Metabolism and Biotransformation

Phase one reactions fall into three categories: oxidation (the most common, accounting for the majority of cytochrome P450-mediated reactions), reduction, and hydrolysis. The cytochrome P450 superfamily — heme-containing monooxygenases located predominantly in the endoplasmic reticulum of hepatocytes — drives the large majority of phase one oxidative drug metabolism. Significant additional cytochrome P450 expression occurs in enterocytes of the small intestinal wall, which accounts for intestinal first-pass metabolism as described in Module 1. The cytochrome P450 reaction uses molecular oxygen and inserts one oxygen atom into the substrate, typically forming a hydroxyl group, converting a lipophilic drug into a more polar hydroxylated product.

Of more than 50 human cytochrome P450 isoforms, six enzymes metabolize the large majority of clinically used drugs: cytochrome P450 3A4, cytochrome P450 2C9, cytochrome P450 2C19, cytochrome P450 2D6, cytochrome P450 1A2, and cytochrome P450 2B6. These isoforms have overlapping but distinct substrate specificities, different tissue expression patterns, and different susceptibilities to induction and inhibition by other drugs — which is why cytochrome P450-mediated drug-drug interactions are so common and clinically consequential.

Non-cytochrome P450 phase one enzymes include monoamine oxidase (which metabolizes catecholamines, serotonin, and tyramine, and is the target of monoamine oxidase inhibitor antidepressants), xanthine oxidase (which metabolizes azathioprine and 6-mercaptopurine, relevant to allopurinol interactions), alcohol dehydrogenase and aldehyde dehydrogenase (which metabolize ethanol and are inhibited by disulfiram), and flavin-containing monooxygenases (which metabolize several drugs including the prodrug clozapine's metabolite pathways).

Cytochrome P450 3A4 is the most abundant hepatic cytochrome P450 and metabolizes approximately 50 percent of clinically used drugs. Its substrates include statins (simvastatin, lovastatin, atorvastatin), calcineurin inhibitors (cyclosporine, tacrolimus), benzodiazepines, most calcium channel blockers, and many other drugs. It is expressed in both the liver and intestinal wall, making it the dominant contributor to both intestinal and hepatic first-pass metabolism for many drugs. Grapefruit juice irreversibly inhibits intestinal cytochrome P450 3A4, increasing bioavailability of sensitive substrates.

Cytochrome P450 2C9 metabolizes warfarin (the active S-enantiomer), many nonsteroidal anti-inflammatory drugs, phenytoin, and sulfonylureas. This isoform is particularly susceptible to competitive inhibition by fluconazole and amiodarone, which can dramatically potentiate warfarin anticoagulation. Cytochrome P450 2C19 metabolizes the proton pump inhibitors (omeprazole, pantoprazole, lansoprazole), clopidogrel (prodrug activation, discussed in Section 6), diazepam, and certain antidepressants. Cytochrome P450 2D6 metabolizes approximately 20 to 25 percent of commonly prescribed drugs, including codeine (prodrug activation to morphine), tramadol, most tricyclic antidepressants, most selective serotonin reuptake inhibitor antidepressants, and beta-blockers including metoprolol and carvedilol. Cytochrome P450 1A2 metabolizes theophylline, caffeine, clozapine, and some antidepressants, and is induced by cigarette smoking — a clinically relevant reason why plasma theophylline and clozapine concentrations fall when patients begin smoking and rise when they stop.

Phase one metabolism does not always inactivate a drug. Metabolites can be pharmacologically active (contributing to the overall drug effect), inactive (the desired outcome for elimination), or toxic (an unwanted and sometimes dangerous consequence of metabolism).

Acetaminophen is the paradigmatic example of toxic metabolite formation. At therapeutic doses, acetaminophen is metabolized primarily by glucuronidation and sulfation (phase two reactions) to non-toxic conjugates that are excreted in urine. A small fraction is oxidized by cytochrome P450 2E1 (and to a lesser extent cytochrome P450 3A4 and 1A2) to a reactive, highly electrophilic metabolite called N-acetyl-para-benzoquinone imine. Under normal circumstances, this toxic metabolite is immediately conjugated with glutathione in the liver, forming a harmless mercapturic acid conjugate. In overdose, or when glutathione stores are depleted (as in chronic alcohol use, malnutrition, or fasting), N-acetyl-para-benzoquinone imine accumulates, covalently binds hepatocyte proteins, and causes hepatocellular necrosis. The antidote N-acetylcysteine works by replenishing glutathione precursors, allowing conjugation of N-acetyl-para-benzoquinone imine before it injures hepatocytes.

A drug with low hepatic extraction is one where the liver's enzyme activity is far greater than needed to metabolize the drug presented to it at typical portal blood flow rates. In this situation, the liver clears only a small fraction of the drug per pass, and plasma concentrations after oral dosing are relatively close to those after parenteral dosing — because first-pass elimination is modest. The key feature of low-extraction drugs is that their clearance depends entirely on how efficiently the metabolizing enzymes can process the drug: enzyme activity (and therefore the free drug concentration available to the enzymes) is the rate-limiting factor. These drugs are therefore called capacity-limited or enzyme-limited.

For low-extraction drugs, enzyme inhibitors substantially reduce clearance and raise plasma concentrations. If an inhibitor halves the activity of the metabolizing cytochrome P450, plasma concentrations of a low-extraction substrate can double. Changes in hepatic blood flow, by contrast, have little effect on clearance of low-extraction drugs — the liver is already far from its capacity limit, and delivering more drug per minute does not change how much gets metabolized.

Examples of low-extraction drugs include warfarin, phenytoin, diazepam, tolbutamide, theophylline, and most proton pump inhibitors. The plasma protein binding of these drugs also influences effective clearance, since only the unbound fraction is accessible to hepatic enzymes.

A drug with high hepatic extraction is one where the liver's enzymes are so efficient that they metabolize nearly all of the drug presented during a single hepatic pass. The liver is not the bottleneck — delivery of drug to the liver is. Hepatic clearance of these drugs therefore approximates hepatic blood flow (approximately 1,500 milliliters per minute in a healthy adult), and their clearance is called flow-limited. Because the liver removes drug so efficiently, oral bioavailability of high-extraction drugs is low, as discussed in Module 1.

For high-extraction drugs, enzyme inhibitors have less effect than might be expected on clearance, because the enzyme was already operating far above the drug concentration needed for near-complete extraction — reducing enzyme activity modestly still leaves capacity for extensive extraction. However, changes in hepatic blood flow substantially affect the clearance of high-extraction drugs. Conditions that reduce hepatic blood flow — propranolol and other drugs that reduce cardiac output, portosystemic shunting in cirrhosis, or hypotension — decrease the delivery rate of drug to the liver and correspondingly reduce clearance, raising plasma concentrations. Conversely, conditions that increase hepatic blood flow, such as meals (which increase splanchnic blood flow), may increase the clearance of high-extraction drugs.

High-extraction drugs include lidocaine, morphine, propranolol, verapamil, most opioids, and many other drugs discussed in Module 1's first-pass section. In cirrhosis with significant portosystemic shunting, blood from the portal circulation bypasses the liver almost entirely before reaching systemic circulation — dramatically increasing the oral bioavailability of high-extraction drugs and requiring dose reduction.

Genetic polymorphisms in cytochrome P450 genes produce four phenotypic categories. Poor metabolizers have two non-functional alleles and essentially no enzymatic activity for that isoform. Intermediate metabolizers have one functional and one reduced-function or non-functional allele, producing enzyme activity between poor and normal. Extensive metabolizers (sometimes called normal metabolizers) have two functional alleles and represent the population average. Ultra-rapid metabolizers carry gene duplications that produce multiple copies of functional enzyme, metabolizing substrates much faster than the typical individual.

The clinical consequences of metabolizer phenotype depend on whether the drug is a substrate being cleared (in which case poor metabolizers have higher plasma concentrations and more toxicity) or a prodrug being activated (in which case poor metabolizers have less active drug produced and reduced efficacy, while ultra-rapid metabolizers generate more active drug and risk toxicity from it).

Cytochrome P450 2D6 is one of the most polymorphic cytochrome P450 genes, with more than 100 known alleles producing a wide spectrum of enzyme activity. Approximately 7 to 10 percent of European populations are poor metabolizers of cytochrome P450 2D6, while 1 to 2 percent are ultra-rapid metabolizers. The clinical consequences are starkest for codeine, which is a prodrug that requires cytochrome P450 2D6 to convert it to morphine.

In poor metabolizers of cytochrome P450 2D6, codeine is not converted to morphine, producing no meaningful analgesia. These patients receive only the side effects of codeine (constipation, nausea) without the therapeutic benefit and may be inappropriately labeled as opioid non-responders. In ultra-rapid metabolizers, codeine is converted to morphine at an accelerated rate, producing morphine plasma concentrations far above those expected from a standard codeine dose. This can cause respiratory depression, particularly in breastfed infants whose mothers are ultra-rapid metabolizers receiving codeine for postpartum pain — a scenario that led to infant deaths and resulted in the United States Food and Drug Administration issuing a black box warning against codeine use by breastfeeding mothers. The interaction between ultra-rapid cytochrome P450 2D6 metabolizer genotype and codeine use represents one of the most important examples of pharmacogenomics affecting patient safety.

Cytochrome P450 2D6 metabolizer status also affects dosing of many antidepressants (particularly tricyclic antidepressants and certain selective serotonin reuptake inhibitors) and antipsychotics. Poor metabolizers of cytochrome P450 2D6 taking tricyclic antidepressants are at substantially higher risk of cardiac arrhythmias and anticholinergic toxicity from drug accumulation at standard doses. Cytochrome P450 2D6 is also relevant for tamoxifen (discussed in Section 6).

Cytochrome P450 2C19 polymorphisms have two major clinical consequences. Clopidogrel, the antiplatelet prodrug used after coronary stent placement, requires cytochrome P450 2C19 to generate its active thienopyridine metabolite. Poor metabolizers of cytochrome P450 2C19 — approximately 2 to 3 percent of European populations and up to 15 to 20 percent of Asian populations — cannot adequately activate clopidogrel. Clinical studies show that these patients have significantly higher rates of major adverse cardiovascular events after percutaneous coronary intervention with stent placement. The United States Food and Drug Administration added a black box warning to clopidogrel noting reduced effectiveness in poor cytochrome P450 2C19 metabolizers, and the label recommends testing for genotype or considering alternative antiplatelet agents (prasugrel, ticagrelor) in high-risk patients.

For proton pump inhibitors metabolized by cytochrome P450 2C19, poor metabolizers actually have higher drug exposure and potentially greater acid suppression. This is generally clinically beneficial — poor metabolizers have higher eradication rates of Helicobacter pylori when using proton pump inhibitor-based regimens. Ultra-rapid metabolizers of cytochrome P450 2C19 may have reduced proton pump inhibitor efficacy and may require dose escalation for adequate acid suppression.

Warfarin's highly variable dose requirement across patients is partly explained by cytochrome P450 2C9 polymorphisms. The active S-enantiomer of warfarin is primarily metabolized by cytochrome P450 2C9, and variants with reduced enzyme activity (the most important being cytochrome P450 2C9*2 and cytochrome P450 2C9*3 alleles) lead to reduced warfarin clearance and markedly elevated anticoagulation at doses that would be appropriate for patients with normal activity. Poor metabolizers of cytochrome P450 2C9 require substantially lower warfarin doses to maintain the international normalized ratio within the therapeutic range, and they are at higher risk of serious bleeding during warfarin initiation if standard loading approaches are used. Commercial warfarin dosing algorithms now routinely incorporate cytochrome P450 2C9 genotype alongside cytochrome P450 vitamin K epoxide reductase complex subunit 1 (VKORC1) genotype, which determines warfarin's pharmacodynamic target sensitivity.