ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 12, No. 1 Copyright © 1982, Institute for Clinical Science, Inc.

Mechanisms of Antibiotic-Induced Nephrotoxicity THOMAS W. SEALE, Ph .D. and OWEN M. RENNERT, M.D. Departments of Pediatrics and Biochemistry and Molecular Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73190

ABSTRACT Each of the major classes of clinically useful antibiotics can cause neph­ rotoxicity. Major differences exist among classes, and these are reviewed with regard to the frequency of occurrence of nephrotoxicity, the direct or indirect toxic effects of the drugs, their site(s) of action within the kidney, current views of their pathophysiological effects and observations which provide clues to the biochemical mechanisms underlying their adverse effects on kidney structure and function. Even for a single abnormal kidney function, such as non-oliguric reduction of glomerular filtration rate by gentamicin, the physiological and biochemical mechanisms may be com­ plex. In no instance has the biochemical mechanism underlying antibiotic- induced nephrotoxicity been definitely established.

Introduction of antibiotics with relatively low Nephrotoxicity has long been recog­ therapeutic indices and significant neph­ nized as a serious complication resulting rotoxic potential. In addition, some from antibiotic administration. The vast widely employed antibiotics with low po­ array of antibiotic types and derivatives tential for inducing renal disease will in­ available to the clinician and the in­ frequently have significant detrimental creased monitoring for early signs of renal effects on the kidneys of certain patients. impairment have reduced the risk of kid­ These idiosyncratic drug responses may ney damage compared to the early days of reflect intrinsic or environmentally- sulfonamide and neomycin therapy. induced changes in renal function or drug However, the improved technical capac­ metabolism. The intent of this paper is to ity to maintain the severely debilitated review concisely current insights into the patient who is markedly more susceptible mechanisms underlying the most fre­ to a variety of opportunistic pathogens and quently encountered types of antibiotic- the increasing incidence of microbial induced nephrotoxicity. The aminoglyco­ drug resistance often necessitate the use sides, amphotericin, and the polymixins 0091-7370/82/0100-0001 $01.50 © Institute for Clinical Science, Inc. 2 SEALE AND RENNERT are reviewed in detail based on a litera­ As shown in table I, these antibiotics ture search complete through December, include a diverse group of chemically dis­ 1980. Additional aspects of this topic can similar compounds which exert their be found in excellent reviews by Appel antimicrobial effects through site-specific and Neu1 and Sanders and Sanders.56 action on several distinct biological targets. The structural and functional dis­ General Features of Antibiotic- similarities among these agents, e.g., Induced Nephrotoxicity polymyxin and penicillin, suggested the likelihood that differences would occur in Several general considerations help to their target sites within the kidney, in provide an overview of the patho­ their biochemical mechanisms of action physiological effects of antibiotics on on renal tissue, and in their frequencies of the kidney. (1) Toxicity has been doc­ inducing renal damages. The order of ar­ umented to occur upon administration of rangement of the antibiotics listed in table the major classes of therapeutically use­ I reflects in general their propensity for ful antibiotics. (2) The frequency of occur­ impairing renal function. Thus, most pa­ rence of antibiotic-induced nephrotox­ tients receiving amphotericin B have icity varies greatly and depends upon the evidence of some degree of nephrotox­ class of antibiotic, the properties of par­ icity.7,13,68 The high frequency (approxi­ ticular derivatives within one antibiotic mately 20 percent) of occurrence of renal class, and the physiological status of the damage associated with colistimethate patient. (3) Several anatomically and func­ (polymyxin E)35 and other polymyxins51 tionally distinct kidney sites are suscep­ limits their clinical application to situa­ tible targets for the adverse effects of tions in which other antibiotics cannot be antibiotics. (4) Antibiotic-induced nephro­ employed. Nephrotoxicity associated toxicity may be indirectly or directly with the aminoglycosides ranges from 2 to mediated. 25 percent of treated patients, depending TABLE I upon the specific aminoglycoside ad­

Commonly Employed Antibiotics ministered, the criteria for defining Causing Nephrotoxicity toxicity, and the patient population.9,15,59 The remaining classes of antibiotics Antibiotic Class Chemical Structure Microbial Target listed in table I are much less frequently Amphotericin Large double bonded Membrane associated with untoward renal effects. ring structure permeability (polyene)-amino (fungal) From the search for improved therapeutic sugar conjugate indices and more useful antimicrobial ac­ Polymyxins Small cationic Membrane tion spectra, many analogs of classic an­ polypeptides permeability tibiotic types have become available. Aminoglycosides Three covalently Protein bonded amino sugars synthesis These derivatives often differ markedly

Sulfonamides Derivatives of Folic acid in their renal effects. For example, the p-aminobenzene- synthesis aminoglycosides, tobramycin and amika­ sulfonic acid cin, are markedly less nephrotoxic than Rifampicin Complex double Ribonucleic acid bonded ring synthesis gentamicin in both animal models20,31,42 structure and in man.59,60,61 The relationship be­ Penicillins Thiazolide (five Cell wall tween empirically determined nephrotox­ membered) - synthesis 0-lactam ring icity and chemical modifications of the conjugates parent antibiotic can help to identify par­ Cephalosporins Thiazolide (six Cell wall membered) - synthesis ticular portions of the antibiotic molecule 3-lactam ring which function in the induction of renal conjugates damage and may suggest possible MECHANISMS OF ANTIBIOTIC-INDUCED NEPHROTOXICITY 3 physiological mechanisms of action of many antibiotics on renal functions, cer­ these drugs. tain antibiotics may act indirectly. In addition to the intrinsic toxicity of an In table II is indicated the association of antibiotic, its nephrotoxicity depends each of the major classes of antibiotics upon physiological influences and phar­ with either immunity-mediated phenom­ macokinetic considerations, such as the ena or direct toxic action. In general, the maximum concentration achieved and the immunity-related nephrotoxic activities duration of significant levels of the an­ of antibiotics occur infrequently. Intersti­ tibiotic. Pre-existing renal dysfunction, tial nephritis brought about by methicillin the disease process rendering the indi­ is a rare complication that appears to be vidual susceptible to infection, changes in associated with prolonged exposure to renal blood flow, electrolyte imbalance high levels of antibiotic.3’10 Fever, and pre- or concomitant administration of eosinophilia, and rash, indicators of an other nephrotoxic drugs, increases the immune reaction, often precede or occur likelihood of deleterious effects of an­ concomitantly with the penicillin- tibiotics on the kidney. These factors and induced renal lesion.3 Penicillins, serving the problem of distinguishing between as haptens, produce allergic reactions the patient’s underlying illness and drug- through their ability to stimulate anti­ induced nephrotoxicity make difficult the body production and to initiate hapten- determination of the incidence of antibody reactions. In patients, circulat­ antibiotic-induced renal disease in pa­ ing antibodies to penicillins are not di­ tients. Such factors further complicate the rected to the intact penicillin molecule determination of the mechanisms respon­ but to its degradation products, e.g., the sible for the deleterious action of antibio­ penicilloyl group, that can form covalent tics on the kidney. bonds to proteins.40’50 With the possible exception of the Loop Target organ specificity, as for example of Henle, each region of the nephron can might be suggested to explain the occur­ be damaged by one or more antibiotic rence of interstitial nephritis, could result classes. The proximal tubule, site of the from at least four biochemical factors. (1) quantitatively most significant reabsorp- The intrinsic activity of the metabolic tive functions, is particularly vulnerable. pathway converting the antibiotic to the If generalized effects, such as phenomena activated hapten may be restricted to a related to hypersensitivity, are excluded, specific tissue and thereby elevate its most antibiotics affect only one or a few local concentration. (2) Conjugation of the targets. For example, the major site of ac­ tion of the aminoglycosides is the proxi­ mal tubule,2,16,32 the site of potential toxic action of the currently used sulfonamides TABLE I I is the collecting duct. Site specificity of Antibiotic - Induced Nephrotoxicity Arises antibiotic-induced nephropathy implies By Two Different Mechanisms several possibilities: (1) there exist cell Indirect Action (Immunity-Medicated) types with distinctly susceptible targets Penicillins (expecially methicillin) Cephalosporins (espec ially cephaloth in) for the deleterious effects of certain Sulfonamides specific antibiotics; (2) the drug may Rifampicin achieve toxic concentrations in only cer­ Direct Action tain kidney regions; and (3) toxic metabo­ Amphotericin lites may be generated in only certain Polymyxins Aminoglycosides kidney cells. However, in addition to the Sulfonamides apparently direct deleterious action of 4 SEALE AND RENNERT hapten to a particular tissue-specific pro­ Mechanisms of Nephrotoxic tein may generate antibodies which react Antibiotics with the protein directly. This may lead to localized autoimmune disease. (3) One Consideration of the spectrum of chem­ tissue may possess the ability to concen­ ical structures, the known mechanisms of trate the antibiotic to levels sufficient for antimicrobial actions (table I), and the di­ the generation of significant quantities of verse sites of action of the nephrotoxic the hapten. (4) The specific capacity of a antibiotics suggests that a variety of dis­ tissue to bind antigen-antibody com­ tinctly different mechanisms must ac­ plexes formed elsewhere may predispose count for their direct toxic action on renal it to immunologically-mediated damage.49 structure and function. Although much It is unlikely that an interstitial concen­ progress has been made in identifying the trating mechanism contributes to the oc­ anatomical sites which are the targets for currence of interstitial nephritis induced renal damage by antibiotics, there cur­ by penicillins because immune-mediated rently exists only a partial description of nephrotoxicity occurs with penicillins the physiological events that are as­ that differ from one another markedly sociated with the derangement of renal with regard to protein binding and serum functions by antibiotics. In no case has the half life.1,12,63 The kidney is one of the biochemical mechanism underlying target organs in drug-induced serum sick­ specific renal physiological alterations ness, the glomerulus being an especially been established. For certain antibiotics, frequent site of non-specific localization however, descriptions of their action are of antigen-antibody complexes. Although sufficiently detailed that one may begin to this phenomenon would seem to be a log­ predict their probable mechanism(s) of ac­ ical consequence in many penicillin re­ tion. The known nephrotoxic actions of cipients who possess antibodies to the several antibiotic classes are summarized drug, the absence of the characteristic in the following sections. glomerular lesion seen in other immune complex disease and the extreme rarity of A mphotericin B penicillin-induced glomerulitis suggest Nephrotoxicity is the most serious side that immune mechanisms account for effect of amphotericin B, a polyene penicillin-induced nephrotoxicity.4 Cer­ macrolide antifungal agent. Most patients tain cephalosporins also act indirectly on receiving the drug demonstrate some the kidney. Some instances of cephalothin degree of dose-dependent nephrotox­ nephropathy have a strong resemblance to icity.7,13 Urinalysis reveals hematuria, penicillin-induced nephropathy.4 Neph­ pyuria, presence of tubular cells, and rotoxicity associated with cephalothin is cylindruria but little proteinuria.8,33 His­ uncommon even with large doses of this tologic changes in blood vessels and glo­ antibiotic. It too is frequently charac­ meruli occur infrequently and are not terized by signs of hypersensitivity in­ specific for the drug.13,33 In animals, am­ cluding rash, eosinophilia, and interstitial photericin causes intratubular calcium nephritis. In neither penicillin-induced deposits accompanied by necrosis of the nor cephalosporin-induced renal lesions proximal and distal convoluted tubules is the exact mechanism of immune- and thickening of the tubular basement mediation established. membranes.65,66 Much of the effect of With this general background, the di­ amphotericin on the kidney involves its rect toxic actions of antibiotics as they per­ action on the distal tubule. The drug tain to the kidney will be discussed. causes a distal tubular acidosis charac­ MECHANISMS OF ANTIBIOTIC-INDUCED NEPHROTOXICITY 5 terized by net acid excretion and reduced polyene macrolide concentrations that do acidification of urine in response to an not elicit changes in membrane permea­ acid load.11,18 Kaliuresis and resultant bility. Alternatively, immediate mem­ potassium deficiency also frequently oc­ brane damage may occur in the absence of cur.7,11 Both inulin and p-aminohippuric cell lethality.25 It appears, then, that the acid clearance decrease during acute ad­ acute effect of amphotericin B on the dis­ ministration of amphotericin. Decreased tal tubule is a reflection of a rapid per­ glomerular filtration rate and ischemic meability change to chloride ions. With damage appear to be secondary to drug- extended therapy, this may be accom­ mediated renal vasoconstriction.54 These panied by a more generalized cellular effects are probably the result of the action toxicity which results from continued ex­ of amphotericin on the distal tubule. It has posure to the drug or from exposure to an been established that an increased deliv­ increased concentration of this drug. ery of chloride ion to the distal tubule can lead to a profound decrease in the glo­ merular filtration rate of the nephron.58 T h e P o l y m y x in s This suppression of glomerular function The polymyxins—polymyxin B, colis- by the distal tubule is called tubulo-glo- tin, and colistimethate—are basic, cyclic merular feedback. Both this tubulo-glo- polypeptide antibiotics with a broad merular feedback mechanism and the range of activity against Gram-negative acute ischemic change caused by am­ bacteria. All are associated with signifi­ photericin are attenuated by preloading cant nephrotoxicity typically charac­ animals with sodium or by pre-treating terized by decreased creatinine clear­ them with furosemide.27 These findings ance, proteinuria and cellular casts in the are consistent with the hypothesis that the urine.35,47,55 In fatal cases, histological acute amphotericin-mediated response studies have established proximal tubular includes an increased chloride permea­ damage and acute tubular necro­ bility of the distal tubule followed by acti­ sis.53, 55PhysiologicaI evidence for the ac­ vation of tubulo-glomerular feedback. tion of polymixins on the proximal convo­ How can this site-specific action be luted tubule has been obtained from in reconciled with the known biochemical vitro studies employing hippurate uptake activities of amphotericin? Polyene mac- by rabbit renal cortical slices.46 (Hippu­ rolide antibiotic toxicity toward eukary­ rate transport is a biochemical marker of otic cells in generally attributed to mem­ proximal tubule function.) In man, the de­ brane permeability changes resulting gree of inhibition of this organic acid from polyene macrolide-sterol interac­ transport system by the polymyxins is tions.17 Amphotericin is slowly excreted well correlated with their relative ability and has a half life of about 24 hours.24 It is to induce nephrotoxicity.46 While the ori­ almost completely protein bound; im­ gin of the relatively specific effects of paired renal function does not signifi­ polymyxins on the proximal tubule may in cantly alter the serum concentration.24 some manner relate directly to the proxi­ Thus, generalized cellular toxicity result­ mal tubule organic base transport sys­ ing from membrane alterations might be tem,39 this hypothesis remains to be expected in the kidney. However, under verified. certain circumstances it is possible to dis­ It seems reasonable to assume that the sociate toxicity and permeability changes. kidney is susceptible to the surfactant ac­ In some in vitro experimental systems, tion of the polymyxins through which cell toxicity has been demonstrated at their antibacterial action is mediated.23,48 6 SEALE AND RENNERT The interaction of these compounds with problems of extrapolating to man conclu­ the phospholipid component of bacterial sions based on observations in animal cell membranes, especially phospha- models are numerous, but the agreement tidylethanolamine groups, leads to is surprisingly good for those amino­ changes in membrane permeability and glycosides rigorously evaluated in man cell death.23 Polymyxins bind to the cellu­ (e.g., comparing nephrotoxicity of gen­ lar components of the kidney,37 and tamicin and tobramycin).44,60 polymyxin B accumulates intracellularly Clinical observations and animal exper­ to high levels in the kidney.34 The drug is iments demonstrate that aminoglycosides retained for an extended period in an un­ can induce both structural and functional metabolized form. With regard to the kidney damage.2,26,32,36 Typical findings chemical structure of the polymyxins, the include proteinuria, enzymuria (derived degree of tissue binding, nephrotoxicity, from proximal tubule lysosomal en­ inhibitory action on proximal tubular hip- zymes), hematuria, cylindruria, elevated purate transport, and antibacterial activity blood urea nitrogen, reduced glomerular are directly related to the number of free filtration rate and impaired urinary con­ amino groups in this antibiotic.46,48,52,64 It centrating ability.1,2,16 These abnor­ has been suggested that a balance be­ malities may or may not be associated tween protonated and unprotonated with oliguria, although the disease may amino groups in the molecule is important progress to oliguric renal failure should in both antibacterial and nephrotoxic ef­ aminoglycoside administration be con­ fects since there is a pH optimum for hip- tinued. Proximal tubule damage progress­ purate transport inhibition and phos­ ing to acute tubular necrosis typifies the pholipid binding by polymyxins.23,46 histological picture associated with From these observations it can be conjec­ aminoglycoside-induced renal dam­ tured that proximal tubular damage by age.14,32,36 The glomeruli and blood ves­ polymyxins may result from their rela­ sels are not histologically altered. tively selective accumulation in cells of Myeloid body formation is a prominent the proximal tubule where their concen­ feature of aminoglycoside nephrotoxicity tration becomes sufficient to produce di­ and occurs most prominently in lyso- rectly significant membrane damage. somes of proximal tubule epithelial cells.20,32,36 Aminoglycosides are not metabolized T h e A minoglycosides to a significant extent, and the major route Nephrotoxicity and ototoxicity are major of excretion is by glomerular filtra­ factors limiting the clinical utility of tion.1,2,32 In patients with severely im­ the aminoglycosides. Included among the paired renal function, the serum half life group of clinically useful aminoglycosides of these drugs increases to 25 to 40 times are neomycin, gentamicin, kanamycin, that of the two hours28,38'41 seen in normal tobramycin, netilmicin, amikacin, and individuals. One feature of amino­ streptomycin. The preceeding list is given glycoside antibiotics which distinguishes in order of decreasing nephrotoxicity and is them from most other antibiotics is their based upon a composite of data derived accumulation in the renal cortex42,43,44,67 from experiments in animal models19,20,22, (consistent with the tubular action of 32,42,44an(j clinjcal observations.9,59,60,61 aminoglycosides since the tubules are lo­ Neomycin, sometimes employed orally to cated in the renal cortex). The ratio of reduce the gut flora, is not administered cortex-bound to serum aminoglycoside intravenously because of its high fre­ concentration ranges from 5 to 20, de­ quency of nephrotoxic complications. The pending on experimental conditions. MECHANISMS OF ANTIBIOTIC-INDUCED NEPHROTOXICITY 7 Other classes of antibiotics, e.g., penicil­ binding results in inhibition of protein lins or cephalosporins, have cortical to synthesis and mis-translation during serum ratios ranging from one to three. polypeptide chain elongation. Although The capacity of the kidney to accumulate these effects usually have not been dem­ aminoglycosides is large, and repeated in­ onstrated to occur with eukaryotic cyto­ jections result in progressively increased plasmic ribosomes, there are reports that cortical levels. in eukaryotes inhibition of protein syn­ In dogs, at stable therapeutic serum thesis and messenger RNA misreading levels, the aminoglycosides can be se­ can occur in vitro. The proximal tubule quentially ordered with regard to their specificity observed in vivo may result levels of cortical accumulation. From from the unusually high levels of amino­ highest to lowest these are—neomycin > glycoside achieved and intrinsic or en­ gentamicin > kanamycin > tobramycin > vironmentally induced factors. Such fac­ streptomycin—an order which bears a tors appear to make certain tubular cells remarkable similarity to their relative sensitive to the drug, perhaps at the level nephrotoxic potencies.67 Streptomycin is of the protein synthesis. Other mecha­ not accumulated to a significant extent nisms may, however, be involved. For and is the least toxic. Results generally example, neomycin has major effects similar to those in the dog are found in the upon polyphosphoinositide metabolism rat model. One exception, netilmicin, an in guinea pig kidney both in vivo and in aminoglycoside with low nephrotoxicity, vitro.57 In view of the probable function of accumulates to an extent similar to highly polyphosphoinositides in membrane toxic gentamicin.42,43’45 The toxic effects permeability and ion transport,30,57 altera­ of all aminoglycosides are known to be tion of their metabolism could have prp- both dose- and duration-dependent. In found effects on the cell. Finally, there addition to being concentrated in the may be unexpected effects of amino­ renal cortex, these antibiotics also are re­ glycosides on particular metabolic func­ tained there in chemically unmodified tions, for example, the inhibition of the form for an extended period of Embden-Meyerhof pathway by kanamy­ time.29’43,44’67 In the rat, for example, the cin has been shown in the guinea pig.62 half-time of gentamycin in serum is 0.5 A less generally appreciated second hours, whereas in the renal cortex the target for aminoglycosides is the half-time is 109 hours. A single dose may glomerulus. Reduction of glomerular fil­ persist for days and, after cessation of drug tration rate often occurs after amino­ administration following repeated doses, glycoside administration and is fre­ aminoglycoside may be detectable in quently of the non-oliguric type.60 Baylis urine for weeks. These results may ex­ has investigated the mechanisms respon­ plain delayed nephrotoxicity and de­ sible for the impairment of glomerular fil­ velopment of nephrotoxicity in patients tration by gentamicin in the rat.5,6 Several receiving a second course of therapy factors alone or in combination could within a few weeks of previous drug cause a fall in glomerular filtration rate. (1) therapy.67 There is a major and directly proportional There is no definitely known mecha­ dependence on the renal plasma flow rate. nism explaining how these antibiotics (2) Reduction of the hydrostatic pressure cause their deleterious effect on cells difference across the capillary wall has an of the proximal tubule. The amino­ important effect. (3) An increase in plasma glycosides exert their effect on bacteria by protein concentration leading to elevated binding to specific protein components of plasma oncotic pressure will decrease fil­ the small (30S) ribosomal subunit. This tration rate. (4) Reduction in capillary 8 SEALE AND RENNERT water permeability and/or effective sur­ References face area in the glomerulus will reduce 1. A p p e l , G. B. and N e u , H. C.: The neph­ the glomerular filtration rate. At low to rotoxicity of antimicrobial agents. New Eng. J. moderate concentrations of gentamicin, M ed. 296:663-670; 722-728; 784-787, 1977. 2. Ap p e l , G. B. and N e u , H. C.: Gentamicin in micropuncture studies establish the sig­ 1978. Ann. Intern. Med. 89:528-538, 1978. nificant alteration of the last pair of pa­ 3. Ba l d w in , D. S., L e v in e , B. B., M c C l u sk ey , R. T., and Ga l l o , G. R.: Renal failure and intersti­ rameters. At high concentrations of the tial nephritis due to penicillin and methicillin. antibiotic, the reduction in glomerular New Eng. J. M ed. 279:1245-1252, 1968. filtration rate is multifactorial.5,6 The de­ 4. Ba r z a , M.: The nephrotoxicity of cephalo­ sporins: An overview. J. Infect. Dis. 137 tailed molecular mechanisms accounting (Suppl.):560-573, 1978. for any of these physiological changes by 5. Ba y l is, C.: The mechanisms of the decline in gentamicin are completely unknown. glomerular filtration rate in gentamicin- induced acute renal failure in the rat. J. An­ Conclusions timicrobial Chemotherap. 6:381-388, 1980. 6. Ba y l is, C., Re n n k e , H. R., and Br e n n e r , B. Among nephrotoxic agents, the most M.: Mechanisms of the defect in glomerular common offenders are drugs prescribed ultrafiltration associated with gentamicin ad­ for the treatment of clinical conditions not ministration. Kidney Int. 72:344-353, 1977. 7. Be n n e t t , J. E.: Chemotherapy of systemic my­ directly concerned with the kidney. In coses. New Eng. J. Med. 290:30-32, 1974. this category, antibiotics are major induc­ 8. Be n n e t , J. E., Br a n d r iss, M. W., Bu t l e r , W. T., and H il l , G. J.: Amphotericin B toxicity. ers of kidney disease. While the incidence Ann. Int. Med. 61:334-354, 1964. of nephrotoxicity mediated by allergic re­ 9. Bo c k , B. V., E d e l s t e in , P. H., and M e y e r , R. sponses is uncommon, high frequencies D.: Prospective comparative study of efficacy and toxicity of netilmicin and amikacin. An- of renal impairment occur with the admin­ timicrob. Agents Chemother. 17:217-225, istration of the major antibiotic used to 1980. combat systemic fungal infections, am­ 10. B o r d e r , W. A., L e h m a n n , D. H., E g a n , J. D., photericin B, and the aminoglycosides Sa s s , H. J., G l o d e , J. E., and WILSON, C. B.: Antitubular basement membrane antibodies in used to combat broadly antibiotic resis­ methicillin-associated interstitial nephritis. tant organisms such as Pseudomonas N ew E ng. J. M ed. 291:381-384, 1974. aeruginosa. A detailed picture is begin­ 11. B u r g e s s , J. L. and BiRCHALL, R.: Neph­ rotoxicity of amphotericin B with an emphasis ning to be established of the pathological on changes in tubular function. Amer. J. Med. alterations which arise in the kidney upon 53:77-84, 1972. antibiotic administration. A few of the 12. Bu r t o n , J. R., Lichtenstein , N. S., and C o l ­ v in , R. B.: Acute interstitial nephritis of oxacil­ circumstances which may potentiate lin. Johns Hopkins Med. J. 134:58-61, 1974. antibiotic-induced nephrotoxicity are 13. Bu t l e r , W. T., Be n n e t t , J. E., Al l in g , D. W., known, but it is unclear why in the same W e r t l a k e , P. T., U t z , J. P., and H il l , G. J.: Nephrotoxicity of Amphotericin B: Early and kidney some nephrons are damaged while late effects in 81 patients. Ann. Intern. Med. others are not when presumably exposed 61:175-187, 1964. to similar local concentrations of the drug. 14. CUBANILLAS, F., BüRGOS, R. C., RODRIGUEZ, R. C., and Ba l d iz o n , C .: Nephrotoxicity of com­ In no case has the underlying biochemical bined cephalothin-gentamicin regimen. Arch. mechanism been definitively established Intern. Med. 135:850-852, 1975. for antibiotic-induced nephrotoxicity. 15. C o c a , A., M a r t in e z , A., So r ia n o , E., Bl a d e , J., Se g u r a , F., and Rib a s-M u n d o , M.: Tob­ Further clarification of the underlying ramycin nephrotoxicity. A prospective clinical mechanisms of toxicity could have major study. Postgrad. Med. J. 55:791-796, 1979. 16. C r o n in , R. E.: Aminoglycoside nephrotoxicity: impact upon the design of new antibiotic Pathogenesis and prevention. Clin. Nephrol. analogs with reduced nephrotoxic proper­ 11:251-256, 1979. ties and for the establishment of therapeu­ 17. D e k r u ijf f , B., G e r r it s e n , W. J., O e r le m a n s, A., D e m e l , R. A., and Va n D e e n a n , L. L. M.: tic regimens which ameliorate the Polyene antibiotic-sterol interactions in mem­ nephrotoxic effects of antibiotics without branes of Acholeplasma laidawii cells and impairing their antimicrobial activity. lecithin liposomes. I. Specificity of the mem­ MECHANISMS OF ANTIBIOTIC-INDUCED NEPHROTOXICITY 9 brane permeability changes induced by the 33. Iovine, G., Berman, L. B., Halikis, D. N., polyene antibiotics. Biochem. Biophys. Acta M o w re y , F. H., Chappelle, E. H., and 339:30-43, 1974. Gierson, H. W.: Nephrotoxicity of amphoteri­ 18. D o u g l a s, J. B. and H e a ly , J. K.: Nephrotoxic cin B: A clinical-pathologic study. Arch. Intern. effects of amphotericin B including renal tubu­ M ed. 772:853-862, 1963. lar acidosis. Amer. J. Med. 46:154-162, 1969. 34. Ja c o b so n , M., K och, A., K u n tz m an , R., and 19. ElNSPRUCH, B. C. and G o n z a l e z , V. V.: Clini­ B urchall, J.: The distribution and binding of cal and experimental nephropathy resulting tritiated polymixin B in the mouse. J. Phar­ from use of neomycin sulfate. J. Amer. Med. macol. Exp. Ther. 783:433-439, 1972. Assoc, i 73:809-811, 1960. 35. Koch-Weser, J., Sidel, V. W., Federman, E. 20. E n g l e , J. E., Ab t , A. B., Sc h n e c k , D. W., and B., K a n a rc k , P., Finer, D. C., and E a to n , A. Sc h o o l w o e t h , A. C.: Netilmicin and to­ E.: Adverse effects of sodium colistimethate: bramycin: Comparison of nephrotoxicity in dogs. Manifestation and specific reaction rates during Invest. Urol. 77:98-102, 1979. 317 courses of therapy. Ann. Intern. Med. 21. F a b r e , J., Ru d h a r d t , M., and Bl a n c h a r d , P.: 72:857-868, 1970. Persistence of sisomicin and gentamicin in 36. Kosek, J. C., M azze, R. I., and Cousins, M. J.: renal cortex and medulla with other organs and Nephrotoxicity of gentamicin. Lab. Invest. serum of rats. Kidney Int. 10 :444-451, 1976. 30:48-57, 1974. 22. F a l c o , F. G., Sm it h , H . M., and Ar e ie r i, G. 37. K unin, C. M. and B u g g , A.: Recovery of tissue M.: Nephrotoxicity of the aminoglycosides and bound polymyxin B and colistimethate. Proc. gentamicin. J. Infect. Dis. 119:406-409, 1969. Soc. Exp. Biol. Med. 737:786-790, 1971. 23. F e in g o l d , D. S., H s u c h e n , C. C., and Sud, I. 38. K unin, C. M. and F in la n d , M.: Persistence of J.: Basis for the selectivity of action of the antibiotics in blood of patients with acute renal polymyxin antibiotics on cell membranes. Ann. failure. III. Penicillin, streptomycin, eryth­ N. Y. Acad. Sci. 235:480-491, 1974. romycin and kanamycin. J. Clin. Invest. 24. F e l d m a n , H . A., H a m il t o n , J. D., and 38:1509-1519, 1959. G u t m a n , R. A.: Amphotericin B therapy in an 39. L a n t, R. F.: Renal excretion and nephrotoxicity anephric patient. Antimicrob. Agents Chemo­ of drugs. Renal Disease. Black, B., ed. London, ther. 4:302-305, 1973. Blackwell Scientific Publications, 1972, pp. 25. F is h e r , P. B., B r y so n, R., and Sc h a f f n e r , C. 591-614. P.: Polyene macrolide cytotoxicity and mem­ 40. L ev in e, B. B.: Immunological mechanisms in­ brane permeability alterations. J. Cell. Physiol. volved in penicillin hypersensitivity in ex­ 97:345-352, 1978. perimental animals and in human beings. Fed. 26. G a r y , N. E ., B u z z e o , L., Sa l a k i, J., and Proc. 24:45-50, 1965. ElSINGER, R. P.: Gentamicin-associated acute 41. Lockwood, W. R. and B o w er, J. D.: Tobramy­ renal failure. Arch. Intern. Med. 136:1101- cin and gentamicin concentrations in the serum 1104, 1976. of normal and anephric patients. Antimicrob. 27. G e r k e n s, J. F. and Br a n c h , R. A.: The influ­ Agents Chemother. 3:125-129, 1973. ence of sodium status and furosemide on canine 42. L u f t, F. C., Bloch, R., S lo a n , R. S., Yum, M. acute amphotericin B nephrotoxicity. J. Phar­ N., C o s te ll o , R., and M a x w e ll, D.: Com­ macol. Exp. Ther. 274:306-311, 1980. parative nephrotoxicity of aminoglycoside an­ 28. G y sely n c k , A. M., F o r r e y , A., and C u t l e r , tibiotics in rats. J. Infect. Dis. 738:541-545, R.: Pharmacokinetics of gentamicin: Distribu­ 1978. tion and plasma and renal clearance. J. Infect. 43. Luft, F. C. and K le it, S. A.: Renal paren­ Dis. 124( Suppl.):70-77, 1971. chymal accumulation of aminoglycoside an­ 29. H a u g h e y , D. B., Ja n ic k e , D. M., Ad e lm a n , tibiotics in rats. J. Infect. Dis. 730:656-665, M., and Sc h e n t a g , J. J.: High pressure liquid 1974. chromatography analysis and single-dose dis­ 44. L u f t, F. C., P atel, V., Yum, M. N., P a te l, B., position of tobramycin in human volunteers. and K le it, S.A.: Experimental aminoglycoside Antimicrob. Agents Chemother. 77:649-653, nephrotoxicity. J. Lab. Clin. Med. 86:213-220, 1980. 1975. 30. H e n d r ic k so n , H. S. and R e in e r t s e n , J. L.: 45. L u f t, J., Yum, M. N., and K le it, S. A.: Com­ Phosphoinositide interconversion: A model for parative nephrotoxicities of netilmicin and gen­ control of Na+ and K+ permeability in the tamicin in rats. Antimicrob. Agents Chemother. nerve axon membrane. Biochem. Biophys. Res. 70:845-862, 1976. Comm. 44:1258-1264, 1971. 46. M ilm a n , N. and D a h la g e r , J.: Effects of 31. H o t t e n d o r f , G. H . and G o r d o n , L. L.: Com­ polymyxin antibiotics on iodohippurate ac­ parative low-dose nephrotoxicities of gentami­ cumulation in rabbit renal cortical slices. J. An­ cin, tobramycin and amikacin. Antimicrob. tibiotics 37:906-910, 1978. Agents Chemother. 78:176-181, 1980. 47. Moyer, J. H., M ills, L. C., and Y ow , E. M.: 32. H o u g h t o n , D. C., P l a m p, C. E., D e F e h r , J. Toxicity of olymixin B. I. Animal studies with M., B e n n e t t , W. M., P o r t e r , G., and particular reference to renal function. Arch. G il b e r t , D.: Gentamicin and tobramycin Intern. Med. 92:238-247, 1953. nephrotoxicity. A morphologic and functional 48. N o rd , N. M. andH o e p ric h , P. D.: Polymixin B comparison in the rat. Amer. J. Pathol. 93:137- and colistin: A critical comparison. New Eng. J. 152, 1978. Med. 270:1030-1035, 1964. 10 SEALE AND RENNERT 49. PARKER, C. W.: Drug allergy. New Eng. J. Med. trolled comparison of amikacin and gentamicin. 292:511-514; 732-736; 957-960, 1975. New Eng. J. Med. 296:349-353, 1977. 50. PARKER, C. W.: Penicillin allergy. Amer. J. 60. Sm it h , C. R., L ip s k y , J. J., La s k in , O. L. Med. 34:747-752, 1963. H e l l m a n , D . B., M e l l it s , E . D ., L o n g - 51. P a r k er , R. H. and H o e p r ic h , P. D.: Toxicity STRETH, J., and L ie t m a n , P. S.: Double-blind associated with intravenous and intramuscular comparison of the nephrotoxicity and auditory polymyxin B sulfate. Clin. Res. 12:244, 1964. toxicity of gentamicin and tobramycin. New E ng. M ed. 52. P e d e r s e n , M. F., P e d e r s e n , J. F., and J. 320:1106-1109, 1980. M a d se n , P. O.: A clinical and experimental 61. Sm it h , C. R., Ma x w e l l , R. R., E d w a r d s, C. Q., R o g e r s, J. F., and L ie t m a n , P. S.: Neph­ study of sodium colistimethate and polymyxin rotoxicity induced by gentamicin and amikacin. B sulfate. Invest. Urol. 9:234-237, 1971. Johns Hopkins Med. J. 142:85-90, 1978. 53. Ra n d a l l , R. E., Jr ., Br id i, G. S., Se t t e r , J. G., 62. T a c h ib a n a , M ., M iz v k o s h i, O., and and Br a c k e t t , N. C., J r .: Recovery from colis­ Ku riyam a, K .: Inhibitory effects of kanamycin timethate nephrotoxicity. Ann. Intern. Med. on glycolysis in cochlea and kidney—possible 73:491-492, 1970. involvement in the formation of oto- and 54. Rh o a d e s, E. G., G in n , H. E., M u c h m o r e , H. nephrotoxicities. Biochem. Pharmacol. 25: G., Sm it h , W. O., and H a m m a r sten , J. F.: Ef­ 2297-2301, 1976. fect of amphotericin B upon renal function in 63. T a n n e n b e r g , A. M., Wic h e r , K. J., and Ro s e , man. Antimicrobial Agents Annual, 1960. Gray, N. R.: Ampicillin nephropathy. J. Amer. Med. P., Tabenkin, B., and Bradley, S. G., eds. New Assoc. 2i8:449, 1971. York, Plenum Press, 1961, pp. 539-542. 64. Vin n ic o m b e , J. and St a m e y , T. A.: The relative 55. R yan, K. J., Sc h a in u c k , L. L., H ic k m a n , R. O., nephrotoxicities of polymyxin B sulfate, and St r ik e r , G. E.: Colistimethate toxicity. sodium sulfomethyl-polymyxin B, sodium Report of a fatal case in a previously healthy sulfomethyl-colistin (colymycin) and neo­ child. J. Amer. Med. Assoc. 207:2099-2101, mycin sulfate. Invest. Urol. 6:505-519, 1969. 1969. 65. W e l d o n , M. W. and Sc h u l t z , M. E.: Renal 56. Sa n d e r s , W. E., J r . and Sa n d e r s , C. C.: ultrastructure after amphotericin B. Pathology Toxicity of antibacterial agents: Mechanisms of 6:191-200, 1974. action on mammalian cells. Ann. Rev. Phar­ 66. W e r t l a k e , P. T., H il l , G. J., II, and B u t ­ macol. Toxicol. 19:53-83, 1979. l e r , W. T.: Renal histopathology associated 57. S c h ib e c i, A. and S c h a c h t , J.: Action of with different degrees of amphotericin B neomycin on the metabolism of polyphos­ toxicity in the dog. Proc. Soc. Exp. Biol. Med. phoinositides in the guinea pig kidney. 118:472, 1965. Biochem. Pharmacol. 26:1769-1774, 1977. 67. W h e l t o n , A.: Aminoglycoside renal cortical 58. ScHNERMANN, J.: Regulation of a single kinetics: A clue to mechanisms of toxicity. nephron filtration rate by feedback. Facts and Prog. Clin. Biol. Res. 35:33-41, 1979. theories. Clin. Nephrol. 3:75-81, 1975. 68. W in n , W. A.: Coccidiodomycosis and am­ 59. Sm ith , C. R., Baughm an, K. L., E dw ards, C. photericin B. Med. Clin. North Am. 47:1131- Q., Ro g er s, T. F., and Lietm a n , P. S.: Con­ 1149, 1963.