Drug metabolism

Drug interactions

Many clinically important drug interactions result from induction or inhibition of drug-metabolizing enzymes. For example, the prodrugs cyclophosphamide and ifosfamide are activated primarily by CYP2B6 and CYP3A4 into cytotoxic metabolites such as phosphoramide mustard and chloroacetaldehyde. Co-administration of strong CYP inducers such as phenytoin or rifampicin can accelerate bioactivation and increase the risk of toxicity, including severe myelosuppression and hemorrhagic cystitis.^[12][13][14]

Drug interactions are commonly evaluated relative to theoretical models of noninteraction, including Loewe additivity and Bliss independence.^[15][16][17] Interactions may be synergistic, producing greater-than-expected effects or toxicity, or antagonistic, reducing therapeutic efficacy.

Physiological factors

Physiological factors affecting drug metabolism include age, sex, diet, enterohepatic circulation, gut microbiota, and inherited genetic variation (pharmacogenetics).^[18][19]

Regulation by xenobiotic receptors

Expression of drug‑metabolizing enzymes and transporters is dynamically regulated by ligand‑activated xenobiotic‑sensing nuclear receptors, including the pregnane X receptor (PXR), constitutive androstane receptor (CAR), and aryl hydrocarbon receptor (AhR). When activated by therapeutic drugs and other xenobiotics, these receptors upregulate overlapping sets of Phase I and Phase II enzymes and efflux transporters, leading to enzyme induction, altered drug interaction profiles, and changes in clearance and exposure that can necessitate dose adjustment or alternative therapy.^[26][27] The broader roles of these receptors in coordinating xenobiotic defense across different organisms and chemical classes are described in xenobiotic metabolism#Xenobiotic sensing and regulation.

Other factors

Dose, frequency, route of administration, tissue distribution, and protein binding can also influence drug metabolism.^[28]

First-pass metabolism

First-pass metabolism (first-pass effect) refers to presystemic biotransformation, primarily in the gut wall and liver, that occurs after oral administration and reduces the fraction of active drug reaching the systemic circulation.^[32][33][36] The extent of first-pass metabolism influences oral bioavailability and may necessitate higher oral doses, alternative routes of administration, or specific formulation strategies.^[36][37]

Prodrugs

A prodrug is a pharmacologically inactive or less active compound that is designed to undergo enzymatic or chemical conversion in the body to release an active drug.^[32][38] Prodrug strategies exploit predictable metabolic pathways, including first-pass metabolism, to improve oral bioavailability, solubility, tissue targeting, or tolerability of therapeutic agents.^[38]

Active metabolites

Active metabolites are biotransformation products that retain pharmacological activity and can contribute substantially to both therapeutic and adverse effects of a parent drug.^[32][39] Recognition of clinically important active metabolites is critical for interpreting dose–response relationships, understanding prolonged or delayed effects, and anticipating toxicity in patients with impaired clearance or altered metabolism.^[39][40]

Drug toxicity

Drug metabolism can decrease toxicity through detoxification of xenobiotics but can also generate reactive or toxic metabolites that cause organ injury.^[32][41][42] Bioactivation to reactive intermediates, often mediated by cytochrome P450 enzymes, is an important mechanism underlying dose‑related and idiosyncratic hepatotoxicity, hypersensitivity reactions, and other forms of drug‑induced tissue damage.^[42]

Therapeutic index

The therapeutic index (TI) is commonly defined as the ratio of the median toxic dose (TD₅₀) to the median effective dose (ED₅₀).^[43] Metabolic clearance is a key determinant of the therapeutic index because it influences systemic exposure to parent drug and metabolites at a given dose.^[33][35] Alterations in drug metabolism caused by enzyme induction, enzyme inhibition, or genetic variation can therefore significantly change the therapeutic index of a drug, particularly for drugs with narrow safety margins. Reduced metabolism may increase drug concentrations and toxicity, whereas accelerated metabolism may reduce efficacy. Interindividual differences and drug–drug interactions that alter metabolic clearance can narrow or widen a drug’s effective and safe concentration range, sometimes necessitating dose adjustment or therapeutic drug monitoring for medicines with a narrow therapeutic index.^[35]

Personalized medicine

Variability in drug‑metabolizing enzymes provides a major rationale for personalized medicine approaches such as pharmacogenetic testing and individualized dosing regimes.^[32][44] Genotype‑guided selection and dosing of drugs metabolized by polymorphic enzymes, for example CYP2D6, CYP2C19 or thiopurine S‑methyltransferase, can improve efficacy and reduce the risk of serious adverse drug reactions.^[45]

Multidrug resistance

Multidrug resistance is the ability of tumour cells, microorganisms, or other pathological cells to resist the effects of multiple structurally and functionally distinct drugs, often through mechanisms that reduce intracellular concentrations of active drug.^[46][47] Altered expression or function of drug-metabolizing enzymes and associated transporters can contribute to multidrug resistance by lowering intracellular concentrations of active drug.^[32][47] In oncology and infectious diseases, induction of metabolic pathways or efflux transporters in tumour or microbial cells, as well as host variability in drug metabolism, can reduce exposure to active drug and compromise treatment outcomes.^[46][47]

Prodrug activation: clopidogrel

Clopidogrel is a thienopyridine antiplatelet prodrug that requires oxidative bioactivation by hepatic cytochrome P450 enzymes, particularly CYP2C19, to form an active thiol metabolite that irreversibly inhibits the P2Y12 receptor on platelets.^[48][49] Reduced CYP2C19 function, due to loss-of-function genetic polymorphisms or concomitant use of CYP2C19 inhibitors, decreases the formation of the active metabolite and attenuates the antiplatelet effect.^[48][49] Such variability in clopidogrel metabolism has been associated with increased rates of thrombotic events, particularly stent thrombosis and myocardial infarction, and has prompted consideration of alternative P2Y12 inhibitors or genotype-guided therapy in some clinical settings.^[49][50]

Prodrug activation: codeine

Codeine is a prodrug that requires O-demethylation by cytochrome P450 2D6 (CYP2D6) to form morphine, which contributes substantially to its analgesic effect.^[51][52] Individuals who are CYP2D6 poor metabolizers may obtain little or no analgesia from standard codeine doses, whereas ultrarapid metabolizers can produce higher morphine concentrations and are at increased risk of opioid toxicity, including serious respiratory depression.^[51][53][52] Concerns about variable metabolism and reports of fatal toxicity in children, particularly in CYP2D6 fast metabolizers, have led regulatory agencies and professional societies to recommend limiting or avoiding the use of codeine in paediatric patients and to favor alternative analgesic strategies.^[54][55]

Food–drug interaction: grapefruit juice

Grapefruit juice contains furanocoumarins and related compounds that inhibit cytochrome P450 3A4 (CYP3A4) in the intestinal wall, leading to reduced first-pass metabolism of susceptible drugs.^[56] For oral medications that undergo extensive intestinal CYP3A4-mediated presystemic metabolism, co-administration with grapefruit juice can markedly increase systemic exposure and thereby heighten the risk of concentration-dependent adverse effects.^[56] This interaction illustrates the clinical importance of enzyme inhibition and first-pass metabolism in determining the bioavailability and safety of many drugs.^[56]

Enzyme induction: rifampicin

Rifampicin is a rifamycin antibiotic that acts as a potent inducer of several drug-metabolizing enzymes and transporters, including cytochrome P450 3A4 (CYP3A4) and P-glycoprotein, predominantly through activation of the pregnane X receptor (PXR).^[57][58][59] Induction by rifampicin can increase the clearance and reduce the plasma concentrations of many co-administered drugs, including oral contraceptives, anticoagulants, and certain antiretroviral agents, and has been associated with loss of efficacy and therapeutic failure.^[57] This interaction illustrates the clinical relevance of enzyme and transporter induction in altering drug exposure and may necessitate dose adjustment or the selection of alternative therapy.^[57][58]

Reactive metabolite toxicity: acetaminophen

At therapeutic doses, acetaminophen (paracetamol) is primarily cleared by hepatic conjugation pathways, including glucuronidation and sulfation, which produce non-toxic, water-soluble metabolites.^[60][61] A small fraction is oxidized by cytochrome P450 enzymes, particularly CYP2E1, to form N-acetyl-p-benzoquinone imine (NAPQI), a highly reactive intermediate that is normally detoxified by conjugation with glutathione.^[60][61] In overdose or in conditions that deplete glutathione, the accumulation of NAPQI leads to covalent binding to cellular proteins, mitochondrial oxidative stress, and centrilobular hepatocellular injury that can progress to acute liver failure.^[60][61] This example illustrates how reactive metabolites generated during drug metabolism can be central mediators of toxicity.^[61]

Hepatic extraction and clearance

Drug metabolism is a primary determinant of systemic drug exposure, quantified through pharmacokinetic parameters including area-under-the-curve (AUC), clearance, and elimination half-life. The hepatic extraction ratio (E_H) describes the fraction of drug irreversibly removed during a single pass through the liver and ranges from 0 (no extraction) to 1 (complete extraction). This ratio is determined by the relationship between intrinsic clearance (CL_int, the liver's enzymatic capacity to metabolize drug) and hepatic blood flow (Q_H, approximately 90 L/hour in humans).^[64]

The hepatic extraction ratio can be defined experimentally as: $${\displaystyle E_{H}={\frac {C_{in}-C_{out}}{C_{in}}}}$$ where ${\displaystyle E_{H}}$ is the hepatic extraction ratio, ${\displaystyle C_{in}}$ is the drug concentration entering the liver, and ${\displaystyle C_{out}}$ is the drug concentration leaving the liver.

According to the well-stirred model of hepatic clearance, hepatic extraction ratio can be predicted from physiological and drug-specific parameters as: $${\displaystyle E_{H}={\frac {f_{u}\cdot CL_{int}}{Q_{H}+f_{u}\cdot CL_{int}}}}$$ where ${\displaystyle f_{u}}$ is the unbound fraction of drug in blood, ${\displaystyle CL_{int}}$ is intrinsic hepatic metabolic clearance, and ${\displaystyle Q_{H}}$ is hepatic blood flow.

For drugs with high extraction ratios (E_H > 0.7), hepatic clearance becomes blood flow-limited and largely independent of enzyme activity or protein binding; conversely, drugs with low extraction ratios (E_H < 0.3) exhibit clearance limited by intrinsic metabolic capacity rather than perfusion. This distinction has critical implications for first-pass metabolism following oral administration, where hepatic bioavailability may be approximated as 1 − E_H for drugs that are completely absorbed and undergo negligible intestinal first-pass metabolism.^[65]

Physiologically-based pharmacokinetic modeling

Physiologically based pharmacokinetic modelling (PBPK) modeling integrates drug-specific properties (molecular weight, lipophilicity, protein binding) with physiological parameters (organ volumes, blood flows, metabolic enzyme abundances) to mechanistically predict drug disposition in diverse patient populations.^[66] By incorporating tissue-specific expression of drug-metabolizing enzymes (cytochrome P450s, conjugating enzymes) and transporter proteins, PBPK models simulate how metabolic pathways influence systemic exposure, tissue distribution, and elimination across different patient populations.^[67][68] These models enable prediction of untested clinical scenarios including metabolism-mediated drug-drug interactions, effects of hepatic impairment on metabolic clearance, and age-related changes in enzyme activity affecting pediatric dosing without conducting dedicated clinical trials.^[69]

PBPK models commonly represent organs as interconnected compartments governed by mass-balance equations describing drug transport and partitioning between blood and tissues:^[70]

$${\displaystyle V_{T}{\frac {dC_{T}}{dt}}=Q_{T}C_{A}-Q_{T}C_{V,T}}$$

where ${\displaystyle V_{T}}$ is tissue volume, ${\displaystyle C_{T}}$ is tissue drug concentration, ${\displaystyle Q_{T}}$ is tissue blood flow, ${\displaystyle C_{A}}$ is arterial drug concentration, and ${\displaystyle C_{V,T}}$ is venous drug concentration exiting the tissue.

In vitro to in vivo extrapolation

In vitro to in vivo extrapolation (IVIVE) scales metabolic stability data from human liver microsomes or hepatocytes to predict human clearance. Intrinsic clearance measured in vitro is scaled using physiological factors (microsomal protein per gram liver, liver weight) and empirical correction factors to account for differences between simplified in vitro systems and intact liver function. While microsomal data with physiological scaling factors consistently underpredict in vivo clearance, empirical scaling factors (typically 5- to 10-fold corrections) substantially improve prediction accuracy.^[71][72][73]

Intrinsic clearance measured in vitro can be scaled to whole-liver intrinsic clearance as:^[73]

$${\displaystyle CL_{int,H}=CL_{int,in\ vitro}\cdot MPPGL\cdot LW}$$

${\displaystyle CL_{int,H}}$ is whole-liver intrinsic clearance, ${\displaystyle CL_{int,in\ vitro}}$ is intrinsic clearance measured in vitro, ${\displaystyle MPPGL}$ is microsomal protein per gram liver, and ${\displaystyle LW}$ is liver weight.

Predicted hepatic clearance can then be estimated using the well-stirred model:^[71]

$${\displaystyle CL_{H}=Q_{H}\cdot {\frac {f_{u}\cdot CL_{int,H}}{Q_{H}+f_{u}\cdot CL_{int,H}}}}$$

${\displaystyle CL_{H}}$ is hepatic clearance, ${\displaystyle Q_{H}}$ is hepatic blood flow, ${\displaystyle f_{u}}$ is the unbound fraction of drug in blood, and ${\displaystyle CL_{int,H}}$ is whole-liver intrinsic clearance.

Regulatory applications

PBPK models incorporating drug metabolism parameters are increasingly accepted by regulatory agencies including the FDA for supporting clinical pharmacology decisions, justifying clinical study designs, and waiving studies in special populations when adequately validated.^[74][75][76] Common metabolism-related regulatory applications include predicting drug-drug interactions involving metabolic enzyme inhibition or induction, assessing the impact of genetic polymorphisms in drug-metabolizing enzymes on exposure, and evaluating dose adjustments needed in patients with hepatic impairment affecting metabolic capacity.^[74][76] The FDA issued guidance in September 2018 outlining format and content expectations for PBPK analyses submitted to support drug applications.^[74] PBPK model predictions can be used to support decisions on whether, when, and how to conduct certain clinical pharmacology studies, particularly those examining metabolic pathways and their variability, and to support dosing recommendations in product labeling, with acceptance determined on a case-by-case basis considering the quality, relevance, and reliability of the metabolic and pharmacokinetic analyses.^[74][77]