How do drugs cause renal impairment?

Approximately 20% of community- and hospital-acquired episodes of acute kidney injury are caused by drugs, with the incidence among older adults as high as 66%, (Naughton, 2008). There are several reasons why the kidneys are particularly vulnerable to damage by drugs. The kidneys constitute only 0.4% of body weight but receive 20% to 25% of total blood flow. This disproportionate blood flow exposes the kidneys excessively to drugs in the blood. In addition, drugs are concentrated as the tubular filtrate passes through the nephron and water is reabsorbed. Tubular transport systems further concentrate drugs in the filtrate. Enzymes in the kidney may metabolize drugs to metabolites that are nephrotoxic. In renal insufficiency, the remaining functional nephrons are even more susceptible to nephrotoxins.

Nephrotoxicity due to drugs contributes to 8% to 60% of AKI cases in hospitalized patients (Rosner & Okusa, 2008), The most commonly implicated pharmacologic nephrotoxins are antibiotics (aminoglycosides, cephalosporins, pentamidine, amphotericin B), radiocontrast agents used for radiologic studies, cyclosporine, cisplatin, angiotensin-converting enzyme (ACE) inhibitors, and nonsteroidal antiinflammatory drugs (NSAIDs). Because of the development of new agents (e.g., lower osmolar radiocontrast agents), changing drug use patterns (e.g., decreased use of aminoglycosides), and the shift of care from inpatient to outpatient settings, NSAIDs and ACE inhibitors are increasingly predominant causes of transient acute kidney injury. In chronic outpatient settings, chronic kidney disease (CKD) can occur due to combination analgesics, which consist of either aspirin or an NSAID combined with acetaminophen, caffeine, and/or codeine. Although the agents specifically cited here are the most frequent causes of renal damage, numerous other medications from diverse drug categories cause renal damage. The risk is greatest in individuals who already have poor kidney perfusion. Whenever a patient evidences renal impairment, a careful analysis of the drug profile for potential drug nephrotoxicity should be conducted. Nephrotoxins should be avoided in patients in any stage of CKD or used with appropriate dosage adjustments and meticulous monitoring.

How do these drugs usually cause renal damage?

Several mechanisms of renal damage by drug nephrotoxins have been identified (Rosner & Okusa, 2008), but most nephrotoxins damage the kidneys through more than one mechanism. Hemodynamic mechanisms involve inhibition of regulatory and compensatory processes, nonspecific renal vasoconstriction, or altered colloid oncotic pressure. A primary example of hemodynamic mechanisms for nephrotoxicity includes transient acute kidney injury from inhibition of the renin-angiotensin-aldosterone system by ACE inhibitors in patients with renal artery stenosis. A second example is the inhibition of prostaglandin-dependent renal blood flow by NSAIDs in patients with conditions associated with decreased renal blood flow (e.g., volume depletion and congestive heart failure). Renal vasoconstriction is a hypothesized mechanism of renal damage from propranolol, mannitol, a combination of triamterene and indomethacin, and the initial months of cyclosporine therapy. Dextran-40 can elevate oncotic pressure and impair glomerular filtration.

Another mechanism of damage is renal vascular alterations, such as thrombotic microangiopathy, that may result from oral contraceptives, cyclosporine, mitomycin C, cisplatin, and quinine. Glomerular alterations that result in nephrotic syndrome and glomerulonephritis are more often immune effects than toxic effects. The most common drug-induced glomerular alteration is membranous nephropathy that occurs with oral and parenteral gold therapy and penicillamine. Less common glomerular toxicities include the following: minimal change nephrotic syndrome associated with NSAIDs, ampicillin, rifampin, phenytoin, and lithium; focal segmental glomerulosclerosis secondary to heroin abuse; and membranoproliferative glomerulonephritis from hydralazine, interferon-alpha, and interleukin-2. Toxic drug effects resulting in acute tubular necrosis are most often caused by aminoglycosides, radiographic contrast media, cisplatin, amphotericin B, pentamidine, and foscarnet. Tubulointerstitial disease takes several forms and commonly manifests as one of the following conditions: acute allergic interstitial nephritis from antibiotics (e.g., penicillins, cephalosporins, tetracyclines, sulfonamides, fluoroquinolones), NSAIDs, diuretics, and anticonvulsants; chronic interstitial nephritis from lithium and cyclosporine; and papillary necrosis from analgesics (e.g., NSAIDs, aspirin, and acetaminophen used alone or in combination) or high-dose dapsone therapy.

Obstructive nephropathies from drugs include uric acid nephropathy during chemotherapy; rhabdomyolysis from phencyclidine, adrenergic drugs including terbutaline, cocaine, vasopressin infusion, erythromycin, and systemic cholesterol-lowering drugs, especially HMG-CoA reductase inhibitors (e.g., lovastatin, atorvastatin, simvastatin); and urinary tract outflow obstruction from anticholinergic drugs (e.g., tricyclic antidepressants, disopyramide), cyclophosphamide, and methysergide.

How can renal damage from such drugs be minimized or avoided?

Naturally, drugs with potential to cause renal damage should be avoided or used cautiously in patients with high risk for renal impairment. Conditions that predispose to renal damage by drugs include use of multiple nephrotoxins, sodium or fluid depletion, preexisting renal disease, and low renal blood flow in patients with diseases like congestive heart failure and cirrhosis, advanced age, and diabetes mellitus (Rosner & Okusa, 2008 ). Often drug-induced renal damage is reversible if the drug is discontinued and supportive care is initiated before permanent effects occur. Giving saline intravenously may decrease damage by some nephrotoxins such as cyclosporine and cisplatin by diluting the concentration of the drug in the renal tubule. Misoprostol, a prostaglandin analog, may prevent NSAID nephropathy. Drugs with the least nephrotoxic potential should be selected. For example, acetaminophen, aspirin, nonacetylated salicylates, sulindac, or nabumetone may have less nephrotoxicity than other NSAIDs. Finally, drugs should be given in the lowest effective doses for the shortest possible duration.

How does chronic kidney disease itself alter response to medications?

The changes that accompany CKD can alter drug response through two major mechanisms: pharmacodynamics and pharmacokinetics. Medications chemically interact with receptors on cell membranes or on enzymes to cause their effects. This interaction is known as pharmacodynamics and can be thought of as what the drug does to the body. Adverse effects (also called side effects or toxic effects) occur when a drug or metabolite acts at receptors other than the target receptors or when excess drug is present at the target receptor. Uremic substances in the blood or altered electrolyte concentrations resulting from renal failure can modify the drug-receptor interaction, resulting in altered drug effect. Altered receptor sensitivity is thought to be responsible for increased central nervous system effects of narcotics, sedatives, and hypnotics, as well as for the resistance to effects of epinephrine and other catecholamines that occurs in uremic patients. Altered electrolyte and acid-base balance also affect the response to such medications as antiarrhythmics, digoxin, phenothiazines, and antidepressants.

The magnitude and persistence of drug action depend on the duration and concentration of the drug in proximity to the receptor. This relationship of time and drug concentration is known as pharmacokinetics and can be thought of as how the body acts on the drug through the processes of absorption, distribution, metabolism, and excretion. How CKD affects pharmacodynamics is not well understood. On the other hand, many of the effects of altered pharmacokinetics in patients on dialysis have been sufficiently studied to develop mathematical formulas for calculating drug dosage. Decreased renal excretion is the most obvious alteration in pharmacokinetics resulting from renal dysfunction, but each of the other pharmacokinetic processes may also be altered.

What are the pharmacokinetic parameters that reflect alterations caused by chronic kidney disease?

Several measurement parameters can be computed for each drug to represent its pharmacokinetic profile. These include bioavailability, volume of distribution, clearance, and elimination half-life. Bioavailability, which is abbreviated as F (for fraction), is the measure of drug absorption. Bioavailability is the percentage of the administered dose that is absorbed into the systemic circulation.

Each drug has a unique pattern of distribution throughout the body, represented as its apparent volume of distribution (V_d). V_d is the hypothetical volume that would be required to contain the dose of drug at its concentration in the plasma. For example, if 500 mg of a drug were administered to a patient and an hour later the concentration of the drug in a sample of that patient’s plasma was 0.001 mg/mL, the V_d would be 500 L. Stated another way, if each milliliter of plasma had 0.001 mg in it, 500 L would be required to contain 500 mg. Although V_d is an abstraction, it can be interpreted to reflect the distribution characteristics of a drug. If the V_d is 5 L (0.06 L/kg), it is likely that most of the drug stays in the intravascular space. If the V_d is more than 46 L (0.7 L/kg), the drug is sequestered in the peripheral tissues, usually dissolved in fatty tissues or bound to tissues. Drugs with large volumes of distribution are poorly dialyzable because most of the drug is outside the bloodstream and therefore not exposed to the dialysis membrane. V_d is used in the calculation of loading doses.

The two modes of elimination, metabolism and excretion, are measured as drug clearance (CL). Clearance is defined as the rate of removal of a drug in proportion to its concentration in the plasma. Clearance is reported as the volume of plasma cleared of the drug per unit of time. Therefore, if the plasma concentration of a drug is 0.002 mg/mL and the body eliminates 2 mg/h, the clearance is 1 L/h or about 17 mL/min. (This is because, in such cases, 1000 mL or 1 L of plasma is needed to contain 2 mg at a concentration of 0.002 mg/mL.) Like V_d, clearance is a useful abstraction rather than a concrete reality. CL is useful for calculating the maintenance dose of a drug. In the patient with renal impairment and lower clearance, lower maintenance doses are needed.

Elimination half-life (t_1/2) is an indicator of how long a drug stays in the body. It is the time required for half of the drug to be eliminated and is manifested clinically as the time required for the concentration in the blood to decline 50%. Half-life is prolonged by large V_d or by slow CL, which are both pharmacokinetic changes common in renal failure. The relationship among half-life, volume of distribution, and drug clearance is:

What factors affect absorption of medications in patients with chronic kidney disease?

In uremia the breakdown of urea in the gastrointestinal tract may raise pH and slow absorption of acid drugs, such as aspirin, iron preparations, and diuretics. Gastroparesis and gastrointestinal responses to uremia (nausea, vomiting, and diarrhea) may significantly alter the absorption of oral medication. Antacids, commonly used to bind dietary phosphate in CKD, diminish absorption of drugs by forming unabsorbable compounds with drugs like digoxin, iron preparations, and some antibiotics (e.g., tetracyclines and fluoroquinolones). Drugs used to suppress gastric acid secretions, such as H₂ blockers (e.g., cimetidine, ranitidine, famotidine), antacids, and proton pump inhibitors (omeprazole, lansoprazole) affect the absorption of some drugs. For example, ketoconazole, which requires an acid environment for absorption, decreases bioavailability in patients taking drugs that suppress gastric acidity, whereas oral penicillins, which are inactivated by gastric acid, improve absorption when patients take these agents.

Orally administered drugs pass through the liver before reaching the main circulation, and large portions of a dose of drugs like morphine, propranolol, and codeine are metabolized during this “first pass” through the liver. Because the first-past effect reduces the bioavailability (F) of orally administered morphine, the oral dose of morphine must be much larger than the injected dose to achieve the same amount of pain relief. In uremia, the fraction of drug metabolized by this first-pass effect may be decreased, unchanged, or increased because some metabolic by-products of uremia change the activity of liver enzymes. Thus bioavailability of drugs is highly variable in CKD.

Drugs are absorbed across the peritoneum when administered into the dialysate. Antibiotics are frequently given via this route, which results in high concentrations at the site of a peritoneal infection. Some patients have experienced improved diabetic control with intraperitoneal administration of insulin. Administration of drugs by the peritoneal route will result in a different plasma profile than oral or parenteral administration, usually with delayed onset, decreased peak plasma concentration, and prolonged duration of action. Presence of peritonitis will alter the bioavailability of drugs administered peritoneally.

What factors affect distribution of medications in patients with renal failure?

When a drug is absorbed into the blood, some molecules bind to proteins in the plasma. Drugs that are highly bound to plasma proteins in the blood usually have small V_d because most of the drug molecules are attached to plasma protein, which normally cannot exit the blood vessels. Drugs that predominantly exit from the blood and bind to muscles or dissolve in fatty tissue in the periphery have large V_d. In general, drugs with small V_d have short half-lives because they are mostly in the plasma, which frequently passes through the liver and kidney (and dialysis machine), where they are eliminated. Conversely, drugs with large V_d have longer half-lives and less susceptibility to removal by dialysis. Edema and ascites often increase V_d and will increase the half-life of drugs that normally have small distribution volumes.

Plasma protein binding of acidic drugs to albumin may be decreased in renal failure as a result of either decreased concentration of albumin or decreased capacity of the albumin to bind to drugs. Changes in protein binding can alter V_d and the drug effect because only the free drug is pharmacologically active. Decreased albumin binding is thought to contribute to the central nervous system toxicity of acid drugs like theophylline, phenytoin, penicillin, phenobarbital, and salicylates in uremia. Some alkaline drugs (e.g., lidocaine, phenothiazines, propranolol, quinidine, and tricyclic antidepressants) that bind to glycoprotein also undergo increased or decreased binding in renal disease, but the clinical relevance of these changes is not as well studied as albumin binding. Although changes in V_d and plasma protein binding theoretically could have substantial effects on drug response, current research suggests that the effects of altered V_d in renal disease are usually minimal. There are a few exceptions, in which these changes require modification in the approach to patient care. An example is decreased protein binding of phenytoin during renal failure, which must be considered in clinical management. Measured serum concentrations of phenytoin that reflect total (bound + free) drug concentration in the plasma often are reported in the subtherapeutic range in patients with renal failure. This is because the amount of phenytoin bound to albumin is decreased, but the fraction of unbound drug is increased. Because the unbound drug is the active portion, lower drug concentrations of phenytoin are desirable for patients in renal failure to achieve the desired effect (Aweeka, 1995). In some centers both free and total phenytoin drug concentrations are measured to avoid toxicity. In patients without CKD, protein binding is about 90% of the total drug, whereas in patients with CKD binding of phenytoin ranges from 65% to 80%. Another approach to interpretation of phenytoin levels in CKD patients is to use a correction formula that adjusts the reported phenytoin plasma concentration for albumin level, renal function, and decreased affinity of phenytoin for albumin in CKD (Liponi et al., 1984).

What factors affect elimination of medications in patients with chronic kidney disease?

Some drugs are cleared almost exclusively in their original chemical form by renal excretion; these drugs are said to be excreted unchanged. Other drugs undergo alteration of chemical structure by enzymes, a process called biotransformation or metabolism. Most drugs are cleared by a combination of hepatic metabolism and renal excretion. Metabolites (the form drugs take after chemical alteration by metabolism) are usually more water soluble than the original drug and are usually eliminated by the kidneys. Active metabolites retain the ability to bind to a receptor and elicit the same effect as the original drug. Inactive metabolites are usually insignificant because they do not stimulate the target receptor. Toxic metabolites are those that cause an adverse effect at a site different from the target receptor.

When a drug that is normally cleared unchanged by the kidneys is repeatedly administered to a patient with renal insufficiency, it begins to accumulate in the blood and may cause adverse effects. Increased portions of the drug may be eliminated by alternate routes, such as hepatic metabolism or through the lungs, bile, or sweat glands. Metabolites of drugs accumulate in renal insufficiency, and active or toxic metabolites contribute to adverse effects. An example is normeperidine, a metabolite of meperidine, which causes stupor or seizures when it accumulates. Box 17-1 includes examples of drugs with active or toxic metabolites that may accumulate in renal failure. If viable alternatives exist, drugs with active or toxic metabolites are avoided in patients with renal failure. When drugs with active or toxic metabolites are used in patients with renal failure, decreased dosages may be required and clinical monitoring must be vigilant. For example, patients with renal impairment who take allopurinol for gout or cancer require lower dosages than those with normal renal function because an active metabolite of allopurinol can cause exfoliative dermatitis when it accumulates in the body. Although far less important than active metabolites, inactive metabolites may also have consequences. For example, the accumulation of inactive metabolites may cause interference with laboratory tests.

Impaired renal function may also affect liver metabolism, decreasing elimination of some drugs (e.g., morphine, clonidine) and increasing metabolism for a few others (e.g., phenobarbital and phenytoin). Renal impairment alters metabolism through accumulation of uremic substances that can induce (speed up) or inhibit (slow down) drug-metabolizing enzymes in the liver. Insulin is metabolized by enzymes in the kidney, so it is more slowly cleared in severe renal disease. Liver metabolism is dependent upon genetic inheritance, diet, environmental pollution, and concurrent administration of other medications; thus the effects of renal dysfunction are likely to be highly variable from drug to drug and from person to person. Effects on renal elimination are more predictable: the greater the proportion of drug or its active metabolites eliminated by the kidneys, the more likely that altered dosing will be required for patients with renal impairment and those on dialysis.

How does dialysis affect pharmacokinetics of drugs and poisons?

The kidneys eliminate drugs through several processes. Although dialysis is not a substitute for all of these renal processes, some drugs are removed by dialysis. Dialysis may also affect other pharmacokinetic parameters. For example, changes in total body water from predialysis to postdialysis will affect the V_d of some drugs. Characteristics of drugs that promote removal by dialysis are as follows: (1) small molecular size, (2) small V_d, (3) water solubility, and (4) low protein binding. If protein binding exceeds 90%, the drug will be negligibly eliminated by dialysis. Drugs are more likely to be removed when the dialyzer membrane is highly permeable and its surface area is large and when the blood flow rate and dialysate flow rate are high. Peritoneal dialysis generally provides little drug removal because dialysate flow rate is slower than with other methods, although a greater amount of protein-bound drug can be removed due to large protein losses seen with this mode. Continuous therapy with hemofiltration or continuous hemodialysis for critically ill patients can remove substantial fractions of drugs. Removal of drugs by hemofiltration procedures is determined by the ultrafiltration rate and the degree of protein binding. Treatment of drug overdose and poisoning involves application of these principles to decrease serum concentration of the toxic drugs or substance (Winchester & Kriger, 1995). Similarly, poisons with high protein binding or large V_d are not dialyzable. Although many standard references classify a drug as dialyzable or not dialyzable, dialyzability is not an all-or-nothing characteristic. Some drugs are virtually entirely removed by dialysis; others have negligible removal. Many drugs fall somewhere in the middle. The type of dialysis equipment and the length of dialysis greatly influence whether a drug is removed. Classification of dialyzability as “yes” or “no” is based on an expert’s opinion of whether removal is clinically significant—that is, sufficient to remove an overdose—or whether the patient will require dosage replacement.