form a valuable component of salvage regimens, making full virologic suppression in patients with drug-resistant virus an attainable goal once again.

Declaration of interest: Dr Julio Montaner has received grants from, served as an ad hoc advisor to Abbott, Argos Therapeutics, Bioject Inc., Boehringer Ingelheim, BMS, Gilead Sciences, GlaxoSmithKline, Hoffmann-La Roche, Janssen-Ortho, Merck Frosst, Panacos, Pfizer, Schering, Serono Inc., TheraTechnologies, Tibotec (J&J), and Trimeris. Dr Montaner's program and research activities are supported by the Ministry of Health Services and the Ministry of Healthy Living and Sport, from the Province of British Columbia; through a Knowledge Translation Award from the Canadian Institutes of Health Research (CIHR); and through an Avant-Garde Award (No. 1DP1DA026182-01) from the National Institute of Drug Abuse, at the US National Institutes of Health.

Dr Mark Hull has received honoraria for speaking engagements and/or consultancy meetings from Merck Frosst, Pfizer, Janssen-Ortho, Gilead Sciences, and Sepracor Pharmaceuticals Inc. He has been a co-investigator on grants supported in part by Wyeth Pharmaceuticals.