Effects of contrast media on renal function in patients with cirrhosis: A prospective study

Authors

  • Mònica Guevara,

    Corresponding author
    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    3. Instituto Reina Sofía de Investigación Nefrológica, Barcelona, Spain
    • Liver Unit, Hospital Clínic, Villarroel, 170, 08036 Barcelona, Spain
    Search for more papers by this author
    • fax: +34 93 22 75 457

  • Glòria Fernández-Esparrach,

    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    Search for more papers by this author
  • Carlo Alessandria,

    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    3. Instituto Reina Sofía de Investigación Nefrológica, Barcelona, Spain
    Search for more papers by this author
  • Aldo Torre,

    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    3. Instituto Reina Sofía de Investigación Nefrológica, Barcelona, Spain
    Search for more papers by this author
  • Carlos Terra,

    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    3. Instituto Reina Sofía de Investigación Nefrológica, Barcelona, Spain
    Search for more papers by this author
  • Xavier Montañà,

    1. Radiology Department, Hospital Clínic, Barcelona, Spain
    Search for more papers by this author
  • Carlos Piera,

    1. Nuclear Medicine Department, Hospital Clínic, Barcelona, Spain
    Search for more papers by this author
  • Maria Luisa Alvarez,

    1. Biochemistry Laboratory, Hospital Clínic, Barcelona, Spain
    Search for more papers by this author
  • Wladimiro Jiménez,

    1. Hormonal Laboratory, Hospital Clínic, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    Search for more papers by this author
  • Pere Ginès,

    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    3. Instituto Reina Sofía de Investigación Nefrológica, Barcelona, Spain
    Search for more papers by this author
  • Vicente Arroyo

    1. Liver Unit, Hospital Clínic, Barcelona, Spain
    2. Institut d'Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Barcelona, Spain
    3. Instituto Reina Sofía de Investigación Nefrológica, Barcelona, Spain
    Search for more papers by this author

Abstract

Patients with cirrhosis are frequently submitted to radiological procedures that require the administration of contrast media. Contrast media is a well-known cause of renal failure, particularly in the presence of some predisposing conditions. However, it is not known whether cirrhosis constitutes a risk factor for contrast media–induced renal failure. The aim of this study was to assess the possible nephrotoxicity of contrast media in patients with cirrhosis. In a first protocol, renal function was evaluated with sensitive methods (glomerular filtration rate using iothalamate I 125 clearance and renal plasma flow using iodohippurate I 131 clearance) before and 48 hours after the administration of contrast media in 31 patients with cirrhosis (20 with ascites, 5 with renal failure). Solute-free water clearance, urine sodium, prostaglandins, and markers of tubular damage were also measured. The administration of contrast media was not associated with significant changes in renal function tests, neither in the whole group of patients nor in patients with ascites or renal failure. Urinary prostaglandin E2 and N-acetyl-β-D-glucosaminidase increased significantly, but sodium and solute-free water excretion remained unchanged. In a second protocol, a different series of 60 patients with cirrhosis and renal failure were examined prospectively. No patient had renal failure due to contrast media. Only in 1 patient with septic shock was contrast media a possible contributing factor. In conclusion, the administration of contrast media is not associated with adverse effects on renal function in patients with cirrhosis. Cirrhosis does not appear to be a risk factor for the development of contrast media–induced nephrotoxicity. (HEPATOLOGY 2004;40:646–651.)

Renal failure is a common and relevant complication in patients with cirrhosis.1, 2 Prognosis of patients with cirrhosis who develop renal failure is very poor owing to the combined detrimental effect of liver and renal failure.3, 4 In some patients, renal failure is due to etiological factors that cause an impairment in effective blood volume, such as bacterial infections, gastrointestinal bleeding, overdiuresis, or large-volume paracentesis without plasma volume expansion.1, 3–7 In other cases, renal failure is related to the administration of drugs that cause a direct effect on the kidneys, such as nonsteroidal anti-inflammatory drugs or aminoglycosides.8, 9 Evidence has been accumulated indicating that the existence of cirrhosis predisposes to nephrotoxicity due to these latter compounds.

Nephrotoxicity induced by the administration of contrast media is a frequent cause of renal failure in hospitalized patients.10–12 Several factors that predispose to renal failure after contrast media administration have been identified, including advanced age, preexisting renal failure, and diabetic nephropathy.10, 11, 13 Cirrhosis has also been considered a potential predisposing factor for contrast media-induced nephrotoxicity,3, 14 but this possibility has not been specifically investigated. In fact, to our knowledge, no studies have been reported assessing the effect of contrast media administration on renal function in patients with cirrhosis. It is our clinical experience in a large liver unit of a tertiary referral center that renal failure after administration of contrast media to patients with cirrhosis is extremely uncommon. Nevertheless, since this is a clinically important topic because of the relevance of renal failure in cirrhosis and the frequency with which patients with cirrhosis undergo diagnostic and/or therapeutic procedures that require the administration of contrast media, we decided to perform a prospective study to investigate this issue.

Abbreviations:

NAG, N-acetyl-β-D-glucosaminidase; GFR, glomerular filtration rate; RPF, renal plasma flow.

Patients and Methods

Protocol 1.

The aim of this protocol was to investigate the effects of contrast media on renal function in a prospective series of hospitalized patients with cirrhosis undergoing a radiological procedure. The study was approved by the Investigational Review Board of the Hospital Clínic (Barcelona, Spain), and patients gave written informed consent to participate. Criteria for exclusion were as follows: administration of contrast media or potentially nephrotoxic agents within 3 weeks before the study, parenchymal renal disease, cardiac or respiratory failure, hepatic encephalopathy, and gastrointestinal bleeding or bacterial infection within 2 weeks prior to the study. Patients with hepatorenal syndrome and serum creatinine greater than 2 mg/dL were also excluded because these patients were treated with vasopressin analogues to improve renal function before any diagnostic procedure was performed.

All patients were studied while on a low-sodium diet. To determine the effects of contrast media on renal function without associated confounding factors, diuretic therapy was withheld for at least 5 days beforeinclusion in the study. On the day before the radiological procedure, urine was collected for 24 hours to measure the urinary concentration of sodium and potassium, markers of tubular damage (such as α1-microglobulin, β2-microglobulin, and N-acetyl-β-D-glucosaminidase [NAG]), and prostaglandins (prostaglandin E2 and 6-keto-prostaglandin F1α). On the next day, prior to the radiological procedure, blood samples were taken for standard liver and renal tests and to measure plasma renin activity, aldosterone, and norepinephrine; glomerular filtration rate (GFR) and renal plasma flow (RPF) were determined by iothalamate I 125 and iodohippurate I 131 clearances, respectively, as described in detail in Analytical Methods. These measurements were repeated 48 hours later. The radiological procedure was a diagnostic abdominal arteriography in 25 patients, a body CT scan in 5, and insertion of a transjugular intrahepatic portosystemic shunt in the remaining patient The contrast media used was iopaminol in 29 patients and ioexol in 2. The mean dose of contrast media given was 202 ± 12 mL (range, 110–400 mL). Patients were not treated with intravenous fluids or any other therapy before the radiological procedure to prevent the development of nephrotoxicity due to contrast media administration.

Protocol 2.

The objective of this protocol was to prospectively assess the possible role of contrast media as an etiological factor of renal failure in patients with cirrhosis. To this aim, all patients with cirrhosis and renal failure seen in our liver unit over a period of 6 months were evaluated prospectively. Renal failure was defined as 2 consecutive determinations of serum creatinine equal to or greater than 1.5 mg/dL within a 24-hour period. All possible etiological factors that could be associated with the development of renal failure, including the administration of contrast media within the previous 3 weeks, were recorded. The diagnosis of contrast media-induced renal failure was considered definite in patients who received contrast media and had no other etiological factors for the development of renal failure. The diagnosis of contrast media–induced renal failure was considered possible in patients who received contrast media but had other etiological factors of renal failure.

Analytical Methods.

GFR and RPF were measured as follows: a priming dose of iothalamate I 125 (0.37 MBq) or iodohippurate I 131 (0.55 MBq) was given intravenously, followed by a constant infusion of both substances (0.0037 MBq/min of iothalamate I 125 and 0.014 MBq/min of iodohippurate sodium I 131) in saline throughout the study. Following an equilibration period of 1 hour, the urine was collected in 3 periods of 30 minutes; in the middle of each, a blood sample was taken to measure iothalamate I 125, iodohippurate I 131, and osmolality. The urine volume of each period was measured, and aliquots were kept to measure iothalamate I 125, iodohippurate I 131, and osmolality. During the equilibration period, an intravenous water load of 20 mL per kilogram of body weight of 5% dextrose was given. This water load was kept constant throughout the study by infusing a volume of 5% dextrose in each period equal to the urine volume of the previous period. GFR, RPF, and osmolar clearance were calculated using standard clearance formulas, and solute-free water clearance was calculated as urine volume (mL/min) minus osmolar clearance (mL/min). The average value of each of these 4 parameters in the 3 periods was calculated and used as a final value. Plasma renin activity, aldosterone, and norepinephrine were determined using previously described methods.15

Statistical Analysis.

Comparisons of the variables in the same group were performed using the paired Student's t test and Wilcoxon test for continuous data and the chi-square test and Fisher test for categorical data. Comparison between groups was performed using the Mann-Whitney test. Results are given as mean ± SE. Statistical analyses of the data were performed using the SPSS 10 statistical software (SPSS Inc. and Microsoft Corp., Chicago, IL).

Results

Protocol 1.

A total of 31 patients with cirrhosis were enrolled in this protocol. Table 1 shows the demographic data and liver and renal function tests in this group of patients. Eleven patients neither had ascites at the time of the study nor had a history of ascites or edema. The remaining 20 patients had ascites as assessed by physical examination and confirmed by diagnostic paracentesis. Four patients had hepatorenal syndrome type 2, with serum creatinine lower than 2 mg/dL. Six patients had diabetes mellitus, but none of them had diabetic nephropathy.

Table 1. Demographic Data and Liver and Renal Function Tests in Patients Included in Protocol 1
Variable DataMean ± SE (n = 31)Range
Etiology (alcohol/no alcohol)7/24
Age (y)59 ± 237–75
Sex (M/F)20/11
Bilirubin (mg/dL)1.8 ± 0.20.4–6.2
Albumin (g/L)30 ± 119–45
Prothrombin time (%)62 ± 445–100
Child-Pugh score (A/B/C)9/16/6
Blood urea nitrogen (mg/dL)18 ± 28–60
Serum creatinine (mg/dL)0.9 ± 0.020.3–1.9
Serum sodium (mEq/L)133 ± 1116–142
Urine sodium (mEq/L)67 ± 170.8–140

The effects of the administration of contrast media on renal function parameters, vasoactive systems, urinary excretion of markers of tubular damage, prostaglandins, and vasoactive systems is shown in Table 2. GFR did not change significantly; RPF increased slightly. Figure 1 shows the individual values of GFR and RPF before and after contrast media administration. No relationship was found between changes in GFR and the amount of contrast media given to each patient. No significant differences were observed between serum creatinine and blood urea nitrogen obtained at baseline and 2 days after the administration of contrast media (0.9 ± 0.06 vs. 0.8 ± 0.03 mg/dL and 18 ± 2 vs. 17 ± 2 mg/dL, respectively). There were no significant differences between urinary sodium excretion and solute-free water clearance obtained before and after the administration of contrast media. The urinary excretion of prostaglandin E2 increased significantly, while no significant changes were observed in that of 6-keto-prostaglandin F1α. There was no relationship between changes in urinary prostaglandin E2 excretion and those in RPF (r = 0.2; not significant). Among markers of tubular damage, there was a significant increase in the urinary excretion of NAG, while those of α1-microglobulin and β2-microglobulin did not change significantly. Finally, no significant changes were found in the parameters estimating the activity of the renin-aldosterone system and sympathetic nervous system. Only 2 patients showed a decline in GFR after administration of contrast media of 20% or greater compared with baseline values (from 98 and 129 mL/min to 72 and 103 mL/min, respectively). RPF also decreased in these 2 patients (from 635 and 576 mL/min to 511 and 388 mL/min, respectively). No changes in serum creatinine concentration were observed in these cases. The urinary excretion of prostaglandin E2 increased in these 2 patients from 2,506 and 911 pg/min at baseline to 3,516 and 1,353 pg/min after administration of contrast media; that of 6-keto-prostaglandin F1α increased from 240 and 479 pg/min to 726 and 566 pg/min, respectively.

Table 2. Renal Function Tests, Vasoactive Factors, Urinary Excretion of Markers of Tubular Damage, and Prostaglandins at Baseline and at 48 Hours After Administration of Contrast Media
VariableBaseline48 Hours
  • *

    P < .05 versus baseline.

Glomerular filtration rate (mL/min)92 ± 7101 ± 8
Renal plasma flow (mL/min)474 ± 34561 ± 41*
Serum sodium (mEq/L)133 ± 1133 ± 1
Urine sodium (mEq/L)67 ± 1743 ± 8
Solute-free water clearance (mL/min)3.7 ± 0.82.9 ± 0.6
Plasma renin activity (ng/mL/h)2.5 ± 13.4 ± 2
Aldosterone (ng/dL)60 ± 2261 ± 20
Norepinephrine (pg/mL)334 ± 42362 ± 34
Urine prostaglandin E2 (pg/min)1730 ± 2083135 ± 595*
Urine 6-keto-prostaglandin F1α (pg/min)728 ± 246922 ± 213
Urine β2-microglobulin (μg/g creatinine)172 ± 30223 ± 28
Urine α1-microglobulin (mg/g creatinine)8 ± 211 ± 2
Urine N-acetyl-β-D-glucosaminidase;(U/gl creatinine)10 ± 220 ± 3*
Figure 1.

Individual values of glomerular filtration rate (GFR) and renal plasma flow (RPF) at baseline conditions and at 48 hours after the administration of contrast media.

To assess whether the presence of ascites or renal impairment (GFR lower than 50 mL/min) could predispose to nephrotoxicity induced by the administration of contrast media, patients with these conditions were analyzed separately. As shown in Table 3, neither the presence of ascites nor the existence of impaired renal function was associated with significant changes in renal function after the administration of contrast media.

Table 3. Renal Function Tests, Vasoactive Factors, Urinary Excretion of Markers of Tubular Damage, and Prostaglandins in Patients With Ascites and Impaired Renal Function
VariablePatients With Ascites (n = 20)Patients With Impaired Renal Function (n = 5)
Baseline48 HoursBaseline48 Hours
  • *

    P < .05 with respect to the baseline of patients with ascites.

  • P < .05 with respect to the baseline of the same group of patients.

Glomerular filtration rate (mL/min)75 ± 888 ± 940 ± 4*49 ± 9
Renal plasma flow (mL/min)422 ± 51475 ± 57266 ± 42*324 ± 50
Serum sodium (mEq/L)131 ± 1.5132 ± 2127 ± 3127 ± 4
Urine sodium (mEq/L)20 ± 109 ± 314 ± 88 ± 6
Solute-free water clearance (mL/min)2.8 ± 12.5 ± 11.2 ± 0.72.7 ± 1.6
Plasma renin activity (ng/mL/h)4 ± 15 ± 18 ± 414 ± 9
Aldosterone (ng/dL)87 ± 3287 ± 27197 ± 114182 ± 88
Norepinephrine (pg/mL)385 ± 56395 ± 38332 ± 135376 ± 96
Prostaglandin E2 (pg/min)1540 ± 2413108 ± 8881576 ± 5293183 ± 1819
Urine 6-keto prostaglandin F1α (pg/min)996 ± 416947 ± 318274 ± 40947 ± 318
Urine β2-microglobulin (μg/g creatinine)208 ± 45210 ± 4981 ± 31*157 ± 86
Urine α1-microglobulin (mg/g creatinine)6 ± 18 ± 15 ± 0.59 ± 2
Urine N-acetyl-β-D-glucosaminidase (U/g creatinine)12 ± 224 ± 417 ± 637 ± 10

Protocol 2.

During a 6-month period, 60 patients with cirrhosis and renal failure were identified. Mean serum creatinine at diagnosis of renal failure was 2.6 ± 0.2 mg/dL (range, 1.6-8.2 mg/dL) and reached a maximum level of 3.5 ± 0.2mg/dL (range, 1.9-9.9 mg/dL). None of the 60 patients had received contrast media within a period of 3 weeks before the diagnosis of renal failure. Causes of renal failure were bacterial infection without septic shock in 15 patients; volume depletion in 15 (gastrointestinal bleeding without shock in 8 patients, severe diarrhea in 4, and overdiuresis due to excessive diuretic therapy in 3); parenchymal renal failure in 9; shock in 9 (septic in 8 and hypovolemic in 1); hepatorenal syndrome in 7; nonsteroidal anti-inflammatory drugs in 2; and multifactorial in 3. Three patients received contrast media after the diagnosis of renal failure during radiological procedures performed for the diagnosis workup (abdominal CT scan in patients and head CT in 1). In 1 patient with septic shock who had an abdominal CT scan performed at the time of admission, serum creatinine was 1.6 mg/dL at admission and 2.2 mg/dL 24 hours later; it subsequently decreased to 1.1 mg/dL at discharge from hospital. In the other 2 patients with septic shock and gastrointestinal bleeding, respectively, serum creatinine did not change after the administration of contrast media (1.9 and 3.2 mg/dL at diagnosis vs. peak values of 2.0 and 3.2 mg/dL, respectively). Therefore, in the whole population of patients with renal failure, no patient had definite contrast media-induced renal failure, and only 1 patient (1.7%) had possible renal impairment due to a combination of both septic shock and contrast media administration.

Discussion

Although the occurrence of renal failure after administration of contrast media has decreased markedly in recent years due to the use of less toxic contrast agents, it still represents an important problem in clinical practice.10–14 In the current study, the possibility that contrast media could cause nephrotoxicity in patients with cirrhosis was explored prospectively using 2 different approaches. In the first (Protocol 1), renal function was assessed with the use of sensitive methods before and after the administration of contrast media in a group of patients with cirrhosis undergoing a radiological procedure. In the second approach (Protocol 2), potential causes of renal failure, including the administration of contrast media, were sought in a group of patients with cirrhosis and renal failure. The purpose of the first approach was to determine whether contrast media could cause impairment of renal function in patients with cirrhosis that could be overlooked by the standard renal function tests used to estimate renal function in normal clinical practice (i.e., serum creatinine). The objective of the second approach was to assess whether the possible nephrotoxicity associated with the use of contrast media could be severe enough to cause clinically significant renal failure. It should be emphasized that the current study was not designed to assess the incidence of contrast media-induced nephrotoxicity in patients with cirrhosis. This would have required the inclusion of a larger patient population and a control group of patients not receiving contrast media.

The findings of the current study indicate that administration of contrast media does not impair renal function in patients with cirrhosis. GFR was measured with a very sensitive method and did not show significant changes after the administration of contrast media. It is important to state that with a study design similar to that used in the current study, we and others have shown a remarkable reduction in GFR after the administration of several drugs, including nonsteroidal anti-inflammatories, vasodilators, and α-adrenergic agonists.16–21 Therefore, it is highly unlikely that a significant negative effect of contrast media on GFR was missed, should one exist. Besides GFR, sodium excretion or solute-free water excretion did not change significantly after the administration of contrast media. The significant improvement in RPF observed in our patients was intriguing, and we have no explanation for this finding. To assess whether specific situations such as ascites or preexisting renal impairment could predispose to the development of contrast media-induced nephrotoxicity, we classified our patients into different subsets. Neither patients with ascites nor those with impaired GFR at baseline showed significant changes in renal function parameters after the administration of contrast media.

The patient population included in Protocol 1 of the current study was representative of the general population of patients with cirrhosis, with respect to both liver and renal function abnormalities. Half of the patients included had moderate liver failure (Child-Turcotte-Pugh class B), and the remaining patients had either good liver function (Child-Turcotte-Pugh class A, 30%) or severe liver failure (Child-Turcotte-Pugh class C, 20%). Almost two-thirds of patients had ascites; among those, 12 patients had marked sodium retention, dilutional hyponatremia, or renal failure. It could be argued that patients with severe hepatorenal syndrome (serum creatinine > 2 mg/dL) were not included. However, these patients were intentionally excluded because they have an extremely poor outcome and were treated with vasopressin analogues to improve their renal function.22, 23 Because of the low renal production of prostaglandins in hepatorenal syndrome,1, 2 patients with this condition may theoretically have a high risk of development of renal failure secondary to different renal insults, including administration of contrast media. This possibility was not tested in the current study because patients with a serum creatinine higher than 2 mg/dL were not included. On the other hand, patients in the current study were studied without the influence of diuretic treatment. Therefore, it is not known whether the concomitant use of diuretics could increase the risk of nephrotoxicity induced by contrast media.

The results of Protocol 2 indicate that, of 60 cases with cirrhosis and clinically significant renal failure (as defined by a serum creatinine equal to or greater than 1.5 mg/dL), none could be attributed definitely to the administration of contrast media. Only in 1 patient could renal failure possibly be related to the administration of contrast media as it occurred in combination with a septic shock, a well-known etiological factor of renal failure in cirrhosis. These findings are in keeping with those of Protocol 1, showing lack of significant changes in GFR after the administration of contrast media. Therefore, although we cannot completely exclude the possibility that contrast media administration could cause renal impairment in some patients with cirrhosis, it would seem that the frequency is quite low.

Besides studying the effects of contrast media on renal perfusion and GFR, we also investigated the effects of contrast media on tubular function—specifically, sodium and water excretion—as well as on some important renal vasoactive mediators (i.e. prostaglandins) and markers of tubular damage.24–26 The urinary excretion of prostaglandin E2 increased significantly after the administration of the contrast media.10, 27, 28) This could be the consequence of either a direct effect of contrast media on tubular cells, which are a major source of prostaglandin E2 synthesis in the kidney, or could represent a defensive mechanism of the kidney to maintain renal hemodynamics and/or tubular function against the possible increase of some intrarenal vasoconstrictors, such us endothelin.28 No changes were found in the urinary excretion of 6-keto-prostaglandin F1α, a stable metabolite of prostaglandin I2, which is thought to play an important role in the regulation of GFR in cirrhosis.29–31 Finally, the urinary excretion of NAG, a marker of tubular dysfunction, increased significantly, indicating that the administration of contrast media was associated with some degree of tubular injury.32 Elevation of NAG in urine provides a marker of tubular damage due to its lysosomal location. It has been shown that NAG is a very sensitive indicator of tubular damage, and false positive values are rare.32 Nevertheless, such tubular injury was not associated with a clinically significant tubular dysfunction because no significant effects were observed in sodium and solute-free water excretion. Therefore, the increase in NAG observed in our study might indicate an incipient tubular injury without measurable impairment of renal function.

In conclusion, the results of our study indicate that the administration of contrast media is not associated with adverse effects on renal function, thus suggesting that cirrhosis per se does not represent a risk factor for contrast media-induced nephrotoxicity.

Acknowledgements

The authors thank Raquel Cela for her technical assistance and the nursing staff of the Liver Unit for their help.

Ancillary