HIV Treatment Bulletin

Drug interactions and pharmacokinetic studies: CROI 2007

Professor David Back, Liverpool University

This year’s conference saw the presentation of some new data on drug interactions. The following studies are summarised in this report.

Use of probe drugs to assess drug metabolism, transport and interactions

In this excellent review lecture, Dr Kashuba described the use of ‘cocktail’ studies (i.e., the simultaneous administration of more than one probe compound) to determine the activity of multiple enzymes and transporters. This procedure is now being used in drug development but there is ongoing debate as to the utility of the approach. Dr Kashuba reviewed current probes and cocktails and the potential value for understanding and predicting antiretroviral drug interactions.

Kashuba A. Phenotyping for Drug Interactions: Cocktails Anyone? Oral abstract 50.
The oral overview presentation can be viewed online from the CROI website (see Monday, 4.00pm Novel Approaches for Pharmacokinetic Assessment).

Plasma concentrations of efavirenz and lopinavir in children with and without rifampicin-based antituberculosis treatment

Rifampicin interactions with antiretroviral therapy are complex in adults. There is a sense in which they can be regarded as more complex in children, due to the changing metabolic potential during the developmental process.

The problem is often the limited amount/lack of data. Here the authors studied children (mean age of 7 years) taking EFV or LPV/r with and without rifampicin-based treatment.

Rifampicin did not significantly reduce EFV concentrations, although of concern 50% of children had EFV Cmin below recommended MEC. LPV concentrations were effectively maintained (above MEC) by adding additional ritonavir.

However there is always a trade-off of increased adverse events when ritonavir is increased.

Ren Y et al. Plasma Concentrations of Efavirenz and Lopinavir in Children with and without Rifampicin-based Antit-TB Treatment. Oral abstract 77.
The oral overview presentation can be viewed online from the CROI website (see Monday, 4.00pm Novel Approaches for Pharmacokinetic Assessment).

Significant interaction between antimalarial drugs and efavirenz

This is an important interaction in the context of developing world. Healthy subjects were enrolled to determine if efavirenz impacted the pharmacokinetics of amodiaquine (a known CYP450 substrate).

The study was terminated after the first two subjects developed asymptomatic but significant elevations of liver transaminases. Addition of efavirenz to amodiaquine resulted in AUC increase of 114% (subject 1) and 302% (subject 2).

Although the authors concluded that efavirenz inhibited amodiaquine metabolism the possibility of an interaction with transporter(s) should not be discounted to explain the increase in plasma levels. Amodiaquine is known to form a toxic metabolite and it is logical to consider that efavirenz increased the turnover of this metabolite.

German P, et al. Drug interaction between antimalarial drugs and efavirenz. Poster abstract 577.

Abacavir plasma levels reduced by 32% with lopinavir/r and 17% by atazanavir/r

The study design was complex in HIV+ subjects on antiretroviral therapy:

  1. Group A – Subjects on ABC 600 mg once daily plus 2 NRTIs had 24 PK on day 1. On day 2 they added ATV/r 300/100 mg once daily or LPV/r 400/100 mg twice daily and had repeat PK on day 15.
  2. Group B – Subjects on ATV/r or LPV/r + 2 NRTIs had 24 h PK or 12 h PK on day 1. On day 2 they added ABC 600 mg once daily and on day 15 had repeat PK.

There was no change in ATV, LPV or ritonavir exposure following addition of ABC. However there were small to moderate decreases in ABC exposure following addition of ATV/r (17%) or LPV/r (32%). The mechanism of the latter interaction is unclear at present.

Waters L et al. Abacavir (ABC) plasma pharmacokinetics (PK) in the absence and the presence of atazanavir/r (ATV/r) or lopinavir/r (LPV/r) and vice versa in HIV+ patients. Poster abstract 557.

Efavirenz significantly reduces diltiazem and itraconazole in HIV-negative subjects

Colleagues from Britol-Myers Squibb described interaction data for efavirenz (600 mg once daily) and diltiazem (240 mg once daily) and for efavirenz (600 mg once daily) and itraconazole (200 mg twice daily). Both studies were of crossover design in 25 and 34 subjects (who completed) respectively.

Coadministration of efavirenz and diltiazem resulted in a significant decrease in exposure of diltiazem (60%) and its 2 major metabolites. An increase in efavirenz exposure was modest (16%).

Coadministration of efavirenz and ITR gave a 1-way interaction of a decrease in ITR concentration.

The authors concluded that no change in efavirenz dose was required. diltiazem dose adjustment should be guided by clinical response. Itraconazole dose recommendations are difficult due to lack of data with higher doses of itraconazole.

Kaul S, et al. Pharmacokinetic interaction between efavirenz and diltiazem or itraconazole after multiple-dose administration in adult healthy subjects. Poster abstract 561.

Effects of tipranavir/ritonavir (TPV/r) on the activity of hepatic and intestinal cytochrome P450 3A4/5 and P-glycoprotein (Pgp): implications for drug interactions

This is an example of the approach discussed by Dr Kashuba in Abstract 50. The probe drugs were midazolam (substrate for CYP3A4/5) and digoxin (substrate for Pgp).

The interacting drug was tipranavir/ritonavir (TPV/r; 500/200 mg twice daily). The study was performed in 16 healthy volunteers with each subject studied at baseline, and 10 hours after a first dose and at steady state of TPV/r. To summarise a complex data set:

  • After the first dose – TPV/r moderately inhibited hepatic CYP3A4/5 and intestinal P-pg and potently inhibited intestinal CYP3A4/5 (deduced from data of both oral and i.v probes).
  • Over time, TPV/r induced CYP3A4/5 and P-gp activity. This approach can help with understanding complex drug interactions.

Vourvahis M, et al. Effects of tipranavir/ritonavir (TPV/r) on the activity of hepatic and intestinal cytochrome P450 3A4/5 and P-glycoprotein (Pgp): implications for drug interactions. Poster abstract 563.

Lopinavir/r (LPV/r) increases rosuvastatin AUC and Cmin by 2.1- and 4.7-fold

Rosuvastatin is not a substrate for CYP3A4 and therefore it was postulated that there would be bioequivalence of rosuvastatin when given alone and in combination with LPV/r.

The study was a single arm, 3-phase study in HIV negative healthy volunteers (20 mg rosuvastatin once daily for 7 days, LPV/r 400/100 mg twice daily alone for 10 days, then rosuvastatin + LPV/r for 7 days).

The main findings were that rosuvastatin AUC and Cmin were unexpectedly increased 2.1- and 4.7-fold in the presence of LPV/r. The mechanism of the interaction is not known.

The authors suggest the possible dose separation of LPV/r and rosuvastatin to mitigate the interactions.

Hoody et al. Drug-drug Interaction between Lopinavir/Ritonavir and Rosuvastatin. Poster abstract 564.

Effect of atazanavir and atazanavir/ritonavir on UGTIA4 using lamotrigine

Lamotrigine (an antiepileptic drug) is a phenotypic probe for the conjugating enzyme UGT1A4.

The authors had previously shown that LPV/r induces glucuronidation (reduction of lamotrigine AUC of 55%). Since atazanavir (ATV) inhibits glucuronidation (via UGT1A1), the objective of this study was to examine effects of ATV alone and ATV/r on lamotrigine conjugation.

The study was in 21 healthy volunteers who were given 100 mg lamotrigine alone and then following ATV (400 mg once daily) and ATV/r (300/100 mg once daily). The main finding was that ATV alone had no effect on single dose lamotrigine conjugation; in contrast ATV/r caused a moderate decrease (32%) in exposure indicating induction.

The effect of atazanavir and atazanavir/ritonavir on UGTIA4 using lamotrigine as a phenotypic probe. Poster abstract 556.

Minocycline reduces atazanavir/r AUC by 33% and Cmin by 50%

There is interest in the use of valproic acid (VA) and minocycline (MC) in HIV-associated cognitive impairment, but is there an interaction with PIs?

This study looked at the potential interaction with ATV/r. The study was in 12 HIV-positve subjects stable on ATV/r (300/100 mg once daily). They received minocycline (100 mg twice daily) and then minocycline + valproic acid (250 mg twice daily). PK was performed at baseline and then following MC alone, MC + VA.

MC administration resulted in decreased ATV exposure (AUC 33%; Cmin 50%), and similar decreases were seen after addition of VA. The mechanism of the interaction was not discussed.

DiCenzo R, et al. Effects of minocycline (MC) and valproic acid (VA) coadministration on atazanavir plasma concentrations. Poster abstract 567.

Dose separation required when using 40mg famotidine with atazanavir/ and tenofovir

The interaction of acid reducing agents and PIs has been at the forefront of antiretroviral drug interactions for the past year or so. Here, colleagues from Bristol-Myers Squibb investigated two doses of famotidine on the bioavailability of ATV/r (300/100 mg once daily) given in combination with tenofovir to HIV-negative individuals.

Following 10 days of ATV/r/TDF subjects were randomised to receive famotidine (20 mg twice daily) either with or 2 hours separated from the antiretrovirals OR famotidine (40 mg twice daily), 2 or 12 hours separated ATV/r/TDF.

Decreases in ATV exposure (AUC 21%, Cmin 28%) were seen when famotidine 40 mg was given two hours after ATV/r/TDF.

However, the study suggested it was possible to administer famotidine 20mg with both ATV/r and TDF but that with the 40mg the dose should be separated by 12 hours.

Agarwala S, et al. Effect of famotidine 20 and 40 mg dosing regimens on the bioavailability of atazanavir with ritonavir in combination with tenofovir in healthy subjects. Poster abstract 568.

Interactions between AMD11070 and substrates of CYP3A4 and 2D6 enzymes

Two posters looked at drug interaction with the CXCR4 receptor blocker AMD11070.

AMD11070 a substrate for CYP3AQ4 and has moderate inhibition of CYP2D6. A study by Nyunt and colleagues [1] of the probe drugs midazolam (a CYP3A4 substrate) and dextromethorphan (a CYP2D6 substrate) in the absence and presence of AMD11070 showed increases in the AUC for both probes and Cmax for the CYP2D6 probe as follows:

  • Dextromethorphan AUC increased 265%
  • Midazolam AUC increased 32%.

Since AMD11070 is a substrate of CYP3A4 and P-gp it is likely to be administered with ritonavir. In a study by Coa and colleagues [2] AMD11070 concentrations were increased with concomitant RTV (100 mg twice daily) dosed to steady state in healthy volunteers as follows:

  • The initial dose of RTV increased AUC by 55% (18-103%).
  • RTV at steady state increased AUC by 24% (-1-82%).

The interaction studies of AMD11070 and individual drugs used in HIV patients are clearly important.


  1. Nyunt M, et al. Pharmacokinetic interaction between AMD11070 and substrates of CYP3A4 and 2D6 enzymes in healthy volunteers. Poster abstract 569.
  2. Cao YJ, et al. Ritonavir increases concentrations of the CXCR4 antagonist AMD070 in healthy volunteers. Poster abstract 570.

Drug interactions with ARVs in New York City HIV clinics

This was an important abstract documenting potential drug interactions from New York State Medicaid records over a 6 month period. Drug interactions were screened using Micromedex and Data from 342 HIV+ patients on ART were available. Among 49 patients they identified 60 potentially significant drug interactions with 14 individuals having 1 or more drug interactions that could compromise HIV suppression by causing suboptimal ART levels.

Shah S et al. Identification of drug interactions involving antiretroviral therapy (ART) in New York City HIV speciality clinics. Poster abstract 573.

Significant interaction between rifampin and twice-daily atazanavir

The results of this ACTG trial in 15 healthy volunteers are important to demonstrate the magnitude of the interaction between atazanavir and rifampicin. However note this was unboosted ATV. This was a 3-period sequential study with PK sampling on the last day of each period.

  • Period 1. ATV 300 mg every 12h for 8-11 days.
  • Period 2. ATV 300 mg every 12h + RIF 600 mg every 24h for 11-15 days.
  • Period 3. ATV 400 mg every 12h + RIF 600 mg every 24h for 8-11 days.

Mean ATV AUC was 77% and 53% lower in Periods 2 and 3 respectively, which was not unexpected. The authors suggest the need to study twice daily ATV with twice daily ritonavir in individuals receiving rifampicin.


Ritonavir boosting may only partially overcome this interaction. A paper on this interaction from David Burger’s group showed that with boosted ATV you still get a marked decrease in ATV levels (Antimicrob Agents Chemother. 2006 Oct;50 (10):3336-42 ).

Acosta E et al. Effect of rifampin on pharmacokinetics and safety of twice-daily atazanavir: ACTG Protocol A5213. Poster abstract 575.

Nevirapine: 400 mg versus 600 mg per day in Thai patients with active TB receiving rifampicin

30 HIV-infected Thai subjects with active TB and on rifampicin were randomised to receive nevirapine 400 mg (arm 1) or nevirapine 600 mg (arm 2).

A NVP lead in was performed in both arms of 200 and 400 mg/day. 77% of patients in the 400 mg arm had suboptimal levels (<3.1 mg/L) at week 2 (after lead in period) compared to 13% of patients in the 600 mg arm. However the higher dose was associated with a high rate of hypersensitivity (4 patients).

There was no difference in the 12 week efficacy data between the 2 arms (<50 copies; 53% vs 43% in arms 1 and 2 respectively). The authors concluded that NVP 400 mg/day may be sufficient in Asian (low body weight) patients receiving rifampicin but the 200 mg lead in should be avoided.

Avihingsanon A, et al. Pharmacokinetic and 12 weeks efficacy of nevirapine: 400 mg versus 600 mg per day in HIV infected patients with active tuberculosis receiving rifampicin: a multicenter study. Poster abstract 576.