Elimination

Each kidney contains approximately one million nephrons, the functional filtration units. Blood enters the glomerulus — a tuft of fenestrated capillaries enclosed within Bowman's capsule — at high hydrostatic pressure. The high pressure forces unbound (free) drug molecules and other small solutes through the glomerular filtration membrane into the tubular fluid. Protein-bound drug cannot pass through this barrier: only the free drug fraction is filtered, because plasma proteins are too large to cross the glomerular membrane.

This selectivity for free drug has an important implication that is often overlooked: glomerular filtration rate does not equal drug filtration rate. For a drug that is 99 percent bound to albumin, only 1 percent of the plasma drug concentration is free and available for filtration. The filtered load of drug equals the product of glomerular filtration rate, the free drug fraction in plasma, and the plasma drug concentration.

Glomerular filtration rate is the volume of plasma filtered by the kidneys per unit time, normally approximately 100 to 125 milliliters per minute in a healthy adult. Because it cannot be measured directly, it is estimated from the plasma concentration of creatinine — a muscle metabolism byproduct that is freely filtered at the glomerulus, minimally secreted, and not reabsorbed. When the kidneys filter less effectively, plasma creatinine rises and the estimated glomerular filtration rate falls proportionally.

For drugs whose elimination is primarily through glomerular filtration (or filtration plus tubular secretion), the elimination rate is directly proportional to renal function. When renal function falls to 50 percent of normal, the filtration-dependent clearance of such a drug falls by approximately 50 percent, causing plasma drug concentrations to approximately double at the same dose. This proportionality is the pharmacokinetic rationale for dose adjustment in renal impairment: reducing the dose or extending the dosing interval in proportion to the reduction in glomerular filtration rate maintains the same average plasma exposure as in a patient with normal renal function.

Drugs that require dose adjustment in renal impairment share two characteristics: they are substantially excreted by the kidneys (typically more than 30 to 50 percent of unchanged drug eliminated renally), and they have a narrow therapeutic index where drug accumulation produces toxicity. Drugs that are primarily eliminated by hepatic metabolism and have wide therapeutic indexes generally do not require renal dose adjustment, because the accumulated drug concentration, though higher than in patients with normal renal function, remains within a safe range.

The most clinically important drugs requiring careful renal dose adjustment include aminoglycoside antibiotics (gentamicin, tobramycin, amikacin), vancomycin, beta-lactam antibiotics at high doses, digoxin, lithium, metformin, enoxaparin and other low-molecular-weight heparins, direct oral anticoagulants (apixaban, rivaroxaban, dabigatran have different degrees of renal dependence), gabapentin and pregabalin, allopurinol, and many antiretroviral agents. Conversely, azithromycin, doxycycline, linezolid, clindamycin, and most macrolide antibiotics do not require renal dose adjustment because they are primarily eliminated by non-renal routes.

The two fundamental approaches to renal dose adjustment are dose reduction (administering a smaller dose at the normal dosing interval) and interval extension (administering the normal dose at a longer dosing interval). These produce different concentration-time profiles: dose reduction lowers both the peak and trough concentrations, while interval extension maintains the same peak but allows the trough to fall lower before the next dose.

Which approach is appropriate depends on the pharmacodynamic properties of the drug. For drugs where efficacy correlates with time above a minimum effective concentration (most beta-lactam antibiotics), maintaining a consistent concentration by reducing the dose and keeping the interval the same is preferred. For drugs where efficacy correlates with peak concentration (aminoglycosides — concentration-dependent bactericidal activity), maintaining a high peak while extending the interval to allow drug to clear is the preferred approach, and once-daily dosing becomes even more rational in renal impairment. For narrow-therapeutic-index drugs, the dose and interval are typically both adjusted empirically based on published guidelines with therapeutic drug monitoring to verify that concentrations are within range.

Renal impairment affects drug pharmacokinetics through multiple mechanisms beyond reduced filtration. Protein binding changes: uremia generates endogenous organic acids that compete with drugs for albumin binding sites, increasing the free fraction of acidic drugs including phenytoin and some nonsteroidal anti-inflammatory drugs. Volume of distribution changes: fluid retention associated with renal failure increases the effective volume of distribution for hydrophilic drugs, requiring higher loading doses. Absorption may be altered: uremia-induced nausea, gastrointestinal dysmotility, and gastroparesis reduce the reliability of oral drug absorption. Metabolism is modestly impaired: some cytochrome P450 enzyme activities are reduced in advanced renal failure, though the magnitude is less than in severe hepatic disease. For drugs with active or toxic metabolites that are renally eliminated (such as morphine-6-glucuronide from morphine, or normeperidine from meperidine), renal failure causes accumulation of the metabolite as well as the parent drug — a consideration that makes meperidine particularly dangerous in renal failure due to normeperidine-induced seizures.

The cycle proceeds as follows: the drug or its conjugated metabolite is secreted from hepatocytes into bile, stored in the gallbladder, and released into the small intestine. In the colon, bacteria express beta-glucuronidase and sulfatase enzymes that cleave the glucuronide or sulfate conjugate, liberating the free (de-conjugated) drug. The free drug, which is now lipophilic enough to be absorbed across the colonic mucosa, is reabsorbed into the portal circulation and returns to the liver, where it re-enters the hepatic processing cycle. Each cycle delays drug elimination and extends the effective half-life.

This recycling is physiologically important for bile salts, which undergo enterohepatic recirculation approximately 6 to 10 times per day, with only a small fraction lost in feces. For drugs, enterohepatic recirculation is clinically relevant when it significantly prolongs drug exposure or when interruption of the cycle is used therapeutically.

Digoxin, morphine, chloramphenicol, ethinyl estradiol (in oral contraceptives), and numerous other drugs undergo clinically meaningful enterohepatic recirculation. For digoxin, the recirculation contributes to its long effective half-life (36 to 48 hours) and means that elimination continues for days after the drug is stopped. For oral contraceptives containing ethinyl estradiol, the estrogen component undergoes hepatic glucuronidation, biliary excretion, and colonic deconjugation before reabsorption — a cycle that was hypothesized to be disrupted by broad-spectrum antibiotics that kill the colonic bacteria responsible for deconjugation.

The antibiotic-oral contraceptive interaction has been a longstanding clinical concern and teaching point. The theoretical mechanism is plausible: if antibiotics substantially deplete the colonic bacteria that deconjugate estrogen glucuronide, less free estrogen would be reabsorbed, potentially reducing the circulating estrogen concentration and diminishing contraceptive efficacy. However, careful pharmacokinetic studies have generally found that most commonly used antibiotics (amoxicillin, tetracyclines) do not substantially reduce ethinyl estradiol plasma concentrations or ovarian suppression. Rifampin is the important exception: it profoundly reduces oral contraceptive efficacy through enzyme induction (cytochrome P450 3A4) rather than enterohepatic mechanism, and alternative or additional contraception is strongly recommended during rifampin use.

Cholestyramine, a bile acid sequestrant resin, actively interrupts enterohepatic recirculation for several drugs by binding them in the intestinal lumen and preventing reabsorption. This is therapeutically useful for accelerating the elimination of digoxin in toxicity (cholestyramine interrupts digoxin's enterohepatic recirculation, accelerating fecal elimination), and it is the mechanism by which cholestyramine and colesevelam lower cholesterol (by interrupting bile salt recirculation, forcing the liver to synthesize new bile salts from cholesterol).