Effects of Drugs in Patients with Reduced Renal Function

Most drugs are weak electrolytes and lipid soluble in the un-ionized state; thus, they are extensively reabsorbed by renal tubular cells (see Chapter 20 and Chapter 37 ). Termination of their action is not dependent on renal excretion; redistribution and metabolism produce this effect. After biotransformation, these drugs are excreted in urine as water-soluble, polar forms of the parent compound. They are usually pharmacologically inactive, and their retention is not harmful.^[17] The majority of drugs with prominent central and peripheral nervous system activity fall into this category, including most narcotics, barbiturates, phenothiazines, butyrophenone derivatives, benzodiazepines, ketamine, and local anesthetics.^[20] However, several drugs are relatively lipid insoluble or are highly ionized in the physiologic pH range and are eliminated unchanged in urine. Their duration of action may be extended in patients with impaired renal function. Drugs in this category include muscle relaxants, cholinesterase inhibitors, thiazide diuretics, digoxin, and many antibiotics ( Table 54-5 ). ^[21]

2183

Opioids

Protein binding of morphine decreases by approximately 10% in CRF (see Chapter 11 ).^[22] This decreased binding does not result in a significant alteration in the free fraction of morphine because it is usually protein bound to only a small extent (23% to 42%) and it has a large volume of distribution.^[23] Morphine is almost completely metabolized in the liver, mostly to the inactive glucuronide, which is then excreted in urine.^{[24] [25]} Thus, administration of morphine to patients with renal failure, particularly in analgesic doses, should not cause prolonged depression. Nevertheless, a report of severe respiratory and cardiovascular depression in a patient with renal failure was attributed to the administration of a single 8-mg dose of morphine.^[26] This depression can be linked to the accumulation of morphine-6-glucuronide, which possesses opioid activity and is excreted by the kidney. The distribution, protein binding, and excretion of meperidine are similar to those of morphine.^{[27] [28]} Accumulation of its metabolite normeperidine can produce excitatory central nervous system (CNS) effects, including convulsions in extreme cases. Fentanyl is also metabolized in the liver, with only 7% excreted unchanged in urine.^{[29] [30]} It is moderately bound to plasma protein (free fraction, 19%), and its volume of distribution is large.^{[23] [30]} Thus, fentanyl should be suitable for the premedication of patients with renal failure. The pharmacokinetics and pharmacodynamics of sufentanil and alfentanil are not significantly different in patients with reduced renal function and normal individuals.^{[31] [32]} An ester linkage in remifentanil renders it susceptible to rapid metabolism by blood and tissue esterases. Therefore, the pharmacokinetics and pharmacodynamics of remifentanil are unaltered in patients with renal disease. Although renal elimination of its principal metabolite is reduced, it is of no clinical significance.^[33]

Inhaled Anesthetics

All inhaled anesthetics (see Chapter 5 and Chapter 8 ) are biotransformed to some extent, with the nonvolatile products of metabolism eliminated almost entirely by the kidney.^[34] However, reversal of the CNS effects of inhaled anesthetics is dependent on pulmonary excretion, so impaired kidney function will not alter the response to these anesthetics. From the viewpoint of selecting an anesthetic that will not be harmful to patients with mild or moderate impairment of renal function, all the modern potent anesthetics are suitable. Enflurane is biotransformed to inorganic fluoride, but levels after 2 to 4 hours of anesthesia average only 19 µM in patients with mild to moderate kidney disease, significantly lower than the nephrotoxic threshold of 50 µM, which is frequently reported after the administration of methoxyflurane.^{[35] [36] [37]} This level of fluoride should not cause further renal impairment. Fluoride levels after isoflurane increase by only 3 to 5 µM^[38] and by only 1 to 2 µM after halothane,^[37] so these agents have no nephrotoxic potential.

Desflurane and sevoflurane, the two new inhaled anesthetics, are remarkably different in their molecular stability and biotransformation. Desflurane is highly stable and resists degradation by soda lime^[39] and the liver. Even in enzyme-induced animals, excretion of organic or inorganic fluoride has been shown to be minimal.^{[40] [41]} The mean inorganic fluoride concentration after 1.0 minimum alveolar concentration (MAC)-hour exposure to desflurane was less than 1 µmol/L.^[42] The safety of desflurane in renal failure patients has been confirmed.^[43] In addition, more sensitive indices of renal function, namely, urine retinol-binding protein and β-N-acetylglucosaminidase, showed no evidence of renal damage. Prolonged exposure to desflurane (7.0 MAC-hours) has been associated with normal renal function.^[43]

Sevoflurane, on the other hand, is not very stable. Soda lime causes it to decompose,^[44] and it is biotransformed by the liver to an extent similar to enflurane. Plasma inorganic fluoride concentrations approaching nephrotoxic levels (50 µmol/L)^{[45] [46]} have been reported after prolonged inhalation of sevoflurane. However, no evidence of gross changes in renal function was found in humans.^[47] In one study, sevoflurane was used at low flow (1 L/min), and there was no relationship to compound A and renal function.

Inhaled anesthetics cause a transient reversible depression in renal function. GFR, renal blood flow, urine output, and urinary excretion of sodium are decreased ( Table 54-6 ). Probable mechanisms include reduced renal blood flow, loss of renal autoregulation, neurohumoral factors (e.g., antidiuretic hormone, vasopressin, renin), and neuroendocrine responses. Although most inhaled anesthetics have been shown to reduce GFR and urinary excretion of sodium, their effects on renal blood flow have yielded conflicting results, which can be explained by differences in experimental methodology. Data suggest that renal blood flow is maintained with halothane, isoflurane, and desflurane^{[48] [49] [50]} but decreased with enflurane and sevoflurane.^{[51] [52]}

Intravenous Anesthetics

Reversal of CNS effects after the administration of ultrashort-acting barbiturates such as thiopental and

2184

TABLE 54-6 -- Effects of various anesthetics on renal function

	RBF	GFR	Urine Output	Urine Solutes
General anesthesia	↓	↓	↓	↓
Intravenous anesthetics
Thiopental	↔	↓	↓	↓
Midazolam	↔	↔	↓	↔
Fentanyl/droperidol	↔	↔	↓	↓
Fentanyl (high dose)	↔	↔	↔	↔
Inhaled anesthetics
Halothane	↔	↓	↓	↓
Enflurane	↓	↓	↓	↓
Isoflurane	↔	↓	↓	↓
PEEP	↓	↓	↓
Regional anesthesia
Epidural (with epinephrine)	↓	↓	↓
Epidural (without epinephrine)	↔	↔	↔
Spinal	↔	↔	↔
Note: Although conflicting reports of anesthetic effects on RBF have been reported because of different investigative methods, the current literature seems to support these data.
Key: ↔, no significant change; , significant data; ↓ = decrease.
GFR, glomerular filtration rate; PEEP, positive end-expiratory pressure; RBF, renal blood flow.
From Hemmings HC Jr: Anesthetics, adjuvant drugs and the kidney. In Malhotra V (ed): Anesthesia for Renal and Genitourinary Surgery. New York, McGraw-Hill, 1996, p 20.

Propofol does not adversely affect renal function as reflected by measurements of creatinine concentration. Prolonged infusions of propofol may result in the excretion of green urine because of the presence of phenols in the urine. This discoloration does not affect renal function. Urate excretion is increased after the administration of propofol and is usually manifested as cloudy urine when urate crystallizes under conditions of low pH and temperature.^[58]

There are no reports of the disposition of narcotics and tranquilizers when used in large dosage for anesthesia in uremic patients. These drugs are extensively metabolized before excretion, so when combined with 30% to 50% nitrous oxide, they should not have a particularly prolonged effect. Of course, the benzodiazepines, especially diazepam,^[21] have a long half-life, so they will tend to accumulate in any case. Because uremic patients are anemic and may require high inspired oxygen concentrations and because of the greater ease of reversibility of the potent inhaled anesthetics than the intravenous drugs, inhaled anesthetics may be preferable for the induction of general anesthesia.

Muscle Relaxants and Their Antagonists

Succinylcholine has been used without difficulty in patients with decreased or absent renal function (see Chapter 13 ). Its metabolism is catalyzed by pseudocholinesterase to yield the nontoxic end products succinic acid and choline. The metabolic precursor of these two compounds, succinylmonocholine, is excreted by the kidney. Thus, large doses of succinylcholine, which might result from prolonged infusion, should be avoided in patients with renal failure. It has been reported that pseudocholinesterase levels are reduced in uremia.^{[59] [60]} However, values are rarely so low that they cause a prolonged block. Hemodialysis has been reported to have no effect on cholinesterase levels.^{[61] [62]}

Administration of succinylcholine causes a rapid, transient increase of 0.5 mEq/L in the serum potassium concentration. In traumatized, burned, or neurologically injured patients, the increase may be as great as 5 to 7 mEq/L, probably as a consequence of denervation supersensitivity of the muscle membrane to succinylcholine and to acetylcholine.^{[63] [64]} In some instances, cardiovascular collapse has occurred.^[63] An exaggerated rise in serum potassium could be particularly dangerous in uremic patients with elevated potassium levels, so the use of succinylcholine is inadvisable unless the patient has undergone dialysis within 24 hours before surgery.

2185

The disposition of nondepolarizing muscle relaxants has been well studied. Renal failure influences the pharmacology of nondepolarizing muscle relaxants by producing either decreased elimination of the drug or its metabolites by the kidney or decreased activity of enzymes that metabolize the drug, such as in the case of mivacurium ( Table 54-7 ). Consequently, the duration of action of muscle relaxants may be prolonged in patients with renal failure. As can be appreciated from examination of the pharmacokinetic data in Table 54-7 , excretion of d-tubocurarine (dTc) is delayed in patients with renal failure.^{[67] [68]} Clearance is reduced and the volume of distribution is unchanged—hence the increased terminal elimination half-life. Because protein binding^[69] and sensitivity of the neuromuscular junction to dTc^[70] are unchanged in patients with renal failure, the consequence of delayed excretion would be a prolonged duration of action. However, this would not be apparent after the administration of a single small dose because redistribution rather than excretion is responsible for termination of action.^[71] On the contrary, recovery after a single large dose or several small doses of dTc is directly related to the terminal elimination half-life. Thus, maintenance doses for patients with decreased renal function should be smaller than for patients with normal renal function, the interval between doses should be increased,^[23] and administration should be monitored with a nerve stimulator.

Pharmacokinetic data for metocurine (see Table 54-7 ) indicate that it differs quantitatively rather than qualitatively from dTc. More than 43% of an injected dose of metocurine is excreted unchanged in urine in 24 hours.^[72] Approximately 40% to 50% of pancuronium is excreted in urine. A portion of this excretion occurs

TABLE 54-7 -- Pharmacokinetic data for nondepolarizing muscle relaxants in normal and anephric patients

Drug	Patients Studied	Elimination Half-Life (hr) (t1/2 B)	Clearance (mL/kg/min)	Volume of Distribution (L/kg)
Vecuronium	Normal	0.9	5.3	0.20
	Anephric	1.4	3.1	0.24
Atracurium	Normal	0.3	6.1	0.18
	Anephric	0.4	6.7	0.22
d-Tubocurarine	Normal	1.4	2.4	0.25
	Anephric	2.2	1.5	0.25
Pancuronium	Normal	2.2	1.8	0.26
	Anephric	4.3	0.9	0.30
Pancuronium	Normal	1.7	1.0	0.14
	Anephric	8.2	0.3	0.14
Metocurine	Normal	6.0	1.2	0.57
	Anephric	11.4	0.4	0.48
Rocuronium	Normal	0.71	2.9	0.207
	Anephric	0.97	2.90	0.264
Cisatracurium	Normal	—	5.2	0.031
	Anephric	—	—	—
Mivacurium	Normal	0.03	106	0.278
	Anephric	0.06	80	0.475

after biotransformation to the less active metabolite 3-hydroxypancuronium.^{[23] [73] [74]} Pancuronium has a prolonged terminal elimination half-life in patients with reduced renal function (see Table 54-7 ),^{[75] [76]} so it should be administered cautiously, particularly when several doses are required.

Two nondepolarizing muscle relaxants, atracurium and vecuronium, were introduced into clinical practice during the early 1980s. After initial reports that the action of neither drug was prolonged in patients with decreased renal function, ^{[77] [78] [79] [80] [81]} it now appears that this is true only for atracurium.^{[82] [83]} Atracurium is broken down by enzymatic ester hydrolysis and by nonenzymatic alkaline degradation (Hofmann elimination) to inactive products and is not dependent on renal excretion for termination of action.^[77] Predictably, their terminal elimination half-life and indices of neuromuscular blockade (onset, duration, and recovery) are the same in patients with normal and absent renal function.^{[78] [79]} Although atracurium may cause release of histamine in patients undergoing renal transplantation, such release does not usually result in clinical signs or symptoms with doses less than 0.4 mg/kg.^[80]

The pharmacokinetics and pharmacodynamics of vecuronium in patients with normal renal function and patients with renal failure have recently undergone reexamination.^{[82] [83]} It now appears to be generally agreed that about 30% of a dose of vecuronium is eliminated by the kidneys. Thus, it is not surprising that Lynam and colleagues ^[82] found that the duration of neuromuscular blockade after the administration of vecuronium was longer in patients with renal failure than in those with normal renal function (99 versus 54 minutes) because of a longer elimination half-life (83 versus 52 minutes) and lower plasma clearance (3.1 versus 5.3 mL/kg/min). In a related area, an interaction between the solvent

2186

Cisatracurium is the single cis isomer of atracurium. Organ-independent mechanisms (Hofmann elimination) account for 77% of the total clearance of cisatracurium. Because renal excretion accounts for only 16% of the elimination of cisatracurium, renal failure should have little impact on its duration of action.^[81]

Doxacurium is a long-acting muscle relaxant whose duration of action is even more prolonged in patients with renal failure.^{[85] [86]} The duration of action of another long-acting muscle relaxant, pipecuronium, shows large variability in patients with renal failure. ^[87] The short-acting drug mivacurium is metabolized by plasma pseudocholinesterase. Its effect has been shown to be lengthened by 10 to 15 minutes in patients with ESRD, most likely because of a decrease in plasma cholinesterase activity in these patients associated with uremia or hemodialysis. ^{[88] [89] [90]} Phillips and Hunter demonstrated a decrease in the mivacurium requirement by infusion in anephric patients.^[89]

The elimination half-life of rocuronium is increased in renal failure because of an increase in the volume of distribution with no change in clearance. This explanation may account for a somewhat longer duration of action in anephric patients, although its clinical significance is uncertain.^[91]

Pharmacokinetic data for the cholinesterase inhibitors neostigmine, pyridostigmine, and edrophonium for normal, anephric, and renal transplant patients are presented in Table 54-8 ; there are no major differences among the three drugs.^{[92] [93] [94]} Renal excretion is of major importance for the elimination of all three agents, with approximately 50% of neostigmine and 70% of pyridostigmine and edrophonium excreted in urine. Excretion of all the cholinesterase inhibitors is delayed in patients with impaired renal function to the same or perhaps to a slightly greater extent than is elimination of muscle relaxants. Thus, "recurarization" after reversal of neuromuscular blockade in a patient with renal failure is, in

TABLE 54-8 -- Pharmacokinetic data for cholinesterase inhibitors in normal, anephric, and renal transplant patients

Drug	Patients Studied	Elimination Half-Life (hr) (t1/2 B)	Clearance (mL/kg/min)	Volume of Distribution (L/kg)
Neostigmine	Normal	1.3	8.4	0.7
	Anephric	3.0 *	3.9 *	0.8
	Renal transplant	1.7	9.4	1.1
Pyridostigmine	Normal	1.9	8.6	1.1
	Anephric	6.3 *	2.1 *	1.0
	Renal transplant	1.4	10.8	1.0
Edrophonium	Normal	1.9	8.2	0.9
	Anephric	3.6 *	2.7 *	0.7
	Renal transplant	1.4	9.9	0.9

*P < .05 versus normal.

Digitalis

Digoxin is the most frequently used digitalis glycoside in both uremic and nonuremic patients. Approximately 72% of a parenteral dose is excreted unchanged in urine.^[95] Thus, administration of digoxin to patients with reduced renal function is potentially dangerous, and maintenance doses must be reduced in proportion to the reduction in renal function. Blood digoxin levels are the time-tested reliable guide to therapy (therapeutic level, >0.8 ng/mL; toxic level, >1.8 ng/mL).^[96] Whenever possible, initial digitalization or changes in digitalis dosage should be avoided immediately before surgery.

Vasopressors and Antihypertensive Drugs

Patients with severe renal disease are frequently given antihypertensive and other cardiovascular medications (see Chapter 16 ). More than 90% of the thiazides^[97] and 70% of furosemide^[98] are excreted by the kidney and have prolonged durations of action in patients with abnormal or absent renal function. Propranolol is almost completely metabolized in the liver, and esmolol is biodegraded by esterases in the cytosol of red blood cells, so their effects are not prolonged in patients with abnormal or absent renal function.^[99] The calcium channel blocking agents nifedipine, verapamil, and diltiazem are extensively metabolized in the liver to pharmacologically inert

2187

During anesthesia, if a reduction in arterial blood pressure is necessary, several drugs may be used safely. Trimethaphan (Arfonad), a ganglionic blocking agent, is suitable because its action is terminated enzymatically rather than by renal excretion. Nitroglycerin is useful because it is metabolized rapidly, with less than 1% excreted unchanged in urine.^[104] Sodium nitroprusside has had a resurgence in use since its initial introduction as a hypotensive drug in the 1920s. Cyanide is an intermediate in the metabolism of sodium nitroprusside, with thiocyanate being the final metabolic product. Although cyanide toxicity as a complication of sodium nitroprusside therapy is well described, ^[105] it is less well appreciated that thiocyanate is also potentially toxic. The half-life of thiocyanate is normally more than 4 days, and it is prolonged in patients with renal failure.^[105] Hypoxia, nausea, tinnitus, muscle spasm, disorientation, and psychosis have been reported when thiocyanate levels exceed 10 mg/100 mL. Thus, sodium nitroprusside is less desirable for prolonged administration than either trimethaphan or nitroglycerin is. Hydralazine is slower acting than the other three drugs discussed previously. Its action is terminated by hydroxylation and subsequent glucuronidation in the liver, with approximately 15% excreted unchanged in urine.^[106] The elimination half-life of hydralazine is prolonged in uremic patients, so caution is required when it is administered.^[107] After a single intravenous dose of 0.5 mg/kg of labetalol, the volume of distribution, clearance, and elimination half-life were similar in patients with ESRD and normal volunteers.^[108] Esmolol is independent of renal function because it is metabolized by red blood cell cytosol esterases.^[109]

If administration of a vasopressor is necessary, a direct α-adrenergic-stimulating drug such as phenylephrine will be effective. Unfortunately, this type of vasopressor causes the greatest interference with renal circulation. Although β-adrenergic-stimulating drugs such as isoproterenol maintain heart and brain perfusion without renal vasoconstriction, they also increase myocardial irritability. Therefore, when possible, it is best to substitute simple measures such as blood volume expansion for drug therapy. If these measures are inadequate, β-adrenergic-stimulating drugs or dopamine should be used.