You are on page 1of 9

Current Drug Safety, 2008, 3, 67-75 67

Gadolinium-Contrast Toxicity in Patients with Kidney Disease: Nephro-


toxicity and Nephrogenic Systemic Fibrosis

Mark A. Perazella*

Section of Nephrology, Department of Medicine, Yale University School of Medicine, New Haven, Connecticut 06520-
8029, USA

Abstract: Gadolinium is widely employed as a contrast agent for magnetic resonance imaging (MRI) and has generally
been considered to be safe. As with iodinated radiocontrast, concern for contrast-induced nephropathy existed with gado-
linium-contrast as it possessed many similar qualities (hyperosmolar, renal excretion via glomerular filtration). Early stud-
ies in low risk patients suggested a benign renal profile, however, recent studies raise the possibility of nephrotoxicity. In
addition, reports of a previously rare condition entitled nephrogenic systemic fibrosis (NSF) have recently emerged in pa-
tients with advanced kidney disease and have been linked to exposure to gadolinium-contrast. Nephrogenic systemic fi-
brosis is a debilitating disorder in which progressive and severe fibrosis of the skin and other systemic organs that leads to
significant disability and is associated with increased mortality. Initially reported most commonly in end stage renal dis-
ease (ESRD) patients receiving dialysis, it is also described in patients with severe acute kidney injury (AKI) and ad-
vanced chronic kidney disease (stages 4 and 5) not requiring dialysis. In addition to underlying kidney disease, the risk of
developing NSF is increased with larger doses of gadolinium (or multiple exposures), exposure to specific gadolinium
chelates (non-ionic, linear), underlying pro-inflammatory states (in particular vascular endothelial dysfunction), and per-
haps some currently unrecognized cofactors. No clearly effective therapies exist for NSF, although recovery from AKI
and establishment of normal kidney function with renal transplantation appear to reverse or stabilize the disease in some
cases. Avoidance of gadolinium exposure appears to be the best approach for patients who maintain risk factors. When
gadolinium exposure occurs, aggressive hemodialysis following exposure may be useful as gadolinium is efficiently re-
moved by this extracorporeal technique. Peritoneal dialysis clearance of gadolinium is poor, but aggressive peritoneal di-
alysis prescriptions have not been studied for gadolinium removal.
Keywords: Gadolinium, nephrotoxicity, nephrogenic systemic fibrosis, chronic kidney disease, end stage kidney disease, mag-
netic resonance imaging, radiocontrast-induced nephropathy.

INTRODUCTION the recognition that Gd-contrast exposure in patients with


significant kidney disease may trigger the development of
Magnetic resonance imaging (MRI) is a commonly used
nephrogenic systemic fibrosis (NSF), a debilitating and often
imaging technique for numerous organ systems including the
devastating systemic fibrosing condition that is most clini-
central nervous system, hepatic structures, and the vascula-
cally prominent in the skin [1-12]. This review will focus on
ture. MR images are significantly enhanced by use of gado-
these clinically important complications of Gd-contrast.
linium-based contrast agents. They often provide images that
are superior to those obtained with computed tomography GADOLINIUM CONTRAST: GENERAL PROPER-
(CT) scan and have the advantage of avoiding iodinated ra-
TIES AND PHARMACOKINETICS
diocontrast agents which have more overall toxicity (allergic
and non-allergic reactions). Thus, MRI has been considered a Gadolinium (Gd) is a metal of the lanthanide series
relatively safe alternative to CT scan in situations where a (atomic number 64 on element chart) which has paramag-
contrast agent is thought required for enhanced image at- netic properties that disturb relaxation of water protons and
tainment. However, two complications of gadolinium- by shortening relaxation times, increases signal intensity.
contrast (Gd-contrast) have come to light in recent times. This quality makes it extremely useful as an intravenous
and/or intra-arterial contrast agent for MRI/MRA to enhance
First, concern for contrast-induced nephropathy from Gd-
images of various body organs and tissues. As it is a metal, it
contrast has been raised as several studies over the past dec-
must be in an ionic form to be soluble in water and allow it
ade demonstrate nephrotoxicity following Gd-contrast ad-
to be injected as a contrast agent that distributes throughout
ministration in patients with underlying kidney disease and
other co-morbidities. Gd-contrast agents are hyperosmolar the body. However, gadolinium in this free ionic form (Gd3+)
is highly toxic to humans (and animals). It precipitates in
and completely eliminated from the body via renal excretion
several tissues including the liver, lymph nodes, and bones.
(glomerular filtration). Second, and even more concerning is
Gd3+ obstructs passage of calcium through ion channels of
muscle cells and nerve tissue cells thereby reducing neuro-
*Address correspondence to this author at the Section of Nephrology, De- muscular transmission, and interferes with intracellular en-
partment of Medicine, Yale University School of Medicine, New Haven, zymes and cell membranes. In order to avoid these toxic ef-
Connecticut 06520-8029, USA; Tel: 203-785-4184; Fax: 203-785-7068; fects, Gd3+ must be sequestered by non-toxic substances
E-mail: mark.perazella@yale.edu [13]. This is achieved by binding Gd3+ to another agent,

1574-8863/08 $55.00+.00 © 2008 Bentham Science Publishers Ltd.


68 Current Drug Safety, 2008, Vol. 3, No. 1 Mark A. Perazella

which is known generically as a “chelate”. Chelates are large healthy subjects. However, the mean terminal T1/2 is longer
organic molecules that form a stable complex with Gd3+, do in moderate (5.6 hours) and severe (9.2 hours) kidney dis-
not readily dissociate in vivo, and make the ion biochemi- ease than in healthy subjects (1.6 hours). Mean blood and
cally inert [13,14]. The properties of the various “Gd- renal clearance are both much lower in moderate (56 ml/min;
chelates” allow classification of Gd-contrast into four main 47 ml/min) and severe (31 ml/min; 22 ml/min) kidney dis-
categories based on their biochemical structure (linear versus ease than in normal subjects (183 ml/min; 118 ml/min).
macrocyclic) and their charge (ionic versus non-ionic). Mac- Comparable pharmacokinetics of Gd-contrasts in patients
rocyclic chelates bind Gd3+ more tightly than linear chelates, with underlying kidney disease have been reported in other
tend to be more stable both in vitro and in vivo, and possess studies [17,18]. Thus, it is predictable that tissue gadolinium
lower dissociation rates [15]. Simply stated, macrocyclic exposure is prolonged in the setting of decreased kidney
chelates stick more tightly to Gd3+ than do linear chelates; function. The relatively small molecular weight (500 Da),
this has implications for possible toxicity. This aspect of small Vd (0.28 L/kg), and negligible protein binding charac-
chelate characteristics is important to prevent liberation of teristics of Gd-contrasts make them ideal for removal with an
free Gd3+ from its chelate, a process known as transmetala- extracorporeal therapy such as hemodialysis. In one study,
tion. This phenomenon entails release of free Gd3+ from its the T1/2 of Gd-contrasts of non-dialyzed CKD stage 5 (CrCl
chelate ligand, which then binds with another endogenous = 2-10 ml/min) patients was quite prolonged at 34.3 hours
metal such as zinc or copper, allowing free Gd3+ to bind an but decreased significantly to 2.6 hours following hemo-
endogenous ligand such as phosphorus. The commonly em- dialysis [19]. In another study the average Gd-contrast elimi-
ployed Gd-contrasts approved by the Food and Administra- nation from serum using hemodialysis was 73.8% with one
tion Drug (FDA) and their characteristics are noted in Table treatment, 92.4% with 2 treatments and 98.9% after the 3rd
1. The recommended dose of a Gd-contrast agent for a non- treatment [20,21]. Peritoneal dialysis on the other hand was
vascular MR study is 0.1 mmol/kg. When imaging of vascu- an ineffective method of Gd-contrast removal (T1/2 of 52.7
lar structures (MRA) is required, higher doses of Gd-contrast hours) [19]. This study suffers by the use of what is now
(0.3-0.4 mmol/kg) are often utilized. Until recently, Gd- considered an inadequate peritoneal dialysis prescription (2.0
contrast agents were FDA approved only for MRI studies at liter volumes x 4 exchanges over a 24 hour period).
the 0.1 mmol/kg dose. Gadoteridol (macrocyclic chelate) is
the only MR chelate approved at the higher 0.3 mmol/kg GADOLINIUM-CONTRAST AND NEPHROTOXIC-
dose. Most MRA studies are often performed using Gd- ITY
contrasts at dose levels that are considered “off-label”.
Iodinated radiocontrast-media induced nephrotoxicity is
Following intravenous injection, Gd-contrasts are rapidly well described and common. However, the issue of Gd-
distributed into the extracellular space, quickly equilibrating contrast induced nephrotoxicity is somewhat controversial
between the plasma and interstitial compartments. The ma- and generally overlooked. Initial pharmaceutical trials exam-
jority are restricted to the extracellular space and have lim- ined the potential adverse renal effects of Gd-contrast. Since
ited protein binding. As a result of their exclusion from the Gd-contrasts have characteristics very similar to those of
intracellular compartment, Gd-contrasts have small volumes iodinated radiocontrast, in particular hyperosmolality and
of distribution (Vd), approximately 0.26-0.28 L/kg body renal clearance entirely dependent upon glomerular filtration,
weight. They do not undergo biotransformation and are nephrotoxicity was an obvious concern of both manufactur-
eliminated unchanged by the kidneys via glomerular filtra- ers and physicians. Animal studies demonstrate nephrotoxic-
tion without any contribution from tubular secretion. Renal ity with gadolinium contrast when given in high doses (0.6-
clearance of Gd-contrasts ranges from 1.1 to 1.6 ml/min/kg 3.0 mmol/kg) to rats with normal kidney function [22].
in individuals with normal renal function (approximating the Histopathology demonstrates vacuolization and necrosis of
creatinine clearance). The Gd-contrasts maintain a mean proximal tubular cells. The lower observed risk of Gd-based
terminal half-life (T1/2) of approximately 1.6 hours. Over contrast as compared with iodinated radiocontrast may relate
95% of an injected dose is eliminated within 24 hours with to mechanism of nephrotoxicity. Iodinated radiocontrast
less than 3% being eliminated in the feces [13,14,16]. The causes renal injury primarily through induction of vasocon-
Vd for intravenous Gd-contrasts (gadobenate dimeglumine striction, whereas Gd-based contrast has not been shown to
studied) is similar for patients with moderate (creatinine promote renal ischemia. Early studies in normal healthy sub-
clearance [CrCl], 31-60 ml/min; n=15) and severe (CrCl, 15- jects as well as small groups of patients with mild to moder-
30 ml/min; n=17) kidney disease when compared with ate levels of underlying kidney disease suggested a reasona-

Table 1. FDA Approved Gadolinium Contrast Agents

Osmolality Molecular Stability Con- Excess Chelate Gado Market


Gado Formulation Charge
(mosm/l) Structure stant (mg/ml) Share* (2006)

Gadodiamide (Omniscan) 900 Non-ionic Linear 1014.9 12 34%


18.1
Gadopentetate (Magnevist) 1960 Ionic Linear 10 0.4 47%
15
Gadoversetamide (OptiMARK) 1110 Non-ionic Linear 10 28.4 8%
18.4
Gadobenate (MultiHance) 1970 Ionic Linear 10 0.1 6%
Gadoteridol (Prohance) 630 Non-ionic Cyclic 1017.1 0.23 5%
Abbreviations: Gado, gadolinium. * Contrast Media Industry Guide data.
Gadolinium-Contrast and Kidney Disease Current Drug Safety, 2008, Vol. 3, No. 1 69

bly favorable renal safety profile [23, 24]. Even up until re- ics and renal safety [26]. Patients received 0.2 mmol/kg of
cently, the Gd-contrasts were considered relatively safe for intravenous gadobenate dimeglumine, which has an osmolal-
use in patients with kidney disease, even with the high doses ity of 1970 mosm/L. 24-hour urine CrCl before and at days
required for renovascular imaging, where most other imaging 1, 2, 3, 5, and 7 following gadobenate dimeglumine exposure
techniques were inadequate (renal scans and ultrasonogra- were measured. No patients received any form of contrast
phy) or were too risky (CT with iodinated contrast). The im- prophylaxis prior to or during the study. There was no sig-
portance of this issue deserves emphasis. Numerous patients nificant change in CrCl at any time point in the study, sup-
considered at high risk for radiocontrast-induced nephropa- porting the absence of any clinically important nephrotoxic-
thy (RCIN) from an iodinated radiocontrast study are ex- ity.
posed to a Gd-contrasts as a "renal safe" alternative. Critical
Patients with kidney disease (defined as a serum creatin-
review of the literature on this subject does not, unfortu- ine concentration greater than 1.5 mg/dl) were studied using
nately, allow a definitive answer to the question of Gd-
gadopentetate dimeglumine (0.4 mmol/kg dose) as an alter-
contrast nephrotoxicity. Some but not all of the studies pub-
native imaging agent in patients who were allergic to iodi-
lished on this subject over the past decade will be reviewed;
nated radiocontrast [26]. 31 patients underwent 34 digital
recognizing that other negative studies were published prior
subtraction angiographies (DSAs) with this hyperosmolar
to and during this time period.
agent (1960 mosm/L). Only one out of 34 studies was com-
plicated by contrast-induced nephropathy, which was defined
Gadolinium-Contrast Studies Supporting Renal Safety
as an increase in serum creatinine concentration greater than
Several studies are available in this time period suggest- 0.5 mg/dl.
ing that Gd-contrast agents lack any significant nephrotoxic- In 2000, a study in patients with CKD (serum creatinine
ity (Table 2). A retrospective study published in 1996 exam- greater than 1.5 mg/dl; mean = 2.2 mg/dl; range = 1.6-3.6
ined a cohort of 64 patients with mild CKD as defined by mg/dl) and peripheral vascular disease was undertaken to
baseline serum creatinine concentration of 2.0 ± 1.4 mg/dl compare nephrotoxicity of nonionic radiocontrast with CO2
[25]. All patients received both Gd-contrast and iodinated supplemented with either Gd-contrast (up to 0.4 mmol/kg of
contrast at separate times, thus serving as their own control. gadodiamide) or nonionic radiocontrast [27]. 40 patients un-
The rate of contrast-induced nephrotoxicity, as defined by a derwent 42 lower extremity angiograms using one of the
rise in serum creatinine concentration of 0.5 mg/dl following following contrast protocols: radiocontrast = 15, gadodia-
each exposure, was compared between the 2 different expo- mide = 20, CO2 = 7. All received 300-5000 ml of normal
sures. The dose of Gd-contrasts administered during the saline prior to contrast exposure as prophylaxis. Contrast-
study ranged from 0.2 to 0.4 mmol/kg. No patient receiving induced nephropathy was defined as an increase in serum
Gd-contrasts developed nephrotoxicity as compared with creatinine concentration greater than 0.5 mg/dl at 48 hours
17% of patients receiving iodinated radiocontrast. post procedure. Contrast-induced nephropathy developed in
A prospective study of 32 patients with moderate (CrCl, 6 out of 15 (40%) radiocontrast studies but only 1 out of 20
31-60 ml/min) and severe (CrCl, 10-30 ml/min) kidney dis- gadodiamide exposures (5%).
ease was undertaken to examine Gd-contrast pharmacokinet-

Table 2. Studies Supporting Renal Safety of Gadolinium-Contrast Agents

Author (Year) Study Contrast Agent Dose (mmol/kg) Renal Function ([Cr] in mg/dl) Result

Prince (1996) Retrospective, N=64 Gadopentetate 0.2 - 0.4 [Cr] > 1.5, CIN:
[Cr] 2d pre and 2d post, Gadodiamide Mean [Cr] = 2.0±1.4 RC- 11/64 (17%)
CIN  0.5 mg/dL Gadoteridol Gado- 0/64 (0%)
Swan (1999) Prospective, double blind random, 32 pts (2: Gabobenate 0.2 CrCl 10-30, No CIN
1), dimeglumine CrCl 31-60,
CIN > 0.5 mg/dL 24 hr urine
Hammer (1999) N = 31, 34 DSAs, Gadopentetate 0.4 [Cr] > 1.5 CIN: 1/34 (3%)
Mean age 53.1
CIN > 0.5 mg/dL
Spinosa (2000) N = 40, LE angiograms Gadodiamide up to 0.4 [Cr] > 1.5, IC- 6/15 (40%)
42 procedures Mean [Cr] = 2.2, Gado- 1/20 (5%)
RC- 15, Gado- 20 Range [Cr] = 1.6-3.6
CIN  0.5 mg/dL at 48 hr
Spinosa (2001) Consecutive patients treated with Gado + Gadodiamide < 0.3 [Cr] > 1.5, CIN: 3/95 (3%)
CO2,
CIN > 0.5 mg/dL at 48 hr
Sancak (2001) N = 16, IV Gado for upper extremity or SVC Gadodiamide 0.3 Mean [Cr] = 1.5, Largest increase in
Range [Cr] = 1.2-1.8 [Cr] = 0.2 mg/dL
Rieger (2002) Prospective, N = 29, 32 procedures (IA & IV) Gadopentetate 0.34 ± 0.06 [Cr] > 1.5 1/29 (atheroemboli)
CIN > 0.5 mg/dL at 72 hr Mean [Cr] 3.6±1.4
Abbreviations: Gado, gadolinium; RC, iodinated radiocontrast; CrCl, creatinine clearance; [Cr], serum creatinine concentration; CIN, contrast-induced nephropathy; ref, reference;
DSA, digital subtraction angiogram; LE, lower extremity; SVC, superior vena cava; IV, intravenous; IA, intra-arterial.
70 Current Drug Safety, 2008, Vol. 3, No. 1 Mark A. Perazella

The same authors published another study again support- contrast prophylaxis was provided. Patients who received
ing Gd-contrast safety in patients with CKD and ischemic intravenous and intra-arterial gadopentetate dimeglumine
nephropathy [28]. 146 consecutive patients with a serum had mean baseline serum creatinine concentrations of 2.1
creatinine concentration greater than 1.5 mg/dl underwent mg/dl (estimated CrCl = 61 ml/min) and 2.6 mg/dl (esti-
renal angiography using a combination of CO2 and gadodia- mated CrCl = 40 ml/min), respectively. Contrast-induced
mide. Contrast-induced nephropathy was defined as an in- nephropathy, defined as an increase in serum creatinine con-
crease in serum creatinine concentration greater than 1.5 centration > 1.0 mg/dl within 48 hours developed in 3.5%
mg/dl at 48 hours; 95 patients had data available for study. (7/195) of the entire population: 1.9% (3/153) with intrave-
Three patients (3.2%) developed nephrotoxicity, less than nous and 9.5% (4/42) with intra-arterial administration. In
what is commonly seen in similar patients exposed to iodi- the 7 patients who developed nephrotoxicity, the average
nated radiocontrast. baseline serum creatinine concentration was 2.5 mg/dl (esti-
mated CrCl = 33 ml/min); 4 had diabetes and 5 hypertension
In 2001, 16 patients with mild CKD (mean serum creatin-
suggesting that these may be risk factors for Gd-contrast
ine concentration = 1.5 mg/dl; range = 1.2-1.8 mg/dl) un-
related nephrotoxicity. Although this study was uncontrolled,
derwent intravenous studies with gadodiamide at a dose of
the high rate of acute kidney injury seen in this population
0.3 mmol/kg [29]. No patient developed clinically significant
was evidence for Gd-contrast nephrotoxicity. While en-
kidney dysfunction. The largest increase in serum creatinine
concentration was 0.2 mg/dl. hanced nephrotoxicity from intra-arterial use of Gd-contrast
is likely, complications (i.e., atheroemboli) from the artery
Lastly, a study published in 2002 prospectively examined manipulation by the procedure confounds the study results.
29 patients with chronic kidney disease (mean serum creatin-
In a prospective study, 21 CKD patients with a serum
ine concentration of 3.6 mg/dl; range, 1.6-7.0 mg/dl) who
creatinine concentration greater than 1.5 mg/dl (eGFR < 50
received 0.34 mmol/kg (range, 0.23-0.44 mmol/kg) of
gadopenetate dimeglumine [30]. In contrast to other studies, ml/min/m2) were randomized to either high dose gadobutrol
(1603 mosm/L, 0.34 to 0.90 mmol/kg) or iohexol (820
intravenous saline was employed as “Gd-contrast prophy-
mOsm/L iodinated radiocontrast) for digital subtraction an-
laxis”. A total of 32 procedures were performed on these
giography [32]. Ten patients receiving gadobutrol had a
patients. None of the patients developed Gd-contrast neph-
baseline eGFR of 34 ml/min and 60% had diabetes while the
roxicity as defined as an increase in serum creatinine concen-
iohexol group (n = 11) had an eGFR of 29 ml/min with 36%
tration > 0.5 mg/dl, over a 3 day period of observation [30].
One patient developed acute kidney injury; however, this of the patients having diabetes; both groups received intrave-
nous fluids prior to contrast. Contrast-induced nephropathy
was attributed to renal atheroemboli rather than Gd-contrast
(50% decrease in eGFR within 48 hours) developed in 45%
injury.
of iohexol patients and 50% of gadobutrol patients; none of
which required renal replacement therapy.
Gadolinium-Contrast Studies Supporting Nephrotoxicity
In a retrospective study, 91 patients with CKD stage 3 (n
In contrast to the prior reports supporting renal safety, = 50) and 4 (n = 41) were examined for nephrotoxicity (in-
four studies suggest that Gd-contrast agents exhibit variable
crease in serum creatinine concentration of 0.5 mg/dl within
degrees of nephrotoxicity (Table 3). An uncontrolled retro-
24-72 hours) of Gd-contrast exposure [33]. The patients re-
spective study published in 2003 examined the effect of both
ceived one of three different Gd-contrast preparations (2 with
intra-arterial (n=42) and intravenous (n=153) gadopentetate
high-osmolality, 1 with low-osmolality) at 0.2 mmol/kg.
dimeglumine (1960 mOsm/L) on kidney function in patients
Contrast prophylaxis was not employed. Approximately 20%
with underlying CKD [31]. The average dose of high osmo- of patients had diabetes and 80% hypertension. Eleven pa-
lality gadopenetate dimeglumine was 0.28 mmol/kg and no
tients (12.1%) developed contrast-induced nephropathy,

Table 3. Studies Supporting Nephrotoxicity of Gadolinium-Contrast Agents

Author (Year) Study Contrast Agent Dose (mmol/kg) Renal Function ([Cr] in mg/dl) Result

Sam (2003) N = 195 with CKD Gadopentetate 0.28 CrCl < 80 ml/min, CIN: 7/195
No control group CrCl = 38.2±16 ml/min MRA: 3/153 (1.9%)
CIN > 1.0 mg/d at 48 hr with DSA: 4/42 (9.5%)
oligoanuria
Erley (2004) Randomized prospective Gadobutrol = 10 0.57±0.17 [Cr] > 1.5 or CIN:
N = 21 Iohexol = 11 CrCl < 50 ml/min/1.73m2 Gado: 5/10 (50%)
CIN >50% decrease in GFR RC: 5/11 (45%)
Briguori (2006) Retrospective, N = 25, (historical Gadodiamide = 8 0.6±0.3 [Cr] > 2 mg/dL or CIN:
controls, N = 32) Gadobutrol = 17 0.28-1.23 CrCl < 40 ml/min Gado: 7/25 (28%)
CIN  0.5 mg/dL within 48 hr or 3: 1 mixture with RC RC: 2/32 (6.5%)
dialysis within 5 days
Ergun (2006) Retrospective, N = 91 Gadopentetate 0.2 Stage 3 and 4 CKD CIN: 11/91 (12.1%)
[Cr] measured pre-Gado, days 1, Gadodiamide Mean [Cr] = 33 ml/min Range CKD Stage 4: 9/11 with
3, and 7, and 1 mo, Dotarem CrCl = 15-58 CIN
CIN  0.5 mg/dL within 72 hr
Abbreviations: CKD, chronic kidney disease; Gado, gadolinium; RC, iodinated radiocontrast; CIN, contrast-induced nephropathy; [Cr], serum creatinine concentration; MRA, mag-
netic resonance angiography; DSA, digital subtraction angiography; CrCl, creatinine clearance; mo, month; hr, hours; GFR, glomerular filtration rate.
Gadolinium-Contrast and Kidney Disease Current Drug Safety, 2008, Vol. 3, No. 1 71

again suggesting that Gd-contrasts can be nephrotoxic. Six of on chronic dialysis (n = 5), and one patient with AKI were
these patients had diabetes mellitus and 9 had stage 4 CKD. noted to develop a previously unrecognized fibrosing disor-
The type of Gd-contrast (i.e., high versus low osmolality) der of the skin [7]. This new disease entity was descriptively
administered to the patients who developed RCIN was not coined nephrogenic fibrosing dermopathy (NFD) after de-
noted. No patient required renal replacement therapy for tailed examination of the clinical and histopathologic data of
acute kidney injury. 14 cases by Cowper and colleagues [7]. A case control study
of 8 patients with NSF undertaken by the Centers for Disease
A prospective study in 25 patients with CKD (mean se-
Control and Prevention (CDC) and California Department of
rum creatinine concentration of 2.3 mg/dl) and a matched
Health could not identify a specific etiology or trigger, but
historical control (n = 32) examined the nephrotoxicity of 2
found advanced kidney dysfunction as a common thread.
different Gd-contrast agents administered during cardiac
catheterization (34). Contrast prophylaxis with 0.45% saline Following the subsequent recognition that fibrosis also oc-
curred in systemic organs and the role of the circulating fi-
and N-acetylcysteine was provided to all patients. In the Gd-
brocyte in the fibrosing reaction, the name was changed to
contrast group, a contrast mixture that contained 0.6
nephrogenic systemic fibrosis (NSF). This process symmet-
mmol/kg of Gd-contrast and 0.4 ml/kg of iso-osmolar non-
rically affects the extremities more so than the trunk, the face
ionic radiocontrast was compared with the same iso-osmolar
is always spared. Initial signs and symptoms include sharp
radiocontrast agent alone in the historical control group. The
2 contrast protocols are considered equivalent based on the pain and burning associated with redness and swelling of the
skin. These changes progress over a matter of weeks to
concept of "X-ray attenuating doses". In the Gd-contrast/iso-
months to extensive dermal fibrosis (entire limbs), often pro-
osmolar contrast group, 28% of patients developed an in-
ducing severe joint contractures and marked limitations in
crease in serum creatinine concentration of 0.5 mg/dl within
mobility. This may lead to a wheelchair dependent or bed
48 hours as compared with 6.5% in the radiocontrast alone
bound state [6]. Involvement of systemic organs such as the
historical control group [34].
liver, heart, lungs, diaphragm, esophagus and skeletal muscle
has also been reported and may be associated with fatal con-
Gadolinium-Contrast Nephrotoxicity: What Should We
sequences [6,7].
Conclude?
Drawing a conclusion from the information reviewed Literature Review on the Gadolinium-NSF Link
above is difficult. The majority of studies suggest safety, but
After the initial report of cases in 2000, the NSF literature
clearly Gd-contrast-induced nephrotoxicity can develop. One
is hampered by data from studies that use variable designs consisted predominantly of case reports/case series with the
cause being ascribed to any of a number of potential agents
with small numbers, patients with varying levels of kidney
or associations. Included were exposure to high-dose
function, wide ranges of Gd-contrast osmolality and dose,
erythropoietin, presence of anti-phospholipid antibodies (and
non-uniform measures of kidney function, erratic use of con-
other hypercoaguable states), vascular injury and vascular
trast prophylaxis, and poor controls or lack of control groups
surgical procedures, ischemia, and liver failure (in particular
altogether. Clearly, better studies are required that employ
adequate numbers of high-risk patients (CKD 3 and 4, dia- hepatorenal syndrome and liver transplantation). Despite
this, no unifying agent or risk factor except for underlying
betic nephropathy) and the use of appropriate contrast pro-
kidney disease was identified. A major breakthrough oc-
phylaxis (intravenous fluids, N-acetylcysteine). That being
curred in 2006 when Grobner reported the development of
said, there appears to be adequate data to suggest that Gd-
NSF in 5 ESRD patients exposed to gadodiamide in the set-
contrast agents have enough of a nephrotoxic potential that
ting of metabolic acidosis [2]. Subsequent to this report, a
caution should be exercised in their use in patients with stage
4/5 CKD, and possibly even more so in patients with ad- number of centers have replicated this association. Marck-
mann and colleagues described 13 patients in Denmark who
vanced CKD and diabetes. Non-contrast studies (ultrasound,
developed symptoms of NSF within 2 to 75 days post expo-
CO2) may be preferred. If this is not possible, it would seem
sure to gadodiamide [3]. Eight patients had ESRD (one on
reasonable to consider some form(s) of contrast prophylaxis
peritoneal dialysis), while 5 had advanced (stage 5) CKD not
in higher-risk patients receiving Gd-contrasts. Higher doses
yet on dialysis. All had courses complicated by vascular in-
(> 0.3-0.4 mmol/kg) and arterial injection of Gd-contrast
agents appear to enhance risk. Whether the use of higher jury and none had acidosis. A small case control study from
Connecticut noted NSF in 3 patients exposed to gadolinium
osmolality Gd-contrast agents increase the nephrotoxicity is
contrast (2 gadodiamide, 1 gadopentetate) demonstrating an
uncertain since the nephrotoxic potential of Gd-contrasts at
incidence of 4.3 cases per 1000 patient years [35]. This was
different osmolalities have not been systematically exam-
associated with an absolute risk of 3.4% for development of
ined. Regardless, it is prudent to employ the lowest dose of
NSF in an exposed patient.
Gd-contrast possible to achieve adequate image quality in
higher-risk patients. There is no evidence that these maneu- In California, another 12 patients were noted to develop
vers would be efficacious, but the similarities of Gd-contrast NSF following gadodiamide exposure (2-11 weeks), with an
“nephrotoxicity” to that of typical iodinated radiocontrast- odds ratio of 22.3 [4]. Four patients were suffering from
induced nephropathy make these suggestions reasonable. AKI, eight had ESRD on dialysis, and 33% of the patients
had some form of vascular injury. Another 6 patients with
GADOLINIUM-CONTRAST AND NEPHROGENIC variable levels of kidney disease were described in Texas to
SYSTEMIC FIBROSIS develop NSF following gadodiamide [5]. The onset of symp-
toms ranged form 19 days to 2 months. The CDC published
In 1997, several renal transplant recipients with failed
their findings in MMWR (3/5/07) of a case control study of
allografts requiring chronic dialysis (n = 9), ESRD patients
72 Current Drug Safety, 2008, Vol. 3, No. 1 Mark A. Perazella

19 patients with confirmed NSF from 2 St. Louis area hospi- the gadolinium chelate [22]. Recently, an abstract was pre-
tals [36]. Of these patients, 11 were maintained on hemo- sented at the International Society of Nephrology/World
dialysis, 6 on peritoneal dialysis, and 2 had temporary hemo- Congress of Nephrology in Rio de Janiero, Brazil [43]. 40
dialysis for AKI. The attack rate for peritoneal dialysis (4.6 rats with normal kidney function were exposed to 2 types of
cases/100 patients) was much higher than for hemodialysis high dose gadolinium (2.5 mmol/kg of either gadodiamide or
(0.61/100 patients), suggesting worse clearance of Gd- gadopentetate), gadodiamide (2.5 mmol/kg) without its ex-
contrast with peritoneal dialysis. The type and dose of Gd- cess sodium caldiamide chelate, or saline control daily for 28
contrast was not noted, but hypothyroidism, edema and deep days. Rats exposed to gadodiamide without excess chelate
venous thrombosis were significant risk factors for develop- developed NSF-like skin lesions within 6 days, those ex-
ment of NSF. In one case; however, they could not document posed to gadodiamide with excess chelate developed NSF-
exposure to Gd-contrast. Physicians from Wisconsin re- like skin lesions at 16 days, while the gadopentetate and sa-
ported another 13 cases of NSF following gadodiamide ad- line exposed rats did not develop skin lesions. Although
ministration noting that an underlying pro-inflammatory state histopathology was note reported, higher gadolinium concen-
(major surgery, infection, vascular event or thrombosis) was trations were found in the tissues of the 2 groups of rats that
an important risk factor [37]. They also confirmed the high developed NSF. This experiment suggests that gadodiamide,
mortality (31%) associated with this disease state. They, due to the process of transmetalation with release of gadolin-
however, incorrectly classified 2 patients with AKI as CKD ium, is more likely to cause NSF than gadopentetate, at least
stage 3. Most recently, a case control study of 19 patients in rats.
with NSF from the Denmark group described that high cu-
In the 7 published reports (total of 58 patients) where the
mulative gadodiamide dose, elevated serum calcium and
specific MR contrast agent was identified, the Gd-contrast
phosphate concentrations and high dose epoietin- (trend)
administered was gadodiamide in all but one, which was
increase the risk of developing NSF [38]. It is notable that
gadopentetate [2-5, 36-38]. While the initial inclination is to
NSF has been reported in most European countries including ascribe NSF to this particular agent, care must be taken be-
Denmark, United Kingdom, Austria, Belgium, the Nether-
fore gadodiamide is blamed as specifically responsible for
lands, Norway, Sweden, and Switzerland [1,39].
NSF since this agent is one of the more commonly used Gd-
The FDA has also linked NSF to Gd-contrast exposure contrasts. In fact, gadodiamide and gadopentetate are the
and has released two Public Health Advisories, one in 6/06 most commonly used Gd-contrast agents claiming 81% of
reporting NSF in 25 ESRD patients after Gd-contrast expo- the market share for 2006 per the Contrast Media Industry
sure [10], and an update in 12/06 increasing this number to Guide (CMIG) report. Also, according to the FDA Med-
90 patients [11]. An NSF registry was created in 2001 to col- Watch reporting system, as of 1/07 there have been more
lect data on all cases of NSF [7] and to date greater than 95% than 100 cases of NSF in which 85 were associated with ex-
of 239 cases of NSF (where data are available) have been posure to gadodiamide (Omniscan), 21 with gadopentetate
linked to exposure to gadolinium [25]. Gadodiamide has (Magnevist), 6 with gadoversetamide (OptiMARK), and one
been implicated in approximately 85% of these cases and with gadobenate (MultiHance), although this patient also
gadopentetate in 15%. The vast majority of patients that de- received gadodiamide. More recently (3/07), Bayer Health
velop NSF are dialysis-dependent (~90%), although it has Care stated that they were aware of 42 cases of NSF associ-
also been described in patients with advanced CKD not yet ated with exposure to gadopentetate (Magnevist) [44]. As
on dialysis, patients with a poorly functioning renal trans- noted above, the NSF registry implicates gadodiamide in
plant and those with acute kidney injury [6,7]. approximately 85% of these cases and gadopentetate in 15%
Further evidence of the importance of Gd-contrast agents [7]. Each of these NSF associated preparations is a linear
Gd-contrast agent. To date, there are no reports of NSF asso-
as a trigger for NSF was provided by documentation of Gd3+
ciated with gadoteridol (ProHance), a cyclic Gd-contrast.
within the tissues of patients with NSF using scanning elec-
tron microscopy and energy dispersive X-ray spectroscopy
Gadodiamide and NSF
[40,41]. In addition to this qualitative evidence, High and
coworkers quantified the concentration of Gd3+ in tissues of If gadodiamide is more likely to cause NSF than the other
the NSF patients previously examined [42]. They found that Gd-contrast agents, what factors make it unique in this re-
the NSF tissues contained 35-150 fold higher amounts (5- gard? One theory relates to its stability, its ability to bind to
106 parts per million [ppm]) than the tissues of healthy sub- and sequester Gd+3. Gadodiamide has the lowest stability
jects (0.477-1.77 ppm) exposed to Gd-contrast. The authors constant and highest dissociation rate of the five Gd-contrast
speculate that phagocytosis of Gd3+ retained in tissues by preparations available in the United States (Table 1). Be-
macrophages results in production of profibrotic cytokines cause of this decreased stability, the Gd+3 ion of gadodiamide
that eventuate in dermal and/or systemic fibrosis. is more likely to dissociate from its stabilizing chelate moi-
ety than other Gd-contrast agents. Because of this fact, ex-
Animal Research Examining Gadolinium-NSF Link cess chelate (12 mg/ml of sodium calcium diamide) is added
to the commercial formulations of gadodiamide in an attempt
An obvious area of research to examine the potential for
to diminish freely circulating Gd3+ [14,15]. The stability con-
gadolinium-contrast agents to promote the development of
stant of gadoversetamide is essentially the same as gadodia-
NSF is studies in animals. Interestingly, rats exposed to
daily, high dose gadolinium-contrast for 28 days developed mide, however the excess chelate in this preparation is even
greater at 28.4 mg/ml (sodium calcium versetamide), and is
numerous skin lesions that included hair loss, erythema,
tempting to use this as an explanation for the decreased cases
thickening, ulceration and crusting. These lesions were at-
of NSF associated with gadoversetamide compared to
tributed to zinc deficiency from transmetalation of zinc by
Gadolinium-Contrast and Kidney Disease Current Drug Safety, 2008, Vol. 3, No. 1 73

gadodiamide. However, the stability constant of gadopen- merous patients in the United States alone with ESRD and
tetate is 1000 times greater than either gadodiamide or gado- advanced CKD, most of which have significant co-
versetamide and this agent is now the second most common morbidities that require radiographic imaging. It is very
Gd-contrast associated with NSF. This may relate to its more likely that MR studies are commonly used in this population
widespread use, but it still calls into question the concept of to assess vascular disease and other end organ pathology. If
relative toxicity being related to chelate stability. At the pre- therefore seems reasonable to assume that many patients
sent time, it is risky to blame this disease on specific agents with advanced kidney disease that have received an MR
and I agree with the FDA [11] that it must be assumed to be study with gadolinium have been spared this abysmal com-
a “class effect” until more data are available. plication. Thus a combination of risk factors or cofactors are
likely required for NSF to occur. Clearly, advanced kidney
The dose of Gd-contrast utilized appears to be impor-
tantly associated with the development of NSF. A typical disease is a requisite. The dose of Gd-contrast appears to
play a role and the specific Gd-contrast (gadodiamide) util-
non-vascular MRI examination employs 0.1 mmol/kg while
ized may also be a factor. All of the published studies and
an MRA often utilizes up to 0.3 mmol/kg of Gd-contrast. In
the data from the NSF registry describe the presence of vas-
the 5 cases reported from Austria, the dose of Gd-contrast
cular/endothelial injury (hypercoaguable states, venous
was approximately 0.26 mmol/kg if one assumes a 70 kg
thrombosis, vascular surgery) and perhaps a proinflamma-
body weight [2]. The 13 cases reported from Denmark re-
ported an average contrast volume of 18.5 mmol, which tory state (infection, major surgery) in NSF patients. A num-
ber of other cofactors may also be important including meta-
would be 0.26 mmol/kg for a 70 kg individual [3]. Only one
bolic acidosis, intravenous iron dosing, increased serum cal-
patient received less than a 10 mmol dose. In the study by
cium and phosphate concentrations, and high dose erythro-
Broome, gadodiamide was administered at “double dose”
poietin, but they are hardly confirmed. Fig. (1) demonstrates
(0.2 mmol/kg) in each of the patients in whom NSF subse-
the factors potentially responsible for the development of
quently developed [4]. During the six-year period of this
report, 559 MRI examinations (301 with gadolinium-contrast NSF following Gd-contrast exposure.
and 258 without contrast) were performed on 168 patients
Prevention and Treatment of NSF
with advanced kidney disease receiving dialysis. Since there
were no cases of NSF in the patients receiving an MR study Currently, there is no effective therapy for NSF. Physical
performed without Gd-contrast, the calculated odds ratio for therapy plays an important role in increasing and maintaining
gadolinium causing NSF was 22.3. Of the 301 MR studies the mobility of affected limbs and joints. Renal transplanta-
utilizing Gd-contrast, 207 were performed using a dose of tion has anecdotally improved or stabilized NSF in patients
0.2 mmol/kg while the remaining 94 received 0.1 mmol/kg. who have excellent graft function. Extracorporeal photo-
Since none of the cases of NSF occurred in patients receiving pheresis has shown some hope while others with variable
0.1 mmol/l of Gd-contrast, the odds ratio of developing NSF response include steroids, pentoxyphylline, plasmapheresis,
with the higher dose of Gd-contrast was 12.1 [4]. Review of and intravenous sodium thiosulfate [45,46]. Since NSF is
all of the published literature generally supports the en- still relatively rare and sporadic, none of these maneuvers
hanced risk with higher Gd-contrast dose. The NSF registry have been tested in a controlled clinical trial.
also supports this contention as the majority of patients with
What recommendations can be made regarding the use of
NSF were exposed to high doses of and multiple exposures
Gd-contrast-based MR studies in “high-risk” patients? The
to Gd-contrast. These data suggest that the risk of NSF is
vast majority of cases of NSF occur in patients with ESRD
significantly correlated to the dose of Gd-contrast adminis-
receiving hemodialysis or peritoneal dialysis. However, ap-
tered and patients would therefore be at much higher risk if
receiving an MRA as opposed to a typical non-vascular MRI proximately 10% of the cases develop in patients with acute
kidney injury (many requiring dialysis), patients with ad-
study.
vanced CKD stage 4 (estimated GFR 15-30 ml/min) and in
Why should Gd-contrast have the potential to trigger the patients with CKD stage 5 not receiving renal replacement
development of NSF in patients with underlying kidney dis- therapy (estimated GFR < 15 ml/min). Thus, for the time
ease? Certainly, reduced kidney function increases the T1/2 of being, it would seem reasonable to try to avoid administra-
Gd-contrast considerably as it is slowly excreted by the kid- tion of all types of Gd-contrast to patients with these charac-
neys in acute kidney injury, advanced stages of chronic kid- teristics. This recommendation is even stronger when con-
ney disease (CKD stages 4 and 5), and dialysis-dependent sidering an MRA in this population, as the higher contrast
ESRD where it requires three hemodialysis treatments to dose required with those studies appears to significantly in-
remove >95% of the administered dose. Thus, significant crease the risk of NSF. Alternative imaging modalities (ul-
renal impairment is associated with increased time for trans- trasonography, PET scan, CO2 angiography, MRI without
metalation and prolonged tissue exposure, which may pro- contrast, CT scan without contrast) should be employed in
mote deposition of toxic Gd3+ leading to fibrosis. In addition, these patients whenever possible, including studies using
the association of higher doses Gd-contrast increasing the iodinated radiocontrast when necessary. Using iodinated
risk of developing NSF also supports this hypothesis as these radiocontrast in patients with advanced CKD not on dialysis
higher doses would further increase and prolong tissue expo- puts patients at risk for radiocontrast-induced nephropathy
sure in the setting of impaired excretion. and may “tip them over” into a situation that may require
Compared with other conditions and complications that temporary or permanent dialysis. Iodinated radiocontrast-
afflict ESRD patients, NSF appears to be a relatively rare induced nephropathy is more likely to occur than is NSF
condition occurring in only 5% of patients receiving induced by Gd-contrast exposure in this group of patients.
gadodiamide in the report from Denmark [3]. There are nu- However, since radiocontrast-induced nephropathy is poten-
74 Current Drug Safety, 2008, Vol. 3, No. 1 Mark A. Perazella

Fig. (1). Speculative mechanism by which gadolinium might trigger nephrogenic systemic fibrosis. In the setting of kidney disease (1), im-
paired renal excretion of high dose gadolinium prolongs the T1/2 and enhances the chance for dissociation of gadolinium from its chelate (2),
allowing increased tissue exposure. Pro-inflammatory conditions (3), vascular trauma and endothelial dysfunction allows free Gd3+ to more
easily enter tissues, where macrophages phagocytose the metal and produce local profibrotic cytokines as well as signals that attract circulat-
ing fibrocytes to the tissues. Other co-factors (4), such as increased serum calcium and phosphorus, metabolic acidosis, intravenous iron ther-
apy and high dose intravenous erythropoietin may also enhance the development of NSF through various effects. Once in tissues, circulating
fibrocytes and perhaps Gd3+ (through direct effects on collagen) induce a fibrosing process that is indistinguishable from normal scar forma-
tion. Abbreviations: Gd3+, gadolinium; cyto, cytokines; cF, circulating fibrocyte.

tially reversible while NSF is not, a Gd-contrast-based study radiocontrast agents, they are clearly not entirely benign.
may be a better choice in this group. If an MR study with Nephrotoxicity can occur in high risk patients who receive a
contrast (especially MRA) were felt to be absolutely required large dose (> 0.3-0.4 mmol/kg) and intra-arterial injection of
in a patient with ESRD or CKD 4 or 5 not on dialysis, the Gd-contrast. NSF can be a catastrophic complication of Gd-
lowest possible dose of Gd-contrast should be used and if contrast exposure and we should make a concerted effort to
feasible, a Gd-contrast other than gadodiamide. Also, until avoid this dreaded condition. There is strong evidence that
further data are available on this topic, it would seem prudent NSF is related to a tissue response to the toxic effects of
to perform hemodialysis both immediately following, and gadolinium. Hopefully, future work in this field will provide
again the day after the MR study to accelerate Gd-contrast more clearly identifiable risk factors and preventative ma-
elimination, in patients already receiving hemodialysis. This neuvers that will allow us to give advice beyond simple
maneuver should also be entertained for CKD patients (not avoidance altogether. This will be important information as
currently on chronic dialysis) and patients maintained on MR Gd-contrast studies play an important role in the diagno-
chronic peritoneal dialysis who receive a Gd-contrast (MRA) sis, treatment and follow-up of a vast array of medical condi-
study. This is obviously a very inconvenient strategy, but this tions in patients with kidney disease.
aggressive approach underscores the concern over, and un-
certainty surrounding, the development of this devastating REFERENCES
condition. Obviously, these are difficult decisions that can [1] Perazella MA, Rodby RA. Gadolinium use in patients with kidney
only be made utilizing input from the physician requesting disease: A cause for concern. Semin Dial 2007; 20: 179-84.
the MR study, the nephrologist and the radiologist. There [2] Grobner T. Gadolinium-a specific trigger for the development of
will be situations that Gd-contrast administration cannot be nephrogenic fibrosing dermopathy and nephrogenic systemic fibro-
avoided in a high-risk patient. sis? Nephrol Dial Transplant 2006; 21: 1104-08.
[3] Marckmann P, Skov L, Rossen, et al. Nephrogenic systemic fibro-
sis: suspected etiological role of gadodiamide used for contrast-
CONCLUSION enhanced magnetic resonance imaging. J Am Soc Nephrol 2006;
17: 2359-62.
In conclusion, Gd-contrast agents can no longer be as- [4] Broome DR, Girguis MS, Baron PW, et al. Gadodiamide-
sumed to be as safe as they have traditionally been consid- associated nephrogenic systemic fibrosis: Why radiologists should
ered when administered to patients with underlying kidney be concerned. Am J Roentgenol 2007; 188(2): 586-92.
disease. Although the toxicity profile of the Gd-contrast [5] Khurana A, Runge VM, Narayanan M, et al. Nephrogenic systemic
fibrosis: A review of 6 cases temporally related to gadodiamide in-
agents may be narrower than that associated with iodinated jection (Omniscan). Invest Radiol 2007; 42: 139-45.
Gadolinium-Contrast and Kidney Disease Current Drug Safety, 2008, Vol. 3, No. 1 75

[6] Galan A, Cowper SE, Bucala R. Nephrogenic systemic fibrosis supplement CO2 angiography in patients with renal insufficiency. J
(nephrogenic fibrosing dermopathy). Curr Opin Rheumatol 2006; Vasc Interv Radiol 2000; 11(1): 35-43.
18(6): 614-17. [28] Spinosa DJ, Matsumoto AH, Angle JF, et al. Safety of CO2- and
[7] Cowper SE. Nephrogenic Fibrosing Dermopathy [NFD/NSF Web- gadodiamide-enhanced angiography for the evaluation and percu-
site]. 2001-2007. Available at http: //www.icnfdr.org Accessed tanwous treatment of renal artery stenosis in patients with chronic
6/1/2007. renal insufficiency. Am J Roentgenol 2001; 176(5): 1305-11.
[8] Thomsen HS, Morcos SK, Dawson P. Is there a causal relation [29] Sancak T, Bilgic S, Sanldilek U. Gadodiamide as an alternative
between the administration of gadolinium based contrast media and contrast agent in intravenous digital subtraction angiography and
the development of nephrogenic systemic fibrosis. Clin Radiol interventional procedures of the upper extremity veins. Cardiovasc
2006; 61: 905-06. Intervent Radiol 2002; 25(1): 49-52.
[9] Thomsen HS. Nephrogenic systemic fibrosis: a serious late adverse [30] Reiger J, Sitter T, Toepfer M, Linsenmaier U, Pfeofer KJ, Schiffl
reaction to gadodiamide. Eur Radiol 2006; 16: 2619-21. H. Gadolinium as an alternative contrast agent for diagnostic and
[10] Food and Drug Administration (2006) Public Health Advisory: interventional angiographic procedures in patients with impaired
Gadolinium-containing contrast agents for magnetic resonance im- renal function. Nephrol Dial Transplant 2002; 17: 824-28.
aging (MRI). http: //www.fda.gov/cder/drug/advisory/gadolinium_ [31] Sam AD, Morasch MD, Collins J, et al. Safety of gadolinium con-
agents.htm. trast angiography in patients with chronic renal insufficiency. J
[11] Food and Drug Administration (2006) Public Health Advisory: Vasc Surg 2003; 38: 313-18.
Gadolinium-containing contrast agents for magnetic resonance im- [32] Erley CM, Bader BD, Berger ED. Gadolinium-based contrast me-
aging (MRI). http: //www.fda.gov/cder/drug/advisory/gadolinium_ dia compared with iodinated media for digital subtraction angi-
agents_20061222.htm ography in azotemic patients. Nephrol Dial Transplant 2004; 19:
[12] Perazella MA. Nephrogenic systemic fibrosis, gadolinium, and 2526-2531.
chronic kidney disease: Is there a link? Clin J Am Soc Nephrol [33] Ergun I, Keven K, Uruc I, et al. The safety of gadolinium in pa-
2007; 2: 200-02. tients with stage 3 and 4 renal failure. Nephrol Dial Transplant
[13] Bellin MF. MR contrast agents, the old and the new. Eur J Radiol 2006; 21: 697-700.
2006; 60: 314-23. [34] Briguori C, Colombo A, Airoldi F, et al. Gadolinium-based con-
[14] Lorusso V, Pascolo L, Fernetti C, et al. Magnetic resonance con- trast agents and nephrotoxicity in patients undergoing coronary ar-
trast agents: from the bench to the patient. Curr Pharm Des 2005; tery procedures. Catheter Cardiovasc Interv 2006; 67: 175-80.
11: 4079-4098. [35] Deo A, Fogel M, Cowper SE. Nephrogenic systemic fibrosis: A
[15] Runge VM. Safety of magnetic resonance contrast media. Top population study eaxmining the relationship of disease develop-
Magn Reson Imag 2001; 12(4): 309-14. ment to gadolinium exposure. Clin J Am Soc Nephrol 2007; 2: 264-
[16] Swan SK, Lambrecht LJ, Townsend R, et al. Safety and pharma- 67.
cokinetic profile of gadobenate dimeglumine in subjects with renal [36] Nephrogenic fibrosing dermopathy associated with exposure to
impairment. Invest Radiol 1999; 34(7): 443-55. gadolinium-containing contrast agents-St. Louis, Missouri, 2002-
[17] Reinton V, Berg KJ, Svaland MG, et al. Pharmacokinetics of 2006. Morb Mortal Wkly Rep 2007; 56 (7): 137-41.
gadodiamide injection in patients with moderately impaired renal [37] Sadowski EA, Bennett LK, Chan MR, et al. Nephrogenic systemic
function. Acta Radiol 1994; 1: 56-61. fibrosis: Risk factors and incidence estimation. Radiology 2007;
[18] Schuhmann-Giampieri G, Krestin G. Pharmacokinetics of Gd- 243: 148-57.
DTPA in patients with chronic renal failure. Invest Radiol 1991; [38] Marckmann P, Skov L, Rossen K, Heaf J, Thomsen HS. Case-
11: 975-79. control study of gadodiamide-related nephrogenic systemic fibro-
[19] Joffe P, Thomsen HS, Meusel M. Pharmacokinetics of gadodia- sis. Nephrol Dial Transplant 2007; Epub doi: 10/1093.
mide injection in patients with severe renal insufficiency and pa- [39] Personal communication, Henrik S. Thomsen, Department of Diag-
tients undergoing hemodialysis or continuous ambulatory perito- nostic Radiology, Copenhagen University, Denmark.
neal dialysis. Acta Radiol 1998; 5: 491-502. [40] High WA, Ayers RA, Chandler J, et al. Gadolinium is detectable
[20] Saitoh T, Hayasaka K, Tanaka Y, et al. Dialyzability of gadodia- within the tissue of patients with nephrogenic systemic fibrosis. J
mide in hemodialysis patients. Radiat Med 2006; 24: 445-51. Am Acad Dermatol 2007; 56: 21-26.
[21] Okada S, Katagirir K, Kumazaki T, et al. Safety of gadolinium [41] Boyd AC, Zic JA, Abraham JL. Gadolinium deposition in nephro-
contrast agent in hemodialysis patients. Acta Radiol 2001; 42: 339- genic fibrosing dermopathy. J Am Acad Dermatol 2007; 56; 27-30.
41. [42] High WA, Eng M, Ayers RA, Cowper, SE. Gadolinium is quantifi-
[22] Wible JH, Troup CM, Hynes MR, et al. Toxicological assessment able within the tissue of patients with nephrogenic systemic fibro-
of gadoversetamide injection (OptiMARK), a new contrast- sis. J Am Acad Dermatol 2007; 56: 1-2.
enhancement agent for use in magnetic resonance imaging. Invest [43] Peitsch H, Sieber M, Frenzel T, Weinmann H. Nephrogenic sys-
Radiol 2001; 36: 401-12. temic fibrosis: Are gadolinium-based MRI contrast agents a possi-
[23] Niendorf HP, Alhassan A, Haustein J, Clauss W, Cornelius I. ble trigger? Poster presentation World Congress of Nephrol-
Safety and risk of gadolinium-DTPA: extended clinical experience ogy/International Society of Nephrology, Rio de Janiero, Brazil,
after more than 5,000,000 applications. Adv MRI Contrast 1993; 2: 2007.
12-19. [44] Magnevist Safety Information: NSF/NFD Update. Available at
[24] Bellin MF, Deray G, Assogba U, et al. Gd-DOTA: evaluation of its http: //www.imaging.bayerhealthcare.com/html/magnevist/nsf_nfd.
renal tolerance in patients with chronic renal failure. Magn Reson html?WT.mc_id=berleximaging.com.
Imaging 1992; 10: 115-18. [45] Yerram P, Saab G, Karuparthi P, et al. Nephrogenic systemic fibro-
[25] Prince MR, Arnoldus C, Friscoli JK. Nephrotoxicity of high-dose sis: A mysterious disease in patients with renal failure-role of gado-
gadolinium compared with iodinated contrast. J Magn Reson Imag- linium-based contrast media in causation and the beneficial effect
ing 1996; 6 (1): 162-66. of intravenous sodium thiosulfate. Clin J Am Soc Nephrol 2007; 2:
[26] Hammer FD, Goffette PP, Malaise J, Mathurin P. Gadolinium 258-63.
dimeglumine: An alternative contrast agent for digital subtraction [46] Gilliet M, Cozzio A, Burg G, et al. Successful treatment of three
angiography. Eur Radiol 1999; 9(1): 128-36. cases of nephrogenic fibrosing dermopathy with extracorporeal
[27] Spinosa DJ, Angle JF, Hagspiel KD, et al. Lower extremity arteri- photopheresis. Br J Dermatol 2005; 152: 531-36.
ography with use of iodinated contrast material or gadodiamide to

Received: June 1, 2007 Revised: July 10, 2007 Accepted: July 11, 2007

You might also like