Aminoglycoside GERALD
B.
APPEL,
Nephrotoxicity
M.D. New York, New York
The high incidence of associated nephrotoxicity represents an important concern in the use of aminoglycoside antibiotics, which have been implicated as one of the primary causes of druginduced acute renal failure: Several factors, including the underlying health of the patient, criteria used to define nephrotoxicity, and the specific aminoglycoside administered, may contribute to the nephrotoxic potential of these agents. The development of aminoglycoside-induced nephrotoxicity is a complex problem. These drugs aupear to be only minimally metabolized within the body and undergo nearly exclusive renal excretion, primarily by glomerular filtration. Ultimately, reabsorption and accumulation within the kidney results in proximal tubular cell damage; several possible mechanisms have been proposed, both for the development of such.tiell damage and for its subsequent role in the evolution of nephrotoxicity. The pathology and the clinical pattern of aminoglycoside-induced kidney damage have been extensively studied in animal models and in humans. Although the data often conflict, many of these studies have attempted to identify some of the factors associated virith a higher risk for aminoglycoside nephrotoxicity. Of the factors generally agreed upon to influence risk, correction of volume depletion and diminished renal perfusion, as well as dose adjustment for level of renal function, have been identified as critical measures for prevention of renal damage by aminoglycosides. Recent studies have indicated that newer agents, such as third-generation cephalosporins and aztreonam, often may be as therapeutic and cost-effective as the aminoglycosides without the nephrotoxicity associated with the latter agents. Clearly, such a safe and effective alternative, if confirmed, would be preferable to aminoglycoside therapy in patients at high risk for nephrotoxicity.
From the Department of Clinical Nephroiogy, Columbia-Presbyterian Hospital, and the Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York. Requests for reprints should be addressed to Gerald 6. Appel, M.D., Department of Clinical Nephrology, Columbia-Presbyterian Medical Center, 622 West 168th Street, New York, New York 10032.
3C16S
March 23, 1990
The American Journal of Medicine
minoglycoside antibiotics remain among the most A widely used antimicrobial agents. Their broad spectrum of activity against aerobic gram-negative and gram-positive organisms, their chemical stability, and their rapid bactericidal action have often made them first-line drugs in a variety of life-threatening situations [1,2]. Unfortunately, adverse side effects, especially nephrotoxicity; have been associated with their use [3-g]? To clarify the clinical significance of this aminoglycoside-associated nephrotoxicity, its incidence, development, pathology, clinical features, and potential prevention are discussed in this presentation. INCIDENCE OF DRUG-INDUCED NEPHROTOXICITY In virtually all recent studies of acute renal failure (ARF) in humans, medication-induced nephrotoxicity has been cited as a major cause, and antibioticsespecially aminoglycoside antibiotics-have been among the most common offending agents. For example, in one study of over 2,000 hospitalized medical and surgical patients at risk for renal damage, 4.9 percent experienced renal insufficiency [lo]. Aminoglycosides were one of the two primary nephrotoxic causes reported and,accounted for 7 percent of the episodes of renal insufficiency. Two other studies have similarly identified aminoglycosides as the antibiotic class producing the most nephrotoxicity [ll, 121. Finally, two recent studies that analyzed risk factors for ARF found exposure to aminoglycoside antibiotics to be a major risk factor [13,14]. Although the precise contribution of aminoglycoside antibiotics to the renal damage reported in some of these studies is debatable, there can be no doubt that the aminoglycosides possess considerable nephrotoxic potential in clinical use. FACTORS INFLUENCING NEPHROTOXIC INCIDENCE In studies of populations of patients treated with aminoglycoside antibiotics, the incidence of nephrotoxicity has ranged from less than 2 percent to almost 50 percent [3,4,15-1’71. There are several reasons for these discrepancies. First, the study of different types of patient populations will expectedly produce markedly different incidences of nephrotoxicity. Relatively healthy patients receiving a short course of an aminoglycoside for a resistant urinary infection, for example, may be expected to have a low incidence of nephrotoxicity, whereas hypotensive, septic patients and those receiving a prolonged, high-dose course of therapy for endocarditis or osteomyelitis are likely to have a much higher incidence of adverse renal reactions. The combined results of large preclinical trials showed that a variety of aminoglycosides have produced renal damage in only 1 to 3 percent of patients [3,4]. It has been noted, however, that the highest incidence of renal damage has always occurred in severely ill populations [16,17]. The second major factor affecting the incidence of nephrotoxicity is the criteria used to judge renal dam-
Volume 88 (suppl 3C)
SYMPOSIUM
age. In the past, nephrotoxicity was defined in some trials by the presence of cylindruria and other urinary sediment changes, the excretion of renal tubular cell enzymes, or small changes in the serum creatinine [3,4]. Not all investigators., however, agreed that such changes represented sigmficant alteration of the glomerular filtration rate (GFR). Subsequent welldesigned trials have used significant changes in the serum creatinine as the marker for renal damage and consequently report a lower incidence of adverse renal reactions than did prior studies that used more sensitive but less specific tests. Finally., the specific aminoglycoside used may influence the incidence of nephrotoxicity. For example, of the commercially available aminoglycosides, neomycin is the most nephrotoxic, and streptomycin, the least nephrotoxic [3]. The nephrotoxic potential of many other currently available aminoglycosides has been widely debated. Clearly, multiple factors can influence the incidence of nephrotoxicity reported in any given study. Despite the variability, however, an overall incidence of nephrotoxicity in between 5 and 10 percent of patient courses has been reported in the majority of recent studies. Given the number of courses of aminoglycosides used in the United States annually, the total number of adverse renal reactions is in the many thousands [3,4,6]. DEVELOPMENT OF AMINOGLYCOSIDE NEPHROTOXICITY Extensive data from a number of experimental in vitro and in vivo animal models are available to clarify the development of nephrotoxicity attributable to aminoglycoside therapy [3 75 78 ,9 718,191. Overall, these studies show that aminoglycosides are not significantly metabolized within the body and only a small percentage of the drug is plasma-protein bound. In addition, many clearance studies in animals and humans indicate that these agents are largely excreted by the kidney. The major pathway of excretion appears to be glomerular filtration, whereas tubular secretion plays, at most, a minor role in aminoglycoside elimination. Once filtered, a portion of the cationic aminoglycoside antibiotic binds to phospholipid receptors on the brush border of cells of the proximal convoluted tubule and pars recta; the drug is subsequently reabsorbed, largely by pmocytosis, and accumulates in the lysosomes of the proximal tubular cell as well as in other subcellular compartments. The net reabsorption of aminoglycoside by the kidney results in high concentrations of the drug within the renal cortex. This forms a poorly exchangeable drug pool, and whereas the serum half-life of aminoglycosides is in the range of several hours, the renal tissue half-life of these drugs is several hundred hours. Thus an aminoglycoside may be slowly excreted in the urine for weeks following completion of a course of the drug, even after serum levels are undetectable. In animal studies, there has not always been a good correlation between the renal tissue concentration of aminoglycoside and nephrotoxicity. This has been attributed both to compartmentalization of the aminoglycoside within the proximal tubular cells and to varying intrinsic nephrotoxic potential of the individual aminoglycosides administered [ZO-221. Ultimately, proximal tubular cell damage occurs.
ON AZTREONAM/APPEL
Hypothesized mechanisms for this damage include alterations in plasma membrane structure and function via binding of aminoglycosides to plasma phospholipids; activation of cellular phospholipases; alterations of phospholipid accumulation and metabolism, resulting in an impaired ability of the “stuffed” lysosomes to perform their normal function of degrading and phagocytosing cellular debris; damage to sensitive mitochondrial respiratory mechanisms, ultimately leading to a decline in cellular adenosine triphosphate production; and alterations of sodium-potassium adenosine triphosphatase activity [5,8,191. Although all of these actions may occur, it is not yet clear which mechanism is primary and which are epiphenomena in the critical process of cell damage. Similarly, the ultimate step to irreversible cell death may involve changes in subcellular calcium compartmentalization, but the precise mechanism is unknown. Once proximal cell damage occurs, it must be translated into a decline in GFR to produce ARF. Again, experimental studies provide conflicting explanations of the mechanism of this decline in GFR [5,9,18,23,24]. Some investigators have proposed that a release of vasoconstrictive hormones results in alterations in renal blood flow; others hypothesize that damage to proximal renal tubular integrity leads to back-leak of fluid and waste products across the damaged epithelium. There are good micropuncture data supporting the proposal that obstruction to the individual nephrons results in increased hydrostatic pressure within the tubular lumen and thereby causes a decline in the net ultrafiltration pressure within the glomerular capillaries [23]. Finally, some investigators have found a decline in the glomerular ultrafiltration coefficient in experimental models of aminoglycoside nephrotoxicity [24]. This may be due to either diminished surface area for filtration or altered permeability of the glomerulus to the passage of filtrate. It should be stressed that none of these mechanisms is mutually exclusive; i.e., obstruction of the tubule will cause a greater decline in glomerular filtration in the presence of back-leak and an altered permeability of the glomerulus. It is also possible that in different settings, i.e., in different experimental models, the noted decline in GFR may be attributable to different components of these proposed mechanisms. Recent studies of the mechanism of nephrotoxicity focus on the subcellular sites of damage as well as on the way in which this cellular damage translates into a decline in GFR. PATHOLOGY OF AMINOGLYCOSIDE NEPHROTOXICITY Numerous experiments using laboratory animals, as well as several studies of aminoglycoside-damaged human kidneys, have examined the pathology involved in nephrotoxicity due to aminoglycosides [35,8,19,25]. Light-microscopy studies of both animal and human kidneys indicate, for example, that necrosis of the proximal tubular cells is dose related. In animal models, this damage is most prominent in the Sl and S2 segments of the proximal tubule. Earliest changes noted by electron microscopy include alterations in the size, number, and structure of the lysosomes. Cytosegregosomes, which are altered secondary lysosomes containing electron-dense lamellated whirl-like configurations known as myeloid bodies,
March 23, 1990
The American Journal of Medicine
Volume 88 (suppl 3C)
3c17s
SYMPOSIUM
ON AZTREONAM
/ APPEL
TABLE I Clinical Pattern of Aminoglycoside
Nephrotoxicity
Early reversible alterations Lysosomal enzymuria Urinary sediment changes (e.g., cylindruria) Polyuria Nonoliguric ARF Decreased GFR Rise in BUN, plasma creatinine High urinary Na+ concentration and FE Nat BUN = blood urea nitrogen: Na = sodium; FE Na = fractional excretion of sodium.
TABLE II Proposed Risk Factors for Aminoglycoside
Nephrotoxicity
Initial patient status Advanced age Prior renal dysfunction Female gender Prior courses of aminogfycoside therapy Liver disease Volume depletion, hypotension, shock Drug administration Large total dose Long duration of therapy Frequent dosing intervals Choice of aminoglycoside in high-risk patient Course of therapy High peak or trough serum levels Concurrent nephrotoxins
TABLE Ill Concurrent Nephrotoxins That May Potentiate Aminoglycoside Nephrotoxicity Furosemide and other potent diuretics-only if they induce volume depletion Cephalosporins-conflicting data in animal models and humans Others Radiographiccontrast agents Vancomycin Amphotericin B Methoxyflurane Cisplatin Cyclosporine Nonsteroidal anti-inflammatory drugs
appear in significant numbers in the cytoplasm of the proximal tubular cells. With further progression, alterations of the brush-border microvilli appear, along with vacuolation of the cytoplasm and alterations of the cisternae of the endoplasmic reticulum. At this point, the lumen of the tubules contains fragments of degenerating cells, myeloid bodies, and cytoplasmic debris. Mitochondrial swelling becomes prominent, and epithelial cell necrosis ensues. Ultimately, the tubular lumen fills with sloughed degenerating cells and cellular debris. Few changes noted in the glomeruli are notable on light microscopy or routine electron microscopy, but alterations of the endothelium and its fenestrae have been noted on scanning electron microscopy. By light microscopy, patchy inflammatory infiltrates may then be seen in the renal cortex, and foci of regenerating proximal tubular cells are often noted. CLINICAL FEATURES OF NEPHROTOXICITY The clinical pattern of aminoglycoside nephrotoxicity has been well studied, both in animal models and in humans (Table I) [3-51. The initial appearance of 3C-18s
March 23, 1990
The American Journal of Medicine
brush-border and lysosomal enzymes in the urine and granular casts in the urinary sediment is followed by the development of polyuria, perhaps resulting from aminoglycoside interference with vasopressin activity. A subsequent decline in the GFR is associated with a rise in both blood urea nitrogen concentration and plasma creatinine. The urinary sodium concentration and fractional excretion of sodium also are typically high (greater than 40 mEq/liter and greater than 1 percent, respectively), as is common with other forms of acute tubular necrosis. Renal failure is typically nonoliguric, without any decline in the urinary volume, and often occurs seven to 10 days into the course of treatment. It is often detected by an asymptomatic rise in the serum creatinine. Patients who are septic and hypotensive are more likely to suffer oliguric renal failure early in the course of treatment. RISK FACTORS FOR ARF A number of recent studies have tried to define populations at high risk for development of ARF during a course of aminoglycoside therapy [26-301. Of those features studied, the following have been associated with a higher incidence of renal damage: prolonged duration of therapy; previous courses of aminoglycoside therapy; frequent dosing intervals; greater total aminoglycoside dose; advanced patient age; female gender; presence of hypotension, hypovolemia, or shock; presence of liver disease; and pre-existing renal disease (Table II) K&14,15,26-341. A number of other therapeutic agents may potentiate the nephrotoxicity when given concurrently with an aminoglycoside (Table III), and finally, specific aminoglycosides have been associated with a higher or lower incidence of nephrotoxicity. Unfortunately, the results of many studies raise questions about all of these crucial clinical risk factors [20]. It has been proposed, for example, that preexisting renal disease and advancing age of the patient may increase risk of nephrotoxicity only if the dose of the aminoglycoside is not appropriately adjusted to reflect the diminished renal function of this population. Similarly, a prior course of aminoglycosides might increase risk only if the therapy was recent, which would allow rapid saturation of tissue sites as a result of the prolonged tissue half-life of these drugs. The impact of some other identified risk factors is also uncertain due to conflicting data among the studies. For example, although female gender proved to be a risk factor in one clinical study of aminoglycoside nephrotoxicity [27], this was not confirmed by other studies in human or animal models, and men have certainly not been found to be immune from nephrotoxicity when treated with this class of drugs [30-33,35521. In many studies, large total doses of aminoglycosides given for prolonged time intervals appeared to correlate with nephrotoxicity. But whether more frequent dosing rather than higher doses at less frequent intervals would be less nephrotoxic in humans, as it has been shown to be in some animal studies, has yet to be proved [343. Similarly, although both high peak and trough serum levels of aminoglycoside have been claimed to correlate with nephrotoxicity, the correlation is clearly imprecise, and in many cases, serum levels are poor predictors of renal outcome. Of those risk factors associated with the patient’s status at the time of drug administration, volume depletion, hypotension, and/or shock seem to be most consistently
Volume 88 (suppl 3C)
SYMPOSIUM
associated with the development of aminoglycoside nephrotoxicity. The presence of liver disease is also a risk factor for many forms of renal failure, perhaps because of alterations in renal profusion, it may result in greater reabsorption of aminoglycoside by the proximal tubule. Nephrotoxic Impact of Combination Therapy
Concurrent use of a number of nephrotoxic medications and therapeutic agents also has been claimed to increase the risk of aminoglycoside nephrotoxicity (Table III). Of these, the use of cephalosporin antibiotics has raised the most controversy. Clearly, cephaloridine and a number of other cephalosporins act as dose-related nephrotoxins to the proximal tubule, and a number of studies have claimed synergistic nephrotoxicity between aminoglycosides and cephalosporins [3]. However, newer cephalosporins are far less nephrotoxic than older agents [31. Moreover, in animal studies it has been difficult to prove synergistic nephrotoxicity between these agents [531. In fact, a protective effect from concurrent cephalosporin use has been noted in some animal studies [531. Likewise, aminoglycoside nephrotoxicity has not been shown to be increased significantly by concurrent cephalosporin use in humans [3]. Thus, although it is clearly advisable to exercise caution when combining an aminoglycoside with a cephalosporin, the nephrotoxic potential appears to reside primarily with the aminoglycoside. It has been suggested that loop diuretics and other potent diuretics also may enhance aminoglycoside nephrotoxicity. Based on well-studied animal models, it appears that this occurs only with diuretics that induce volume contraction. Another potential nephrotoxic drug, vancomycin, causes renal damage in less than 5 percent of treated patients. Recent retrospective studies, however, suggest that as many as 35 percent of patients treated with the combination of vancomycin plus an aminoglycoside will experience nephrotoxicity, confirming animal data of synergistic damage to the kidneys between these two classes of drugs 154,551. Other agents associated with an increased risk of nephrotoxicity include amphotericin B, cisplatin, cyclosporine, nonsteroidal anti-inflammatory drugs, and radiographic contrast agents. All are known nephrotoxins, and when possible, their use should be avoided in combination with aminoglycoside therapy. Comparative Nephrotoxicity of Various Aminoglycosides
Probably the most debated issue concerning aminoglycoside antibiotics and the kidneys is which agents of this class of drugs produce significantly less clinical renal damage than others [15,26,35,36]. Streptomycin, for example, clearly has only minimal nephrotoxicity potential. Neomycin, recognized as the most nephrotoxic agent, is only administered for irrigation, topically, and as a bowel-sterilizing agent. Sisomicin and kanamycin, although more nephrotoxic than gentamicin, are not often used in the United States. Of the widely available aminoglycosides, gentamicin has appeared to be more nephrotoxic than tobramycin or netilmicin in a vast number of in vitro and in vivo animal studies [3-5,30-39, 53-551. In clinical trials, netilmicin has been comparable to tobramycin in degree of nephrotoxicity [40,41]. Amikacin is at least no more nephrotoxic than gentamicin and is perhaps less so [32,42,43]. Since the former is restricted for use in
ON AZTREONAM / APPEL
patients with resistant organisms, any slight differences in nephrotoxic potential are probably of little clinical importance. A number of well-controlled clinical trials have confirmed these differences among the aminoglycosides, but other studies have been unable to establish different degrees of nephrotoxicity for some of the individual agents [‘7,17,26,31-33,36,40,42,44--521. In low-risk patients., there is probably little significant difference in the risk of renal damage between gentamicin and tobramycin or netilmicin. In these patients, any renal dysfunction is usually mild, reversible, and associated with little residual damage. In contrast, in patients at high risk for nephrotoxicity, even small differences in nephrotoxic potential between individual drugs may potentiate other risk factors and add to the burden of renal damage. PREVENTION OF AMINOGLYCOSIDE-INDUCED RENAL DAMAGE
Correction of all of those factors identified as potentially leading to a greater risk of renal dysfunction is critical to the prevention of renal damage associated with aminoglycoside therapy. Of these factors, correction of volume depletion and/or congestive heart failure and diminished renal perfusion is probably the most significant. Also important are adjustment of dosage to compensate for the patient’s level of renal function and awareness of the decreased GFR that may be present in elderly patients despite plasma ereatinine in the normal range for the laboratory. Measurement of serum levels of the aminoglycoside may also be useful. Since these drugs are excreted only by glomerular filtration, anything that impairs renal function can be expected to also cause a rise in serum levels. Thus a rising trough level of drug (drawn just prior to administration of the next dose) indicates renal dysfunction, and the subsequent dose must be adjusted accordingly. Clearly, to minimize the risk of renal damage, using the shortest appropriate course of the aminoglycoside and avoiding concurrent treatment with other nephrotoxic agents also are advisable. It has yet to be determined whether employing less-nephrotoxic aminoglycosides or altering dose administration schedules in higher-risk patients will truly minimize renal damage, although such an effect has been suggested by recent studies 1341. Several recent controlled studies have shown that newer third-generation cephalosporins, as well as other agents, such as aztreonam, are less nephrotoxic than aminoglyeosides when used in similar situations 156-581. Recent studies also have examined the important question of the financial cost of aminoglycoside-related nephrotoxicity [30,59-611. In a recent study of 1,756 patients at six Philadelphia hospitals, there was a 7.3 percent incidence of aminoglycoside nephrotoxicity [611. The additional cost to the 129 patients in whom nephrotoxicity occurred was over $2,500 per patient. When this cost of nephrotoxicity was factored among all 1,756 patients receiving the aminoglycoside, the added cost for each patient was $183. Although alternative treatments may seem to be more expensive than aminoglycosides at first look, consideration of this cost of nephrotoxicity, added to the expense of measuring serum aminoglycoside levels and renal function to monitor for nephrotoxicity, results in competitive cost-effectiveness for the two
March 23, 1990
The American Journal of Medicine
Volume 88 (suppl 3C)
3c19s
SYMPOSIUM
ON AZTREONAM
/ APPEL
[61]. Clearly, if non-nephrotoxic antibiotics, such as third-generation cephalosporins or aztreonam, are equally efficacious and equally cost-effective in treating a given infection in a patient at high risk for renal damage, use of the alternative agent would be preferable to use of the aminoglycoside.
COMMENTS The aminoglycosides remain effective and widely used antimicrobial agents, although their risk of associated nephrotoxicity in large unselected populations is probably in the range of 5 to 10 percent. Extensive study of patient risk factors for nephrotoxicity has resulted in a better understanding of its development and recognition that some features are potentially correctable. Correction of such risk factors, careful monitoring of patients, and consideration of the use of newer antimicrobial agents in patients at high risk for nephrotoxicity may reduce the incidence of this common form of renal damage.
REFERENCES 1. Sigenthaler WE, Bonetti A, Lughy R: Aminoglycoside antibiotics in infectious disease: an overview. Am J Med 1986; 80 (suppl 68): 2-14. 2 Moellering RC: Have the new beta-lactams rendered the aminoglycosides obsolete for the treatment of serious nosocomial infections? Am J Med 1986; 80 (suppl 68): 44-47. 3. Appel GB, Neu HC: The nephrotoxicity of antimicrobial agents. N Engl J Med 19n; 296: 663-670, 722-728, 784-781. 4. Appel GB, Neu HC: Gentamicin 1978. Ann Intern Med 1978; 89: 528-538. 5. Humes MD, Weinberg JM, Knaus TC: Clinical and psychophysiologic aspects of aminoglycoside antibiotics in infectious disease: an overview. Am J Kidney Dis 1982; 2: 5-29. 6. Whelton A: Therapeutic initiative for avoidance of aminoglycoside toxicity. J Clin Pharmacol 1985; 25: 67-81. 7. Lietman PS, Smith CR: Aminoglycoside nephrotoxicity in humans. Rev Infect Dis 1983; 5 (suppl 2): 284-292. 8. Bennett WM, Lull F, Porter GA: Pathogenesis of renal failure due to aminoglycosides and contrast media used in roentgenography. Am J Med 1980; 69: 767-774. 9. Appel GB, Siegel NJ, Appel AS, Hayslett JP: Studies on the mechanism of nonoliguric experimental acute renal failure. Yale J Biol Med 1981; 54: 273-281. 10. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT: Hospital-acquired renal insufficiency: a prospective study. Am J Med 1983; 74: 243-248. 11. McMurray SD, Luft RC, Maxwell DR, eta/: Prevailing patterns and predictor variables in patients with acute tubular necrosis. Arch Intern Med 1978; 138: 950-955. 12. Kleinknect D, Landais P, Goldfarb B: Pathophysiology and clinical aspects of druginduced tubular necrosis in man. Contrib Nephrol 1987; 55: 145-158. 13. Rasmussen MH, lbels LS: Acute renal failure. Multivariate analysis of causes and risk factors. Am J Med 1982; 73: 211-218. 14. Shusterman N, Strom BL, Murray TG, Morrison G, West SL, Maislin G: Risk factors and outcome of hospital-acquired acute renal failure. Am J Med 1987; 83: 65-71. 15. Smith CR, Lietman PS: Comparative clinical trials of aminoglycosides. In: Whelton A, Neu HC, eds. The aminoglycosides. New York, Basel: Marcel Dekker, Inc., 1982; 497-509. 16. Tablan OC, Reyes MD, Rintelmann WF, Lerner AM: Renal and auditory toxicity of high dose, prolonged therapy with gentamicin and tobramycin in Pseudomonas endocarditis. J Infect Dis 1984; 149: 257-263. 17. Keys TF, Kurtz SB, Jones JD, Muller SM: Renal toxicity during therapy with gentamicin or tobramycin. Mayo Clin Proc 1981; 56: 556-598. 18. Appel GB: Aminoglycoside nephrotoxicity: physiologic studies of the sites of nephron damage. In: Whelton A, Neu HC, eds. The aminoglycosides. New York, Base1 Marcel Dekker Inc., 1982. 19. Silverblatt F: Pathogenesis of nephrotoxicity of cephalosporins and aminoglycosides: a review of current concepts. Rev Infect Dis 1982; 4 (suppl): 360-365. 20. Tulkens PM: Experimental studies on nephrotoxicity of aminoglycosides at low doses. Am J Med 1986; 80 (suppl 68): 105-114. 21. DeBroe ME, Giuliano RA, Verpoeten GA: Choice of drug and dosage regimen. Two important factors for aminoglycoside nephrotoxicity. Am J Med 1986; 80 (suppl 68): 114-118. 22. Brier ME, Mayer PR, Brier RA, et a/: Relationship between rat renal accumulation of gentamicin, tobramycin, and netilmicin and their nephrotoxicities. Antimicrobial Agents Chemother 1985; 27: 812-816. 23. Neugarten J, Aynedjian HS, Bank N: The role of tubular obstruction in acute renal failure due to gentamicin. Kidney Int 1983; 24: 330-335. 24. Baylis C, Rennke HR, Brenner BM: Mechanisms of the defect in glomerular ultrafiltration associated with gentamicin administration. Kidney Int 19R 12: 344-353. 25. Olsen S: Acute tubular necrosis and toxic renal injury. In: Tisher CC, Brenner BM, eds. Textbook of renal pathology. Philadelphia: JB Lippincott, 1989.
3C-20s
March 23, 1990
The American Journal of Medicine
26. Meyer RD: Risk factors and comparisons of clinical nephrotoxlcity of aminoglycosides. Am J Med 1986; 80 (suppl 6B): 119-125. 27. Moore RD, Smith CR, Lipsky JJ, Mellits ED, Lietman PS: Risk factors for nephrotoxicity in patients treated with aminoglycosides. Ann Intern Med 1984; 100: 352-357. 28. Sawyers CL, Moore RD, Lerner SA, Smith CR: A model for predicting nephrotoxicity in patients treated with aminoglycosides. J Infect DIS 1986; 153: 1062-1068. 29. Lam YWF, Arana CJ, Shikuma LR, Rotshafer JC: The clinlcal utility of published monogram to predict aminoglycoside nephrotoxicity. JAMA 1986; 255: 639-642. 30. Gelfand J, Neu HC, Morelli H, Appel GB: Safety and cost efficacy of a restricted policy for aminoglycoside antibiotic usage (abstr). Kidney Int 1985; 27: 138. 31. Kahlmeter G, Hallberg T, Kamme C: Gentamicin and tobramycin in patients with various infections-nephrotoxicity. J Antimicrob Chemother 1978; 4 (suppl A): 47-52. 32. Matzke GR, Lucarotti RL, Shapiro HS: Controlled comparison of gentamicln and tobramycin nephrotoxicity. Am J Nephrol 1983; 3: 11-17. 33. Lerner SA, Schmitt BA, Seligsohn R, Metz GL: Comparative study of toxicity and nephrotoxicity in patients randomly assigned treatment with amikacin and gentamicin. Am J Med 1986; 80 (suppl 6B): 98-104. 34. Kapusnik JE, Sands MA: Challenging conventional amlnoglycoside dose regimen. Am J Med 1986; 80 (suppl 68): 179-181. 35. Evans DA, Buring J, Mayrent S, Rosner B, Colton T, Mennekens C: Qualitative overview of randomized trials of aminoglycosides. Am J Med 1986; 80 (suppl 68): 39-43. 36. Cone LA: A survey of prospective controlled clinical trials of gentamicin, tobramycin, amikacin, and netilmicin. Clin Ther 1982; 5: 155-162. 37. Soberon L, Bowman RL, Pastoriza-Munoz E, et at Comparative nephrotoxicities of gentamicin, netilmicin, and tobramycin in the rat. J Pharmacol Exp Ther 1979; 210: 334343. 38. Gilbert DN, Plamp C, Starr P, et at Comparative nephrotoxicity of gentamicin and tobramycin in rats. Antimicrob Agents Chemother 1979; 13: 34-40. 39. Luft FC, Bloch R, Sloan RS, et al: Comparative nephrotoxicity of aminoglycoside antibiotics in rats. J Infect Dis 1978; 138: 541-545. 40. Gatell JM, San Miguel JG, Aranago V, et at Prospective and randomized double-blind comparison of nephrotoxicity and auditory toxicity of tobramycin and netilmicin. Antimicrob Agents Chemother 1984; 26: 766-769. 41. Lerner AM, Cone LA, Jansen W, et a/; Randomized, controlled trial of the comparative efficacy, auditory toxicity and nephrotoxicity of tobramycin and netilmicin. Lancet 1983; I: 1123-1125. 42. Gatell JM, San Miguel JG, Zamora L, et at Comparison of the nephrotoxicity and auditory toxicity of tobramycin and amikacin. Antimicrob Agents Chemother 1983; 23: 897-901. 43. Smith CR, Baughman KL, Edwards CQ, et at Controlled comparison of amikacin and gentamicin. N Engl J Med 19n; 296: 349-354. 44. Frimodt-Moller, Maigaard S, Madsen PO: Comparative nephrotoxicity among aminoglycosides and beta-lactam antibiotics. Infection 1980; 8 (suppl): 283-289. 45. Plaut ME, Schentag JJ, Jusko WJ: Aminoglycoside nephrotoxicity: comparative assessment in critically ill patients. J Med 1979; 10: 257-286. 46. Schentag JJ, Plaut ME, Cerra FB: Comparative nephrotoxicity of gentamicin and tobramycin: pharmacokinetics and clinical studies in 201 patients. Antimicrob Agents Chemother 1981; 19: 859-866. 47. Cerra FB, Plaut ME: Clinical and pharmacokinetic characteristics of amlnoglycoside nephrotoxicity in 201 critically ill patients. Antimicrob Agents Chemother 1982; 21: 721726. 48. Kumin GD: Clinical nephrotoxicity of tobramycin and gentamicin. JAMA 1980; 244: 1808-1810. 49. Smith CR, Lipsky JJ, Laskin OL, et at Double-blind comparison of the nephrotoxicity and auditory toxicity of gentamicin and tobramycin. N Engl J Med 1980; 302: 1106-1109. 50. Fong IN, Fenton RS, Bird R: Comparative toxicity of gentamicin versus tobramycln: randomized prospective study. J Antimicrob Chemother 1981; 7: 81-88. 51. Bock BV, Edelstein PH, Meyer RD: Prospective comparative study of efficacy and toxicity of netilmicin and amikacin. Antimicrob Agents Chemother 1980; 17: 217-225. 52. Kahlmeter G, Dahlager JI: Aminoglycoside toxicity-a review of clinical studies published between 1975 and 1982. J Antimicrob Chemother 1984; 13 (suppl A): 9-22. 53. Luft FC: Cephalosporin and aminoglycoside interactions: clinical and toxicologic implications. In: Whelton A, Neu HC, eds. The aminoglycosides. New York: Marcel Dekker Inc., 1982. 54. Appel GB, Given DB, Levine LR, Cooper GL: Vancomycin and the kidney. Am J Kidney Dis 1986: 8: 75-80. 55. Wold TJ, Turnipseed SA: Toxicology of vancomycin in laboratory animals. Rev Infect Dis 1981: 3: 274-235. 56. Smith CR, Ambinder R, Lipsky JJ, eta/: Cefotaxime compared with nafcillin plus tobramycin for serious bacterial infections. Ann Intern Med 1984; 101: 469-477. 57. Neu HC: Aztreonam: a novel monocyclic beta lactam antibiotic. Am J Med 1985: 78 (suppl 2A): 57-65. 58. Henry SA. Bendush CB: Aztreonam: worldwide overview of the treatment of patients with gram-negative infections. Am J Med 1985; 78 (suppl 2A): 57-65. 59. Gladen HE: Cost-effect of aminoglycoside therapy in surgical patients. Am J Med 1986; 80 (suppl 68): 228-233. 60. Holloway JJ, Smith CR, Moore RD, eta/; Comparative cost effectiveness of gentamicin and tobramycin. Ann Intern Med 1984; 101: 764-769. 61. Eisenberg JM, Koffer H, Glick HA, et at What is the cost of nephrotoxicity associated with aminoglycosides? Ann Intern Med 1987; 107: 900-909.
Volume 88 (suppl 3C)
B.
APPEL,
Nephrotoxicity
M.D. New York, New York
The high incidence of associated nephrotoxicity represents an important concern in the use of aminoglycoside antibiotics, which have been implicated as one of the primary causes of druginduced acute renal failure: Several factors, including the underlying health of the patient, criteria used to define nephrotoxicity, and the specific aminoglycoside administered, may contribute to the nephrotoxic potential of these agents. The development of aminoglycoside-induced nephrotoxicity is a complex problem. These drugs aupear to be only minimally metabolized within the body and undergo nearly exclusive renal excretion, primarily by glomerular filtration. Ultimately, reabsorption and accumulation within the kidney results in proximal tubular cell damage; several possible mechanisms have been proposed, both for the development of such.tiell damage and for its subsequent role in the evolution of nephrotoxicity. The pathology and the clinical pattern of aminoglycoside-induced kidney damage have been extensively studied in animal models and in humans. Although the data often conflict, many of these studies have attempted to identify some of the factors associated virith a higher risk for aminoglycoside nephrotoxicity. Of the factors generally agreed upon to influence risk, correction of volume depletion and diminished renal perfusion, as well as dose adjustment for level of renal function, have been identified as critical measures for prevention of renal damage by aminoglycosides. Recent studies have indicated that newer agents, such as third-generation cephalosporins and aztreonam, often may be as therapeutic and cost-effective as the aminoglycosides without the nephrotoxicity associated with the latter agents. Clearly, such a safe and effective alternative, if confirmed, would be preferable to aminoglycoside therapy in patients at high risk for nephrotoxicity.
From the Department of Clinical Nephroiogy, Columbia-Presbyterian Hospital, and the Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York. Requests for reprints should be addressed to Gerald 6. Appel, M.D., Department of Clinical Nephrology, Columbia-Presbyterian Medical Center, 622 West 168th Street, New York, New York 10032.
3C16S
March 23, 1990
The American Journal of Medicine
minoglycoside antibiotics remain among the most A widely used antimicrobial agents. Their broad spectrum of activity against aerobic gram-negative and gram-positive organisms, their chemical stability, and their rapid bactericidal action have often made them first-line drugs in a variety of life-threatening situations [1,2]. Unfortunately, adverse side effects, especially nephrotoxicity; have been associated with their use [3-g]? To clarify the clinical significance of this aminoglycoside-associated nephrotoxicity, its incidence, development, pathology, clinical features, and potential prevention are discussed in this presentation. INCIDENCE OF DRUG-INDUCED NEPHROTOXICITY In virtually all recent studies of acute renal failure (ARF) in humans, medication-induced nephrotoxicity has been cited as a major cause, and antibioticsespecially aminoglycoside antibiotics-have been among the most common offending agents. For example, in one study of over 2,000 hospitalized medical and surgical patients at risk for renal damage, 4.9 percent experienced renal insufficiency [lo]. Aminoglycosides were one of the two primary nephrotoxic causes reported and,accounted for 7 percent of the episodes of renal insufficiency. Two other studies have similarly identified aminoglycosides as the antibiotic class producing the most nephrotoxicity [ll, 121. Finally, two recent studies that analyzed risk factors for ARF found exposure to aminoglycoside antibiotics to be a major risk factor [13,14]. Although the precise contribution of aminoglycoside antibiotics to the renal damage reported in some of these studies is debatable, there can be no doubt that the aminoglycosides possess considerable nephrotoxic potential in clinical use. FACTORS INFLUENCING NEPHROTOXIC INCIDENCE In studies of populations of patients treated with aminoglycoside antibiotics, the incidence of nephrotoxicity has ranged from less than 2 percent to almost 50 percent [3,4,15-1’71. There are several reasons for these discrepancies. First, the study of different types of patient populations will expectedly produce markedly different incidences of nephrotoxicity. Relatively healthy patients receiving a short course of an aminoglycoside for a resistant urinary infection, for example, may be expected to have a low incidence of nephrotoxicity, whereas hypotensive, septic patients and those receiving a prolonged, high-dose course of therapy for endocarditis or osteomyelitis are likely to have a much higher incidence of adverse renal reactions. The combined results of large preclinical trials showed that a variety of aminoglycosides have produced renal damage in only 1 to 3 percent of patients [3,4]. It has been noted, however, that the highest incidence of renal damage has always occurred in severely ill populations [16,17]. The second major factor affecting the incidence of nephrotoxicity is the criteria used to judge renal dam-
Volume 88 (suppl 3C)
SYMPOSIUM
age. In the past, nephrotoxicity was defined in some trials by the presence of cylindruria and other urinary sediment changes, the excretion of renal tubular cell enzymes, or small changes in the serum creatinine [3,4]. Not all investigators., however, agreed that such changes represented sigmficant alteration of the glomerular filtration rate (GFR). Subsequent welldesigned trials have used significant changes in the serum creatinine as the marker for renal damage and consequently report a lower incidence of adverse renal reactions than did prior studies that used more sensitive but less specific tests. Finally., the specific aminoglycoside used may influence the incidence of nephrotoxicity. For example, of the commercially available aminoglycosides, neomycin is the most nephrotoxic, and streptomycin, the least nephrotoxic [3]. The nephrotoxic potential of many other currently available aminoglycosides has been widely debated. Clearly, multiple factors can influence the incidence of nephrotoxicity reported in any given study. Despite the variability, however, an overall incidence of nephrotoxicity in between 5 and 10 percent of patient courses has been reported in the majority of recent studies. Given the number of courses of aminoglycosides used in the United States annually, the total number of adverse renal reactions is in the many thousands [3,4,6]. DEVELOPMENT OF AMINOGLYCOSIDE NEPHROTOXICITY Extensive data from a number of experimental in vitro and in vivo animal models are available to clarify the development of nephrotoxicity attributable to aminoglycoside therapy [3 75 78 ,9 718,191. Overall, these studies show that aminoglycosides are not significantly metabolized within the body and only a small percentage of the drug is plasma-protein bound. In addition, many clearance studies in animals and humans indicate that these agents are largely excreted by the kidney. The major pathway of excretion appears to be glomerular filtration, whereas tubular secretion plays, at most, a minor role in aminoglycoside elimination. Once filtered, a portion of the cationic aminoglycoside antibiotic binds to phospholipid receptors on the brush border of cells of the proximal convoluted tubule and pars recta; the drug is subsequently reabsorbed, largely by pmocytosis, and accumulates in the lysosomes of the proximal tubular cell as well as in other subcellular compartments. The net reabsorption of aminoglycoside by the kidney results in high concentrations of the drug within the renal cortex. This forms a poorly exchangeable drug pool, and whereas the serum half-life of aminoglycosides is in the range of several hours, the renal tissue half-life of these drugs is several hundred hours. Thus an aminoglycoside may be slowly excreted in the urine for weeks following completion of a course of the drug, even after serum levels are undetectable. In animal studies, there has not always been a good correlation between the renal tissue concentration of aminoglycoside and nephrotoxicity. This has been attributed both to compartmentalization of the aminoglycoside within the proximal tubular cells and to varying intrinsic nephrotoxic potential of the individual aminoglycosides administered [ZO-221. Ultimately, proximal tubular cell damage occurs.
ON AZTREONAM/APPEL
Hypothesized mechanisms for this damage include alterations in plasma membrane structure and function via binding of aminoglycosides to plasma phospholipids; activation of cellular phospholipases; alterations of phospholipid accumulation and metabolism, resulting in an impaired ability of the “stuffed” lysosomes to perform their normal function of degrading and phagocytosing cellular debris; damage to sensitive mitochondrial respiratory mechanisms, ultimately leading to a decline in cellular adenosine triphosphate production; and alterations of sodium-potassium adenosine triphosphatase activity [5,8,191. Although all of these actions may occur, it is not yet clear which mechanism is primary and which are epiphenomena in the critical process of cell damage. Similarly, the ultimate step to irreversible cell death may involve changes in subcellular calcium compartmentalization, but the precise mechanism is unknown. Once proximal cell damage occurs, it must be translated into a decline in GFR to produce ARF. Again, experimental studies provide conflicting explanations of the mechanism of this decline in GFR [5,9,18,23,24]. Some investigators have proposed that a release of vasoconstrictive hormones results in alterations in renal blood flow; others hypothesize that damage to proximal renal tubular integrity leads to back-leak of fluid and waste products across the damaged epithelium. There are good micropuncture data supporting the proposal that obstruction to the individual nephrons results in increased hydrostatic pressure within the tubular lumen and thereby causes a decline in the net ultrafiltration pressure within the glomerular capillaries [23]. Finally, some investigators have found a decline in the glomerular ultrafiltration coefficient in experimental models of aminoglycoside nephrotoxicity [24]. This may be due to either diminished surface area for filtration or altered permeability of the glomerulus to the passage of filtrate. It should be stressed that none of these mechanisms is mutually exclusive; i.e., obstruction of the tubule will cause a greater decline in glomerular filtration in the presence of back-leak and an altered permeability of the glomerulus. It is also possible that in different settings, i.e., in different experimental models, the noted decline in GFR may be attributable to different components of these proposed mechanisms. Recent studies of the mechanism of nephrotoxicity focus on the subcellular sites of damage as well as on the way in which this cellular damage translates into a decline in GFR. PATHOLOGY OF AMINOGLYCOSIDE NEPHROTOXICITY Numerous experiments using laboratory animals, as well as several studies of aminoglycoside-damaged human kidneys, have examined the pathology involved in nephrotoxicity due to aminoglycosides [35,8,19,25]. Light-microscopy studies of both animal and human kidneys indicate, for example, that necrosis of the proximal tubular cells is dose related. In animal models, this damage is most prominent in the Sl and S2 segments of the proximal tubule. Earliest changes noted by electron microscopy include alterations in the size, number, and structure of the lysosomes. Cytosegregosomes, which are altered secondary lysosomes containing electron-dense lamellated whirl-like configurations known as myeloid bodies,
March 23, 1990
The American Journal of Medicine
Volume 88 (suppl 3C)
3c17s
SYMPOSIUM
ON AZTREONAM
/ APPEL
TABLE I Clinical Pattern of Aminoglycoside
Nephrotoxicity
Early reversible alterations Lysosomal enzymuria Urinary sediment changes (e.g., cylindruria) Polyuria Nonoliguric ARF Decreased GFR Rise in BUN, plasma creatinine High urinary Na+ concentration and FE Nat BUN = blood urea nitrogen: Na = sodium; FE Na = fractional excretion of sodium.
TABLE II Proposed Risk Factors for Aminoglycoside
Nephrotoxicity
Initial patient status Advanced age Prior renal dysfunction Female gender Prior courses of aminogfycoside therapy Liver disease Volume depletion, hypotension, shock Drug administration Large total dose Long duration of therapy Frequent dosing intervals Choice of aminoglycoside in high-risk patient Course of therapy High peak or trough serum levels Concurrent nephrotoxins
TABLE Ill Concurrent Nephrotoxins That May Potentiate Aminoglycoside Nephrotoxicity Furosemide and other potent diuretics-only if they induce volume depletion Cephalosporins-conflicting data in animal models and humans Others Radiographiccontrast agents Vancomycin Amphotericin B Methoxyflurane Cisplatin Cyclosporine Nonsteroidal anti-inflammatory drugs
appear in significant numbers in the cytoplasm of the proximal tubular cells. With further progression, alterations of the brush-border microvilli appear, along with vacuolation of the cytoplasm and alterations of the cisternae of the endoplasmic reticulum. At this point, the lumen of the tubules contains fragments of degenerating cells, myeloid bodies, and cytoplasmic debris. Mitochondrial swelling becomes prominent, and epithelial cell necrosis ensues. Ultimately, the tubular lumen fills with sloughed degenerating cells and cellular debris. Few changes noted in the glomeruli are notable on light microscopy or routine electron microscopy, but alterations of the endothelium and its fenestrae have been noted on scanning electron microscopy. By light microscopy, patchy inflammatory infiltrates may then be seen in the renal cortex, and foci of regenerating proximal tubular cells are often noted. CLINICAL FEATURES OF NEPHROTOXICITY The clinical pattern of aminoglycoside nephrotoxicity has been well studied, both in animal models and in humans (Table I) [3-51. The initial appearance of 3C-18s
March 23, 1990
The American Journal of Medicine
brush-border and lysosomal enzymes in the urine and granular casts in the urinary sediment is followed by the development of polyuria, perhaps resulting from aminoglycoside interference with vasopressin activity. A subsequent decline in the GFR is associated with a rise in both blood urea nitrogen concentration and plasma creatinine. The urinary sodium concentration and fractional excretion of sodium also are typically high (greater than 40 mEq/liter and greater than 1 percent, respectively), as is common with other forms of acute tubular necrosis. Renal failure is typically nonoliguric, without any decline in the urinary volume, and often occurs seven to 10 days into the course of treatment. It is often detected by an asymptomatic rise in the serum creatinine. Patients who are septic and hypotensive are more likely to suffer oliguric renal failure early in the course of treatment. RISK FACTORS FOR ARF A number of recent studies have tried to define populations at high risk for development of ARF during a course of aminoglycoside therapy [26-301. Of those features studied, the following have been associated with a higher incidence of renal damage: prolonged duration of therapy; previous courses of aminoglycoside therapy; frequent dosing intervals; greater total aminoglycoside dose; advanced patient age; female gender; presence of hypotension, hypovolemia, or shock; presence of liver disease; and pre-existing renal disease (Table II) K&14,15,26-341. A number of other therapeutic agents may potentiate the nephrotoxicity when given concurrently with an aminoglycoside (Table III), and finally, specific aminoglycosides have been associated with a higher or lower incidence of nephrotoxicity. Unfortunately, the results of many studies raise questions about all of these crucial clinical risk factors [20]. It has been proposed, for example, that preexisting renal disease and advancing age of the patient may increase risk of nephrotoxicity only if the dose of the aminoglycoside is not appropriately adjusted to reflect the diminished renal function of this population. Similarly, a prior course of aminoglycosides might increase risk only if the therapy was recent, which would allow rapid saturation of tissue sites as a result of the prolonged tissue half-life of these drugs. The impact of some other identified risk factors is also uncertain due to conflicting data among the studies. For example, although female gender proved to be a risk factor in one clinical study of aminoglycoside nephrotoxicity [27], this was not confirmed by other studies in human or animal models, and men have certainly not been found to be immune from nephrotoxicity when treated with this class of drugs [30-33,35521. In many studies, large total doses of aminoglycosides given for prolonged time intervals appeared to correlate with nephrotoxicity. But whether more frequent dosing rather than higher doses at less frequent intervals would be less nephrotoxic in humans, as it has been shown to be in some animal studies, has yet to be proved [343. Similarly, although both high peak and trough serum levels of aminoglycoside have been claimed to correlate with nephrotoxicity, the correlation is clearly imprecise, and in many cases, serum levels are poor predictors of renal outcome. Of those risk factors associated with the patient’s status at the time of drug administration, volume depletion, hypotension, and/or shock seem to be most consistently
Volume 88 (suppl 3C)
SYMPOSIUM
associated with the development of aminoglycoside nephrotoxicity. The presence of liver disease is also a risk factor for many forms of renal failure, perhaps because of alterations in renal profusion, it may result in greater reabsorption of aminoglycoside by the proximal tubule. Nephrotoxic Impact of Combination Therapy
Concurrent use of a number of nephrotoxic medications and therapeutic agents also has been claimed to increase the risk of aminoglycoside nephrotoxicity (Table III). Of these, the use of cephalosporin antibiotics has raised the most controversy. Clearly, cephaloridine and a number of other cephalosporins act as dose-related nephrotoxins to the proximal tubule, and a number of studies have claimed synergistic nephrotoxicity between aminoglycosides and cephalosporins [3]. However, newer cephalosporins are far less nephrotoxic than older agents [31. Moreover, in animal studies it has been difficult to prove synergistic nephrotoxicity between these agents [531. In fact, a protective effect from concurrent cephalosporin use has been noted in some animal studies [531. Likewise, aminoglycoside nephrotoxicity has not been shown to be increased significantly by concurrent cephalosporin use in humans [3]. Thus, although it is clearly advisable to exercise caution when combining an aminoglycoside with a cephalosporin, the nephrotoxic potential appears to reside primarily with the aminoglycoside. It has been suggested that loop diuretics and other potent diuretics also may enhance aminoglycoside nephrotoxicity. Based on well-studied animal models, it appears that this occurs only with diuretics that induce volume contraction. Another potential nephrotoxic drug, vancomycin, causes renal damage in less than 5 percent of treated patients. Recent retrospective studies, however, suggest that as many as 35 percent of patients treated with the combination of vancomycin plus an aminoglycoside will experience nephrotoxicity, confirming animal data of synergistic damage to the kidneys between these two classes of drugs 154,551. Other agents associated with an increased risk of nephrotoxicity include amphotericin B, cisplatin, cyclosporine, nonsteroidal anti-inflammatory drugs, and radiographic contrast agents. All are known nephrotoxins, and when possible, their use should be avoided in combination with aminoglycoside therapy. Comparative Nephrotoxicity of Various Aminoglycosides
Probably the most debated issue concerning aminoglycoside antibiotics and the kidneys is which agents of this class of drugs produce significantly less clinical renal damage than others [15,26,35,36]. Streptomycin, for example, clearly has only minimal nephrotoxicity potential. Neomycin, recognized as the most nephrotoxic agent, is only administered for irrigation, topically, and as a bowel-sterilizing agent. Sisomicin and kanamycin, although more nephrotoxic than gentamicin, are not often used in the United States. Of the widely available aminoglycosides, gentamicin has appeared to be more nephrotoxic than tobramycin or netilmicin in a vast number of in vitro and in vivo animal studies [3-5,30-39, 53-551. In clinical trials, netilmicin has been comparable to tobramycin in degree of nephrotoxicity [40,41]. Amikacin is at least no more nephrotoxic than gentamicin and is perhaps less so [32,42,43]. Since the former is restricted for use in
ON AZTREONAM / APPEL
patients with resistant organisms, any slight differences in nephrotoxic potential are probably of little clinical importance. A number of well-controlled clinical trials have confirmed these differences among the aminoglycosides, but other studies have been unable to establish different degrees of nephrotoxicity for some of the individual agents [‘7,17,26,31-33,36,40,42,44--521. In low-risk patients., there is probably little significant difference in the risk of renal damage between gentamicin and tobramycin or netilmicin. In these patients, any renal dysfunction is usually mild, reversible, and associated with little residual damage. In contrast, in patients at high risk for nephrotoxicity, even small differences in nephrotoxic potential between individual drugs may potentiate other risk factors and add to the burden of renal damage. PREVENTION OF AMINOGLYCOSIDE-INDUCED RENAL DAMAGE
Correction of all of those factors identified as potentially leading to a greater risk of renal dysfunction is critical to the prevention of renal damage associated with aminoglycoside therapy. Of these factors, correction of volume depletion and/or congestive heart failure and diminished renal perfusion is probably the most significant. Also important are adjustment of dosage to compensate for the patient’s level of renal function and awareness of the decreased GFR that may be present in elderly patients despite plasma ereatinine in the normal range for the laboratory. Measurement of serum levels of the aminoglycoside may also be useful. Since these drugs are excreted only by glomerular filtration, anything that impairs renal function can be expected to also cause a rise in serum levels. Thus a rising trough level of drug (drawn just prior to administration of the next dose) indicates renal dysfunction, and the subsequent dose must be adjusted accordingly. Clearly, to minimize the risk of renal damage, using the shortest appropriate course of the aminoglycoside and avoiding concurrent treatment with other nephrotoxic agents also are advisable. It has yet to be determined whether employing less-nephrotoxic aminoglycosides or altering dose administration schedules in higher-risk patients will truly minimize renal damage, although such an effect has been suggested by recent studies 1341. Several recent controlled studies have shown that newer third-generation cephalosporins, as well as other agents, such as aztreonam, are less nephrotoxic than aminoglyeosides when used in similar situations 156-581. Recent studies also have examined the important question of the financial cost of aminoglycoside-related nephrotoxicity [30,59-611. In a recent study of 1,756 patients at six Philadelphia hospitals, there was a 7.3 percent incidence of aminoglycoside nephrotoxicity [611. The additional cost to the 129 patients in whom nephrotoxicity occurred was over $2,500 per patient. When this cost of nephrotoxicity was factored among all 1,756 patients receiving the aminoglycoside, the added cost for each patient was $183. Although alternative treatments may seem to be more expensive than aminoglycosides at first look, consideration of this cost of nephrotoxicity, added to the expense of measuring serum aminoglycoside levels and renal function to monitor for nephrotoxicity, results in competitive cost-effectiveness for the two
March 23, 1990
The American Journal of Medicine
Volume 88 (suppl 3C)
3c19s
SYMPOSIUM
ON AZTREONAM
/ APPEL
[61]. Clearly, if non-nephrotoxic antibiotics, such as third-generation cephalosporins or aztreonam, are equally efficacious and equally cost-effective in treating a given infection in a patient at high risk for renal damage, use of the alternative agent would be preferable to use of the aminoglycoside.
COMMENTS The aminoglycosides remain effective and widely used antimicrobial agents, although their risk of associated nephrotoxicity in large unselected populations is probably in the range of 5 to 10 percent. Extensive study of patient risk factors for nephrotoxicity has resulted in a better understanding of its development and recognition that some features are potentially correctable. Correction of such risk factors, careful monitoring of patients, and consideration of the use of newer antimicrobial agents in patients at high risk for nephrotoxicity may reduce the incidence of this common form of renal damage.
REFERENCES 1. Sigenthaler WE, Bonetti A, Lughy R: Aminoglycoside antibiotics in infectious disease: an overview. Am J Med 1986; 80 (suppl 68): 2-14. 2 Moellering RC: Have the new beta-lactams rendered the aminoglycosides obsolete for the treatment of serious nosocomial infections? Am J Med 1986; 80 (suppl 68): 44-47. 3. Appel GB, Neu HC: The nephrotoxicity of antimicrobial agents. N Engl J Med 19n; 296: 663-670, 722-728, 784-781. 4. Appel GB, Neu HC: Gentamicin 1978. Ann Intern Med 1978; 89: 528-538. 5. Humes MD, Weinberg JM, Knaus TC: Clinical and psychophysiologic aspects of aminoglycoside antibiotics in infectious disease: an overview. Am J Kidney Dis 1982; 2: 5-29. 6. Whelton A: Therapeutic initiative for avoidance of aminoglycoside toxicity. J Clin Pharmacol 1985; 25: 67-81. 7. Lietman PS, Smith CR: Aminoglycoside nephrotoxicity in humans. Rev Infect Dis 1983; 5 (suppl 2): 284-292. 8. Bennett WM, Lull F, Porter GA: Pathogenesis of renal failure due to aminoglycosides and contrast media used in roentgenography. Am J Med 1980; 69: 767-774. 9. Appel GB, Siegel NJ, Appel AS, Hayslett JP: Studies on the mechanism of nonoliguric experimental acute renal failure. Yale J Biol Med 1981; 54: 273-281. 10. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT: Hospital-acquired renal insufficiency: a prospective study. Am J Med 1983; 74: 243-248. 11. McMurray SD, Luft RC, Maxwell DR, eta/: Prevailing patterns and predictor variables in patients with acute tubular necrosis. Arch Intern Med 1978; 138: 950-955. 12. Kleinknect D, Landais P, Goldfarb B: Pathophysiology and clinical aspects of druginduced tubular necrosis in man. Contrib Nephrol 1987; 55: 145-158. 13. Rasmussen MH, lbels LS: Acute renal failure. Multivariate analysis of causes and risk factors. Am J Med 1982; 73: 211-218. 14. Shusterman N, Strom BL, Murray TG, Morrison G, West SL, Maislin G: Risk factors and outcome of hospital-acquired acute renal failure. Am J Med 1987; 83: 65-71. 15. Smith CR, Lietman PS: Comparative clinical trials of aminoglycosides. In: Whelton A, Neu HC, eds. The aminoglycosides. New York, Basel: Marcel Dekker, Inc., 1982; 497-509. 16. Tablan OC, Reyes MD, Rintelmann WF, Lerner AM: Renal and auditory toxicity of high dose, prolonged therapy with gentamicin and tobramycin in Pseudomonas endocarditis. J Infect Dis 1984; 149: 257-263. 17. Keys TF, Kurtz SB, Jones JD, Muller SM: Renal toxicity during therapy with gentamicin or tobramycin. Mayo Clin Proc 1981; 56: 556-598. 18. Appel GB: Aminoglycoside nephrotoxicity: physiologic studies of the sites of nephron damage. In: Whelton A, Neu HC, eds. The aminoglycosides. New York, Base1 Marcel Dekker Inc., 1982. 19. Silverblatt F: Pathogenesis of nephrotoxicity of cephalosporins and aminoglycosides: a review of current concepts. Rev Infect Dis 1982; 4 (suppl): 360-365. 20. Tulkens PM: Experimental studies on nephrotoxicity of aminoglycosides at low doses. Am J Med 1986; 80 (suppl 68): 105-114. 21. DeBroe ME, Giuliano RA, Verpoeten GA: Choice of drug and dosage regimen. Two important factors for aminoglycoside nephrotoxicity. Am J Med 1986; 80 (suppl 68): 114-118. 22. Brier ME, Mayer PR, Brier RA, et a/: Relationship between rat renal accumulation of gentamicin, tobramycin, and netilmicin and their nephrotoxicities. Antimicrobial Agents Chemother 1985; 27: 812-816. 23. Neugarten J, Aynedjian HS, Bank N: The role of tubular obstruction in acute renal failure due to gentamicin. Kidney Int 1983; 24: 330-335. 24. Baylis C, Rennke HR, Brenner BM: Mechanisms of the defect in glomerular ultrafiltration associated with gentamicin administration. Kidney Int 19R 12: 344-353. 25. Olsen S: Acute tubular necrosis and toxic renal injury. In: Tisher CC, Brenner BM, eds. Textbook of renal pathology. Philadelphia: JB Lippincott, 1989.
3C-20s
March 23, 1990
The American Journal of Medicine
26. Meyer RD: Risk factors and comparisons of clinical nephrotoxlcity of aminoglycosides. Am J Med 1986; 80 (suppl 6B): 119-125. 27. Moore RD, Smith CR, Lipsky JJ, Mellits ED, Lietman PS: Risk factors for nephrotoxicity in patients treated with aminoglycosides. Ann Intern Med 1984; 100: 352-357. 28. Sawyers CL, Moore RD, Lerner SA, Smith CR: A model for predicting nephrotoxicity in patients treated with aminoglycosides. J Infect DIS 1986; 153: 1062-1068. 29. Lam YWF, Arana CJ, Shikuma LR, Rotshafer JC: The clinlcal utility of published monogram to predict aminoglycoside nephrotoxicity. JAMA 1986; 255: 639-642. 30. Gelfand J, Neu HC, Morelli H, Appel GB: Safety and cost efficacy of a restricted policy for aminoglycoside antibiotic usage (abstr). Kidney Int 1985; 27: 138. 31. Kahlmeter G, Hallberg T, Kamme C: Gentamicin and tobramycin in patients with various infections-nephrotoxicity. J Antimicrob Chemother 1978; 4 (suppl A): 47-52. 32. Matzke GR, Lucarotti RL, Shapiro HS: Controlled comparison of gentamicln and tobramycin nephrotoxicity. Am J Nephrol 1983; 3: 11-17. 33. Lerner SA, Schmitt BA, Seligsohn R, Metz GL: Comparative study of toxicity and nephrotoxicity in patients randomly assigned treatment with amikacin and gentamicin. Am J Med 1986; 80 (suppl 6B): 98-104. 34. Kapusnik JE, Sands MA: Challenging conventional amlnoglycoside dose regimen. Am J Med 1986; 80 (suppl 68): 179-181. 35. Evans DA, Buring J, Mayrent S, Rosner B, Colton T, Mennekens C: Qualitative overview of randomized trials of aminoglycosides. Am J Med 1986; 80 (suppl 68): 39-43. 36. Cone LA: A survey of prospective controlled clinical trials of gentamicin, tobramycin, amikacin, and netilmicin. Clin Ther 1982; 5: 155-162. 37. Soberon L, Bowman RL, Pastoriza-Munoz E, et at Comparative nephrotoxicities of gentamicin, netilmicin, and tobramycin in the rat. J Pharmacol Exp Ther 1979; 210: 334343. 38. Gilbert DN, Plamp C, Starr P, et at Comparative nephrotoxicity of gentamicin and tobramycin in rats. Antimicrob Agents Chemother 1979; 13: 34-40. 39. Luft FC, Bloch R, Sloan RS, et al: Comparative nephrotoxicity of aminoglycoside antibiotics in rats. J Infect Dis 1978; 138: 541-545. 40. Gatell JM, San Miguel JG, Aranago V, et at Prospective and randomized double-blind comparison of nephrotoxicity and auditory toxicity of tobramycin and netilmicin. Antimicrob Agents Chemother 1984; 26: 766-769. 41. Lerner AM, Cone LA, Jansen W, et a/; Randomized, controlled trial of the comparative efficacy, auditory toxicity and nephrotoxicity of tobramycin and netilmicin. Lancet 1983; I: 1123-1125. 42. Gatell JM, San Miguel JG, Zamora L, et at Comparison of the nephrotoxicity and auditory toxicity of tobramycin and amikacin. Antimicrob Agents Chemother 1983; 23: 897-901. 43. Smith CR, Baughman KL, Edwards CQ, et at Controlled comparison of amikacin and gentamicin. N Engl J Med 19n; 296: 349-354. 44. Frimodt-Moller, Maigaard S, Madsen PO: Comparative nephrotoxicity among aminoglycosides and beta-lactam antibiotics. Infection 1980; 8 (suppl): 283-289. 45. Plaut ME, Schentag JJ, Jusko WJ: Aminoglycoside nephrotoxicity: comparative assessment in critically ill patients. J Med 1979; 10: 257-286. 46. Schentag JJ, Plaut ME, Cerra FB: Comparative nephrotoxicity of gentamicin and tobramycin: pharmacokinetics and clinical studies in 201 patients. Antimicrob Agents Chemother 1981; 19: 859-866. 47. Cerra FB, Plaut ME: Clinical and pharmacokinetic characteristics of amlnoglycoside nephrotoxicity in 201 critically ill patients. Antimicrob Agents Chemother 1982; 21: 721726. 48. Kumin GD: Clinical nephrotoxicity of tobramycin and gentamicin. JAMA 1980; 244: 1808-1810. 49. Smith CR, Lipsky JJ, Laskin OL, et at Double-blind comparison of the nephrotoxicity and auditory toxicity of gentamicin and tobramycin. N Engl J Med 1980; 302: 1106-1109. 50. Fong IN, Fenton RS, Bird R: Comparative toxicity of gentamicin versus tobramycln: randomized prospective study. J Antimicrob Chemother 1981; 7: 81-88. 51. Bock BV, Edelstein PH, Meyer RD: Prospective comparative study of efficacy and toxicity of netilmicin and amikacin. Antimicrob Agents Chemother 1980; 17: 217-225. 52. Kahlmeter G, Dahlager JI: Aminoglycoside toxicity-a review of clinical studies published between 1975 and 1982. J Antimicrob Chemother 1984; 13 (suppl A): 9-22. 53. Luft FC: Cephalosporin and aminoglycoside interactions: clinical and toxicologic implications. In: Whelton A, Neu HC, eds. The aminoglycosides. New York: Marcel Dekker Inc., 1982. 54. Appel GB, Given DB, Levine LR, Cooper GL: Vancomycin and the kidney. Am J Kidney Dis 1986: 8: 75-80. 55. Wold TJ, Turnipseed SA: Toxicology of vancomycin in laboratory animals. Rev Infect Dis 1981: 3: 274-235. 56. Smith CR, Ambinder R, Lipsky JJ, eta/: Cefotaxime compared with nafcillin plus tobramycin for serious bacterial infections. Ann Intern Med 1984; 101: 469-477. 57. Neu HC: Aztreonam: a novel monocyclic beta lactam antibiotic. Am J Med 1985: 78 (suppl 2A): 57-65. 58. Henry SA. Bendush CB: Aztreonam: worldwide overview of the treatment of patients with gram-negative infections. Am J Med 1985; 78 (suppl 2A): 57-65. 59. Gladen HE: Cost-effect of aminoglycoside therapy in surgical patients. Am J Med 1986; 80 (suppl 68): 228-233. 60. Holloway JJ, Smith CR, Moore RD, eta/; Comparative cost effectiveness of gentamicin and tobramycin. Ann Intern Med 1984; 101: 764-769. 61. Eisenberg JM, Koffer H, Glick HA, et at What is the cost of nephrotoxicity associated with aminoglycosides? Ann Intern Med 1987; 107: 900-909.
Volume 88 (suppl 3C)
