Renal toxicity

Renal toxicityIntroduction:

The extraordinary growth of interventional radiology techniques has led us to regularly expose patients with the nephrotoxic potential of iodinated contrast media (ICP).Acute renal failure (AKI) secondary to PCI is part of the leading trio of ARI acquired in hospital. It is a necessity to know the best ways to prevent the toxicity of these products. The rigorous clinical evaluation of the renal toxicity of the PCI and its methods of prevention is relatively recent, it was practically non-existent until about fifteen years ago. Today, the medical literature in this field is so abundant that it is difficult to extract a clear message. The objective of this synthesis is to try to clarify the sometimes contradictory positions that the reading of the journals has in the prevention of the nephrotoxicity of the PCI.

Definition, clinical forms and diagnosis:

There is no truly consensual definition of ARI at ICPs. The two most commonly accepted definitions used in recent studies are an increase of more than 25% in baseline serum creatinine or an increase of 44 μmol / l (0.5 mg / dL) in serum creatinine. within 72 hours after the injection of the product. These changes in serum creatinine, usually reversible, are not mere “biological” variations. Indeed, studies have shown that the non-renal morbidity and mortality of patients undergoing this type of acute reduction in glomerular filtration was significantly greater than that of unaffected patients. IRA at PCI often begins immediately after the radiological procedure. The increase in creatinine is evident before the 48th hour, and the decrease in serum creatinine often begins between the 3rd and the 5th day, to be complete in 10 to 14 days. Diuresis is usually preserved. The use of dialysis is rarely necessary and is aimed at the patients most at risk. The need to implement a dialysis technique concerns a subgroup of patients in whom the vital prognosis is frequently engaged, as in any IRA. In a study of 1,800 consecutive coronary angiography patients, the incidence of ARF was 14.4% and dialysis was required in 0.8% of patients.

The hospital mortality of dialysis patients was 36%, and their 2 year survival was only 19%.

The notion of exposure to the toxic and the occurrence of accidents in at-risk patients makes the diagnosis of ARI to PCI usually simple. However, care should be taken not to make this diagnosis too much and not to ignore some common causes of ARI occurring in hospitals: functional ARI (hypotension, dehydration, rhythm disturbances, converting enzyme inhibitors) or septic, other toxic causes (NSAIDs [non-steroidal anti-inflammatory drugs], platinum salts, aminoglycosides, antivirals), myeloma, hypercalcemia, obstruction of the urinary tract, cholesterol emboli disease.

This last diagnosis can also be considered after a vascular radiology procedure, but has some characteristics that make it possible to distinguish it from IRA to PCI: the renal functional impairment is later (a few days or weeks) and associates with livedo and frequent abnormalities of distal perfusion (blue or purple toe).

Eryosinophilia and hypocomplementemia are possible. The general state can be frankly altered, realizing a picture of systemic angiitis. As a rule, the progression of cholesterol emboli disease is neither spontaneously nor rapidly favorable.

The proven IRA treatment of IBO does not have a great deal of particularity. As with any ARF, care must be taken to maintain a satisfactory hemodynamic state and hydroelectrolyte balance, no new nephrotoxic aggression should occur. The early use of hemodialysis, even if it purifies a certain amount of PCI, has no favorable effect on the evolution of renal function.

Frequency and risk factors:

In a study of 16,248 patients receiving venous or arterial iodine, the incidence of ARI was close to 1% (183 cases).The mortality of patients who developed this accident was five times higher than that of patients of the same age with the same initial level (34% versus 7%). IBR ARIs were associated with significant over-morbidity, including infections, bleeding events, impaired alertness, and respiratory dysfunction. Overall, in an unselected population, the frequency of ARI to PCI is around 2%. This figure does not reflect the great variability of the frequency of this accident according to the population concerned. The most important risk factor is the existence of chronic renal insufficiency (CRF), the risk of ARI at PCI being very strongly related to the degree of renal failure. If renal failure is absolutely isolated (rare situation) and creatinine clearance (Cl.C) is greater than 50 ml / min, with no other comorbidity, then the risk of ARI is very low. But many other risk factors may increase the risk associated with the degree of CKD, and make it very significant for C-ones greater than 50 ml / min. These factors include: diabetes mellitus, age greater than 70 years, heart failure, effective dehydration and hypovolemia, myeloma, administration of other nephrotoxic agents (ciclosporin, aminoglycosides, NSAIDs, platinum salts, antivirals, IVIG, repetition of PCI), and the existence of proteinuria.

Thus, the risk is almost zero in a population with strictly normal kidney function and no other risk factors; it progressively increases to exceed 50% in a population characterized by advanced renal failure (Cl.C less than 15 ml / min) or severe renal insufficiency (Cl.C less than 30 ml / min) associated with diabetes mellitus or heart failure. The peculiarity of patients with severe CKD (especially if associated with diabetes and cardiac dysfunction) is that the use of extrarenal treatment may become definitive after an acute accident, precipitating a few months or more the onset of chronic dialysis.

It seems that diabetes mellitus without nephropathy is not a significant risk factor. However, diabetes is such a risk factor for nephropathy that one should be cautious about diabetics. Myeloma, without renal dysfunction, is associated with a low risk, close to 2%, which does not prohibit a very careful use of PCI.

Mechanisms of acute renal failure secondary to iodinated contrast media:

Several mechanisms interact to explain secondary IRA to IBOs. In laboratory animals as in humans, the IRA only occurs in the case of associated risk factors, represented in this case by experimental conditioning. Mechanisms of this ARF include: renal vasoconstriction, decreased nitric oxide (NO) -adjusted vasodilatation or some prostaglandins, direct tubular cell toxicity, excessive production of free radicals, increased oxygen requirements during the ischemic period, multifactorial increase in intratubular pressure (diuretic effect, tubular obstruction, urinary hyperviscosity).Functional alterations of red blood cells, cells carrying strong antioxidant equipment, are not to be neglected. The reactive derivatives of oxygen could be essential mediators of cytotoxicity on tubular cells, partly explaining the protective effect of antioxidants such as acetylcysteine.

The main mediators of renal vasoconstriction are probably endothelin, adenosine, and the hyperosmolality of PCI.Depending on the experimental conditions, the effects of PCI on the renal circulation may be opposite. Thus, in rats, iothalamate increases renal medullary blood flow basally but results in pronounced ischemic phenomena and necrosis of tubular cells of the broad ascending limb if the vasodilator tone transmitted by prostaglandins and NO is blocked.The tubular cells of the outer medullary zone of the kidney are very sensitive to hypoxia. In this zone, the oxygen consumption perfusion scale is easily unbalanced towards relative or absolute cellular hypoxia. The vascularization of this zone is very finely regulated by multiple mediators (prostaglandins, NO, angiotensin II, adenosine, endothelin, atrial natriuretic peptide [PAN], etc.). These mediators are all potential targets for the pharmacological prevention of renal toxicity of IBD. However, the differences between animal and man concerning the regulation of the renal circulation, associated with the inhomogeneity of clinical situations, make it often difficult or even impossible to identify protective effects in humans. many molecules with experimental preventive interest. Some agents have been shown to be nephroprotective in animals (ANP, endothelin antagonist) but may aggravate renal dysfunction induced by PCI in humans. Any hypothesis concerning the nephroprotection must be validated in the man on a large scale, and not leave too quickly the experimental field.


The best treatment for the nephrotoxicity of IBD is preventive. It has many aspects.

Screening of patients at risk:

The screening of patients at risk involves a measure of creatinine in all patients at risk of renal damage (hypertensive, diabetic, proteinuric, history of nephropathy or any urinary tract disease, presence of one or more factors vascular risk, age greater than 70, history of neoplasia, infection, etc.).

This amounts to performing a creatinine test in a large number of patients. A patient assessment sheet, distributed or sent at the time of the appointment, can facilitate this screening.

It has the added benefit of screening patients at risk of allergic or allergic reactions, women who may be pregnant, patients treated with biguanides, or frequent users of nephrotoxic drugs (NSAIDs).

Alternatives to iodinated contrast media: CO2 and gadolinium

Whenever possible and when they offer good diagnostic performance, exams without PCI are preferred in patients at risk. Except in a life-threatening emergency, no iodine test should be performed in a patient with a creatinine greater than 200 μmol / l or a Cl.C less than 30 ml / min, and without consideration being given to consider all alternatives to the injection of the PCI. Particular emphasis must be placed on peripheral vascular magnetic resonance imaging techniques. The magnetic resonance technique is also suitable for imaging the urinary system (uro-RM). The scan may not include PCI injection. Although injection is often essential for tumor imaging, it is rarely used for lithiasis imaging of the urinary system (spiral CT).

For vascular imaging of the upper limbs (before creating an arteriovenous fistula), fine vascular mapping performed by ultrasound or magnetic resonance imaging (MRI) can sometimes substitute for iodine examinations. If direct vascular opacification is mandatory, the use of contrast agents presumed to be less nephrotoxic than IBD can be discussed.CO2 can be used for venous imaging. However, the product is not free of significant non-renal toxicity. Not all centers are good at gas phlebography. Gadolinium (MRI contrast agent) has been tested as an alternative to PCI in vascular radiology. It appears less nephrotoxic, without being without a significant nephrotoxicity. Gadolinium does not have authorization for

radiological procedures other than MRI. On the other hand, its non-renal toxicity is significant. It can not today be recommended for widespread use in patients at risk for renal toxicity of IBD. When used, the dose should not be greater than 0.3 mmol / kg. Studies are needed to clarify the value of non-iodinated contrast agents in this situation.

Dose of iodinated contrast agents and repetition of injections:

The PCI dose and the frequent repetition of injections are risk factors for nephrotoxicity. Formulas have been developed to adapt the maximum dose of PCI to kidney function. One of these formulas is to divide the maximum theoretical dose (5 ml / kg) by the value of serum creatinine in mg / dl (example: 2.5 ml / kg of iodine for a serum creatinine at 2 mg / dl or 176 μmol / l). Common sense suggests limiting the dose of PCI as much as possible, knowing that it is often difficult, if not impossible (in the case of pulmonary computed tomography), to perform an effective examination with less than 1 ml / kg of PCI. Cardiac ventriculography, which may accompany a diagnostic coronary angiography, may be replaced by isotopic measurement of the ventricular ejection fraction. Whenever possible, an interval of at least 3 days (and if possible 5) between two PCI injections must be respected. This allows in particular to confirm that the first injection was not responsible for an IRA, the second injection can then have catastrophic consequences.

Limitation of fasting before injection of iodinated contrast media:

Fasting and duration are a prescription of the radiologist. Fasting must not be systematic. It is contradictory with the need for hydration

correct, it promotes lipothymia, hypoglycemia and vomiting. It is a source of discomfort and unjustified postponement of examinations. In rare cases, and for a limited time, fasting can be useful: to limit the biliary emptying, to favor the visualization of the digestive wall or the pancreas, to facilitate the realization of gestures that can justify a sedation or a general anesthesia. In the latter case, the prescription of fasting is the responsibility of the physician anesthesiologist-master: as a rule, solid foods are then interrupted 6 hours before the examination, and the intake of clear liquids is possible until 2 hours before the examination.

Osmolality of iodinated contrast agents and risk of nephrotoxicity:

The nature of the iodinated contrast product is important. Numerous studies have shown a better tolerance of low osmolality products (600 to 800 mOsm / kg) compared to high osmolality products (greater than or equal to 1200 mOsm / kg): the former are systematically chosen from patients at risk. A debate exists about the possible superiority, in terms of renal tolerance, of iso-osmolar plasma products (close to 300 mOsm / kg) compared to low osmolality products. Two studies suggest a better tolerance of an iso-osmolar agent, iodixanol, but others are divergent. The inconsistent results of studies, the high cost of the product (iodixanol) and a possible proarrhythmic effect do not lead to recommend a wide use, until clarification of the question. The use of the least nephrotoxic PCI does not dispense with their use, under the best possible conditions (see below).

Water and sodium intake:

Accurate control of the hydration status of patients with water and sodium intake before and after iodine examination is still an essential part of the prevention of renal toxicity of IBD, even if the evaluation from this point has not always been very rigorous. The modalities of intake of water and sodium are varied. Today, it is only in patients at least risk that we can limit ourselves to a simple increase in oral intake of water and salt (wide water ration, salty food, bicarbonated drinks within 48 hours surrounding the PCI injection). In other situations, this measure should be combined with intravenous water and sodium. An addition of sodium chloride in isotonic solution is preferable to a semiisotonic one. For very high risk patients, it is prudent to start intravenous administration at least 12 hours before the examination, at a rate of 100 ml / h. The correction of a possible dehydration is thus obtained, and the cardiac tolerance of the contribution could be verified. For intermediate-risk patients, an intravenous infusion started in the 3 to 6 hours before the examination (and continued as soon thereafter) and providing a total of 15 to 20 ml / kg of isotonic saline is an acceptable alternative to infusion. started the day before.

Recently, one study showed that intravenous sodium bicarbonate was more effective than sodium chloride (both in semi-isotonic), in preventing IRA at PCI. Although this study is unique, it probably allows sodium bicarbonate to be considered as an alternative at least equivalent to sodium chloride. The last attractive aspect of this trial was the close character of the intravenous injection around the radiological procedure (1 hour before and 6 hours later) that can be used for this formula in case of emergency.

Pharmacological agents:

The number of tests that have tested the possible preventive effect of various pharmacological agents on the renal toxicity of IBD is considerable. Two recent reviews analyze them in detail. Overall, the results are very disappointing.

Diuretic, natriuretic or dopaminergic agents:

Mannitol does not improve the prevention of PCI nephrotoxicity (coronary angiography), already provided by a semi-osmolar sodium chloride infusion. In the same study, furosemide is ineffective and may even slightly increase the risk of nephrotoxicity. Dopamine and atrial natriuretic peptide are also ineffective and likely to significantly increase the risk of ARI in high-risk patients such as severe renal failure with or without diabetes. In a controlled study of 247 patients, anaritide (PAN) was ineffective, also associated with a slight increase in risk.

In the Contrast study of 351 patients (half of whom were diabetic), fenoldopam (dopaminergic agent) was found to be ineffective in preventing nephrotoxicity (34% versus 30% for placebo) secondary to cardiovascular angiography.

Other agents:

Many agents have been tested, sometimes in methodologically rigorous studies. For all the following agents, it is possible to conclude that an inefficiency or a benefit-risk ratio goes against their use. The list includes: short or long-acting calcium channel blockers, captopril, theophylline, prostaglandin E1. In one study, the use of a non-selective endothelin antagonist almost doubled the risk of nephrotoxicity in patients undergoing coronary angiography (56% vs. 29%). This result highlights, as the results obtained with the atrial natriuretic peptide, the differences between experimental prevention and clinical prevention of PCI toxicity. Theophylline, the leader in nonspecific adenosine antagonists, has been extensively studied, and several studies do not give it a preventive value, limiting the scope of positive studies. Its potential side effects (arrhythmia, convulsions) and low therapeutic index are other arguments that are not in favor of its use. Studies with prostaglandin E1 do not clearly establish a renal protective effect. Its side effects are not in favor of wide use.

Case of N-acetylcysteine:

Acetylcysteine ​​is an antioxidant, free radical scavenger. It increases the biological effects of NO by combining to form S-nitrosothiol, a more stable and more “vasoactive” agent than NO. It also increases the expression of NO-synthase.For the record, this drug is a bronchial thinning agent whose effectiveness as an antidote for severe intoxications with paracetamol is demonstrated. Since the first study of the New England Journal of Medicine suggesting its effectiveness in the prevention of PCI, this inexpensive drug with a favorable tolerance profile has been the subject of a craze with a proliferation of studies, reviews and meta-analyzes . They are not analyzed in detail here. As usual, the multiplicity of negative studies, even if a slightly higher number of studies is in favor of efficiency, is against a very dramatic effect of the product.

Overall, associated with hydration and sodium intake, acetylcysteine ​​seems to bring a benefit compared to the only water-soluble contributions. Today, the arguments for using N-acetylcysteine ​​extensively are not all rigorous and combine a relatively positive efficiency-ineffectiveness balance with low cost and excellent tolerance. A large controlled study has confirmed the value of N-acetylcysteine ​​in patients with acute coronary syndrome requiring coronary angiography. The dependence of the protective effect of N-acetylcysteine ​​reinforces the message of interest. Finally, its favorable effect goes beyond the scope of simple nephroprotection since the group of patients receiving the highest dose of N-acetylcysteine ​​was marked by a lower mortality.

The use of N-acetylcysteine ​​is therefore recommended, especially when the risk is important or very important, and when the exploration is a coronarography. The use of N-acetylcysteine ​​should not replace any other recommendations already formulated and summarized in conclusion.

Extracorporeal preventive treatment:

“Preventive” hemodialysis, aimed at eliminating iodinated contrast medium after angiographic procedure and thus preventing possible nephrotoxicity, is ineffective in preventing renal risk. Apart from a therapeutic situation (overload, hyperkalemia, decompensation of a latent uremic state), it should not be used. Prophylactic haemofiltration was found to be of interest in a study of 114 patients with renal insufficiency (creatinine greater than 176 μmol / l) to undergo coronary angiography. Extrarenal cleansing for IRA was required in 25% of control subjects compared to 3% of subjects treated preventively with haemofiltration. The authors found a better survival at 1 year of the group of patients treated with hemofiltration. But is not hemofiltration an extrarenal treatment technique? The use of systematic haemofiltration involves imposing an extracorporeal treatment technique to 100% of patients to avoid 25% of them. As part of a prophylactic procedure, it appears very heavy, especially for the most familiar resuscitators and nephrologists of central venous catheterization and extracorporeal circulation in patients at high risk of bleeding.

The sub-group of patients (perhaps at very high risk) that could benefit from this procedure is yet to be defined.

Summary of recommendations:

They are adapted according to the recommendations formulated recently by the French Society of Radiology.

Identification of patients at risk:

Patients at risk are identified: clinical and biological evaluation, self-administered questionnaire. All alternatives to PCI injection are studied. Nephrotoxic drugs (NSAIDs) are discontinued, diuretics also (48 hours before) when possible. A low osmolality PCI is chosen and the radiologist (or cardiologist) knows that he must use the smallest possible amount of PCI. When feasible, do not fast or fast as fast as possible.

Very low risk patients:

Very low risk patients have a Cl.C greater than 50 ml / min with no other risk factor, or a single risk factor without renal dysfunction (risk less than 2%).

We can then limit ourselves to an abundant hydration (bicarbonate drinks), a normosoded diet in the 48 hours preceding the examination and a very fast recovery of intake rich in water and salt after the examination.

Patients at low or medium risk:

Patients at low or medium risk have a Cl.C between 30 and 50 ml / min, without diabetes (risk ranging from 2 to 10%).

The recommendations are the same as above, associated with an infusion of isotonic saline in the hours preceding the examination, according to the modalities defined above. If the patient is not in a conventional hospital setting, it is preferable that the iodine test be performed only at the end of the morning, which may eventually lead to intravenous delivery to the day hospital. A dose of 1200 mg of N-acetylcysteine ​​is recommended the day before and the day of the examination.

It does not dispense with all the other preventive measures.

Creatinine is imperatively dosed within 48 hours after the examination.

Patients at high or very high risk:

Patients at high or very high risk have a Cl.C. less than 30 ml / min, or between 30 and 50 ml / min but with multiple comorbidities (risk greater than 10%).

Only an absolute necessity justifies the injection of iodine. The opinion of the nephrologist is solicited because he may have to take charge of a secondary IRA in the procedure. Patients most at risk (creatinine greater than 200 μmol / l and diabetic and / or heart failure) are warned about the possibility of dialysis treatment in case of severe ARI.

Hospitalization is systematic, at least the day before the examination. The isotonic saline infusion can be performed at a rate of 100 ml / h in the 12 hours before and 12 hours after the radiological procedure. If there is severe cardiac dysfunction, nephrotic syndrome or decompensated cirrhosis, diuretics may be retained to maintain natriuresis. A dose of 1200 mg of N-acetylcysteine ​​is recommended the day before and the day of the examination.

It does not dispense with all the other preventive measures.

Saline intake is adapted to cardiac function.

The amount of PCI is as low as possible (less than 1 ml / kg if possible). Creatinine, urinary volume and ionogram are regularly evaluated within 5 days of the examination.

Emergency procedure in the presence of risk factors:

The use of emergency coronarography in at-risk patients is common. Possible dehydration can be corrected quickly with isotonic saline. In the absence of dehydration, sodium bicarbonate intake is chosen one hour before and up to 6 hours after the examination according to the protocol described by Merten. Minimizing the dose of PCI remains appropriate. N-acetylcysteine ​​is administered the same day and the day after the test.


There are effective measures to prevent the renal toxicity of iodinated contrast media. Properly implemented, they can avoid a large number of iatrogenic accidents. This review is not a plea against the use of PCI. In interventional radiology, they are generally used in conditions where the vital or functional prognosis is engaged. Failure to perform these exams to remove patients at risk from the effects of PCI may be a poor choice for patients’ cardiovascular and vital future. But to carry out an iodine examination in a patient at risk without prophylaxis, and a fortiori by keeping him in a situation increasing the risk (dehydration, prescription of toxic, unjustified fasting), is a fault, both the information concerning the prevention of the nephrotoxicity IBOs are now being released.

This field illustrates the necessary close communication between the prescribers (all the doctors), the actors of the gesture (radiologists and cardiologists), and those who will have to take care of the ARI if the prevention was not effective or badly carried out (resuscitators, nephrologists). Prevention studies are still needed; they must be stratified on the risk (low or high) and the intensity of the renal accident considered (moderate or severe).