Dye-Induced Nephropathy

By Samuel N. Heyman and Seymour Rosen

The expanding use of imaging and interventional studies with iodinated radiologic contrast agents underscores the potential risk for dye nephropathy. Currently, dye-induced nephropathy is one of the leading causes of iatrogenic acute kidney failure, accounting for about 10% of renal failure in intensive care units. In this review, the patho- physiology of radiocontrast nephropathy is discussed, with a special emphasis on the importance of medullary hypoxic damage. The risk factors and clinical course of dye nephropathy, as well as its prevention or potential therapeutic interventions, are discussed in this perspective. © 2003 Elsevier Inc. All rights reserved.

HE EXPANDING USE of imaging and inter- renal dysfunction should be considered including Tventional procedures with iodinated radio- nephrotoxic injury produced by other agents, sep- logic contrast agents underscores the potential risk sis, rhabdomyolysis, cholesterol emboli after arte- for dye (radiocontrast) nephropathy. In the late rial studies, shock, and other ischemic insults. 1960s, this iatrogenic disease entity was consid- This review focuses on the pathophysiology of ered rare and exotic. Its incidence, however, radiocontrast nephropathy, with a special emphasis quickly surged to involve some 40% of patients on tubular hypoxic damage. In this perspective we undergoing dye studies in the early 1980s, rapidly analyze the mechanisms by which recognized risk leading to the recognition of high-risk groups, such factors might predispose to renal dysfunction after as diabetic patients and individuals with preexist- contrast studies, and critically evaluate the preven- ing renal failure, low cardiac output, or dehydra- tive/therapeutic strategies currently applied or in- tion. Subsequently, the incidence of dye nephrop- vestigated. athy gradually declined to less than 3% of contrast PATHOPHYSIOLOGIC BACKGROUND studies, predominantly because of patient selec- tion, because of the recognition of the importance Although recognized for some 40 years, the of hydration status, and because of the introduction pathophysiology of contrast nephropathy remains a of the less-nephrotoxic low osmolar and nonionic controversial issue. It is classified as acute tubular agents. Nevertheless, dye-induced nephropathy re- necrosis or toxic nephropathy but, in fact, very mains one of the leading causes of iatrogenic acute little is known about its true renal morphology in kidney failure, accounting for about 10% of cases humans. Proximal tubular vacuolization has been of acute renal failure in intensive care units. encountered in renal biopsy examinations, but this Among high-risk patients its incidence exceeds change does not correlate with renal failure. In- 20%, despite adherence to well-accepted protective deed, this alteration is present after dye adminis- measures, providing a formidable challenge to fur- tration under experimental settings, irrespective of ther improve preventive strategies. kidney function. Urinary markers of tubular injury Dye nephropathy is defined as a decline in glo- appear inconsistently, or are absent in humans with merular filtration rate (GFR) after the intravascular contrast nephropathy. Many toxic nephropathies are believed to be administration of iodinated contrast agents. Kidney mediated primarily through well-characterized di- dysfunction is manifested with an increasing rect tubular-cell toxicity, such as heavy metals, plasma creatinine level that peaks at 2 to 5 days, cisplatin, or aminoglycosides in experimental and and usually returns to baseline values by 5 to 10 days. Dye-associated renal failure usually is nono- liguric and renal sediment is unremarkable in most From the Department of Medicine, Hadassah Hospital, Mt. cases. Fractional sodium excretion is often low, Scopus, and the Hebrew University Medical School, Jerusalem, Israel; and the Department of Pathology, Beth Israel Deacon- suggesting prerenal failure. Contrast nephropathy ess Medical Center, and Harvard Medical School, Boston, MA. usually is self-limited, but some patients may run a Address reprint requests to Samuel N. Heyman, MD, Depart- protracted course and require renal replacement ment of Medicine, Hadassah University Hospital, Mount Sco- therapy, associated with increased morbidity and pus, PO Box 24035, Jerusalem 91240. E-mail: heyman@ cc.huji.ac.il mortality, and prolonged hospitalization. The clin- © 2003 Elsevier Inc. All rights reserved. ical diagnosis usually is evident under most cir- 0270-9295/03/2305-0009$30.00/0 cumstances, but other potential causes of acute doi:10.1016/S0270-9295(03)00092-5

Seminars in Nephrology, Vol 23, No 5 (September), 2003: pp 477-485 477 478 HEYMAN AND ROSEN clinical settings. Renal handling of the nephrotoxin oxygen balance is manifested by high levels of predominantly determines the distribution pattern cytochrome AA3 in its redox state, and of unsat- of tubular damage. The injury is dose dependent, urated hemoglobin, detected by blood oxygen level and the contribution of perturbations is relatively dependent (BOLD) magnetic resonance imaging2 marginal, with the exception of the hydration state. in humans. On the other hand, data regarding direct tubular Efficient mechanisms closely match medullary toxicity of radiologic dyes is characterized poorly, oxygen supply and demand by the determination of and is occasionally (but not necessarily) attributed regional blood flow and metabolic activity. The to hypertonicity. Contrast agents, once filtered latter effect is mediated by the control of GFR and through the glomerulus, are not reabsorbed or me- proximal tubular reabsorption (both governing sol- tabolized by tubular cells. Furthermore, although ute delivery to the distal nephron) and by the the dose of radiocontrast has been recognized as an regulation of distal tubular reabsorption. Locally independent risk factor for renal dysfunction, dye produced prostaglandins and nitric oxide, and the nephropathy rarely is encountered in the absence generation of adenosine from the breakdown of of predisposing factors. Whereas the incidence of adenosine triphosphate, are major participants in dye nephropathy in the aged population (a predis- these mechanisms. The location of mediator pro- posing factor by itself) was only 1.3%, in the duction and the distribution and density of its re- absence of other risk factors it exceeded 60% in the ceptors are important in maintaining medullary presence of 3 or more confounding conditions.1 oxygenation. For instance, adenosine, released dur- These data suggest that factors other than direct ing adenosine triphosphate consumption, exerts tubular toxicity predominantly are responsible for cortical vasoconstriction and inhibits transport ac- dye nephropathy. We present evidence indicating tivity in mTALs through A1 receptors, and induces that among high-risk individuals, altered renal mi- medullary vasodilation through A2 receptors. crocirculation and medullary tubular hypoxic dam- Comparably, endothelin-1 exerts cortical vasocon- age play a pivotal role in the pathogenesis of dye striction through ET-A receptors, but enhances nephropathy. medullary blood flow activating ET-B receptors. In this perspective, corticomedullary redistribution of RENAL MEDULLARY OXYGENATION blood flow and the activation of the tubuloglo- Renal parenchymal oxygen supply is not ho- merular feedback mechanism may be regarded as mogenous: although the renal cortex is well oxy- measures that maintain medullary oxygenation and genated, a sharp decline in tissue PO2 is noted at the prevent tubular hypoxic damage (see reference 3 corticomedullary junction, reaching 20 to 30 mm for a detailed bibliography). Hg under normal physiologic conditions. Medul- lary hypoxia is the price paid for the renal diluting/ RADIOCONTRAST AND MEDULLARY concentrating capacity, attributed to the unique OXYGENATION architecture of the countercurrent system. Some The administration of a radiocontrast agent re- 10% only of renal blood flow and oxygen supply sults in a marked decline of outer-medullary oxy- are directed to the renal medulla through vascular genation. Using oxygen microelectrodes, we and bundles (vasa recta), originating from efferent ar- others found that renal parenchymal PO2 decreases, terioles of juxtamedullary nephrons. Oxygen coun- both in the cortex and in the medulla. This effect is tercurrent diffusion from descending to ascending most pronounced and dramatic in the outer me- vasa recta further compromises medullary oxygen dulla, where low ambient PO2 declines further to supply to some 8 mL/min/100 g tissue. This lim- values as low as 8 mm Hg after the injection of ited oxygen delivery is marginally sufficient for the ionic high osmolar dye, as well as with nonionic metabolic needs for tubular transport, predomi- low osmolar agents. This dye-induced decline in nantly performed by medullary thick ascending medullary oxygenation can be restored by the in- limbs (mTALs) and S3 (straight) segments of the hibition of mTAL reabsorptive activity with the proximal tubule. Consequently, oxygen extraction loop diuretic furosemide (see reference 3 for a by the renal medulla is near-maximal, reaching detailed bibliography). 79% of the regional oxygen supply, with only a Hypoxia-induced factor 1 recently has been de- small oxygen reserve left. This critical medullary tected within the medulla and medullary rays DYE-INDUCED NEPHROPATHY 479 shortly after the administration of contrast (C. aggregation within papillary vessels. In contrast, Rosenberger, unpublished data). This constitu- Agmon, using laser Doppler needle probes inserted tively formed heterodimer rapidly accumulates through the renal capsule, reported a continued during hypoxia because of the inhibition of the ␣ marked increase in outer-medullary flow after ra- subunit breakdown by oxygen-dependent prolyl diocontrast (see reference 3 for a detailed bibliog- hydroxylases. Thus, detection of hypoxia-induced raphy). These conflicting data may represent dif- factor, which initiates the cellular response to hy- ferent microcirculatory anatomy and behavior at poxic stress, serves as an early indicator of dye- the outer and inner medulla. Thus, although papil- induced medullary hypoxia. lary hypoxia may simply reflect altered regional Radiocontrast-associated medullary hypoxia is blood flow, the decline in outer-medullary ambient mediated through several mechanisms: extrarenal PO2 may be related predominantly to enhanced effects consist of decreased cardiac output that oxygen consumption. may compromise renal blood flow; pulmonary shunts that lead to arterial hemoglobin desatura- RISK FACTORS FOR DYE NEPHROPATHY tion; and a leftward shift of the oxygen-hemoglo- PREDISPOSE TO MEDULLARY HYPOXIA bin dissociation curve, decreasing oxygen avail- Most risk factors for contrast nephropathy are ability to tissues. Blood fluidity may be altered as characterized by structural or functional changes well. Renal effects can be divided into those com- that can predispose to aggravation of hypoxic promising the renal microcirculation and those stress during dye administration.3 Enhanced sys- mechanisms that increase oxygen demand for tu- temic vasoconstrictive stimuli (angiotensin II, bular reabsorptive work. vasopressin, catecholamines) accompany effec- Reabsorptive workload is enhanced by the sol- tive volume depletion, for instance, in patients ute load, and with the early, although transient, with heart failure, nephrosis, hypovolemic increase of GFR. Solute delivery to the distal shock, or dehydration. The medullary microcir- nephron increases markedly, leading to enhanced culation is deformed in pre-existing renal paren- distal tubular transport. chymal disease (the result of structural paren- Altered microcirculation is attributed to abrupt chymal changes), and local oxygen insufficiency changes in various mediators of the renal vascular may be enhanced by increased reabsorbtive tone. Dye administration increases plasma atrial workload in hypertrophic remnant nephrons. En- natriuretic peptide (ANP), endothelin-1, vasopres- hanced tubular reabsorption and oxygen con- sin, renal adenosine, and prostaglandin E2. Sym- sumption may accompany early or uncontrolled pathetic activity probably is enhanced as well. In diabetes owing to increased GFR and osmotic contrast, prostacyclin (PGI2) production declines. diuresis, respectively. Diabetes also is associated Cortical nitric oxide synthesis also is reduced and with defective nitrovasodilation, as is aging, hy- urinary cyclic guanosine monophosphate de- pertension, atherosclerosis, or hypercholesterol- creases. In addition, altered red-cell deformability emia. Defective prostaglandin synthesis is en- and an abrupt increase in renal volume and inter- countered in aging and in patients receiving stitial pressure owing to the forced diuresis also nonsteroidal anti-inflammatory agents. Addi- may alter the renal microcirculation. tional exposure to other nephrotoxins and other The net effect of all these changes on the renal comorbid states may intensify medullary hyp- microcirculation is biphasic. Total renal blood flow oxia by the induction of renal vasoconstriction initially increases for about 10 to 15 minutes, as- (cyclosporine), by the increase in metabolic sociated with enhanced GFR and diuresis. Subse- workload (mannitol), or by their combination quently, an extended decline in renal blood flow is (amphothericin, hypercalcemia). Rhabdomyoly- noted to about 30% below baseline. These changes sis, urine outflow obstruction, and sepsis are are not homogenous within the renal parenchyme. among other renal perturbations characterized by Although cortical flow in general follows the same altered renal microcirculation and predisposition pattern, conflicting data exist regarding medullary to medullary hypoxic damage. Finally, the more flow. Liss, examining exposed papillae with laser nephrotoxic ionic and high osmolar radiocon- Doppler and video-microscopic techniques, found trast agents exert more profound renal microcir- a substantial decline of blood flow and red-cell culatory alterations in a dose-dependent fashion. 480 HEYMAN AND ROSEN

Fig 1. In radiocontrast models of renal failure, ex- tensive vacuolar changes are observed in proximal tu- bules as shown. Such alter- ations also are present in human biopsy specimens of contrast nephropathy and do not correlate with the de- gree of renal failure either in humans or animals. Magnifi- cation: 1 ␮m plastic section .؋570

Thus, dysregulation of medullary oxygen bal- remnant kidney, acute inhibition of prostaglandin ance is encountered in most circumstances con- or nitric oxide synthesis, enhancement of GFR by sidered to predispose to dye nephropathy. amino-acid infusion, and urine outflow obstruc- tion.3,5 These insults, applied alone or in different ANIMAL MODELS OF CONTRAST combinations, were found to contribute to renal NEPHROPATHY dysfunction independently, with the magnitude of With the limited knowledge of the human pa- renal failure (increasing plasma creatinine and urea thology of contrast nephropathy, appropriate ani- levels, and declining creatinine clearance and tu- mal models are essential for the understanding of bular sodium reabsorption) proportional to the its pathophysiology. In animals given radiocontrast number of applied confounding factors. In all these dyes, proximal tubular vacuolization is a consistent protocols, renal morphology revealed variable finding, even when kidney function is well pre- medullary tubular necrosis that appeared as early served (Fig 1). Electron microscopy reveals that as 15 minutes after the administration of the radio- these vacuoles represent membrane outpouching contrast (Figs 2 and 3), predominantly affecting from infoldings of the basolateral membranes with mTALs and, to a much lesser extent, S3 segments other cellular elements being intact. Taken together in the outer medulla and medullary rays. These with the irregular occurrence or absence of urinary morphologic changes range from reversible injury markers of proximal tubular damage, such as kid- ney injury molecule-1 (KIM-1),4 proximal tubular (mitochondrial swelling with maintained cellular vacuolization should be regarded as a marker for integrity) to a more severe damage pattern (nuclear dye exposure, rather than an indicator of contrast pyknosis and disruption of cell membranes). A nephropathy. gradient of damage was noted, maximal among Similar to humans, healthy animals do not de- mTALs at the midinterbundle zone, most remote velop renal dysfunction when exposed to dye from vasa recta. By 24 hours, tubular necrosis was alone. Consequently, we developed experimental maximal (Fig 4), affecting up to 80% of mTALs in models of contrast nephropathy in rats using both the most severe models, occasionally associated dye administration and additional perturbations with injured collecting ducts and damaged inner- that mimic known clinical risk factors. Situations medullary structures. Apoptotic cell death also was that were found to predispose to experimental dye- noted with a pattern of distribution comparable induced renal dysfunction include chronic salt de- with that of tubular necrosis. pletion, angiotensin II infusion, prior unineph- An additional morphologic hallmark of dye ad- rectomy with compensatory hypertrophy of the ministration was outer-medullary vascular conges- DYE-INDUCED NEPHROPATHY 481

Fig 2. Shortly (within 15 min) after administration of contrast to indomethacin- treated, salt-depleted, uni- nephrectomized rats, both apoptotic type (condensed dark cells) and very focal ne- crotic changes (arrow) were noted in medullary thick as- cending limbs (*). Collecting ducts (CD) are unremarkable. Magnification: 1 ␮m plastic .section ؋600 tion that appeared as early as 10 minutes after the contrast nephropathy. The close functional-struc- administration of contrast, representing altered mi- tural correlation, found in the more severe models crocirculation. Tubular necrosis correlated with the of combined insults, is less evident in partial pro- extent of medullary congestion/stasis. It is conceiv- tocols with sparse focal damage. Indeed, renal dys- able that hypoxic endothelial dysfunction further function with minimal or absent tubular damage perpetuated medullary hypoxic stress and renal noted in these animals implies that reversible cell dysfunction. injury has been restored, or that renal microvascu- Taken together, these findings in animal models lar response predominates. Such a hemodynamic indicated that dye-induced altered medullary oxy- reaction, probably mediated by tubuloglomerular genation may evolve into apoptotic or necrotic feedback, helps in restoring medullary oxygenation tubular injury when mechanisms designed to main- by decreasing reabsorbtive workload. Yet, it is tain medullary oxygenation are hampered. In dif- manifested as diminished GFR with a low frac- ferent laboratories, other experimental models of tional sodium excretion. Additional confounding dye nephropathy have been developed, superim- factors may include tubular obstruction by tubular posed on predisposing factors, such as hypercho- cell casts, by oxalate and uric acid crystals, and by lesterolemia,6 diabetes or short periods of renal radiocontrast-precipitated Tamm Horsfall or Bence arterial clamping in rats, salt depletion and indo- Jones proteins in patients with myeloma. Direct methacin in rabbits, and salt depletion or conges- tubular cell toxicity also may coexist, as suggested tive heart failure in dogs.3 Detailed outer-medul- predominantly by in vitro experiments involving lary morphology was not available in most of these free radical formation and lipid peroxidation. Tu- bular obstruction and reduced GFR caused by vol- studies, but the experimental situations suggest an ume depletion may increase the transit time of the important adverse interaction between radiocon- radiocontrast within the tubular lumen, and may trast agents and medullary oxygen insufficiency. intensify direct cellular toxicity. Finally, as stated earlier, contrast studies often are performed in crit- OTHER POTENTIAL MECHANISMS OF RENAL ically ill patients treated with other nephrotoxins or DYSFUNCTION IN DYE NEPHROPATHY displaying comorbid states such as sepsis or myo- Possibly, synergism exists between hypoxic, he- hemoglobinuria that may exert additional tubulo- modynamic, toxic, and obstructive components in toxic and hypoxic insults. 482 HEYMAN AND ROSEN

Fig 3. Higher magnification of 1 ␮m plastic sections from animals cited in Fig 2. (A) A fragmented necrotic cell (arrow) with swollen mitochondria can be seen adjacent to the condensed dark cells of a MTAL. (B) Electron microscopy shows the marked differences between the MTAL condensed cellular elements (upper left) and unremarkable collecting duct (lower right). (C) Necrotic cell death with the associated mitochondrial swelling (arrows) is quite different than the condensed .adjacent cell. Magnification: 1 ␮m plastic section ؋700; electron microscopy ؋760; ؋ 1,800

THERAPEUTIC IMPLICATIONS interventions among high-risk patients include vig- Contrast nephropathy can be predicted to a large orous hydration and the use of low-osmolar and extent in high-risk patients, who often require im- nonionic agents with the most minimal dose pos- age analysis. Being an expected sequela of an sible. Proper hydration decreases vasoconstrictive elective medical intervention, dye nephropathy has stimuli in volume-depleted individuals, may en- been a target for numerous trials of therapeutic or hance dye elimination, and could increase GFR preventive measures. Most important are careful through natriuretic mediators. Intravenous admin- identification of patients at risk, avoidance of un- istration of 0.45% saline at a rate of 1 mL/kg/h for necessary contrast imaging in this population, and 12 to 24 hours currently is recommended, starting removal of coexisting perturbations, such as non- 6 to 12 hours before the contrast study and extend- steroidal anti-inflammatory drugs and other neph- ing up to 12 hours afterward. Hydration strategies rotoxins. So far, the only additional proven active with shorter periods of fluid administration and DYE-INDUCED NEPHROPATHY 483

Fig 4. Whether in the iso- lated perfused kidney or in vivo in the radiocontrast model of renal failure, the dominant change is necro- sis of the MTAL. At 24 hours after induction of this model, the changes are most evi- dent. In this photograph, the tubules adjacent to the vasa recta (upper right, lower left) are very well preserved, in contrast to those in the cen- tral interbundle zone. Mag- nification: 1 ␮m plastic sec- .tion ؋200 partial substitution of the preceding regimen with Restoration of Renal Hemodynamics oral hydration are suggested. The use of 0.9% This approach has been studied extensively in 7 saline may be superior among high-risk patients. experimental settings and in clinical trials among The substitution of high osmolar ionic agents with high-risk patients. Its logic is to reverse radiocon- the newer generation of low osmolar and nonionic trast-induced decline in renal blood flow and GFR, dyes is associated with a two-thirds reduction of thus washing out radiocontrast and removing oblit- the risk for dye nephropathy among high-risk pa- erating tubular casts. L-arginine prevented a dye- tients. This predominantly reflects milder renal he- associated decrease in renal blood flow and GFR in modynamic alterations, but attenuated direct tox- hypercholesterolemic animals,6 perhaps through icity has been suggested as well. the amelioration of endothelial dysfunction. The Because intravenous hydration protocols and the administration of dopamine increases renal blood use of low-osmolar and nonionic agents are time flow, but its protective effects among high-risk consuming and expensive, it is reasonable to sug- patients have been inconsistent and perhaps dele- gest that they should be mandatory for high-risk terious among diabetic patients9 and during estab- patients, although cost-benefit implications may be lished contrast nephropathy.10 Conflicting results taken into account for their application in low-risk also characterize studies with the dopamine-1 re- individuals. ceptor agonist fenoldepam, irrespective of whether Other principal therapeutic interventions, cur- the patients were diabetic.11,12 Prevention of dye rently considered experimental, have been directed nephropathy with theophylline also has been in- at (1) the restoration of renal hemodynamics, (2) consistent,8,10,13,14 and data regarding the use of the improvement of medullary oxygenation by re- calcium-channel blockers are sparse.8 The effect of duction of tubular transport, and (3) the ameliora- the ANP analogue anaritide was found to be absent tion of toxic damage. The reader is referred to a or inconsistent, with a deleterious outcome in the thorough review by Murphy et al,8 updated by diabetic subpopulation.8,15 The administration of a some recent reports quoted later. It is noteworthy nonselective endothelin-receptor antagonist ad- that most of these studies were performed in high- versely affected the clinical outcome.16 However, risk and well-hydrated patients who were given animal studies indicate that selective endothelin-A different classes of dye. receptor antagonists may have a protective effect,17 484 HEYMAN AND ROSEN probably underscoring the importance of endothe- sorption, and may itself result in hypoxic acute lin-B receptorÐmediated medullary vasodilation. tubular necrosis.8 Finally, the administration of prostaglandin E an- alogue may be beneficial,18 particularly after the Measures to Attenuate Toxicity use of nonsteroidal anti-inflammatory drugs, or in N-acetylcysteine (NAC), given before and after the aged population, characterized by defective radiocontrast studies, recently has been reported to renal prostaglandin synthesis.2 improve renal outcome among high-risk well-hy- On the whole, the inconsistent effect of renal drated patients.19 This finding was supported by vasodilators and the deleterious outcome in some some20-22 but not all23 subsequent clinical trials. In subpopulations suggest that the potential beneficial one of these studies, NAC-associated renal protec- outcome of enhanced cortical flow and GFR is tion was restricted to patients given small volumes counterbalanced by the enhancement of solute de- of dye.21 The beneficial effects of this agent have livery for reabsorption in the distal nephron, which been attributed primarily to its antioxidant proper- leads to intensification of outer-medullary hypoxia. ties but as suggested by Safirstein et al,24 recent findings indicate that it also attenuates endothelial Reduction of Metabolic Workload dysfunction25 and may improve renal hemodynam- ics after the administration of contrast.26 Neverthe- In uninephrectomized salt-depleted rats given less, an experimental model of dye nephropathy in indomethacin and radiocontrast, furosemide mark- rats subjected to concomitant inhibition of cyclo- 2,7 edly attenuated hypoxic mTAL damage. How- oxygenase and nitric oxide synthase revealed an ever, kidney dysfunction was improved only mod- unexpected finding: renal dysfunction tended to be estly, probably reflecting the prerenal component milder with NAC administration, but this was as- of dehydration. When cosupplemented with saline, sociated paradoxically with a trend toward en- indices of dehydration resolved, and kidney func- hanced tubular hypoxic damage.26 Again, this tion was preserved as well (see references 5 and 8 might imply that the beneficial effects of restora- for a detailed bibliography). The rational of this tion of renal hemodynamics and tubular cytopro- approach is to prevent dye-associated medullary tection can be offset by the enhancement of reab- hypoxia. However, in 2 clinical studies by So- sorptive workload that may increase medullary lomon et al and by Weinstein et al, as compared hypoxic damage. with hydration protocol, hydration supplemented with furosemide adversely affected kidney func- CONCLUSIONS tion in high-risk patients given dye. Weinstein et al Radiocontrast agents intensify medullary hyp- found that furosemide-treated subjects lost 0.7 kg oxia. Experimental data and clinical characteristics on average, whereas a 1.3-kg weight gain was of predisposing factors indicate that dye nephrop- noted in patients randomized to hydration alone, athy is mediated predominantly by outer-medul- suggesting that in furosemide-treated subjects the lary hypoxic tubular damage, probably combined hydration protocol has been insufficient (see refer- with and accentuated by endothelial dysfunction ences 5 and 8 for a detailed bibliography). A lack and renal microcirculatory alterations. Tubular ob- of ameliorating effects of furosemide also might be struction and direct toxic effect of the dye also may expected in the aged population owing its failure to play a role. Therapeutic interventions directed at improve medullary oxygenation, determined non- renal vasodilation and enhancement of GFR yield invasive with BOLD magnetic resonance imaging. inconsistent results and may have deleterious ef- This may reflect diminished medullary activity of fects, particularly among diabetic patients, perhaps cyclooxygenase or nitric oxide synthase in this age reflecting further intensification of medullary hy- group.2 It is noteworthy that, as shown by Solomon poxemia. Furosemide, given to improve medullary et al, kidney dysfunction after dye exposure in hypoxemia, adversely affected renal function, per- high-risk patients was more pronounced also in a haps owing to ensuing dehydration. Noninvasive mannitol-hydration group. As with furosemide, technologies, such as positron emission tomogra- this also may be related to ensuing volume deficit, phy and BOLD magnetic resonance imaging are but in addition mannitol aggravates medullary hy- critically needed for the real-time determination of poxemia owing to augmented distal tubular reab- intrarenal oxygenation and metabolic status. With DYE-INDUCED NEPHROPATHY 485 such technologies available we may better under- failure after coronary angioplasty in patients with preexisting stand the relative contribution of hypoxia and renal insufficiency. Am J Cardiol 83:260-263, 1999 altered renal microcirculation. Such technical 11. Tumlin JA, Wang A, Murray PT, et al: Fenoldopam mesylate blocks reductions in renal plasma flow after radiocon- progress may enable a more logical therapeutic trast dye infusion: A pilot trial in the prevention of contrast approach (ie, to improve hemodynamics or to re- nephropathy. Am Heart J 143:894-903, 2002 duce metabolic activity). 12. Allaqaband S, Tumuluri R, Malik AM, et al: Prospective Proper selection of patients, elimination of con- randomized study of N-acetylcysteine, fenoldopam, and saline current use of nephrotoxic agents, appropriate hy- for prevention of radiocontrast-induced nephropathy. Cathet dration, and the use of low osmolar and nonionic Cardiovasc Intervent 57:279-283, 2002 13. Kapoor A, Kumar S, Gulati S, et al: The role of theoph- dye in high-risk patients at the lowest volumes ylline in contrast induced nephropathy: A case-control study. required are currently the only proven preventive Nephrol Dial Transplant 17:1936-1941, 2002 measures. It is conceivable that to further reduce 14. Huber W, Ilgmann K, Page M, et al: Effect of theoph- the risk for dye nephropathy in high-risk patients, ylline on contrast material-induced nephropathy in patients with combined strategies should be tested. Suggested chronic renal insufficiency: Controlled, randomized, double- combinations are the inhibition of tubular transport blinded study. Radiology 223:772-779, 2002 15. Kurnik BR, Allgren RL, Genter FC, et al: Prospective by loop diuretic (with sufficient fluid replacement), study of atrial natriuretic peptide for the prevention of radio- in addition to agents that improve renal microcir- contrast-induced nephropathy. Am J Kidney Dis 31:674-680, culation (theophylline, fenoldepam, prostaglandin 1998 E analogues, L arginine, or endothelin-A receptor 16. Wang A, Holcslaw T, Bashore TM, et al: Exacerbation antagonists), perhaps combined with NAC, which of radiocontrast nephrotoxicity by endothelin receptor antago- might provide additional vasodilatory and cytopro- nism. Kidney Int 57:1675-1680, 2000 17. Bird JE, Giancarli MR, Megill JR, et al: Effects of tective effects. endothelin in radiocontrast-induced nephropathy in rats are mediated through endothelin-A receptors. J Am Soc Nephrol REFERENCES 7:1153-1157, 1996 1. Rich MW, Crecelius CA: Incidence, risk factors, and 18. Koch JA, Plum J, Grabensee B, et al: Prostaglandin E1: clinical course of acute renal insufficiency after cardiac cathe- A new agent for the prevention of renal dysfunction in high risk terization in patients 70 years of age or older: A prospective patients caused by radiocontrast media? PGE1 Study Group. study. Arch Intern Med 150:1237-1242, 1990 Nephrol Dial Transplant 15:43-49, 2000 2. Epstein FH, Prasad P: Effects of furosemide on medullary 19. Tepel M, van der Giet M, Schwarzfeld C, et al: Preven- oxygenation in younger and older subjects. Kidney Int 57:2080- tion of radiographic-contrast-agent-induced reductions in renal 2083, 2000 function by acetylcysteine. N Engl J Med 343:180-184, 2000 3. Heyman SN, Reichman J, Brezis M: Pathophysiology of 20. Diaz-Sandoval LJ, Kosowsky BD, Losordo DW: Ace- radiocontrast nephropathy: A role for medullary hypoxia. Invest tylcysteine to prevent angiography-related renal tissue injury Radiol 34:685-691, 1999 (the APART Trial). Am J Cardiol 89:356-358, 2002 4. Han WK, Bailly V, Abichandani R, et al: Kidney injury 21. Briguori C, Manganelli F, Scarpato P, et al: Acetylcys- molecule-1 (KIM-1): a novel biomarker for human proximal teine and contrast agent-associated nephrotoxicity. J Am Coll tubule injury. Kidney Int 62:237-244, 2002 Cardiol 40:298-303, 2002 5. Heyman SN, Lieberthal W, Rogiers P, et al: Animal 22. Kay J, Chow WH, Chan TM, et al: Acetylcysteine for models of acute tubular necrosis. Curr Opin Crit Care 8:526- prevention of acute deterioration of kidney function following 534, 2002 elective coronary angiography and intervention: a randomized 6. Andrade L, Campos SB, Seguro AC: Hypercholesterol- emia aggravates radiocontrast nephrotoxicity: Protective role of controlled trial. JAMA 289:553-558, 2003 L-arginine. Kidney Int 53:1736-1742, 1998 23. Durham J, Fishbane S, Caputo C, et al: A randomized 7. Mueller C, Buerkle G, Buettner HJ, et al: Prevention of controlled trial of N-acetylcysteine to prevent contrast nephrop- contrast media-associated nephropathy: Randomized compari- athy in cardiac angiography. Kidney Int 62:2202-2207, 2002 son of 2 hydration regimens in 1620 patients undergoing cor- 24. Safirstein R, Andrade L, Vieira JM: Acetylcysteine and onary angioplasty. Arch Intern Med 162:329-336, 2002 nephrotoxic effects of radiographic contrast agents: a new use 8. Murphy SW, Barrett BJ, Parfrey PS: Contrast nephropa- for an old drug. N Engl J Med 343:210-212, 2000 thy. J Am Soc Nephrol 11:177-182, 2000 25. Conesa EL, Valero F, Nadal JC, et al: N-acetyl-L-cys- 9. Weisberg LS, Kurnik PB, Kurnik BRC: Risk of radiocon- teine improves renal medullary hypoperfusion in acute renal trast nephropathy in patients with and without diabetes mellitus. failure. Am J Physiol 281:R730-R737, 2001 Kidney Int 45:259-265, 1994 26. Heyman SN, Goldfarb M, et al: N-acetylcysteine (NAC) 10. Abizaid AS, Clark CE, Mintz GS, et al: Effects of ameliorates compromised renal microcirculation: Studies in dopamine and aminophylline on contrast-induced acute renal rats. Kidney Int (in press)