The kidney is the primary organ of elimination for the majority of hydrophilic drugs and for water-soluble metabolites of lipophilic drugs. Renal drug elimination involves three processes operating simultaneously at different nephron sites: glomerular filtration, tubular secretion, and tubular reabsorption. Understanding each process individually and their combined contribution to net renal drug excretion is essential for dosing drugs safely in patients with renal impairment.
Glomerular filtration is the passive, non-selective process by which blood is ultrafiltered across the glomerular basement membrane into Bowman's capsule. The glomerular filtration rate (GFR) represents the volume of plasma filtered per unit time and is approximately 120 to 125 mL/min in a healthy young adult. Only the unbound (free) fraction of drug in plasma is filtered at the glomerulus; protein-bound drug cannot pass through the glomerular basement membrane because the protein-drug complex is too large. For a drug that is 95% protein-bound, only 5% of the plasma drug concentration is available for filtration, substantially limiting the contribution of glomerular filtration to that drug's total renal clearance. The filtered load of drug per unit time equals the product of the free plasma drug concentration and the GFR: filtered load = fu × Cplasma × GFR.1
Clinical Estimation of GFR. Direct measurement of GFR requires exogenous markers such as inulin or iohexol, which are not practical in routine clinical care. Endogenous serum creatinine is used to estimate GFR through validated equations. The Cockcroft-Gault equation estimates creatinine clearance (CrCl) as: CrCl (mL/min) = ((140 − age) × weight in kg) / (72 × serum creatinine in mg/dL), multiplied by 0.85 for females. This equation has historically been used for drug dosing adjustments because most pharmacokinetic studies validating dose adjustment guidelines used Cockcroft-Gault to characterize their patient populations; drug labeling dose adjustment tables therefore specifically reference Cockcroft-Gault CrCl rather than the more modern chronic kidney disease epidemiology collaboration (CKD-EPI) equation. The CKD-EPI equation provides superior accuracy for estimating GFR across the full range of kidney function, particularly at higher GFR values, and is preferred for clinical staging of chronic kidney disease (CKD); however, for drug dosing, Cockcroft-Gault remains the standard specified in most regulatory guidance and product labeling.12
Limitations of Serum Creatinine. Serum creatinine is an imperfect GFR marker because its generation depends on muscle mass, and its plasma concentration reflects the balance between production and filtration. Several clinical situations cause serum creatinine to misrepresent GFR. In elderly, malnourished, or paraplegic patients with markedly reduced muscle mass, serum creatinine may be deceptively low (for example, 0.6 mg/dL), suggesting preserved GFR when actual GFR is substantially reduced. Applying Cockcroft-Gault to such patients without accounting for reduced muscle mass underestimates renal impairment and risks drug accumulation and toxicity. Conversely, patients with very high muscle mass (bodybuilders, athletes) may have elevated serum creatinine without true GFR reduction. Acute changes in kidney function also lag behind creatinine: in acute kidney injury (AKI), creatinine takes 24 to 48 hours to rise appreciably even when GFR has fallen precipitously, and the creatinine at any given moment reflects the GFR of the preceding 24 to 48 hours rather than the current moment. In the recovery phase of AKI, falling creatinine may lag behind improving GFR, making drug dose adjustments based on rising creatinine clearance appropriately gradual. Cystatin C, a low-molecular-weight protein filtered freely at the glomerulus and not secreted or reabsorbed, is a more accurate GFR marker in patients with extreme muscle mass abnormalities and is increasingly used in clinical practice.2
Determinants of Filtration-Dependent Clearance. For drugs eliminated primarily by glomerular filtration with minimal tubular secretion or reabsorption, renal clearance (CLrenal) approximates fu × GFR. Changes in protein binding therefore directly alter filtration-dependent clearance: hypoalbuminemia increases the free fraction of highly bound drugs, increasing the filtered load and potentially increasing renal clearance of the free drug, though total drug clearance effects depend on whether hepatic clearance is also affected. For low-molecular-weight drugs filtered predominantly at the glomerulus, the ratio of renal clearance to GFR indicates the net effect of tubular processes: a ratio of 1 indicates pure filtration; a ratio greater than 1 indicates net tubular secretion; a ratio less than 1 indicates net tubular reabsorption.1
Only free (unbound) drug is filtered: 99% protein-bound drug has only 1% available for glomerular filtration. GFR estimation: use Cockcroft-Gault for drug dosing (matches regulatory labels); CKD-EPI for CKD staging. Beware low serum creatinine in elderly/malnourished: 0.6 mg/dL in an 80-year-old woman weighing 45 kg gives a CrCl of approximately 20 mL/min, not the 90+ mL/min a young person with the same creatinine would have. Acute kidney injury: creatinine lags GFR by 24–48 hours; dose-adjust based on trajectory, not just current creatinine.
Glomerular filtration alone cannot fully account for the renal elimination kinetics of many drugs. Active tubular secretion can drive renal clearance far above the filtration rate, allowing the kidney to eliminate even highly protein-bound drugs efficiently. Conversely, passive tubular reabsorption can dramatically reduce the net urinary excretion of lipophilic drugs, regardless of how much is filtered. The interplay between these processes determines whether a drug's renal clearance exceeds, equals, or falls short of the glomerular filtration rate (GFR).
Active Tubular Secretion. Active tubular secretion occurs in the proximal tubule and is mediated by transport proteins expressed on both the basolateral (blood-facing) and apical (lumen-facing) membranes of tubular epithelial cells. The two most clinically important secretory systems are the organic anion transporter (OAT) system, primarily the organic anion transporter 1 (OAT1) and organic anion transporter 3 (OAT3) isoforms on the basolateral membrane, and the organic cation transporter (OCT) system, primarily the organic cation transporter 2 (OCT2) isoform on the basolateral membrane. On the apical membrane, the multidrug resistance protein 4 (MRP4), multidrug and toxin extrusion transporter 1 (MATE1), and MATE2-K (multidrug and toxin extrusion transporter 2-K) export drugs into the tubular lumen. The OAT system secretes acidic (anionic) drugs including penicillins, cephalosporins, methotrexate, furosemide, and most non-steroidal anti-inflammatory drugs (NSAIDs). The OCT system secretes basic (cationic) drugs including metformin, cimetidine, ranitidine, procainamide, and trimethoprim. Because tubular secretion is an active, carrier-mediated process, it can secrete both free and, indirectly, protein-bound drug: as free drug is secreted into the lumen, the reduction in free drug concentration drives dissociation of drug from plasma protein, which then becomes available for secretion. Drugs with high tubular secretion can therefore have renal clearances substantially higher than GFR, sometimes approaching or exceeding total renal plasma flow (approximately 660 mL/min).3
Drug-Drug Interactions at Tubular Secretion Sites. Competition between two drugs for the same tubular secretion transporter reduces the renal elimination of both, with the higher-affinity drug disproportionately inhibiting the secretion of the lower-affinity drug. The classic pharmacokinetic interaction between probenecid and penicillin, exploited historically to prolong penicillin's duration of action by blocking OAT-mediated tubular secretion, is the founding example. Clinically important contemporary examples include the inhibition of renal OCT2 (organic cation transporter 2) and MATE (multidrug and toxin extrusion) transporters by cobicistat and ritonavir (human immunodeficiency virus (HIV) pharmacokinetic boosters), which reduce the renal secretion of metformin and creatinine (the latter causing a rise in serum creatinine that does not reflect true GFR reduction, a source of clinical confusion). Trimethoprim inhibits OCT2-mediated creatinine secretion as well, accounting for the predictable and clinically benign rise in serum creatinine seen with trimethoprim-containing antibiotics in patients with normal kidney function. NSAIDs compete with methotrexate for OAT-mediated tubular secretion and can dramatically increase methotrexate concentrations, a potentially life-threatening interaction particularly in patients receiving high-dose methotrexate chemotherapy.34
Tubular Reabsorption. Once filtered or secreted into the tubular lumen, a drug may be reabsorbed across the tubular epithelium back into the peritubular capillaries. Passive reabsorption is the dominant mechanism for lipophilic drugs and follows the same physicochemical principles as gastrointestinal absorption: un-ionized, lipophilic, low-molecular-weight molecules cross tubular epithelial membranes by passive diffusion, while ionized or polar molecules remain in the tubular fluid and are excreted. The extent of passive reabsorption therefore depends directly on the drug's pKa and the pH of the tubular fluid. For weak acids, alkalinization of the urine (raising urine pH) increases the ionized fraction in the tubular lumen, reducing passive reabsorption and increasing urinary drug excretion. For weak bases, acidification of the urine increases ionization and thereby reduces reabsorption. This pH-dependent reabsorption is the pharmacokinetic basis for urinary alkalinization with sodium bicarbonate in the management of salicylate overdose: raising urine pH to above 7.5 substantially increases salicylate ionization in the tubular lumen, trapping it there and accelerating its urinary elimination by five- to ten-fold compared to acidic urine. The same principle applies to phenobarbital overdose management with urinary alkalinization.45
Active Tubular Reabsorption. Several physiologically important endogenous substances are actively reabsorbed from the tubular fluid, and certain drugs exploit these same transporters to achieve tubular reabsorption. Uric acid is reabsorbed by the urate transporter URAT1 (SLC22A12) in the proximal tubule; low-dose aspirin inhibits URAT1-mediated uric acid secretion (paradoxically increasing serum uric acid) while high-dose aspirin blocks reabsorption and is uricosuric. Sodium-glucose cotransporter 2 (SGLT2) inhibitors (dapagliflozin, empagliflozin, canagliflozin) block active glucose reabsorption in the proximal tubule, a clinically exploited mechanism for the treatment of type 2 diabetes mellitus (DM) and heart failure. Active tubular reabsorption of drugs by nutrient transporters is relevant for certain nucleoside analogs and amino acid-derived prodrugs that exploit these pathways.3,5
NSAIDs + methotrexate: OAT competition reduces methotrexate secretion → potentially fatal toxicity; avoid in high-dose MTX therapy. Trimethoprim + creatinine: OCT2 inhibition raises serum creatinine without true GFR reduction; do not adjust drug doses based on this creatinine rise. Ritonavir/cobicistat + metformin: MATE inhibition reduces metformin secretion; monitor for lactic acidosis. Probenecid + penicillin: historical use to prolong penicillin; now rarely employed deliberately. Salicylate overdose: sodium bicarbonate to alkalinize urine to pH >7.5 accelerates elimination 5–10×.
Renal impairment reduces the clearance of drugs eliminated by the kidney, causing drug and active metabolite accumulation at standard doses. The relationship between degree of renal impairment and required dose adjustment is not linear for all drugs, and the clinical approach must account for the therapeutic index of the drug, the contribution of renal elimination to total clearance, the pharmacological activity of any renally excreted metabolites, and the clinical consequences of both under- and over-dosing.
Chronic kidney disease is staged by glomerular filtration rate (GFR): stage G1 (GFR above 90 mL/min/1.73 m2, normal or mildly increased), stage G2 (GFR 60 to 89, mildly decreased), stage G3a (GFR 45 to 59, mildly to moderately decreased), stage G3b (GFR 30 to 44, moderately to severely decreased), stage G4 (GFR 15 to 29, severely decreased), and stage G5 (GFR below 15, kidney failure, including patients on dialysis). Drug dose adjustments are generally not required until GFR falls below 60 mL/min for most drugs, and become progressively more important as GFR falls below 30 mL/min. However, for drugs with narrow therapeutic indices (NTIs), adjustments may be warranted at milder degrees of impairment, and therapeutic drug monitoring (TDM) is essential whenever GFR falls significantly below normal. The fraction of drug dose that requires reduction for a given level of renal impairment can be estimated as: dose fraction = 1 − fe(1 − KF), where fe is the fraction of drug eliminated unchanged in urine and KF is the ratio of the patient's CrCl to the normal CrCl (approximately 120 mL/min).12
Dose Reduction Versus Interval Extension. Two general strategies are used to adjust dosing in renal impairment: reducing the dose while maintaining the normal dosing interval, or maintaining the normal dose while extending the dosing interval. The choice between strategies depends on the pharmacodynamic target of the drug. For drugs where efficacy depends on maintaining plasma concentrations above a minimum effective concentration (MEC) throughout the dosing interval (time-dependent antibiotics such as beta-lactams, vancomycin, and antiepileptics), dose reduction with unchanged interval is preferred to avoid troughs falling below the MEC. For drugs where efficacy depends on peak concentration (concentration-dependent antibiotics such as aminoglycosides) or where toxicity is related to trough concentrations, interval extension may be more appropriate, allowing adequate peaks and drug-free periods. In practice, combinations of both strategies are often applied. Aminoglycosides in renal failure provide a clear example: the once-daily dosing strategy used in normal renal function must be modified in severe renal impairment (CrCl below 20 mL/min), where extended intervals of 48 to 72 hours or more are used, with TDM of both peak and trough concentrations guiding subsequent doses.25
Active Metabolite Accumulation. Renal impairment can cause toxicity through accumulation of pharmacologically active or toxic metabolites that are renally eliminated, even when the parent drug itself is not significantly renally excreted. Morphine is primarily metabolized by the liver, but its active metabolite morphine-6-glucuronide (M6G), which is more potent at mu-opioid receptors than morphine itself, is renally eliminated. In patients with renal failure, M6G accumulates and can cause prolonged, potentially fatal opioid toxicity. Normeperidine, the active demethylated metabolite of meperidine (pethidine), accumulates in renal failure and is a central nervous system (CNS) neurostimulant that causes tremor, myoclonus, and seizures; standard meperidine doses that are safe in normal renal function can cause status epilepticus in patients with significant chronic kidney disease (CKD). Gabapentin is almost entirely renally eliminated and accumulates in CKD, requiring substantial dose reduction and extended intervals, with severe CNS depression reported at standard doses in patients with end-stage renal disease (ESRD). Allopurinol's active metabolite oxypurinol also accumulates in CKD, increasing the risk of allopurinol hypersensitivity syndrome.24
Drug Removal by Dialysis. Hemodialysis (HD) and continuous renal replacement therapy (CRRT) can remove drugs from the blood if the drug's physicochemical properties are favorable. Dialyzability is predicted by a small volume of distribution (Vd below 1 L/kg, ensuring most drug is in plasma rather than sequestered in tissues), low protein binding (ensuring the drug is free in plasma and can pass across the dialysis membrane), low molecular weight (allowing passage across most dialysis membranes; high-flux membranes accommodate larger molecules), and water solubility (lipophilic drugs distribute into tissues and are not efficiently removed by dialysis). Lithium (Vd approximately 0.7 L/kg, low protein binding, low molecular weight) is effectively removed by dialysis, making HD a therapeutic option in severe lithium toxicity. Digoxin (Vd approximately 7 L/kg, extensive tissue binding) is not effectively dialyzed despite its low molecular weight. The practical implication is that drugs dialyzable by HD require supplemental dosing after each HD session to replace drug removed during dialysis; reference guides such as Aronoff's Drug Prescribing in Renal Failure provide specific supplemental dose recommendations for commonly encountered drugs.25
Morphine: M6G (morphine-6-glucuronide) accumulates in CKD → prolonged respiratory depression; use fentanyl or hydromorphone instead (better renal safety profiles). Meperidine: normeperidine accumulates → seizures; avoid in CKD, contraindicated in ESRD. Gabapentin: renally eliminated; dose-reduce by 50% for CrCl 30–59, by 75% for CrCl 15–29; use with extreme caution in ESRD. LMWH (low-molecular-weight heparins): renally eliminated; monitor anti-Xa levels in CrCl <30 mL/min; consider unfractionated heparin. NSAIDs: reduce renal blood flow via prostaglandin inhibition; worsen CKD acutely; avoid or use with extreme caution in CKD.
While the kidney is the primary excretory organ for hydrophilic small molecules, the liver serves as the primary elimination route for many large, lipophilic, or conjugated drugs through biliary excretion. Understanding which drugs rely substantially on biliary elimination helps predict the consequences of hepatic disease and biliary obstruction on drug accumulation, and explains the pharmacokinetic basis of enterohepatic recirculation.
Biliary excretion is the process by which the liver actively transports drugs and metabolites from the hepatocyte into the bile canaliculus, from which they flow with bile into the duodenum. The rate-limiting step in biliary excretion is the active transport across the canalicular membrane of the hepatocyte, mediated by a set of adenosine triphosphate (ATP)-binding cassette (ABC) transporters. The multidrug resistance protein 2 (MRP2, encoded by ABCC2) is the primary canalicular transporter for anionic drug conjugates (glucuronides, glutathione conjugates, sulfates). The bile salt export pump (BSEP, encoded by ABCB11) handles bile salts. P-glycoprotein (P-gp, encoded by ABCB1) on the canalicular membrane exports certain amphipathic, lipophilic drugs including digoxin, vincristine, and certain human immunodeficiency virus (HIV) protease inhibitors. The breast cancer resistance protein (BCRP, encoded by ABCG2) exports sulfate conjugates and certain chemotherapeutic agents. Inhibition of MRP2 or other canalicular transporters can cause cholestatic drug reactions: the drug or its metabolites accumulate within hepatocytes and in bile, causing hepatocellular and cholestatic liver injury.6
Molecular Weight Threshold for Biliary Excretion. There is a well-established molecular weight threshold above which drugs are preferentially excreted in bile rather than urine. In humans, this threshold is approximately 400 to 500 Daltons: drugs with molecular weights above approximately 500 Da are more likely to be excreted in bile, while smaller molecules are more readily filtered and excreted by the kidney. This relationship is not absolute, as charge, lipophilicity, and transporter substrate specificity also govern the route of excretion, but molecular weight is a useful initial predictor. Glucuronide conjugates, which add approximately 176 Daltons to the parent drug, may push drugs across this threshold: a drug with a molecular weight of 350 Da that undergoes glucuronidation produces a conjugate of approximately 526 Da, which will be preferentially excreted in bile rather than urine. This is the basis for the biliary excretion of morphine glucuronides, irinotecan's 7-ethyl-10-hydroxycamptothecin (SN-38) glucuronide, and numerous steroid hormone conjugates.67
Hepatic Disease and Biliary Obstruction. Hepatic disease disrupts biliary drug excretion through several mechanisms. In cirrhosis, reduced hepatocyte functional mass and downregulation of canalicular transporters reduce biliary drug secretion, causing accumulation of drugs that rely substantially on biliary elimination. Cholestasis, whether intrahepatic or extrahepatic (obstructive), impairs bile flow and causes accumulation of bile acids and biliary-excreted drugs in the liver and systemic circulation. Drugs that are primarily eliminated by biliary excretion require dose reduction or substitution in patients with significant cholestatic liver disease. Rifampicin, a drug predominantly eliminated via biliary excretion of its desacetyl and glucuronide metabolites, reaches markedly elevated plasma concentrations in cholestatic patients. Rosuvastatin, unlike most statins, depends substantially on hepatic uptake by organic anion transporting polypeptide 1B1 (OATP1B1) and biliary excretion for its elimination; OATP1B1 inhibition by cyclosporine dramatically increases rosuvastatin plasma concentrations by reducing hepatic uptake and biliary clearing, a pharmacokinetic interaction that can cause rhabdomyolysis.36
Molecular weight above ~500 Da: expect biliary rather than renal excretion. Glucuronide conjugates: typically excreted in bile; hydrolysis by gut bacteria regenerates parent drug (enterohepatic recirculation). Cholestasis: accumulation of biliary-excreted drugs; dose-reduce or substitute in significant cholestatic disease. OATP1B1 inhibitors (cyclosporine, gemfibrozil, rifampicin): impair hepatic drug uptake for statins and other OATP substrates; risk of statin myopathy from cyclosporine co-administration. MRP2 inhibitors (some NSAIDs, probenecid): may reduce biliary conjugate excretion; monitor for drug accumulation.
Enterohepatic recirculation is a pharmacokinetic phenomenon in which a drug or its metabolite, excreted in bile into the intestinal lumen, is hydrolyzed by intestinal bacterial enzymes and reabsorbed across the intestinal mucosa, re-entering the systemic circulation. The net effect is prolongation of the drug's effective half-life and duration of action beyond what would be predicted from its primary elimination kinetics alone. Understanding enterohepatic recirculation is essential for interpreting the prolonged pharmacological effects of several clinically important drugs and for predicting drug-drug interactions involving disruption of gut flora.
The cycle of enterohepatic recirculation proceeds through the following sequence. A drug or its conjugated metabolite is excreted in bile from hepatocytes into the bile canaliculus, flows through the biliary tree, and enters the duodenum at the sphincter of Oddi. In the ileum and colon, intestinal bacteria express beta-glucuronidase and sulfatase enzymes that hydrolyze glucuronide and sulfate conjugates, regenerating the free, lipophilic parent drug (the aglycone). The regenerated parent drug is absorbed across the intestinal mucosa, enters the portal circulation, and returns to the systemic circulation. This cycle can repeat multiple times before final elimination, effectively creating a reservoir of drug in the enterohepatic circuit that delays ultimate elimination. The plasma concentration-time profile of drugs undergoing enterohepatic recirculation typically shows one or more secondary peaks in plasma concentration 4 to 12 hours after an oral or IV dose, corresponding to cycles of reabsorption from the gut. These secondary peaks may be mistaken for slow or delayed absorption or for drug-drug interactions.78
Clinically Important Drugs Undergoing Enterohepatic Recirculation. Estrogens are the most clinically relevant example. Estradiol and synthetic estrogens in oral contraceptives are conjugated in the liver and excreted in bile as glucuronide and sulfate conjugates; bacterial beta-glucuronidase in the gut regenerates free estrogen, which is reabsorbed and contributes to sustained systemic estrogen exposure. Disruption of gut flora by broad-spectrum oral antibiotics reduces intestinal beta-glucuronidase activity, theoretically reducing enterohepatic recirculation of estrogens and thereby reducing oral contraceptive efficacy. This interaction has been debated extensively; clinical evidence for a significant reduction in contraceptive efficacy from most antibiotics is limited and inconsistent, and current guidance from major reproductive health organizations does not recommend routine backup contraception for most antibiotic courses. However, rifampicin is a special case, because it both disrupts enterohepatic recirculation and potently induces hepatic CYP3A4 (cytochrome P450 3A4)-mediated estrogen metabolism, producing documented contraceptive failure; backup contraception is mandatory during rifampicin therapy.89
Other Drugs with Clinically Significant Enterohepatic Recirculation. Mycophenolate mofetil (MMF), after hydrolysis to mycophenolic acid (MPA) in the intestinal wall and plasma, undergoes hepatic glucuronidation to MPA glucuronide (MPAG), which is excreted in bile. Intestinal beta-glucuronidase regenerates MPA, which is reabsorbed, producing a characteristic secondary plasma MPA peak at 6 to 12 hours post-dose. Concomitant oral antibiotics that disrupt gut flora can reduce MPA re-absorption from enterohepatic recirculation, reducing total MPA exposure by 10 to 40% and potentially increasing transplant rejection risk; this pharmacokinetic effect explains why monitoring MPA area under the plasma concentration-time curve (AUC) is recommended when broad-spectrum antibiotics are added to or removed from the regimen of a transplant patient on MMF. Methotrexate undergoes enterohepatic recirculation of its polyglutamylated metabolites, contributing to its prolonged antifolate activity and to the cyclical nature of its gastrointestinal (GI) toxicity. Cholestyramine and colestipol, bile acid sequestrant resins, interrupt enterohepatic recirculation by binding bile acids in the intestinal lumen and preventing their reabsorption; they also bind other drugs excreted in bile, including levothyroxine, warfarin, digoxin, and fat-soluble vitamins, reducing their effective oral bioavailability and systemic exposure.9
Therapeutic Exploitation of Enterohepatic Recirculation Interruption. The ability of cholestyramine to interrupt enterohepatic recirculation is therapeutically exploited in the management of leflunomide toxicity. Leflunomide, an immunosuppressant used in rheumatoid arthritis, has an active metabolite (teriflunomide) with a very long half-life of 1 to 4 weeks due to extensive enterohepatic recirculation. When leflunomide must be rapidly eliminated from the body (for example, before conception in women of childbearing age, or in cases of severe adverse effects), the standard washout procedure involves administering cholestyramine 8 grams three times daily for 11 days, which interrupts enterohepatic recirculation and accelerates elimination, reducing the washout time from weeks to months down to approximately 11 days, verified by plasma teriflunomide concentrations below 0.02 mg/L on two separate measurements at least 14 days apart. Activated charcoal can serve a similar role in some drug toxicity settings by adsorbing drugs secreted into the intestinal lumen, preventing their reabsorption and creating a continuous gut elimination pathway even for drugs administered intravenously.9
Secondary plasma concentration peaks 4–12 hours after dose: suspect enterohepatic recirculation. Oral contraceptives + rifampicin: documented failure; backup contraception mandatory; rifampicin also induces hepatic estrogen metabolism. Oral contraceptives + other antibiotics: theoretical interaction; evidence for clinically significant failure is limited; backup not mandated by most guidelines except as precaution. MMF in transplant patients: broad-spectrum antibiotics reduce MPA AUC by 10–40%; monitor if adding or stopping antibiotics. Leflunomide washout: cholestyramine 8 g TID for 11 days interrupts enterohepatic recirculation of teriflunomide; confirm elimination by plasma level below 0.02 mg/L.
Beyond renal and hepatic-biliary elimination, drugs may be excreted by several additional routes whose clinical relevance ranges from directly therapeutic (pulmonary elimination of volatile anesthetics) to safety-critical (breast milk transfer to nursing infants) to occasionally diagnostically useful (salivary drug monitoring). Understanding these alternative routes is important for managing drug therapy in specific patient populations and for interpreting pharmacokinetic data appropriately.
Pulmonary Elimination. Volatile anesthetic gases and vapors are eliminated primarily through the lungs by the reverse of their absorption process. Exhaled anesthetic concentration decreases as blood and tissue concentrations fall toward zero following discontinuation of the inhaled agent. The rate of pulmonary elimination depends on the blood-gas partition coefficient: agents with low blood-gas partition coefficients (desflurane, sevoflurane, nitrous oxide) have low solubility in blood and are eliminated rapidly, producing fast emergence from anesthesia; agents with high blood-gas partition coefficients (halothane, diethyl ether) are more soluble in blood and tissues, requiring longer washout times. Pulmonary function impairment (reduced alveolar ventilation, high ventilation-perfusion mismatch) slows elimination and prolongs emergence. Some solvents, including acetone (endogenously produced in diabetic ketoacidosis), toluene, and other volatile organic compounds, are partially eliminated by exhalation and detected by breath analysis. Alcohol breath testing exploits the predictable relationship between alveolar ethanol concentration and blood ethanol concentration (approximately 1:2100 volume ratio) to estimate blood alcohol concentration non-invasively.9
Salivary Drug Excretion. Many drugs are excreted in saliva at concentrations that may reflect free drug plasma concentrations, making saliva a potential non-invasive alternative to blood sampling for therapeutic drug monitoring (TDM) of certain drugs. Saliva drug concentrations are determined by the same pH-partition principles governing other membrane transfers: un-ionized, lipophilic drugs diffuse passively from plasma into saliva across the salivary gland epithelium, and the saliva-to-plasma ratio approaches the free drug plasma fraction for drugs at or near their un-ionized form at physiological pH. For drugs where the salivary concentration reliably tracks free plasma concentration (theophylline, phenytoin, carbamazepine, lithium, several antiretrovirals), saliva offers a practical, painless alternative to venipuncture for monitoring, particularly in pediatric patients and for point-of-care testing in resource-limited settings. Salivary monitoring of methadone and buprenorphine in opioid treatment programs is well established. The practical limitation is that oral hygiene, salivary pH, salivary flow rate, and local oral drug administration (sublingual buprenorphine leaving residual drug in oral mucosa) can confound salivary drug concentrations.5
Breast Milk Transfer. Drugs excreted in breast milk represent a passive transfer route of clinical significance because of potential exposure to nursing infants. Transfer of drugs into breast milk follows the same physicochemical principles as other membrane transfers: lipophilic, un-ionized drugs transfer readily, while hydrophilic or ionized drugs transfer poorly. However, human milk has a slightly lower pH than plasma (approximately 7.0 to 7.2 versus 7.4) and a higher fat content, causing ion trapping and lipid sequestration of basic and lipophilic drugs respectively, sometimes producing milk-to-plasma ratios substantially above 1.0 for basic lipophilic drugs. The infant's relative drug dose is estimated as the milk-to-plasma ratio multiplied by average daily milk intake (approximately 150 mL/kg/day) divided by the therapeutic dose for the infant's weight; a relative infant dose (RID) below 10% is generally considered acceptably safe by most pharmacological references. Highly protein-bound drugs, very hydrophilic drugs, and drugs with short half-lives that can be timed around feeds represent lower-risk options. The LactMed database maintained by the National Institutes of Health (NIH) provides current, evidence-based guidance on specific drugs and lactation safety.10
Renal elimination: glomerular filtration (free drug only) + tubular secretion (active, can exceed GFR) − tubular reabsorption (passive, pH-dependent for ionizable drugs). Use Cockcroft-Gault for drug dosing. Biliary: MW above ~500 Da; glucuronide conjugates; active transporters (MRP2, P-gp, BCRP). Enterohepatic recirculation: biliary conjugate → bacterial hydrolysis → reabsorption; prolongs half-life; secondary plasma peaks; disrupted by antibiotics and bile acid resins. Breast milk: RID below 10% is generally safe threshold; consult LactMed. Pulmonary: volatile anesthetics; low blood-gas partition = fast elimination. Dialysis removes drugs with small Vd, low protein binding, low MW only.
Rowland M, Tozer TN. Clinical Pharmacokinetics and Pharmacodynamics: Concepts and Applications. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2011. ISBN 9780781750462.
Aronoff GR, Bennett WM, Berns JS, et al. Drug Prescribing in Renal Failure: Dosing Guidelines for Adults and Children. 5th ed. Philadelphia: American College of Physicians; 2007. ISBN 9781930513693.
Giacomini KM, Huang SM, Tweedie DJ, et al. Membrane transporters in drug development. Nat Rev Drug Discov. 2010;9(3):215-236.
doi:10.1038/nrd3028Verbeeck RK, Musuamba FT. Pharmacokinetics and dosage adjustment in patients with renal dysfunction. Eur J Clin Pharmacol. 2009;65(8):757-773.
doi:10.1007/s00228-009-0678-8Shargel L, Wu-Pong S, Yu ABC. Applied Biopharmaceutics and Pharmacokinetics. 7th ed. New York: McGraw-Hill; 2016. ISBN 9780071830935.
Keppler D. Multidrug resistance proteins (MRPs, ABCCs): importance for pathophysiology and drug therapy. Handb Exp Pharmacol. 2011;(201):299-323.
doi:10.1007/978-3-642-14541-4_8Doherty MM, Charman WN. The mucosa of the small intestine: how clinically relevant as an organ of drug metabolism? Clin Pharmacokinet. 2002;41(4):235-253.
doi:10.2165/00003088-200241040-00001Roberts MS, Magnusson BM, Burczynski FJ, Weiss M. Enterohepatic circulation: physiological, pharmacokinetic and clinical implications. Clin Pharmacokinet. 2002;41(10):751-790.
doi:10.2165/00003088-200241100-00005Brunton LL, Hilal-Dandan R, Knollmann BC, eds. Goodman & Gilman's: The Pharmacological Basis of Therapeutics. 13th ed. New York: McGraw-Hill; 2018. ISBN 9781259584732.
Hale TW. Medications and Mothers' Milk. 19th ed. New York: Springer; 2021. ISBN 9780826190550.