The scale-up of antiretroviral treatment (ART) in resource-poor countries has been impressive, both in its scope and in its impact. In 2003, ~400,000 people in resource-limited settings were receiving antiretroviral drugs (ARVs); by 2006, this number had increased to >2 million people [1]. Treatment of common infections occurring with HIV is also increasing, as most HIV-infected persons live in areas where infectious diseases such as tuberculosis (TB) and malaria are prevalent. The use of over-the-counter and natural health products is also pervasive [2]. Therefore, the management of HIV infection in these settings will require multiple concurrent medications, with a potential for drug interactions and overlapping toxicities. Although in recent years there has been a significant expansion in access to ARVs in resource-poor countries, options are limited in most areas. In 2006, 95% of individuals receiving first-line ARVs in resource-poor settings were being treated with 1 of 4 fixed-dose regimens: stavudine (d4T) + lamivudine (3TC) + nevirapine (NVP); zidovudine (AZT) + 3TC + NVP; d4T + 3TC + efavirenz (EFV); or AZT + 3TC + EFV [1]. Currently, the World Health Organization (WHO) recommends a first-line regimen that includes a thiacytadine analogue (either 3TC or emtricitabine), a companion nucleoside reverse-transcriptase inhibitor (NRTI) (either AZT, tenofovir [TDF], abacavir [ABC], or d4T), and a non-NRTI (NNRTI) (either EFV or NVP) [3]. Second-line regimens include 2 previously unused NRTIs (didanosine, TDF, ABC, 3TC, and/or AZT) coupled with either a ritonavir (RTV)-boosted protease inhibitor (PI) (either lopinavir, saquinavir, indinavir, atazanavir, or fosamprenavir) or unboosted nelfinavir. Available antimicrobial drugs such as itraconazole, ketoconazole, clarithromycin, and rifampin have serious interactions or overlapping toxicities with ARVs. Cost, as well as political and infrastructural constraints, limits access to drugs—such as fluconazole, azithromycin, and rifabutin—that have fewer interactions [4]. Thus, even when drug interactions are well characterized, clinicians caring for HIV-infected patients have few options. Now that HIV treatment is available in many resource-limited countries, clinicians must understand the basic principles of drug metabolism. In this review, we describe new insights into mechanisms of CYP450 induction and inhibition. In addition, we provide a comprehensive review of the major interactions between ARVs recommended by WHOas first- or second-line therapy for HIV in resource-limited settings and drugs used to treat TB, malaria, and fungal, bacterial, and parasitic diseases. The CYP450 family of heme-containing monooxygenases are responsible for the oxidative metabolism of endogenous and exogenous compounds. Drugs that induce or inhibit CYP450 enzymes may decrease or increase, respectively, concentrations of concurrently administered drugs that are CYP450 substrates. Changes in drug concentrations resulting from drug interactions can lead to treatment failure or toxicities. New insights into CYP450 induction: pregnane X receptor (PXR)/retinoic acid receptor (RXR) and xenobiotic response element (XRE) as parallels to the pathways of the constitutive androstane receptor and antioxidant response element. Until the late 1990s, it was known only that structurally different molecules could induce the same metabolic enzymes and, further, that a single compound could lead to the induction of multiple metabolic pathways as well as drug transporters. However, the molecular mechanisms underlying these findings were unclear. In 1998, a novel orphan nuclear receptor involved in steroid hormone-signaling pathways was discovered [5]. This nuclear receptor, PXR, was subsequently shown to be activated by compounds that induce CYP3A4 and to bind to CYP3A4's promoter response element [6]. PXR is a member of the superfamily of nuclear receptors that act as transcription factors and regulate the expression of down-stream genes. In contrast to other members of the nuclear receptor- gene super-family, PXR binds ligands promiscuously, with low affinity and low specificity [7]. After a ligand has bound to PXR, it forms a heterodimer with the 9-cis RXR and translocates to the nucleus (figure 1). There, it recruits a coactivator and binds to response elements in the promoter or enhancer regions of target genes, leading to increased transcription of those genes [8]. The PXR/RXR heterodimer, for example, can bind to 2 areas in CYP3A4's regulatory region: a proximal promoter and the XRE [9]. PXR targets a plethora of genes, including those for CYP3A4, other CYP450 enzymes, phase II drug-metabolizing enzymes, UDP-glucurosyltransferases, sulfotransferases, and the drug transporters P-glycoprotein (P-gp) and multidrug-resistance pump 1. This targeting helps explain why rifampin, a drug with a high affinity for PXR's ligand-binding domain, can induce multiple metabolic pathways, leading to interactions between the drug and numerous compounds. The constitutive androstane receptor and the antioxidant response element are parallel regulatory pathways [10, 11]. The role that pregnane X receptor (PXR) plays in human metabolism. A ligand such as rifampin (RIF) binds PXR, then forms a heterodimer with the 9-cis retinoic acid receptor (RXR). The complex then migrates to the nucleus, recruits a coactivator, and subsequently can bind to multiple possible target-gene response elements, including the xenobiotic-response element (XRE) in CYP3A4's regulatory region. The metabolism of commonly used ARVs. NRTIs serve as the backbone of ART regimens. AZT and ABC, which are commonly used NRTIs, are metabolized by hepatic 5′-glucuronidation and are phosphorylated intracellularly to their active triphosphate form. These NRTIs are substrates of phase II metabolizing enzymes, enzymes involved in conjugation reactions that do not involve CYP450; and, because the other drugs in the NRTI class largely bypass hepatic metabolism, this class of drugs may be less prone to drug interactions. NNRTIs are commonly paired with NRTIs for the treatment of HIV, particularly in resource-limited settings. EFV is largely cleared by CYP2B6 and, to a lesser extent, by CYP3A4. EFV is a moderate inducer of CYP3A4; however, it can act as an inhibitor, leading to increased concentrations of some drugs, including RTV and nelfinavir [12]. NVP is eliminated via CYP3A4 and CYP2B6 and induces CYP3A4. PIs are available as a second-line treatment in some resource-limited settings. RTV is an exceptionally potent inhibitor of CYP3A4 and frequently is used with other PIs to pharmacologically boost their concentrations. Less commonly, RTV can serve as an inducer of CYP3A4 and can cause decreased exposure to drugs metabolized by glucuronosyl S-transferase, for example. RTV also acts as a mixed competitive/noncompetitive inhibitor of CYP2D6 [13]. Of the PIs recommended by WHO, SQV, LPV, IDV, and ATV are metabolized by CYP3A4 and act as weak inhibitors of CYP3A4. NFV is the only PI that is metabolized by both CYP2C19 and CYP3A4 [14, 15], and it is a moderate inhibitor of CYP3A4. Entry and integrase inhibitors are unavailable or are cost prohibitive in most resource-limited settings and are not included in this review. Currently, rifamycins serve as the corner-stone of TB therapy, largely because of their unique ability to kill persistent intracellular mycobacteria. Concurrent treatment of TB and HIV is nettlesome owing to interactions between antituberculosis drugs and ARVs, shared drug toxicities, and immune-reconstitution syndrome [15]. Rifampin, the most commonly used rifamycin, is a potent inducer of cytochrome P450 enzyme activity, as well as of P-gp and phase 2 enzyme activity [16]. Because most PIs and NNRTIs are metabolized by CYP3A4, induction of CYP3A4 by rifampin can lead to reduced serum concentrations of ARVs, with the consequent risk of treatment failure or emergence of resistance to ARVs. As an inducer, rifabutin is 40% as potent as rifampin and can be used safely for TB treatment in patients receiving ARVs [15]. However, it is currently unavailabl in most resource-limited settings. Rifabutin's active metabolite is in part metabolized by CYP3A4 isoenzymes, so dose adjustment is necessary when it is given with NNRTIs or PIs, except for NVP. Rifapentine, another rifamycin, is under investigation for TB treatment; however, the FDA-approved dose leads to high relapse rates, and the optimal dose has not been determined [17]. With regard to concomitant use of NRTIs with rifampin, early studies showed diminished serum concentrations of AZT when it was given with rifampin (table 1) [18, 19]. However, AZT concentrations in serum are not predictive of intracellular concentrations of the active anabolite, AZT-triphosphate [41]. Overlapping toxicities are more of a concern for patients taking NRTIs and antituberculous medications; for example, the use of either d4T or didanosine with isoniazid should be avoided because of the increased risk of peripheral neuropathy [42]. Selected clinical studies of interactions between rifampin (RIF) and World Health Organization—recommended first- or second-line antiretroviral drugs (ARVs). NNRTIs are part of all WHO-recommended first-line HIV regimens. There has been considerable debate regarding the use of NNRTIs with rifampin, a debate that has been focused largely on the dose of EFV. In 2002, a group in Spain showed that the pharmacokinetics of EFV at a dose of 600 mg daily were similar to those of EFV at a dose of 800 mg daily when coadministered with rifampin daily in HIV/TB-coinfected patients [21]. Partly on the basis of these pharmacokinetic data, the Centers for Disease Control issued recommendations that the EFV dose be increased to 800 mg daily when given with rifampin. However, in resource-limited settings, dose adjustment may be impractical and expensive, because most ARVs are available only as fixed-dose formulations. Pharmacodynamic studies evaluating a dose of 600 mg of EFV in HIV/TB-coinfected patients in Brazil, South Africa, Thailand, and India who were receiving rifampin have shown durable virologic and immunologic responses [22–26, 43]. Furthermore, the 800-mg dose may cause CNS toxicity [44]. These findings suggest that a daily dose of 600 mg of EFV may suffice in patients who are concurrently being treated with a rifampin-based regimen for HIV and TB. For white patients >60 kg, a dose of 800 mg is well tolerated, with favorable pharmacokinetic parameters, and can be considered [27]. NVP, the other commonly used NNRTI, is widely available in fixed-dose combinations, but it causes more liver toxicity than does EFV. When the standard dose of 200 mg twice daily is coadministered with rifampin, serum concentrations of NVP are decreased by 20%-30% [28, 29]. However, HIV/TB-coinfected patients receiving a highly active ART regimen containing NVP at 200 mg twice daily coupled with a rifampin-containing TB regimen have demon-strated appropriate viral suppression, TB cure rates, and immunologic recovery [30–32]. At this point, there are insufficient data to conclude definitively that rifampin and NVP can be given together without dose adjustment. We hope that further studies will substantiate the safety of this approach, especially in light of the dangers of administering EFV to women of childbearing age. Although rifampin's induction of metabolism makes coadministration with NNRTIs complicated, it makes combined treatment with most PIs impossible. Not only are PIs substrate of CYP3A4, which is induced by rifampin, but rifampin's upregulation of P-gp may lead to diminished PI absorption. Several studies of healthy volunteers or HIV-infected patients have shown markedly diminished area-under-the-time-concentration curve (AUC) and trough concentration (Cmin) for PIs at standard doses coadministered with rifampin, even when boosted with RTV (table 1). Increased doses of PI or boosting with increased doses of RTV fails to produce adequate PI concentrations and sometimes leads to unacceptable toxicity [34, 37–40, 45]. In a small study, high-dose SQV + RTV coadministered with rifampin was well tolerated, produced adequate concentrations, and was recommended for use [46]. However, in a study of 20 HIV/TB-coinfected patients who started this regimen after 30 days of rifampin-containing TB treatment, 15 dropped out because of adverse events [39]. Furthermore, in a study of 14 healthy volunteers who first were given 600 mg of rifampin daily for 14 days and then also were given 1000/100 mg of SQV/r twice daily, 9 had grade 3 or 4 hepatotoxicities; the trial was stopped early [38]. Currently, there are no HIV/TB treatment regimens containing a PI and rifampin that are known to be safe and effective. For patients in whom NNRTI- or PI-based therapy cannot be used, triple-or quadruple-nucleoside therapy may be considered, although there are concerns about the potency of triple-nucleoside regimens [47]. In a study in Uganda and Zimbabwe, 61% of patients receiving AZT/3TC/TDF had a viral load of <50 copies of HIV RNA/mL of blood at 48 weeks [48]. TDF-containing quadruple-nucleoside regimens such as TDF/AZT/3TC/ABC may perform even better [49]. It is important to note that, in HIV/TB-coinfected patients, rifampin may induce the metabolism of other antimicrobial drugs as well. The interactions between rifampin and other antimicrobial drugs have been summarized by Niemi et al. [16]. Azole antifungal drugs act as competitive inhibitors of CYP3A4 isoenzymes. They vary in their CYP-inhibitory capacity, with ketoconazole being a much more potent inhibitor than itraconazole, followed by fluconazole [50]. Ketoconazole also is a moderate to strong inhibitor of CYP2C9, 2C19, and 2D6, whereas fluconazole is a strong inhibitor of 2C9 and 2C19. The metabolism of azole antifungal drugs varies significantly as well. Itraconazole and ketoconazole are metabolized extensively by CYP3A4, whereas fluconazole is excreted largely unchanged [51]. Significant interactions between azoles and ARVs can occur via the CYP450 pathway (table 2). Dose recommendations for concomitant use of World Health Organization—recommended first- and second-line antiretroviral drugs (ARVs) and fluconazole, ketoconazole, or itraconazole. Treatment of cryptococcosis, a common complication of AIDS, requires initial treatment with amphotericin or high-dose fluconazole, followed by long-term fluconazole therapy. Although prior to 2001 fluconazole was scarcely available in resource-poor countries that recognize drug patents, a public-private donation program has led to broader access to this drug in sub-Saharan Africa [69]. Also, manufacturers in Thailand, India, Colombia, and Spain make less-expensive generic formulations of the drug [70]. EFV and fluconazole can be used together without dose adjustment [35]. However, NVP concentrations were increased 2-fold in a single-arm study of 24 HIV-infected subjects concomitantly taking fluconazole, and 25% of the subjects developed serious hepatotoxicity during the coadministration of NVP and fluconazole [71]. Despite that finding, 2 studies of patients with advanced AIDS have found that increased concentrations of NVP in patients receiving fluconazole for treatment or prophylaxis of cryptococcal meningitis has not led to increased NVP-associated toxicity [58, 72]. Therefore, pending further studies, monitoring for NVP-associated liver toxicity may be prudent. Coadministration of fluconazole and a PI may lead to higher PI concentrations; the clinical relevance of this interaction is under investigation [67]. Other azole antifungal drugs are commonly used for the treatment of systemic mycoses. For example, itraconazole is used for lifelong prophylaxis in patients with HIV who have a history of histoplas-mosis. EFV can decrease itraconazole concentrations in serum and cause treatment failure [73]. Similarly, when NVP is coadministered with 200 mg of itraconazole daily, the latter's AUC is decreased by 61% [59]. No trials of higher-dose itraconazole coadministered with NNRTIs have been performed, so we can provide no recommendations for dose adjustments; substitution of another antifungal, if available, is advised. PIs, on the other hand, may increase itraconazole concentrations, prolong its half-life, and lead to its accumulation (table 2) [65]. Ketoconazole has significant interactions with NNRTIs: one study has reported a 72% reduction in the AUC of ketoconazole when the latter is coadministered with EFV [60], and another study has found a bidirectional interaction between ketoconazole and NVP, consisting of a 72% decrease in ketoconazole concentration and a 15%-30% increase in NVP concentration [74]. Until further studies are undertaken, it will remain unclear whether increasing doses of ketoconazole may overcome enzyme induction by NNRTIs; therefore coadministration is not advised. Ketoconazole increases PI concentrations by and PIs can lead to increased concentrations as well (table 2). for example, RTV increases AUC [50]. Ketoconazole has been as an pharmacokinetic for SQV in a recent study of HIV-infected patients, was found to be to RTV and are substrates of CYP3A4, whereas is a substrate and, for the most is excreted in Also, whereas and inhibit CYP3A4, does so only interactions are much less common than those other In HIV-infected patients, and are widely used for the treatment and of When is used concomitantly with ARVs, there are no significant drug except when it is coadministered with in which the has been found to in healthy a decrease in NFV AUC and a increase in AUC Although in some settings, it is available for treatment of and of infections and may more available in the however, has significant interactions with NNRTIs and with When is coadministered with its AUC is decreased by and is common [35]. NVP induces the of clarithromycin, the latter's whereas increases NVP concentrations by In where is available, it is the treatment; when it is not available, patients being treated with a of and NNRTIs should be interactions between and PIs can lead to increased serum concentrations of both drugs, with the increase in concentration being of coadministration of unboosted ATV and clarithromycin, for example, in of the and the dose of must be decreased by that SQV can be coadministered safely with clarithromycin, without dose adjustment [35]. For patients with the dose of must be decreased if it is coadministered with regimens containing As in the of substitution of is if it is has an interaction similar to that of because it is used for there are that can be for and are highly in resource-poor countries, together the of million people in [2]. Although their their and is common In with malaria, with may lead to an increase in a higher of malaria and and an increase in treatment failure Treatment of HIV with ARVs and of standard prophylaxis with can decrease malaria has activity whereas ART can decrease the either via the activity of PIs, or by boosting the A of between HIV and malaria the potential for drug interactions. Of concern are interactions between NNRTIs or PIs and drugs that are metabolized by P450 the data regarding these potential interactions are (table or interactions between drugs and World Health Organization—recommended second-line antiretroviral drugs (ARVs). The class of drugs includes the drug of for treatment of for treatment of for treatment of malaria, and an for malaria prophylaxis and drugs are metabolized by CYP3A4 and cause significant toxicities. In a study of healthy pharmacokinetics were only by CYP3A4 and coadministration led to a decrease in at For 2 HIV-infected patients taking a NFV + ART prophylaxis not PI concentrations, and accumulation of was not studies evaluating the potential interactions between NNRTIs or PIs and are are the most is by CYP3A4, to its more active It is commonly with either or to the emergence of and are metabolized by CYP3A4 is commonly with a active is largely metabolized by with some from CYP3A4, whereas is metabolized by In a recent study, coadministration of EFV with a of led to increases in and the study was stopped after the first 2 developed significant of liver Other such as SQV and LPV, may increase concentrations has a and can cause a so its use with PIs is It that should be when is used with either PIs or NNRTIs In addition, drugs may induce CYP3A4 and enzymes Although a of an may or may not significantly concentrations, increased concentrations of the drugs with which are may in serious toxicities. In a study of however, the of to increased AUC by with no significant toxicities with is used for prophylaxis and treatment of Although is expensive, in some countries it is available donation for malaria Although one study found that increases concentrations by concentrations of at doses are to be high to produce significant of AZT PI combinations, such as may induce of leading to diminished concentrations, but the clinical of this reduction is [35]. concentrations are reduced by in healthy volunteers taking unboosted with but the of this is because unboosted is not widely used data on interactions between ARVs and drugs are currently of interactions between ARVs and other are The metabolic pathways of many of these have not been so regarding potential drug interactions must be from limited pharmacokinetic and toxicity data (table interactions between antiretroviral drugs (ARVs) and As of 2006, >2 million people in resource-limited settings were taking ARVs for the treatment of Treatment of HIV is by not commonly in the between ARVs and used to treat these infections are formulations of ARVs make dose adjustment when drug interactions the is to a for the In resource-poor settings, such as fluconazole, and may be unavailable or cost it is hope that these can be less and more widely studies of interactions between ARVs and used in resource-limited settings are We for the use of and and and for their with the and
No takes yet. Share an insight, caveat, or question.
Dooley et al. (2008) studied this question.
Synapse has enriched 4 closely related papers on similar clinical questions. Consider them for comparative context: