Clinical Pharmacokinetics

, Volume 38, Issue 1, pp 41–57

Pharmacokinetic-Pharmacodynamic Consequences and Clinical Relevance of Cytochrome P450 3A4 Inhibition


  • George K. Dresser
    • Department of MedicineLondon Health Sciences Centre and The University of Western Ontario
    • Department of Pharmacology & ToxicologyLondon Health Sciences Centre and The University of Western Ontario
  • J. David Spence
    • Department of MedicineLondon Health Sciences Centre and The University of Western Ontario
    • Department of Pharmacology & ToxicologyLondon Health Sciences Centre and The University of Western Ontario
    • Department of NeurologyLondon Health Sciences Centre and The University of Western Ontario
    • Department of MedicineLondon Health Sciences Centre and The University of Western Ontario
    • Department of Pharmacology & ToxicologyLondon Health Sciences Centre and The University of Western Ontario
Review Article Drug Interactions

DOI: 10.2165/00003088-200038010-00003

Cite this article as:
Dresser, G.K., Spence, J.D. & Bailey, D.G. Clin Pharmacokinet (2000) 38: 41. doi:10.2165/00003088-200038010-00003


Drug interactions occur when the efficacy or toxicity of a medication is changed by administration of another substance. Pharmacokinetic interactions often occur as a result of a change in drug metabolism. Cytochrome P450 (CYP) 3A4 oxidises a broad spectrum of drugs by a number of metabolic processes. The location of CYP3A4 in the small bowel and liver permits an effect on both presystemic and systemic drug disposition. Some interactions with CYP3A4 inhibitors may also involve inhibition of P-glycoprotein.

Clinically important CYP3A4 inhibitors include itraconazole, ketoconazole, clarithromycin, erythromycin, nefazodone, ritonavir and grapefruit juice. Torsades de pointes, a life-threatening ventricular arrhythmia associated with QT prolongation, can occur when these inhibitors are coadministered with terfenadine, astemizole, cisapride or pimozide. Rhabdomyolysis has been associated with the coadministration of some 3-hydroxy-3-methylglutaryl-coenzyme Areductase inhibitors (‘statins’) and CYP3A4 inhibitors. Symptomatic hypotension may occur when CYP3A4 inhibitors are given with some dihydropyridine calcium antagonists, as well with the phosphodiesterase inhibitor sildenafil. Excessive sedation can result from concomitant administration of benzodiazepine (midazolam, triazolam, alprazolam or diazepam) or nonbenzodiazepine (zopiclone and buspirone) hypnosedatives with CYP3A4 inhibitors. Ataxia can occur with carbamazepine, and ergotism with ergotamine, following the addition of a CYP3A4 inhibitor.

Beneficial drug interactions can occur. Administration of a CYP3A4 inhibitor with cyclosporin may allow reduction of the dosage and cost of the immunosuppressant. Certain HIV protease inhibitors, e.g. saquinavir, have low oral bioavailability that can be profoundly increased by the addition of ritonavir.

The clinical importance of any drug interaction depends on factors that are drug-, patient- and administration-related. Generally, a doubling or more in plasma drug concentration has the potential for enhanced adverse or beneficial drug response. Less pronounced pharmacokinetic interactions may still be clinically important for drugs with a steep concentration-response relationship or narrow therapeutic index. In most cases, the extent of drug interaction varies markedly among individuals; this is likely to be dependent on interindividual differences in CYP3A4 tissue content, pre-existing medical conditions and, possibly, age.

Interactions may occur under single dose conditions or only at steady state. The pharmacodynamic consequences may or may not closely follow pharmacokinetic changes. Drug interactions may be most apparent when patients are stabilised on the affected drug and the CYP3A4 inhibitor is then added to the regimen. Temporal relationships between the administration of the drug and CYP3A4 inhibitor may be important in determining the extent of the interaction.

A drug interaction occurs when the effectiveness or toxicity of a drug is altered by the administration of another drug or substance. Mechanisms of drug interactions fall into 2 broad categories: pharmacokinetic and pharmacodynamic. Pharmacokinetic alterations are the most common and include changes in drug absorption (rate and/or extent), distribution (plasma protein binding displacement), metabolism and excretion (pulmonary, renal and biliary).

A change in the metabolism of a drug by the coadministration of another substance is a frequent cause of clinically important drug interactions. This can occur from the induction of new protein synthesis, which accelerates drug metabolism and decreases the magnitude and duration of drug response, or from inhibition, which results in elevated plasma drug concentrations with increased potential for enhanced beneficial or, in most cases, adverse effects.

The cytochrome P450 (CYP) mixed function oxidases are a family of enzymes which account for the majority of oxidative biotransformations of xenobiotics and endogenous biochemicals. Over 30 different human CYP enzymes have been identified. CYP3A4 appears to be one of the most important human enzymes as approximately 60% of oxidised drugs are biotransformed, at least in part, by it.[1]

The purpose of this article is to review drug interactions resulting from the inhibition of CYP3A4 metabolic activity in humans. Following a brief outline of the CYP3A subfamily, this article will focus on CYP3A4 substrates and their predicted pharmacokinetic characteristics of the interaction, CYP3A4 inhibitors, clinically important CYP3A4 substrate-inhibitor interactions and factors determining the clinical relevance of the interaction.

1. Cytochrome P450 (CYP) 3A Subfamily

The isoforms of CYP3A in humans include 3A3, 3A4, 3A5 and 3A7. Each of these enzymes share at least 85% amino acid sequence homology.[2] CYP3A3 is so similar to CYP3A4 that the distinction between the 2 may be artificial and for the purpose of this review, no attempts will be made to distinguish between them.[3] CYP3A5 is the predominant isoform in the lung[4] and stomach[2] and is present in the small bowel and renal tissue.[5] Its contribution to drug metabolism is uncertain. CYP3A7 is found in fetal liver but does not appear to be present in adults.

CYP3A4 is the predominant isoform of CYP3A in adult humans. It can catalyse a remarkable number of metabolic processes including aliphatic oxidation, aromatic hydroxylation, N-dealkylation, O-demethylation, S-demethylation, oxidative deamination, sulfoxide formation, N-oxidation and N-hydroxylation, to mention a few. This usually produces inactivation and elimination of most pharmaceuticals. However, it can also activate carcinogenic substances such as the aflatoxins and polycyclic aromatic hydrocarbons.

Although CYP3A4 drug metabolising activity varies widely among individuals, it has a unimodal population distribution and does not appear to be subject to genetic polymorphism as is seen with other CYP isoforms (2D6, 2C9 and 2C19). The wide interindividual variability is likely, in part, to be caused by ethnic or cultural differences, perhaps related to an interaction between race and diet. Other factors known to play a role in activity are age[6] and the presence of small bowel or liver disease.[7,8] There may be modest gender differences, perhaps related to the hormonal milieu in which the enzyme functions, although this is controversial.[9]

CYP3A4 content is highest in the liver. Hepatic CYP3A4 content has been shown to vary at least 20-fold among individuals and activity, as measured by the erythromycin breath test, to range 10-fold.[10] Small bowel CYP3A4 is found in the apical enterocytes and its content varies 11-fold among individuals. Hepatic and enteric CYP3A4 content appear to be regulated independently of each other.[5] The location of CYP3A4 in the small bowel and liver makes it well suited to play a significant role in first-pass (or presystemic) drug metabolism. This is illustrated by the dihydropyridine calcium antagonist, felodipine, which is a CYP3A4 substrate that normally undergoes sequential metabolism in the apical enterocytes of the small bowel and then in the hepatocytes of the liver following oral administration, resulting in a mean 15% absolute bioavailability (fig. 1).[11,12]
Fig. 1

Sequential presystemic metabolism of felodipine by cytochrome P450 (CYP) 3A4 in apical enterocytes of the small bowel (A) and in the hepatocytes of the liver (B) following oral drug administration. The white sectors of the symbols representing felodipine, and the adjacent percentages, show the proportion of felodipine unmetabolised before and after passage through the gut wall and the liver.

2. CYP3A4 Substrate Drugs

A number of drugs from a broad range of therapeutic categories are CYP3A4 substrates (table I). With regard to orally administered drugs, knowledge of the extent of, and factors determining, the bioavailability of these substrates allows for prediction of the changes that could occur to their pharmacokinetics with administration of a CYP3A4 inhibitor. Oral bioavailability (i.e. the fraction of the total dose that reaches the systemic circulation unchanged) is determined by 2 major processes: the proportion of the dose that is absorbed from the gut and the fraction that does not undergo presystemic metabolism, mainly in the small bowel and liver. The contributions of enteric and hepatic metabolism can be estimated separately by determining drug clearance following both oral and intravenous administration.
Table I

Drugs metabolised by cytochrome P450 (CYP) 3A4 and extent of presystemic metabolism

If a substrate normally has high presystemic elimination (low oral bioavailability) and is primarily dependent upon CYP3A4 for elimination, then administration of an inhibitor of its metabolism can be expected to produce substantial change in substrate pharmacokinetics under single dose conditions. The interaction would be characterised by a higher drug peak plasma drug concentration (Cmax) from reduced presystemic metabolism and a greater area under the drug concentration-time curve (AUC) possibly from both lower presystemic and systemic elimination.

Furthermore, there seems to be an inverse relationship between the extent of inherent presystemic elimination and magnitude of the increase in Cmax and AUC among medications. This has been observed with dihydropyridines and grapefruit juice,[13] benzodiazepines and ketoconazole,[14] and 3-hydroxy- 3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors with itraconazole.[15,16]

The interaction appears to be different if the CYP3A4 substrate normally has a high oral bioavailability. In this setting, inhibition of first-pass metabolism in the small bowel and liver would not be expected to have much effect on the Cmax under single dose conditions. The AUC may be somewhat augmented as a result of prolongation of the apparent elimination half-life (t½β) from reduced systemic drug clearance. In this situation, repeat administration of CYP3A4 substrate and inhibitor may produce a cumulative increase in plasma substrate concentrations. Thus, the clinical importance of the inhibitory interaction for CYP3A4 substrates with high inherent oral bioavailability may be realised only during steady-state administration.

With regard to intravenous therapy, this route of administration, by definition, results in complete drug bioavailability. Consequently, the pharmacokinetic characteristics of the interaction would probably be similar to those observed with a CYP3A4 substrate with high oral bioavailability.

It should be noted that certain substrates metabolised by CYP3A4 are also substrates for P-glycoprotein, a transmembrane adenosine triphosphate (ATP)-dependent active transport protein found in a number of organs, including the gut, brain, liver and kidney.[17] P-glycoprotein acts as an efflux transporter to decrease drug absorption into the portal circulation and CNS and to increase drug elimination into the bile and urine. CYP3A4 and P-glycoprotein in the gut and liver appear capable of acting in concert to decrease plasma drug concentrations. For those drugs that are substrates for both CYP3A4 and P-glycoprotein, inhibition of CYP3A4 or P-glycoprotein might be expected to produce similar changes to the pharmacokinetics of the parent drug. Additionally, certain substances appear capable of inhibiting both CYP3A4 and P-glycoprotein.[18,19] Consequently, some interactions originally thought to be caused solely by the inhibition of CYP3A4-mediated drug metabolism are likely to be mediated, in part, by the inhibition of P-glycoprotein transport.

3. CYP3A4 Inhibitor Drugs

A wide variety of endogenous and exogenous substances has been identified as inhibiting CYP3A4 activity. The most clinically relevant of these are listed in table II. Although these substances all share the common feature of interfering with CYP3A4 activity, they vary in potency and mode of action.
Table II

Clinically relevant inhibitors of cytochrome P450 (CYP) 3A4

Many known inhibitors of CYP3A4 are of uncertain clinical relevance.[20] However, specific inhibitors of CYP3A4 that deserve mention because they are potent and have been associated with clinically relevant interactions include azole antifungals, macrolide antibacterials, nefazodone, the HIV protease inhibitors and grapefruit juice.

Several mechanisms of inhibition are possible. Azole antifungals and first generation HIV protease inhibitors appear to act via competitive inhibition by rapid, reversible binding of the inhibitor or its metabolite to CYP3A4. Macrolide antibacterials produce slowly reversible, noncompetitive inhibition.[20] This has been interpreted by some as mechanism-based inhibition,[21,22] but this does not appear to be the case. The furanocoumarins in grapefruit juice, dihydroxybergamottin and bergamottin, cause irreversible, mechanism-based (suicide) inhibition.[23,24] This presumably involves CYP3A4-mediated formation of a reactive metabolite that covalently binds to the enzyme, leading to its inactivation.[20,25]

Most potent orally administered inhibitors act at the level of the small bowel and liver. However, grapefruit juice is an example of an inhibitor that appears to be clinically active against only enteric CYP3A4[26] and may be useful as a probe of enteric CYP3A4 activity. Conversely, intravenous administration of a CYP3A4 inhibitor probably produces more selective action in the liver.[27]

4. Adverse Clinical Consequences of CYP3A4 Interactions

4.1 Torsades de Pointes

Torsades de pointes is a life-threatening ventricular arrhythmia which occurs in the setting of electrocardiographic QT interval prolongation. It has been shown to occur with several medications that are CYP3A4 substrates: the nonsedating antihistamines terfenadine and astemizole,[28,29] the gastrointestinal prokinetic agent cisapride[30] and the antipsychotic drug pimozide.[31] These CYP3A4 substrates appear to act, in part, in a concentration-dependent manner to block potassium rectifier current in cardiac conduction pathways, which is the basis for QT interval prolongation. Therefore, conditions in which plasma concentrations of these drugs are markedly elevated would probably be associated with increased risk of developing this arrhythmia.

QT interval prolongation and/or torsades de pointes have been reported to occur with concomitant administration of the CYP3A4 inhibitors clarithromycin, diltiazem, erythromycin, grapefruit juice, itraconazole or ketoconazole with astemizole, cisapride, pimozide or terfenadine.[28,29,3244] Pharmacokinetic-pharmacodynamic interaction studies demonstrated that these CYP3A4 inhibitors augmented plasma substrate concentrations and prolonged QT interval, supporting the predicted mechanism of action.[31,4552] In addition to avoiding the above CYP3A4 inhibitors, it also appears prudent to avoid coprescription of the other CYP3A4 inhibitors listed in table II with astemizole, cisapride, pimozide or terfenadine because of the seriousness of the adverse effect.

Alternative pharmacotherapy to avoid this effect might include the nonsedating antihistamines cetirizine, fexofenadine (active terfenadine metabolite) or loratidine. These do not appear to prolong QT interval.[53,54] Metoclopramide or domperidone are alternatives to cisapride, and haloperidol could be used in place of pimozide for treatment in certain conditions.

The azole antifungal fluconazole has significantly less inhibitory activity against CYP3A4 compared with ketoconazole or itraconazole, and in most cases does not appear to produce clinically significant inhibition of CYP3A4.[55] It should be noted, however, that in select cases, interactions have been noted with fluconazole, particularly with very low therapeutic index drugs such as cyclosporin.[5659] The macrolide antibacterial azithromycin does not significantly inhibit CYP3A4.[60] Coadministration of these latter drugs is still not generally recommended because of similarities to the contraindicated drugs.

4.2 Rhabdomyolysis

The HMG-CoA reductase inhibitors are an important class of cholesterol-lowering medications. However, their use can be associated with diffuse myalgia and marked elevation of creatine kinase. There have been reports of severe rhabdomyolysis that precipitated acute renal failure.[6164] The mechanism of this adverse effect has not been clearly defined. However, it appears to occur in conditions in which plasma concentrations of parent drug and metabolite are profoundly elevated, suggesting that this adverse effect has a pharmacokinetic basis.

CYP3A4 plays a major role in the metabolism of lovastatin, simvastatin, atorvastatin and cerivastatin. CYP3A4 is involved in a minor metabolic pathway for pravastatin, whereas CYP2C9 is the primary CYP isoform metabolising fluvastatin.[65]

The magnitude of the pharmacokinetic effect on different HMG-CoA reductase inhibitors by CYP3A4 inhibitors is shown in table III. Lovastatin and simvastatin consistently demonstrated profound increases in AUC and Cmax when administered with a number of CYP3A4 inhibitors. On the other hand, fluvastatin and pravastatin appear to be the least susceptible, whereas atorvastatin and cerivastatin are intermediate in susceptibility to a pharmacokinetic interaction with CYP3A4 inhibitors.
Table III

Pharmacokinetic interactions between cytochrome P450 (CYP) 3A4 inhibitors and 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitors. These are all single dose studies unless indicated otherwise

In patients requiring the concurrent use of CYP3A4 inhibitors and HMG-CoA reductase inhibitor therapy, fluvastatin or pravastatin would be reasonable choices to avoid potential drug-drug interaction problems.[65] In this setting, dosage adjustment of the other HMG-CoA reductase inhibitors does not appear to be feasible because of the unpredictable variability of the interaction among individuals. Particular subgroups of patients, such as those with renal dysfunction and those on immunosuppressants, are particularly prone to rhabdomyolysis with these drugs and should receive special consideration for HMG-CoA reductase inhibitors that are not CYP3A4 substrates.

4.3 Symptomatic Hypotension

Hypotension is a dose-dependent adverse effect of many antihypertensive medications. A number of important antihypertensive medications, including most of the dihydropyridine calcium antagonists, are substrates of CYP3A4. Case reports of significant interactions[76] resulting in dose-dependent adverse effects have been published for felodipine and erythromycin, felodipine and itraconazole,[77] and nifedipine and itraconazole.[78]

The effect of several CYP3A4 inhibitors on the pharmacokinetics of dihydropyridines is shown in table IV. Grapefruit juice is the most extensively studied CYP3A4 inhibitor of this class of drugs. It increased the AUC and Cmax for all of the dihydropyridines except amlodipine, which usually undergoes low presystemic elimination and has a high oral bioavailability compared with the other dihydropyridines. The CYP3A4 inhibitors cyclosporin, erythromycin and, particularly, itraconazole produce a clinically important increase in the oral bioavailability of felodipine. It seems likely that these inhibitors will interact with other dihydropyridines.
Table IV

Pharmacokinetic interactions between cytochrome P450 (CYP) 3A4 inhibitors and dihydropyridine calcium antagonists. These are all single dose studies

Perhaps one of the most significant reports of a CYP3A4 drug interaction involves mibefradil, a nondihydropyridine calcium antagonist. Mibefradil was approved in the US in 1997 for the treatment of patients with hypertension. It was a particularly promising drug because of a number of desirable haemodynamic features not normally found in a single drug, such as a long t½β, lack of negative inotropic effects, lack of sympathetic stimulation, selective coronary vasodilation and afterload reduction.[97] Early work with the drug suggested that it had some inhibitory effects on both CYP3A4 and CYP2D6, but in spite of this it survived phase I to III clinical trials with little evidence of problems. However, within months of approval there were reports of significant interactions with both dihydropyridines and β-blockers. In one case, a single dose of nifedipine was consumed 24 hours after the last dose of mibefradil.[98] The ensuing hypotension and bradycardia was refractory to all available therapy, and the patient died within 24 hours of taking the dihydropyridine. The recent release and ensuing withdrawal of this medication is an important lesson in maintaining appropriate vigilance and awareness of drug-drug interactions, particularly for recently released medications.

In general, patients who require therapy with dihydropyridine calcium antagonists should avoid consuming grapefruit juice. Drugs that inhibit CYP3A4 should be avoided whenever possible. If coprescription of a dihydropyridine and an inhibitor is unavoidable, the starting dosage of the antihypertensive should be low and the dihydropyridine should be titrated slowly to antihypertensive effect. Alternatively, the dihydropyridine calcium antagonist amlodipine may be a better choice, although adverse effects may still occur at steady state.

Sildenafil, a selective inhibitor of cyclic guanosine monophosphate (cGMP)-specific phosphodiesterase type 5 (PDE5), is available for the treatment of patients with erectile dysfunction. Sildenafil inhibits the degradation of cGMP from the nitric oxide (NO) pathway, producing selective vasodilation in the corpus cavernosal smooth muscle.[99] The NO-cGMP pathway is also responsible for vasodilation in other systemic vascular beds. Sildenafil is a CYP3A4 substrate with intermediate bioavailability and is, therefore, subject to single dose interactions with CYP3A4 inhibitors.[100] Under normal circumstances, no systemic vasodilation is observed with sildenafil because of its specificity for corpus cavernosal PDE5. However, under conditions of excessive sildenafil exposure, particularly with additional nitrate therapy, hypotension,[101] myocardial infarction and/or death have resulted.[102104]

Sildenafil should not be prescribed to anyone requiring antianginal nitrate therapy. In the event that concomitant administration of sildenafil and CYP3A4 inhibitors cannot be avoided, a lower starting dose of sildenafil 25mg should be considered.[102]

4.4 Excessive Sedation

A number of psychopharmacological agents have been reported to be susceptible to CYP3A4 drug-drug interactions. Case reports of increased sedation related to drug-drug interactions are somewhat lacking, particularly considering the prevalent use of potentially interacting medications. However, pharmacokinetic-pharmacodynamic studies have documented significant interactions with a number of important hypnosedatives, including midazolam,[105] triazolam,[106,107] alprazolam,[108] diazepam,[109] and the nonbenzodiazepines zopiclone[110] and buspirone.[111]

Midazolam is a short-acting benzodiazepine with significant amnesic properties. It is a popular intravenous sedative agent used in anaesthesia as well as for sedation of patients in the intensive care unit. It has an oral bioavailability of approximately 25 to 40% and is extensively metabolised by enteric as well as hepatic CYP3A4.[112] In a study investigating the metabolism of midazolam, the CYP inducer rifampicin (rifampin) decreased the AUC to 2.3% of baseline, whereas itraconazole increased the AUC to 800% of baseline. Altogether there was a 400-fold difference in the AUC of midazolam, depending on the coadministered drug.[113]

Significant interactions resulting in both pharmacokinetic and pharmacodynamic effects, as assessed by a variety of tools, have been observed when midazolam is coadministered with clarithromycin, diltiazem, erythromycin, fluconazole, itraconazole, ketoconazole or verapamil.[27,105,114116] An interaction was also observed with grapefruit juice and orally administered midazolam, although the magnitude of increase in AUC was only 52%.[117] Three benzodiazepines have been shown not to interact significantly with inhibitors of CYP3A4, temazepam[118] and nitrazepam.[119] Lorazepam is primarily metabolised by conjugation and is, therefore, not susceptible to an interaction with CYP3A4 inhibitors.[120]

The susceptibility of midazolam to interactions with inhibitors of CYP3A4 is most problematic in patients in the intensive care settings, where both midazolam and a number of the important inhibitors tend to be used liberally. Fortunately, the single dose interaction of intravenously administered midazolam with CYP3A4 inhibitors is smaller than the interaction of oral midazolam with the same CYP3A4 inhibitors.[121] For patients in this setting requiring therapy with azole antifungals, macrolide antibacterials or any other inhibitors of CYP3A4, alternative choices for sedation include lorazepam or nonbenzodiazepine sedatives such as propofol.

4.5 Ataxia

Carbamazepine is an anticonvulsant that is both a substrate for CYP3A4 and an inducer of the enzyme. It induces the metabolism of a variety of CYP3A4 substrates[122] and also interacts with other inhibitors of its own metabolism. Case reports of carbamazepine toxicity related to the inhibition of carbamazepine metabolism have been reported with clarithromycin, diltiazem, erythromycin, fluoxetine, fluvoxamine, nefazodone or verapamil.[123132]

In a pharmacokinetic study, the addition of fluoxetine to carbamazepine therapy at steady state resulted in an increase in the AUC of carbamazepine to 127% of pre-fluoxetine concentrations.[133] Coadministration of erythromycin with both single and multiple doses of carbamazepine was associated with significant increases in carbamazepine concentrations.[134,135]

Avoidance of toxicity for patients requiring long term carbamazepine therapy involves using knowledge of CYP3A4 inhibitors to avoid coprescription of potentially harmful combinations. When a patient is started on an inhibitor, a carbamazepine dosage adjustment may be required. Monitoring of carbamazepine concentrations may be useful in avoiding toxicity.

Although the selective serotonin reuptake inhibitors (SSRIs) fluoxetine and fluvoxamine have both demonstrated inhibitory effects on CYP3A4 metabolism, the other SSRIs sertraline and paroxetine have not.[136] These may be appropriate alternatives in situations of potential drug interaction with carbamazepine.

4.6 Ergotism

The ergot alkaloids are a family of xenobiotics with a variety of pharmacological effects which are useful in the treatment of patients with migraine headache.[137] Ergotamine, one of the first ergot alkaloids, is a substrate of CYP3A4, with an apparently low oral bioavailability.[138140]

Initially referred to as ‘St Anthony’s Fire’ during the Middle Ages, ergotism is a syndrome of vascular ischaemia and neurological compromise associated with excessive ergotamine intake. This syndrome has been recognised in patients concomitantly receiving standard doses of ergotamine with inhibitors of CYP3A4, including clarithromycin,[141] ritonavir[142,143] and triacetyloleandomycin.[144] Although no pharmacokinetic studies have documented the extent of interaction, it seems plausible that the mechanism at work is again inhibition of ergotamine metabolism, causing excessive tissue concentrations of active ergotamine.

In patients receiving ergotamine, all inhibitors of CYP3A4 should be avoided. Alternatively, better options exist for the treatment of patients with migraine headache, including the triptans (sumatriptan, zolmitriptan, naratriptan, rizatriptan and eletriptan). However, there is currently little information on the possible interactions, and any resulting risks, between these drugs and CYP3A4 inhibitors and caution is advised in these situations.

5. Beneficial Clinical Consequences of CYP3A4 Interactions

Although drug-drug interactions are generally negative occurrences to be avoided, CYP3A4 interactions have also been used to achieve beneficial therapeutic goals. The following creative examples of beneficial interactions are provided.

5.1 Cost Savings

Cyclosporin is a calcineurin inhibitor and immunosuppressive agent metabolised by CYP3A4 that is useful in organ transplantation. Unfortunately, like most immunosuppressive agents, it is expensive.

It has been demonstrated that the CYP3A4 inhibitor ketoconazole more than doubles the oral bioavailability of concurrently administered cyclosporin and reduces the dosage of cyclosporin required to produce adequate immunosuppression by 60 to 80%.[145,146] In view of the potential cost savings, 2 randomised controlled trials of reduced dosage cyclosporin with ketoconazole versus regularly administered cyclosporin have been undertaken. In both studies, the combination therapy proved to be well tolerated, effective and considerably less costly than the standard therapy over a 1- to 4-year period.[146,147] This interaction can be particularly beneficial if the immunosuppressed patient requires antifungal therapy concurrently with immunosuppressive therapy.

It was important for the success of these studies that plasma cyclosporin concentrations were measured routinely as a part of standard clinical practice, making this beneficial interaction feasible.

Given the low cost of grapefruit juice, it is possible that it could be used for this purpose. However, one theoretical difficulty with the use of grapefruit juice is quality control in the content of the active ingredient and its uniformity of effects on cyclosporin bioavailability.[148152]

5.2 Enhanced Efficacy

The protease inhibitors are antiretroviral drugs with a low incidence of dose-related toxicity used in the treatment of patients with HIV-1 infection. These drugs are often prescribed concurrently and with a variety of other antiretroviral and anti-infective medications. Some of them, particularly ritonavir and saquinavir, are pharmacokinetically characterised by low, and often erratic, bioavailability.[153] The protease inhibitors are metabolised by CYP3A4 and inhibit this isoenzyme to varying degrees.[154] They are also substrates of P-glycoprotein and can inhibit its activity.[155]

A number of studies investigating possible combination strategies have been published. The pharmacokinetics of saquinavir in patients with advanced HIV disease have been studied in the presence and absence of ritonavir. The addition of ritonavir to saquinavir therapy resulted in a 33-fold greater Cmax and a 58-fold greater AUC.[156] This profound pharmacokinetic interaction is larger than that seen with pure CYP3A4 inhibitors and most likely reflects both CYP3A4 and P-glycoprotein inhibition. A similar study was conducted using grapefruit juice. It was found that the oral bioavailability of saquinavir was doubled from 0.7% to 1.4%.[157]

These findings from pharmacokinetic studies have now been put into clinical practice. Ritonavir is often coprescribed with other protease inhibitors in order to augment the pharmacokinetic profiles of both medications. Saquinavir may be consumed with grapefruit juice in order to augment its bioavailability.

6. Clinical Relevance

Information on inhibitory drug-drug interactions arises from a number of sources. Preliminary information is often available from in vitro studies, possibly done in the early stages of drug development. Unfortunately, the observed in vitro interactions often do not predict in vivo drug interactions well. However, they can be used as a screen and can provide a rationale for in vivo human investigations. Relevant drug interactions are often based on reports from clinical practice, which are subsequently confirmed in clinical trials. For these reasons, we would suggest that human in vivo data be pursued whenever possible.

The therapeutic importance of any drug interaction in humans depends on several factors. Drug-related factors such as the pharmacokinetic-pharmacodynamic properties of the substrate and inhibitor obviously play a critical role and have been the focus of this review. Patient-related factors might help to determine susceptibility to any potential interaction. Finally, administration-related factors will determine specifically what opportunity for interaction between substrate and inhibitor exists. Each of these factors will be discussed.

6.1 Drug-Related Factors

Since a medication can be administered as a prodrug where the metabolite is primarily active or as an active species with metabolites possessing less or more pharmacological activity, an initial consideration should be the effect on the pharmacokinetics of the active substrate(s).

When an inhibitor of CYP3A4 metabolism augments the concentration of the active substrate in plasma or tissue, the effect on drug pharmacodynamics will be determined by 3 factors: (i) the position of the substrate concentration on the response curve before the interaction; (ii) the slope of the concentration-response curve; and (iii) the magnitude of the increase in concentration of the substrate. In many situations, medications are administered at a dose producing less than the maximum effect. This is often the case for drugs meant for long term administration, as the dose is initially low and can be gradually titrated upwards to achieve the minimum effective dosage and optimum drug safety. Thus, many clinically employed dosages are on the positive slope portion of the drug concentration-response curve, and a marked increase in plasma drug concentration might be expected to produce some augmented (beneficial or adverse) effect.

Increased plasma drug concentrations can produce adverse drug effects from an excessive extension of the primary action of the drug. Adverse drug effect can also result from a different mechanism of action. In this case, there is more than one drug concentration-response curve, one for efficacy and another shifted to the right for toxicity. The difference between the 2 curves is termed the therapeutic index. The smaller the difference between the 2 curves (at 50% of maximum effect) the lower the therapeutic index and the greater the risk that increased plasma drug concentrations will induce toxicity.

Since drug response is often linear with the logarithm of drug concentration, a general rule could be that an inhibitor that produces a doubling or more in plasma drug concentration should alert the physician or pharmacist to the potential for enhanced drug response (beneficial or adverse). Less than a doubling of concentration may also be clinically important for medications with a steep drug concentration-response relationship or narrow therapeutic index, such as carbamazepine or cyclosporin.

6.2 Patient-Related Factors

Patient-related factors can affect the therapeutic importance of the interaction. In most cases, the extent of the pharmacokinetic interaction varies markedly among individuals for the same inhibitor-drug combination. For example, grapefruit juice produced an effect ranging from no change to an 8-fold increase in the AUC of felodipine.[83] This appears to be the result of interindividual differences in CYP3A4 content, possibly dependent on genetic and/or environmental (dietary) factors.

Pre-existing medical conditions can increase a patient’s susceptibility to drug interactions. Dihydropyridines produce an antihypertensive effect dependent on the pretreatment blood pressure, with the greatest reduction occurring in patients with the highest pretreatment pressure.[158] The magnitude and/or rate of blood pressure reductions for drugs of this class may be important because of the apparently greater mortality observed in post-infarction patients managed with regular release dihydropyridines.[159,160] Also, drugs that can produce torsades de pointes (astemizole, cisapride, pimozide and terfenadine) following CYP3A4 inhibition are particularly likely to increase the risk of this serious arrhythmia in patients with pre-existing prolonged QT interval.[37,40]

Inflammatory small bowel or cirrhotic liver disease are known to decrease the content and activity of CYP3A4 in these tissues.[7,8] However, the effect of a CYP3A4 drug-inhibitor interaction is uncertain in these conditions, since the importance of CYP3A4 to drug metabolism in the absence of inhibitor already appears to be reduced.

Few studies have assessed the relationship between age and susceptibility to drug interactions. However, hepatic CYP3A4-mediated clearance of felodipine is decreased in the elderly,[11] which may be the result of age-related changes in liver blood flow, size, drug binding or distribution since hepatic CYP3A4 content appears unaffected by age. This population also showed greater susceptibility to the antihypertensive effects of dihydropyridines compared with younger individuals, possibly because of altered autonomic regulation. Since the elderly represent a group most often consuming medications, the potential for relevant drug interactions in this population appears substantial.

The influence of gender on susceptibility to CYP3A4 drug-drug interactions is uncertain.[9]

6.3 Administration-Related Factors

Administration-related factors also affect the interaction. A marked pharmacokinetic interaction can occur under single dose conditions or only following repeat drug administration (steady state), depending on the normal extent of presystemic drug elimination. Pharmacodynamic effects may closely follow pharmacokinetic changes. However, this may not always be the case, as a delay in the appearance of symptoms can occur, for example with HMG-CoA reductase and CYP3A4 inhibitors and the development of rhabdomyolysis.

If the patient is already taking the CYP3A4 inhibitor, a marked response or untoward effect may occur with the initial dose of the affected drug. However, the cause could be incorrectly attributed to inherent variation in the normal effect of the drug among patients and be overlooked as a drug interaction by the patient, pharmacist or physician. Drug interactions might be more readily recognised in patients stabilised on a particular dosage of the causative agent when the CYP3A4 inhibitor substance is added to the regimen.

Some CYP3A4 inhibitors, such as mibefradil, appear to have a very long duration of action, and administration of a substrate up to 24 hours following the last dose of inhibitor may result in a significant interaction. However, other drug-drug interactions may only occur when the drugs are administered together, and interactions may be avoided by adjusting the times of administration. When grapefruit juice is coingested with lovastatin, there is a 15-fold increase in the AUC of lovastatin,[161] but when the lovastatin is consumed approximately 12 hours after the grapefruit juice, preliminary data suggest only a 2-fold increase in AUC.[162]

7. Conclusions

The potential exists for numerous CYP3A4 drug-inhibitor interactions. The nature of the pharmacokinetic, and possibly the pharmacodynamic, interaction during single dose and steady-state administration can be estimated by knowing the normal extent of presystemic drug elimination for drugs extensively metabolised by CYP3A4. In general, a doubling or more of plasma drug concentrations by an inhibitor should be considered potentially relevant, although a lesser effect may also be important for certain drugs. Clinically important adverse interactions with specific medications may lead to torsades de pointes, rhadomyolysis, symptomatic hypotension, excessive sedation, ataxia and ergotism.

Alternative therapeutic strategies are available for most or all of these scenarios. Potential beneficial interactions include cost savings and enhanced efficacy in certain cases. Patients vary in their susceptibility to any particular CYP3A4 drug-inhibitor combination, depending on drug-, patient- and administration-related factors.

Copyright information

© Adis International Limited 2000