Changes in relative rate of absorption

Although a medicine may eventually be fully absorbed, a significantly slowed absorption rate may mean that it never reaches effective serum levels, or that an unwanted ‘sustained release’ effect causes prolonged activity or a delay in prompt relief. In some clinical situations, decreased rates of absorption are of no concern; however, in others it may be important. Generally, a decreased rate of absorption is less important for medicines that are given in multiple-dose regimens to achieve a steady state serum level than for those that are given as single doses or are required to produce a rapid effect (e.g. analgesics).

Changes in extent of absorption

Altering the extent of absorption may also have significant consequences. Increasing the amount of medicine absorbed may produce higher plasma levels and a higher risk of adverse reactions or toxicity. Alternatively, reducing the amount of medicine absorbed may result in reduced efficacy or therapeutic failure. This is particularly of concern for medicines with a narrow therapeutic index, such as warfarin and digoxin.

Mucilaginous herbs

Although little research has been conducted to determine the effects of herbal medicines on the absorption of other medicines, one double-blind study found that guar gum slowed the absorption rate of digoxin, but did not alter the extent of absorption, whereas penicillin absorption was both slowed and reduced (Huupponen et al 1984). This brings into question the effects of other gums and highly mucilaginous herbal medicines, such as Ulmus fulvus (slippery elm), Althea officinalis (marshmallow) and Plantago ovata (psyllium). Poorly lipid soluble, the mucilaginous content forms an additional physical barrier to absorption, but whether this will have a clinically significant effect on rate and/or extent of absorption of other medicines is uncertain and remains to be tested.

Nutrients

More research has been conducted on the way nutrients interact and alter the absorption of pharmaceutical medicines than with herbal medicines. The interactions between iron and many minerals provide a useful example. Aluminium, calcium bicarbonate or magnesium trisilicate taken in antacid preparations are known to reduce the extent of iron absorption owing to an alteration in gastric pH. This type of interaction is easily avoided by separating the intake of iron by at least 2 hours from the last dose of antacid.

Intrinsic drug transporters

Until recently, when an orally administered medicine exhibited poor absorption, it was generally assumed that this was because of either physicochemical problems or significant first-pass hepatic metabolism. Recently, it has been recognised that for many medicines poor oral bioavailability could be related to the influence of transporter proteins (Benet & Cummins 2001). Transporter proteins are associated with the transfer of some medicines from the intestinal lumen, through the biological barrier of the intestinal mucosa, into the systemic circulation and back again. Transporters fall into two main categories: carriers and pumps. Carriers are involved in three types of transport processes: facilitated diffusion, co-transport and counter-transport. Pumps are distinguished from carriers by the linkage of transport to an external source of energy. Examples of transporters in the intestine are P-glycoprotein (P-gp), members of the multi-drug-resistance-associated protein family (breast-cancer resistance protein, organic cation and anion transporters), and members of the organic anion polypeptide family (Wagner et al 2001). Of these, P-gp is the most studied (see Clinical note).

Herbal and natural medicines affecting P-gp

The influence of herbal and natural medicines on P-gp expression has only recently been considered, so much is still unknown and speculative. To date, most research has centred on St John’s wort, although clinical testing with other herbs has now gained momentum: every few months more studies investigating the likelihood and clinical significance of proposed interactions are published.

In 1999, clinical testing found that St John’s wort significantly reduces serum levels of digoxin after 10 days’ co-administration (Johne et al 1999). The suspected mechanism of interaction was chiefly liver enzyme induction; however, the magnitude of effect seen in this study, and in others, suggested that P-gp induction was involved (Hennessy et al 2002). More recently, several clinical tests have confirmed that St John’s wort has significant P-gp induction effects. One study found up to a 4.2-fold increase in P-gp expression compared with a placebo after 16 days’ administration. Similar

results were seen in another study in which 14 days’ administration of St John’s wort resulted in a 1.4-fold increased expression of duodenal P-gp (Durr et al 2000). Induction of P-gp is attributed to pregnane X receptor activation by the hyperforin constituent (Lin 2003). Low hyperforin products are therefore less likely to induce P-gp.

St John’s wort is not the only natural substance found to influence P-gp. In vitro studies suggest that rosemary extract may have the opposite effect, inhibiting P-gp. Treating multi-drug-resistant mammary tumour cells with rosemary extract produced an increase in intracellular concentrations of doxorubicin and vinblastine (both P-gp substrates) (Plouzek et al 1999). The same effects were not seen in cells that lack P-gp expression, suggesting that rosemary extract inhibits P-gp activity. The isoflavone genistein has also been investigated, with some results suggesting inhibition of P-gp-mediated drug transport (Castro & Altenberg 1997). Other studies have found that different polyphenols, such as green tea polyphenols, resveratrol (a polyphenol from red wine), curcumin, caffeine, theanine and methoxyflavones from orange, may inhibit drug transport via P-gp (Jodoin et al 2002).

Grapefruit juice is well known to interact with a number of medicines, so it is not surprising that it has also been investigated for effects on P-gp. Currently, evidence is conflicting, as several studies have found that components in grapefruit juice inhibit P-gp (Eagling et al 1999, Ohnishi et al 2000, Soldner et al 1999, Spahn-Langguth & Langguth 2001), whereas a randomised, crossover clinical study found no significant effects (Becquemont et al 2001). A more recent in vitro study produced similar results with 5% normal concentration of grapefruit, orange and apple juices; however, inhibition of other transporter proteins was observed (Dresser et al 2002).

New research suggests that quercetin inhibits P-gp expression (Choi & Li 2005, Chung et al 2005, Kitagawa et al 2005, Wang et al 2004). Studies with experimental models have demonstrated that pretreatment with quercetin increases the bioavailability of the calcium-channel blocker, diltiazem (Choi & Li 2005), and of digoxin (Wang et al 2004). Intriguingly, one in vivo study found quercetin significantly decreased the oral bioavailability of cyclosporin, which is the opposite to what would be expected, suggesting that other mechanisms may also be involved (Hsiu et al 2002).

Although in vitro studies have also identified an inhibitory effect on P-gp for silymarin, the active constituent group in St Mary’s thistle (Chung et al 2005, Zhang & Morris 2003), recent clinical testing found no significant effect in vivo (Gurley et al 2006).

Table 8.1 (p 98) lists herbal and natural medicines that have suspected or known effects on P-gp, together with the research that gives evidence for these effects.

TABLE 8.1 Herbal and Natural Medicines with Suspected or Known Effects on P-Glycoprotein

Cytochromes

Cytochrome P450 is a generic term for a super-family of enzymes (haem containing mono-oxygenases) that have existed throughout nature since the beginning of life more than 3.5 billion years ago (Pirmohamed & Park 2003). The P450s are found chiefly in the liver, but also to a lesser extent in the intestines, kidneys, skin and lungs. These enzymes are responsible for foreign compound metabolism, which evolved about 400–500 million years ago to enable animals to detoxify chemicals in plants. The cytochrome P450 (CYP) enzymes are the most powerful in vivo oxidising agents and are able to catalyse the oxidative biotransformation of a wide range of chemically and biologically unrelated exogenous and endogenous substrates. They are named by a root symbol CYP, followed by a number for family, a letter for subfamily, and another number for the specific gene. For example, CYP3A4 would refer to a specific enzyme from the cytochrome P450 system, family 3, subfamily A and gene 4. Three main CYP families (CYP1, 2, 3) are responsible for metabolism of therapeutic drugs. The different P450 isoforms vary in their abundance within the liver. Of these, the cytochromes CYP2C9, CYP2D6 and CYP3A4 are the most abundant in the human body (Pirmohamed & Park 2003).

The CYP2C sub-family accounts for 15–20% of the total P450 content of the liver, and metabolises approximately 20% of all drugs (Pirmohamed & Park 2003). The main member of this sub-family is CYP2C9, which is responsible for the metabolism of a number of compounds, including warfarin, phenytoin and various NSAIDs. Cytochrome 2D6 is responsible for the metabolism of approximately 25% of therapeutically used drugs, although it accounts for only < 5% of the total P450 content. The CYP3A sub-family accounts for 30% of the total P450 content and is responsible for metabolism of about 50% of therapeutic drugs. Table 8.2 provides examples of common drugs and the cytochromes chiefly responsible for their metabolism. For a more complete list that is frequently updated, see Flockhart 2007.

TABLE 8.2 Selected Drugs that Act as Substrates for CYP Enzymes

The expression and activity of many CYP isoenzymes vary enormously between individuals. Part of the inter-individual variability is environmentally determined by the concomitant intake of drugs and foodstuffs that cause induction and inhibition of the different P450 isoforms. P450 gene polymorphism may also influence expression and activity of CYP enzymes, as this can lead to:

For example, one study identified that 30% of Ethiopians had multiple copies of the 2D6 gene (up to 13), resulting in ultra-rapid metabolism of CYP2D6 substrates (Aklillu et al 1996). As a result, standard doses of CYP2D6 substrates (e.g. beta-blockers, some opioids and tricyclic antidepressants) will not produce anticipated or adequate responses, and higher drug doses are necessary to produce therapeutic effects. It has also been estimated that 7% of Caucasians lack CYP2D6. These individuals will not experience the anticipated therapeutic effects of prodrugs that are CYP2D6 substrates, such as codeine, and will appear more sensitive to the side effects of CYP2D6 substrates (e.g. beta-blockers, some opioids and tricyclic antidepressants) (USFDA 2006).

Another example is the polymorphic distribution of CYP2C9, which is absent in 1% of Caucasians. More than 100 currently used drugs are known substrates of CYP2C9, corresponding to approximately 10–20% of commonly prescribed drugs. Of note, CYP2C9 is chiefly responsible for the metabolism of NSAIDs and COX-2 specific inhibitors (e.g. celecoxib).

Many other factors affect CYP activity, such as the ingestion of foreign compounds (e.g. environmental contaminants) or the ingestion of certain constituents found in food, beverages and medicines.

Because of overlapping substrates, many drug interactions involve both P-gp and CYP3A4.

Enzyme inhibition

Competitive CYP inhibition is dose-dependent and occurs when inhibitors compete with other substances for a particular enzyme. Non-competitive inhibition is also possible and occurs when a substance either destroys or binds irreversibly to a CYP enzyme. In both instances, serum levels of those drugs chiefly metabolised by the affected enzyme will become elevated and the inhibition process is rapid. This is of particular concern with medicines that have a narrow therapeutic index, as very small changes in dose or blood levels can produce significant changes in activity. In the case of enzyme inhibition, raised blood levels can lead to increased side effects and toxicity.

Although inhibition of a CYP enzyme is generally regarded as raising the serum levels of an active drug, this is not always the case. If the drug involved is a prodrug, then it is inactive in its administered form and must be converted to its active form, usually via metabolic processes. If metabolism is slowed, then the production of active metabolites will also be slowed. An example is codeine, which is primarily metabolised by CYP2D6 into its active analgesic metabolite morphine; therefore, CYP2D6 inhibition has the potential to slow or reduce its analgesic activity.

Inhibitors do not all have the same strength and can be classified as strong, moderate or weak, depending on their effect on drug clearance compared to normal values. For example, Flockhart describes a strong inhibitor as one that causes more than 80% decrease in clearance, a moderate inhibitor as one that causes a 50–80% decrease in drug clearance, and a weak inhibitor as one that causes a 20–50% decrease in drug clearance (Flockhart 2007).

Enzyme inhibition is not always harmful and has been manipulated to raise serum drug levels without the need to increase the dose administered. The result has obvious cost advantages when extremely expensive drugs are involved and has been used in some hospitals for medicines such as cyclosporin.

To date, the most studied natural substance capable of significantly inhibiting CYP enzymes is grapefruit. The finding that grapefruit juice markedly increases the bioavailability of some orally administered medicines was based on an unexpected observation from an interaction study between the calcium-channel antagonist felodipine and ethanol. In the study, grapefruit juice was used to mask the taste of ethanol, but actually affected the results by reducing CYP3A4 by 62% and significantly raising felodipine levels (Bailey et al 1998). Since then, the constituents of grapefruit juice have been extensively studied and found to affect the transport and metabolism of many other medicines (Eagling et al 1999, Kane & Lipsky 2000).

Increasingly, research is being conducted to determine whether other commonly used herbal medicines have an affect on CYP activity in vivo. For example, one research group in the United States screened Citrus aurantium, Echinacea purpurea, milk thistle (Silybum marianum), and saw palmetto (Serenoa repens) extracts for effects on CYP1A2, CYP2D6, CYP2E1 and CYP3A4 activity in healthy subjects (Gurley et al 2004). Of the four herbs, only E. purpurea was found to have any effect on the CYP enzymes tested, with a minor influence on CYP1A2 and CYP3A4. Using the same method, extracts of the herbs goldenseal (Hydrastis canadensis), black cohosh (Cimicifugaracemosa), kava kava (Piper methysticum) and valerian (Valeriana officinalis) were tested for effects on CYP1A2, CYP2D6, CYP2E1 or CYP3A4/5 activity (Gurley et al 2005). Goldenseal strongly inhibited CYP2D6 and CYP3A4/5 activity in vivo, whereas kava kava inhibited CYP2E1, black cohosh weakly inhibited CYP2D6 and no effect was observed for valerian. In a separate study, testing Panax ginseng and Ginkgo biloba, no effect was observed on CYP activity; however, garlic oil inhibited CYP2E1 activity by 39% (Gurley et al 2002). There are many other examples to be found in the individual monographs of this book.

Enzyme induction

Alternatively, many different medicines and everyday substances have been found to induce the CYP enzymes (e.g. broccoli, Brussels sprouts, char-grilled meat, high-protein diets, tobacco and alcohol). Research is now identifying a number of herbal medicines that cause CYP induction to various degrees; however, the most studied to date is St John’s wort.

Clinical tests have confirmed that long-term administration of St John’s wort has significant CYP inducer activity, particularly CYP3A4 (Durr et al 2000, Roby et al 2000, Ruschitzka et al 2000, Wang et al 2001). It appears that the hyperforin component is a potent ligand for the pregnane X receptor, which regulates expression of CYP3A4 mono-oxygenase. In this way, hyperforin increases the availability of CYP3A4, resulting in enzyme induction (Moore et al 2000).

Lack of in vitro — in vivo correlation

It is interesting to observe an apparent lack of in vitro – in vivo correlation with some studies of CYP. For instance, in vitro investigations implicate milk thistle extract and/or silibinin as inhibitors of human CYP3A4, CYP2C9, CYP2D6 and CYP2E1; however, in vivo evidence for CYP-mediated interactions by milk thistle is less compelling (Gurley et al 2004). This may be owing to poor bioavailability of key constituents, large inter-individual differences in absorption of constituents, inter-product variability in the ratios of constituent, poor dissolution of dosage forms or other mechanisms.