Interactions between protease inhibitors and acid-reducing agents: a systematic review
Abstract
Objective
The purpose of this article is to provide a systematic review of the available pharmacokinetic and clinical data on drug interactions between protease inhibitors (PIs) and acid-reducing agents, and their clinical relevance.
Methods
A literature search was performed using Medline and EMBASE, abstracts of the previous 2 years of major conferences were searched and the drug information service of the manufacturer of every currently available PI was contacted. All data were summarized, and verified by at least two authors.
Results
A total of 1231 references were identified, 22 of which were studies of pharmacokinetic interactions between PIs and acid-suppressive agents and a further 12 of which provided pharmacokinetic and/or clinical data.
Conclusions
Many pharmacokinetic studies show a lack of a drug interaction between at least one acid-reducing agent and most PIs. Little clinical information is available, except on interactions between atazanavir and acid-reducing agents. This is probably a consequence of the complexity of the interaction.
Introduction
Background and rationale
Protease inhibitors (PIs) constitute an important component of highly active antiretroviral therapy. For some PIs, the concomitant use of acid-reducing medications can result in a clinically significant decrease in their plasma concentrations, which can eventually lead to treatment failure.
As demonstrated by the results of several surveys conducted internationally, the use of acid-reducing agents is widespread among patients infected with HIV [1–3]. This widespread use of acid-reducing agents raises important concerns regarding potential drug interactions with PIs, especially given that some of the potent acid-reducing agents, such as omeprazole in the USA, are available without a prescription. A review on the concomitant use of PIs and acid-reducing agents was recently published [4]; however, the lack of rigour in the search strategy resulted in a number of references that were not identified and not considered [5–16].
The purpose of this article is to provide a systematic review of the available pharmacokinetic (PK) and clinical data on drug interactions between PIs and acid-reducing agents, and their clinical relevance.
Pharmacological considerations
It is well established that proton pump inhibitors (PPIs), histamine-2 receptor antagonists (H2RAs) and antacids all raise gastric pH to different degrees and for different durations [17]. The PPIs are the most potent acid-reducing agents and tend to have a longer lasting effect than H2RAs, whereas the antacids are the least potent acid-reducing drugs and tend to have the shortest duration of action [17,18]. The acid-reducing agents also generally show a dose-related effect [18–20]. The type of acid-reducing drug, the dose and the time of administration of each agent are important to consider when comparing the results of different drug interaction studies.
The majority of the PK studies reviewed in this article were conducted in healthy volunteers, and observations are assumed to be the same as those that would be observed in the HIV-infected population. It is important to keep in mind, however, that differences between healthy volunteers and HIV-infected patients, as well as within the HIV-infected population, may cause differences in drug exposures [21,22].
Another point to consider when interpreting the results is that of polymorphisms of the cytochrome P450 2C19, by which all PPIs are metabolized to varying degrees [23–25]. It has been shown that approximately 3% of Caucasians and 15 to 20% of the Asian population are slow metabolizers of substrates of this cytochrome [23]. In slow metabolizers, the concentrations of omeprazole and lansoprazole have been reported to be at least 5 times higher than in the rest of the population [26]. Polymorphisms were not controlled for in the majority of the PK studies, and may have had an impact on the outcomes. An example may be found in a study in 19 healthy volunteers who received atazanavir/ritonavir 300/100 mg once a day (qd) with or without omeprazole 20 mg qd. Five of the subjects showed a >50% decrease in the area under the curve (AUC) and minimum concentration (Cmin) for atazanavir, whereas the other subjects experienced little to no change in atazanavir exposure. The reason for this observation was not elucidated, but could have been related to genetic polymorphism [12].
The clinical assessment and management of a drug interaction may be supported by the use of therapeutic drug monitoring (TDM) of antiretrovirals, which has been the subject of recent review articles [27,28].
Methods
A literature search was performed using Medline (National Library of Medicine) (1966 to November 2006) for indexed and nonindexed journals, and EMBASE (Excerpta Medica) (1980 to November 2006, limited to humans and English). Keywords used were the commercial names and generic names of PPIs (esomeprazole, lansoprazole, omeprazole, pantoprazole and rabeprazole), H2RAs (cimetidine, famotidine, nizatidine and ranitidine), commonly used antacids (aluminium hydroxide, magnesium hydroxide, Magaldrate, Alka-seltzer, Amphogel, Diovol, Gaviscon, Gelusil, Maalox, milk of magnesia, Mylanta, Riopan, Rolaids, Tums, calcium compounds, antacids and buffered agents), didanosine and PIs (amprenavir, atazanavir, darunavir, fosamprenavir, indinavir, lopinavir, nelfinavir, ritonavir, saquinavir and tipranavir). The bibliography of every article was reviewed for relevant references. We also manually searched the abstracts of the previous 2 years of the following conferences: Interscience Conference on Antimicrobial Agents and Chemotherapy, Conference on Retroviruses and Opportunistic Infections, World AIDS Conference, International AIDS Society Conference on Pathogenesis and Treatment, and the International Workshop on Clinical Pharmacology of HIV Therapy. Lastly, we contacted the drug information service of the manufacturer of every currently available PI.
Publications, abstracts and posters, and information listed in the product monograph or investigator's brochure were eligible for review if they provided information on the interaction between a PI and an acid-reducing drug in the adult population.
Pharmacokinetic studies are summarized in Table 1, and the data were verified by at least two authors. An interpretation of the PK data, as well as a summary and interpretation of the clinical data, was provided for each PI.
| Reference | Design* | Patientpopulation | n | Referenceregimen | Test regimen | AUC† | C max † | C min † | Comments† | |
|---|---|---|---|---|---|---|---|---|---|---|
| Protease inhibitor | Acid-suppressive agent | |||||||||
| Atazanavir | ||||||||||
| Agarwala et al. 2005 [29] | R, PG, OL, multiple doses, 24-h PK | Healthy | 48 | ATV/r 300/100 mg qd | ATV/r 300/100 mg qd | OMP 40 mg qd, 2 h before ATV/r | ↓ 76% | ↓72% | ↓ 78% | RTV Cmin↓by up to 30% (P-value NA) |
| ATV/r 300/100 mg qd+cola | OMP 40 mg qd, 2 h before ATV/r | ↓ 70% | ↓66% | ↓ 73% | ||||||
| ATV/r 400/100 mg qd | OMP 40 mg qd, 2 h before ATV/r | ↓ 61% | ↓56% | ↓ 66% | ||||||
| Luber et al. 2006 [12] | R, CO, OL, multiple doses, 24-h PK | Healthy | 19 | ATV/r 300/100 mg qd AM | ATV/r 300/100 mg qd AM | OMP 20 mg single dose PM | NA | NA | ↔ | |
| ATV/r 300/100 mg qd AM | OMP 20 mg qd PM | ↓ 27% | ↓33% | ↓ 27% | ||||||
| Tomilo et al. 2006 [38] | CO, OL, 24-h PK | Healthy | 9 | ATV 400 mg single dose | ATV 400 mg single dose | LANS 60 mg qd × 2 doses | ↓ 94% | ↓91% | NA | Standard LANS dose is 30 mg qd |
| Agarwala et al. 2005 [37] | R, PG, OL, multiple doses, 24-h PK | Healthy | 60 | ATV 400 mg qd AM | ATV 400 mg qd AM | FAM 40 mg bid | ↓ 41% | ↓47% | ↓ 42% | |
| ATV 400 mg qd AM | ATV 400 mg qd AM+cola | FAM 40 mg bid | ↓ 45% | ↓44% | ↓ 39% | |||||
| ATV 400 mg qd AM | ATV/r 300/100 mg qd AM | FAM 40 mg bid | ↑ 79% | ↔ | ↑346% | |||||
| ATV 400 mg qd PM | ATV 400 mg qd PM | FAM 40 mg bid 2 h after ATV PM dose | ↔ | ↔ | ↔ | ATV Cmin↓21% (NS) | ||||
| Agarwala et al. 2005 [37] | R, PG, OL, multiple doses, 24-h PK | Healthy | 48 | ATV/r 300/100 mg qd | ATV/r 300/100 mg qd | FAM 40 mg bid | ↓ 18% | ↓14% | ↓ 28% | No differences observed in RTV. AUC, Cmax, Cmin |
| ATV/r 300/100 mg qd+cola | FAM 40 mg bid | ↓ 29% | ↓21% | ↓ 36% | ||||||
| ATV/r 400/100 mg qd | FAM 40 mg bid | ↔ | ↔ | ↓ 14% | ||||||
| Agarwala et al. 2005 [5] | R, PG, OL, multiple doses | Healthy | 48 | ATV 400 mg qd | ATV 400 mg qd | OMP40 mg qd, 2 h before ATV | ↓ 94% | ↓96% | ↓ 95% | |
| ATV 400 mg qd + cola | OMP 40 mg qd, 2h before ATV | ↓ 93% | ↓94% | ↓ 95% | ||||||
| ATV/r 300/100 qd | OMP 40 mg qd, 2 h before ATV/r | ↓ 50% | ↓73% | ↑ 23% | ||||||
| Klein et al. 2006 [33] | R, PG, OL, multiple doses, 24-h PK | Healthy | 10 | ATV/r 300/100 mg qd | ATV/r 300/100 mg qd | OMP 40 mg qd, 1.5 h before ATV/r | ↓ 62% | ↓59% | ↓ 61% | RTV AUC↓16%, Cmax↔Cmin↔ |
| 12 | ATV/r 300/100 mg qd | RAN 150 mg, 1.5 h before ATV/r single dose | ↓ 48% | ↓52% | ↓ 43% | RTV AUC↓18%, Cmax↔Cmin↔ATV PK as per mixed linear model for RAN | ||||
| Agarwala et al. 2007 [43] | R, PG, OL, multiple doses | Healthy | 20 | ATV/r 300/100 mg+ TDF 300 mg qd | ATV/r 300/100 mg+TDF 300 mg qd | FAM 20 mg bid | ↓ 10% | ↓9% | ↓ 19% | RTV and TDF exposure not affected by FAM |
| ATV/r 300/100 mg+ TDF 300 mg qd | FAM 20 mg bid, AM dose 2 h after ATV/r/TDF | ↔ | ↔ | ↓ 18% | ||||||
| 20 | ATV/r 300/100 mg+ TDF 300 mg qd | FAM 40 mg qd PM, 12 h after ATV/r/TDF | ↓ 12% | ↓11% | ↓ 23% | |||||
| ATV/r 300/100 mg+ TDF 300 mg qd | FAM 40 mg bid; AM dose 2 h after ATV/r/TDF | ↓ 21% | ↓26% | ↓ 28% | ||||||
| Product monograph [21]; O'Mara et al. 2000 [13] | R | Healthy | 31 | ATV 400 mg | ATV 400 mg single dose | ddI 200 mg single dose | ↓ 87% | ↓89% | ↓ 84% | |
| ATV 400 mg single dose | ddI 200 mg, 1 h before ATV, single dose | ↔ | ↔ | ↔ | ||||||
| Darunavir | ||||||||||
| Sekar et al. 2005 [30]; Investigator's brochure [42] | R, 3-way CO, OL,multiple doses, 12-h PK | Healthy | 18 | DRV/r 400/100 mg | DRV 400/100 mg | RAN 150 mg bid | ↔ | ↔ | ↔ | Systemic exposure to RTV not significantly affected by OMP or RAN |
| DRV/r 400/100 mg bid | OMP 20 mg qd | ↔ | ↔ | ↔ | ||||||
| Fosamprenavir | ||||||||||
| Luber et al. 2006 [12] | R, CO, OL, multiple doses, 24-h PK | Healthy | 19 | FPV/r 1400/200 mg qd AM | FPV/r 1400/200 mg qd AM | OMP 20 mg qd PM | ↔ | ↔ | ↔ | |
| FPV/r 1400/200 mg qd AM | OMP 20 mg qd, single dose | NA | NA | ↔ | ||||||
| Shelton et al. 2006 [31] | R, PG, OL, multiple doses, 12-h PK | Healthy | 25 | FPV 1400 mg bid | FPV 1400 mg bid | ESO 20 mg qd | ↔ | ↔ | ↔ | |
| 23 | FPV/r 700/100 mg bid | FPV/r 700/100 mg bid | ESO 20 mg qd | ↔ | ↔ | ↔ | ||||
| Ford et al. 2005 [32] | R, CO, OL, 24-h PK | Healthy | 26 | FPV 1400 mg single dose | FPV 1400 mg single dose | Maalox TC‡ 30 mL single dose | ↓ 15% | ↓35% | ↔ | C min measured at 12 h |
| FPV 1400 mg single dose | RAN 300 mg 1 h before FPV, single dose | ↓ 26% | ↓51% | ↔ | C min measured at 12 h | |||||
| Indinavir | ||||||||||
| Hugen et al. 1998 [39] | OL, multiple doses, 12-h PK | Healthy | 8 | IDV 800 mg single dose | IDV 800 mg single dose | OMP 40 mg qd | ↔ | NA | NA | P-value NAIDV AUC↓>25% in 4 subjectsIDV AUC↑>25% in 1 subject |
| Rublein et al. 2003 [40] | R, PC, CO, DB, multiple doses, 24-h PK | Healthy | 14 | IDV 800 mg single dose | IDV 800 mg single dose | OMP 40 mg qd | ↓ 47% | NA | ↔ | IDV Cmin↓55% (P>0.05) |
| IDV 800 mg single dose | IDV/r 800/200 mg single dose | OMP 40 mg qd | ↑ 55% | NA | NA | |||||
| Shelton et al. 2001 [16] | R, OL, 5-h PK | HIV | 12 | IDV 800 mg q8h × 2 doses | IDV 800 mg q8h × 2 doses | ddI 400 mg single dose, 1 h before IDV | ↔ | ↔ | NA | |
| Lopinavir | ||||||||||
| Klein et al. 2006 [33] | R, PG, OL, multiple doses, 12-h PK (bid) or 24-h PK (qd) | Healthy | 12 | LPV/r 400/100 mg bid (tablets) | LPV/r 400/100 mg bid | RAN 150 mg, 1.5 h before LPV/r, single dose | ↔ | ↔ | ↔ | LPV PK as per mixed linear modelRTV AUC↔, Cmax↔Cmin↔ |
| 11 | LPV/r 400/100 mg bid | OMP 40 mg qd, 1.5 h before LPV/r | ↔ | ↔ | ↔ | RTV AUC↔, Cmax↔Cmin↔LPV PK as per mixed linear model | ||||
| 11 | LPV/r 800/200 mg qd (tablets) | LPV/r 800/200 mg qd | RAN 150 mg, 1.5 h before LPV/r, single dose | ↔ | ↔ | ↔ | LPV Cmin↓20% (NS); RTV AUC↔, Cmax↔Cmin↔ | |||
| 12 | LPV/r 800/200 mg qd | OMP 40 mg qd, 1.5 h before LPV/r | ↓ 8% | ↓6% | ↔ | RTV AUC↔, Cmax↔Cmin↔ | ||||
| Nelfinavir | ||||||||||
| Fang et al. 2006 [9] | OL, multiple doses, 12-h PK | Healthy | 19 | NFV 1250 mg bid | NFV 1250 mg bid | OMP 40 mg qd | ↓ 36%(NFV)↓ 92%(M8) | ↓37%(NFV)↓89%(M8) | ↓ 39%(NFV)↓75%(M8) | |
| Product monograph [35] | NA | NA | 9 | NFV 750 mg single dose | NFV 750 mg single dose | ddI 200 mg single dose | ↔ | ↔ | NA | |
| Ritonavir | ||||||||||
| Cato et al. 1998 [7] | R, 3-way CO, OL, multiple doses, 24-h PK | HIV | 12 | RTV 600 mg bid | RTV 600 mg bid | ddI 200 mg bid2.5 h before RTV | ↔ | ↔ | ↔ | |
| Saquinavir | ||||||||||
| Boffito et al. 2002 [41] | OL, multiple doses, 24-h PK | Healthy | 12 | SQV sgc 1200 mg tid | SQV sgc 1200 mg bid | CIM 400 mg bid | ↑ 120% | ↑179% | ↔ | C min compared SQV q8h to q12h |
| Kakuda et al. 2006 [10] | R, CO, OL, 24-h PK | Healthy | 12 | SQV hgc 600 mg, single dose+food | SQV hgc 600 mg, single dose | RAN 150 mg × 2 doses+food | ↑ 67% | ↑112% | NA | Statistical significance of Cmax not provided |
| Winston et al. 2006 [34] | OL, multiple doses, 12-h PK | Healthy | 18 | SQV/r 1000/100 mg bid (SQV tablets) | SQV/r 1000/100 mg bid (SQV tablets) | OMP 40 mg qd | ↑ 82% | ↑75% | ↑106% | RTV PK unchanged |
| Tipranavir | ||||||||||
| Van Heeswijk et al. 2004 [36] | OL, 12 hr PK | Healthy | 23 | TPV/r 500/200 mg single dose | TPV/r 500/200 mg single dose | 20 mL Maalox PlusExtra Strength‡ single dose | ↓ 27% | ↓25% | ↓ 29% | |
| Baldwin et al. 1998 [6] | R, CO, OL, 48-h PK | Healthy | 12 | TPV 900 mg single dose | TPV 900 mg single dose | 20 mL Maalox TC‡ single dose | ↔ | ↓41% | NA | Not standard TPV dose |
- * The ‘multiple doses’ design indicates that all agents are given for multiple doses, except where otherwise specified.
- † The minimum concentration (Cmin) was used whenever Ctrough, C12 h, C24 h or C0 h was reported in the studies. The area under the curve (AUC) was used whenever AUC12h, AUC24h or AUC0–8 was reported. Percent changes listed are those with a statistically significant P-value or those with a confidence interval that did not cross 1.
- ‡ ‡ Maalox Plus Extra Strength: per 5 mL: 500 mg aluminum hydroxide, 450 mg magnesium hydroxide and 40 mg simethicone; Maalox TC: per 5 mL: 459 mg aluminum hydroxide and 300 mg magnesium hydroxide.
- ↑ increase;↓ decrease;↔ no significant change.
- AM, morning; ATV, atazanavir; AUC, area under the curve; bid, twice a day; CIM, cimetidine; Cmin, minimum concentration; Cmax, maximum concentration; CO, cross-over; d, day(s); ddI, didanosine; DB, double-blind; DRV, darunavir; ESO, esomeprazole; FAM, famotidine; FPV, fosamprenavir; h, hour; hgc, hard-gel capsules; IDV, indinavir; LANS, lansoprazole; LPV/r, lopinavir/ritonavir; n, number of participants; M8, active metabolite of nelfinavir; NA, not available; NFV, nelfinavir; NS, not statistically significant; OL, open label; OMP, omeprazole; PM, evening; PC, placebo-controlled; PG, parallel group; PK, pharmacokinetics; qd, once a day; R, randomized; RAN, ranitidine; RTV or r, ritonavir; sgc, soft-gel capsules; SQV, saquinavir; tid, three times a day; TPV, tipranavir.
Results and Discussion
A total of 1231 references from Medline and EMBASE were identified, 23 of which were studies of PK interactions between PIs and acid-suppressive agents [5–7,10,12,13,16,29–43].
Atazanavir
The interaction between atazanavir and acid-reducing agents was first recognized in 2000, when the co-administration of a single dose of buffered didanosine was shown to reduce the AUC and Cmax of atazanavir by 87% and 89%, respectively [13]. Several studies to assess the impact of other acid-reducing agents on atazanavir have since been undertaken.
Consistently across trials, PPIs had a significant impact on atazanavir exposure. When atazanavir 400 mg was given with a PPI, the magnitude of the interaction was similar to that of the atazanavir-buffered didanosine interaction [5,38]. Attempts to mitigate this interaction by giving a cola beverage at the time of atazanavir administration were unsuccessful, and increasing atazanavir levels by adding ritonavir 100 mg to atazanavir 300 mg only partially overcame the interaction when compared with atazanavir 400 mg alone [5]. Studies performed with atazanavir/ritonavir 300/100 mg showed a similar pattern of interaction, but of a lesser degree, when using atazanavir/ritonavir 300/100 mg as the regimen of reference [29]. Interestingly, and while recognizing the limitations of comparing the results of different studies, it appears that lower PPI doses and/or a long temporal separation in its administration with atazanavir/ritonavir may have less of an effect on the atazanavir PK parameters than higher PPI doses given 2 h before atazanavir/ritonavir [12,29,33].
Studies with the concomitant use of H2RAs have also been conducted [33,37,43]. High-dose famotidine [40 mg twice a day (bid)] reduced the atazanavir 400 mg PK parameters by 41 to 47%, and, once again, the co-administration of a cola beverage was ineffective in counteracting the effects of the H2RA [37]. Another study looked at different temporal separation strategies for famotidine, using atazanavir/ritonavir given with tenofovir as a reference. The concomitant administration of famotidine with atazanavir/ritonavir and tenofovir, or a temporal separation by 12 h (with famotidine 20 mg bid) or by 2 h (with famotidine 40 mg qd) resulted in a similar reduction of the atazanavir PK parameters. However, in the arm using higher famotidine doses (40 mg bid) the atazanavir PK parameters tended to be reduced to a greater extent [43]. Two strategies, however, were successful in correcting for the interaction with atazanavir/ritonavir without tenofovir: the addition of ritonavir 100 mg to atazanavir 300 mg, and a temporal separation in the administration of atazanavir and famotidine [37]. Klein and colleagues studied the effect of a single dose of another H2RA, ranitidine 150 mg, on the atazanavir/ritonavir 300/100 mg levels [33]. This method was selected in order to mimic the ‘as needed’ use of ranitidine. They found that the reduction in the atazanavir exposure was more pronounced than that observed in the previous famotidine study [37], and was close to the reduction seen with a PPI administration [33]. The reason for this finding is unclear and it cannot be explained by a decreased ritonavir exposure.
Several observational studies have been conducted in HIV-infected patients to evaluate the pharmacokinetics and clinical relevance of the interaction. In the first and largest cohort (n=100), 12 patients were identified to have used drugs for dyspepsia. Patients were using atazanavir/ritonavir (n=81) or atazanavir (n=19) in a wide range of dosages. In this cross-sectional study, the use of drugs for dyspepsia was not associated with a single atazanavir trough concentration. Given the small number of patients and the design of the study, these results have to be interpreted with caution [44]. Still from a PK standpoint, two other studies found no difference in the atazanavir Cmin in cohorts of patients taking atazanavir/ritonavir 300/100 mg when patients were stratified either for the use of a PPI [14] or for the use of a PPI or tenofovir [45]. The reference regimen for each cohort was atazanavir/ritonavir without a PPI, an H2RA or tenofovir. From a virological outcome standpoint, one case series showed that 12 out of 13 patients taking atazanavir with or without ritonavir, with an H2RA or a PPI, achieved virological suppression or had a decrease in their viral load [46]. Similar findings were published in a case report [8], and three historical cohort studies that found adequate viral load control in ≥75% of patients [15,47,48]. From both a PK and a virological outcome standpoint, a study of a cohort of 34 patients taking atazanavir with and without ritonavir showed that four of the six patients taking a PPI and two of the four patients taking an H2RA, and who had a low atazanavir level, had a viral load of <400 HIV-1 RNA copies/mL [49]. Two case reports also provided PK and virological information [11,50]. The first described a patient experiencing treatment failure while taking atazanavir/ritonavir and esomeprazole concomitantly [11]. Despite the dose of atazanavir/ritonavir being increased from once to twice daily for 2 weeks, the viral load continued to increase and an adequate atazanavir Cmin was not achieved [11]. The second case report showed that, despite the concomitant administration of atazanavir/ritonavir and lansoprazole, the atazanavir plasma concentrations remained consistent with historical values and the patient's viral load was undetectable [50].
The manufacturer of atazanavir recommends that the co-administration of PPIs and atazanavir be avoided. While it appears that, with the co-administration of acid-reducing therapies, a significant reduction of atazanavir levels can be expected, the clinical impact of this reduction has not been clearly demonstrated. The use of therapeutic drug monitoring may help facilitate the co-administration of atazanavir and acid-reducing agents.
Darunavir
Darunavir is the latest PI to receive approval from the Food and Drug Administration in the USA, in June 2006, and from Health Canada, a month later. Only one study has been conducted with darunavir and acid-reducing agents [30]. The authors showed that the concomitant administration of darunavir/ritonavir and omeprazole or ranitidine had no significant impact on the AUC, Cmax and Cmin of darunavir. Although the dose of darunavir used in this study was lower than the currently licensed dose (400 mg bid instead of 600 mg bid, each combined with ritonavir), these results can probably be extrapolated to the higher dose of darunavir [42].
Fosamprenavir
Three independent studies have evaluated the PK properties of amprenavir following the administration of its prodrug, fosamprenavir [12,31,32]. In the first two studies [12,31], neither esomeprazole 20 mg qd nor omeprazole 20 mg qd had any significant effect on the pharmacokinetics of fosamprenavir or fosamprenavir/ritonavir. A case report showing a lack of PK and clinical interaction in an HIV-infected patient taking fosamprenavir/ritonavir with esomeprazole supports these findings [11]. In the third study [32], a single dose of both an antacid (Maalox®) and an H2RA (ranitidine) caused a modest decrease in the AUC of fosamprenavir, but had no effect on its Cmin. The reason for these findings compared with those of the PPI studies is not clear. The differences may be attributable to different study designs (e.g. single vs. multiple dose studies or studies with different timings of sampling) or suggest that this interaction may not be the result of a suppression of gastric acid. Further studies are needed to determine if antacids or H2RAs with fosamprenavir at steady state interact in the HIV-infected population. Given the body of evidence, it is likely that no dose modifications of fosamprenavir are necessary when gastric pH is raised.
Indinavir
The first observation of an interaction between indinavir and omeprazole came from a retrospective database analysis [51]. Of nine HIV-infected patients who took indinavir 800 mg three times a day (tid) together with omeprazole, four had indinavir plasma concentrations that were lower than those derived from population PK studies. In two of these four patients, the indinavir dose was increased to 1000 mg tid while omeprazole was continued, and the plasma concentrations increased into the expected range [51].
This study provided a basis from which to explore further this potential interaction. Hugen and colleagues found that a decrease in AUC greater than 25% occurred in four of eight subjects, and an increase in AUC of more than 25% occurred in one patient who took indinavir with omeprazole [39]. These changes in AUC suggest a high interpatient variability in the pharmacokinetics of indinavir. In another study, omeprazole decreased the AUC of indinavir by 47% (P=0.001), and the Cmin by 55% (not statistically significant) [40]. The lack of statistical significance in the latter parameter is probably a result of insufficient power to show a difference, and, once again, suggests that there is a large interpatient variability with indinavir. Interestingly, when indinavir was combined with ritonavir and given with omeprazole, the AUC of indinavir increased by 55% (P=0.02), compared with indinavir 800 mg alone [40].
Indinavir was not studied with H2RAs, but its potential interaction with buffered didanosine given 1 h before indinavir was investigated in HIV-infected patients [16]. No change in the indinavir exposure was found. When indinavir and didanosine were given together, however, the indinavir AUC decreased significantly by 84% [52]. Indinavir and didanosine should therefore be administered at different times [52,53].
Although no clinical data were reported for any of the studies, it appears that indinavir without ritonavir should not be co-administered with acid-reducing agents, and, in the case of buffered didanosine, may be given in a staggered manner. Indinavir co-administered with ritonavir, which is more commonly used than indinavir without ritonavir, can probably be given with PPIs.
Lopinavir/ritonavir
The pharmacokinetics of lopinavir when lopinavir/ritonavir tablets, the latest formulation of the drug, are co-administered with acid-reducing therapy has been investigated in one study [33]. In this study, lopinavir/ritonavir tablets were administered either once daily or twice daily, and in combination with ranitidine 150 mg qd or omeprazole 40 mg qd. In all cases, acid-reducing therapy had no significant effect on the lopinavir pharmacokinetics. Given the lack of drug interaction with omeprazole, it can be expected that even higher doses or multiple doses of ranitidine are unlikely to significantly affect lopinavir bioavailability. The only data available on the soft-gel formulation suggest that, over 48 h, there may be an increase in lopinavir Cmin when it is taken once daily and with acid-reducing agents, but this was not observed when it was taken twice daily [54]. The type and dose of acid-reducing agents in this study were not specified [54]. Lopinavir/ritonavir tablets and PPIs or H2RAs can probably be co-administered without compromising PI efficacy.
Nelfinavir
The acid-reducing agents that were evaluated with nelfinavir are omeprazole and buffered didanosine. Omeprazole was shown to decrease nelfinavir and M8 (the active metabolite of nelfinavir) PK parameters by 36 to 39% and by 75 to 92%, respectively [9]. As for buffered didanosine, the product monographs of nelfinavir [35] and didanosine [52] refer to a small single-dose study that evaluated its effect on the nelfinavir pharmacokinetics. It was found that the nelfinavir AUC and Cmax were not affected by didanosine when nelfinavir was administered with a light meal 1 h after, or more than 2 h before, didanosine. Because of this temporal separation, the antacid effect of buffered didanosine was probably minimal or absent. The reason didanosine and nelfinavir were not given at the same time is that didanosine needs to be administered on an empty stomach, while nelfinavir needs to be taken with food.
These data suggest that nelfinavir and omeprazole should not be co-administered. There are no data on the co-administration of nelfinavir and other acid-reducing agents.
Ritonavir
Although ritonavir is now mostly used to boost the pharmacokinetics of other PIs, one study assessed its concomitant use at the therapeutic dosage of 600 mg bid with buffered didanosine [7]. The results indicated that there was no effect on the pharmacokinetics of ritonavir when it was administered 2.5 h after didanosine, a time at which the antacid effect was probably minimal, if not absent. Data on the interaction with other acid-reducing agents are, however, not available.
The effect of acid-reducing agents on the pharmacokinetics of low-dose ritonavir has been reported in various studies of atazanavir/ritonavir with PPIs and H2RAs [29,33,37,43], darunavir/ritonavir with a PPI and an H2RA [30], lopinavir/ritonavir with a PPI and an H2RA [33], and saquinavir/ritonavir with a PPI [34]. In all cases except one [29], there were no changes in the ritonavir pharmacokinetics or the changes were small. In this one study, ritonavir C24h (concentration at 24 h after drug intake) decreased by up to 30% when it was co-administered with atazanavir/ritonavir and cola, but the statistical significance of this change was not reported, and this finding occurred in only one of the four groups tested [29]. Overall in these studies, the lack of changes in the ritonavir PK parameters suggests that ritonavir is unlikely to be a contributing factor to the changes observed in the PK parameters of some of the co-administered PIs.
Saquinavir
Three studies have been published for saquinavir [10,34,41]. One study investigated the possible interaction between saquinavir soft-gel capsules and the H2RA cimetidine [41]. In addition to having an acid-reducing effect, cimetidine is known for its inhibitory effect on cytochrome P450 metabolism, by which saquinavir is metabolized [26]. The results showed that the exposure to saquinavir 1200 mg bid combined with cimetidine was higher than that of saquinavir 1200 mg tid without cimetidine. Despite an increase in the saquinavir exposure, there was no difference in adverse events between the treatment arms [41]. The relative contribution of the cytochrome P450 3A4 inhibitory effect of cimetidine and of the acid-reducing effect of cimetidine on the saquinavir concentrations is unknown. Another study assessed the concomitant use of saquinavir hard-gel capsules with another H2RA, ranitidine, and the results also showed an increase in the exposure of saquinavir, which, interestingly, was apparently not attributable to an increase in gastric pH [10]. The third study evaluated the interaction between saquinavir tablets/ritonavir 1000/100 mg bid and omeprazole 40 mg qd [34]. Although the saquinavir exposure increased, there was no increase in clinically significant adverse events. The mechanism of this interaction is not clear, but it seems that the interaction may have resulted from an increased absorption rather than a decreased elimination, as suggested by a lack of an increase in half-life and time to reach maximum concentration [34].
Overall, the co-administration of cimetidine, omeprazole or ranitidine with different formulations of saquinavir results in an increase in saquinavir exposure. From the three published studies, there is no evidence that the increased exposure to saquinavir resulted in more clinically significant toxicity; however, therapeutic drug monitoring and dose adjustment of saquinavir may be an option if toxicity occurs in clinical practice.
Tipranavir
The interaction between tipranavir and acid-reducing agents has only been evaluated with the antacid Maalox® in two studies [6,36], and only one of these studies used the currently recommended dose of tipranavir/ritonavir of 500/200 mg bid [36]. When combined with Maalox®, the tipranavir AUC, Cmax and Cmin decreased by 25 to 29%, and so spacing out the time of administration of these drugs may be warranted. The mechanism of the interaction is unknown, but if it occurred via an acid-reducing effect, it can be assumed that the extent of the interaction will be of greater significance when tipranavir is given with an H2RA or a PPI. The concomitant use of tipranavir and acid-reducing agents should therefore probably be avoided until these interactions have been studied. Should these combinations still be used in clinical practice, therapeutic drug monitoring of tipranavir may prove to be of value.
Conclusions
The PIs for which there is clear evidence of a lack of a significant unfavourable PK drug interaction with PPIs are darunavir/ritonavir, fosamprenavir, fosamprenavir/ritonavir, lopinavir/ritonavir and saquinavir/ritonavir. There also appears to be no significant unfavourable PK interaction between H2RAs and atazanavir/ritonavir, darunavir/ritonavir, lopinavir/ritonavir, saquinavir and atazanavir if given in a specific staggered manner as per the manufacturer's recommendations. Lastly, the use of antacids or buffered didanosine does not appear to alter the pharmacokinetics of atazanavir, indinavir, nelfinavir or ritonavir at therapeutic doses, if they are taken at different times.
Except for atazanavir, very little or no clinical information is available for interactions between PIs and acid-reducing agents. Although significant PK interactions between atazanavir and acid-reducing agents have been observed in healthy volunteers, the expected negative clinical consequences have not been clearly observed in clinical practice.
Although convincing data may be lacking to support the administration of some PI/acid-reducing agent combinations, the current body of data does not preclude their safe and effective use in clinical practice.
Acknowledgements
An unrestricted educational grant was provided by Bristol-Myers Squibb Canada. Bristol-Myers Squibb provided no assistance in preparing the manuscript and did not have any editorial input.
