Physicians are constantly introduced to new diagnostic and therapeutic modalities developed to enhance patient care. However, adverse effects of these technological advances often come to light only when they are later widely employed in clinical practice. Gadolinium-based contrast agents are utilized with magnetic resonance (MR) imaging and have been considered generally safe when administered to most patients. With the exception of allergic type reactions, and perhaps the potential of nephrotoxicity in high-risk patients (renal insufficiency, diabetic nephropathy), they are well tolerated. Recently however, reports of a previously rare condition, nephrogenic systemic fibrosis, have surfaced at an alarming rate in patients with advanced kidney disease, and especially those already receiving dialysis, which has now been strongly linked with exposure to gadolinium. As gadolinium-based studies are widely used by physicians across many specialties, including interventional and diagnostic radiology, cardiology, and in particular vascular surgery, the goal of this editorial is to review this new concern for the readership of Vascular Disease Management.

In 1997, several renal transplant recipients with failed allografts requiring chronic maintenance dialysis were noted to develop a previously unrecognized fibrosing disorder of the skin.¹ This new disease entity was descriptively coined nephrogenic fibrosing dermopathy (NFD).¹ Following the subsequent recognition that fibrosis also occurred in systemic organs (in addition to skin), the name was changed to nephrogenic systemic fibrosis (NSF). Nephrogenic systemic fibrosis is an acquired condition that develops in patients with acute kidney injury (AKI), advanced chronic kidney disease (CKD) and dialysis-dependent end-stage renal disease (ESRD).^2,3 It is characterized by dermal fibrosis, which can cause severe joint contractures and limitations in mobility, often leading to a wheelchair-dependent or bed-bound state.² Involvement of systemic organs such as the liver, heart, lungs, diaphragm and skeletal muscle has also been described, sometimes with fatal consequences.^2,3 Since its initial description, over 215 cases have been collected in the NSF registry under the stewardship of Shawn Cowper.³ The vast majority of patients that develop NSF are dialysis-dependent (90%), although it has also been described in patients with advanced CKD not yet on dialysis and those with AKI.^2,3 Exposure to gadolinium following a magnetic resonance imaging (MRI) study in patients with kidney disease has been recently linked to the development of NSF.^4–8 Initially reported in 5 dialysis-dependent ESRD patients following gadolinium exposure,⁴ subsequent reports have confirmed this finding in another 35 patients.^5–9 Deo et al in a small population study noted a 2.4% risk of NSF with each gadolinium exposure.⁶

Another group reported 12 cases of NSF, of which 4 had AKI secondary to hepatorenal syndrome (2 were on renal replacement therapy).⁸ Gadolinium exposure in these patients was associated with an odds ratio of 22.3 for the development of this fibrosing disorder.⁸ Khurana and colleagues described NSF in 6 patients with underlying kidney disease who were exposed to gadolinium.⁹ Four patients were initiated on dialysis after gadolinium exposure, while one patient started dialysis the day prior to administration of gadolinium. The NSF registry confirms the association of gadolinium exposure to development of NSF as all patients with data available for analysis have received this contrast agent prior to development of the disease.³ The time from gadolinium exposure to diagnosis of NSF ranged from 2 to 8 weeks.³ Recently, the FDA has also linked NSF to gadolinium and has released two Public Health Advisories, one published in 6/06 reporting NSF in 25 ESRD patients after gadolinium exposure,¹⁰ and an update in 12/06 increasing this number to 90 patients.¹¹ Two recent studies document free gadolinium within the affected fibrotic tissues of 5 NSF patients, further incriminating gadolinium as the pathogenic trigger.^12,13

There appears to be a dose relationship in the development of NSF as higher doses of gadolinium increase the risk of NSF. Most cases occur in patients who have received > 0.1mmol/kg dose. This aspect of gadolinium exposure and development of NSF is very relevant to physicians who care for patients with vascular disease, who often have underlying kidney disease. MR angiograms are commonly employed to study the renal, peripheral, and carotid vascular beds. Higher doses of gadolinium are usually required to perform these studies, which likely increase the risk for NSF. The initial report by Grobner noted that all patients in his series underwent an MR angiogram, which employs gadolinium doses of 0.2–0.3 mmol/kg.⁴ The 13 cases from Denmark used an average gadolinium volume of 18.5 mmol, or 0.26 mmol/kg for a 70 kg individual.⁵ Broome et al noted an odds ratio of developing NSF equal to 12.1 when comparing a 0.2 mmol/kg dose to a 0.1 mmol/kg dose.⁸ Finally, the dose of gadolinium administered to the 6 patients with NSF ranged from 0.11 to 0.36 mmol/kg, higher than the routine non-angiogram dose of 0.1 mmol/kg.⁹ Thus, an MR angiogram, typically using 0.3 mmol/kg of gadolinium would carry a significantly higher risk for NSF than a non-vascular MRI study.

How does gadolinium exposure in patients with kidney disease trigger NSF? Gadolinium contrast is eliminated almost entirely by the kidneys¹⁴ and underlying renal impairment reduces gadolinium clearance, thereby significantly increasing the T1/2 of gadolinium in patients with significant reductions in renal function (GFR < 30 ml/min).¹⁴ In ESRD patients, native renal clearance approaches zero, however, gadolinium is removed by hemodialysis. Three hemodialysis treatments are required to remove > 95% of an administered dose.¹⁵ Thus, tissue exposure to gadolinium is significantly prolonged in patients with advanced CKD, and is increased even further with higher doses.

Besides enhanced tissue exposure to gadolinium in the setting of kidney disease, other factors unique to gadolinium, in particular its chelate binding characteristics likely contribute. Free gadolinium (Gd3+) is extremely toxic to tissues and unsafe for human use.¹⁶ To prevent toxicity, Gd3+ is bound to a chelate, an organic molecule that forms a stable complex around the Gd3+ and gadolinium preparations differ predominantly in the chelate used. Chelates are categorized as either linear or macrocyclic based on their biochemical structure and charge.¹⁶ Linear chelates are thought to be less stable than macrocyclic chelates (i.e., bind less tightly to Gd3+) and more likely to dissociate (increase free Gd3+). The majority of NSF cases described have followed exposure to the specific gadolinium chelate gadodiamide (Omniscan^®).^4,7,8 Gadodiamide employs a linear chelate, is less stable than other chelates¹⁴ and is more likely to dissociate into free Gd3+, potentially leading to NSF. These data need to be viewed with caution, as gadodiamide was the only chelate used for MR studies by the NSF reporting centers. Furthermore, NSF has been occasionally reported with other gadolinium-chelates (gadopenetetate, gadoversetamide). Thus, for the time being, it is reasonable to view all gadolinium formulations as having potential to cause NSF in patients with advanced CKD (eGFR < 30 ml/min or on dialysis).

The data on gadolinium exposure and NSF are far from definitive at present. Despite this, there is ample evidence to make some reasonable suggestions regarding gadolinium use in patients with kidney disease.^17,18 ESRD patients on either hemo- or peritoneal dialysis are at highest risk and should avoid exposure to gadolinium if possible. Non-gadolinium options should be explored and employed when feasible. Judicious use of either low or iso-osmolar iodinated radiocontrast along with standard prophylactic measures (intravenous fluids and N-acetylcysteine) may be a better choice since iodinated radiocontrast-induced nephropathy is generally reversible while NSF is not. If an ESRD patient must receive gadolinium, we recommend using the lowest dose possible, and avoiding gadodiamide if other gadolinium chelates are available. Performance of hemodialysis after gadolinium exposure and again the following day would seem to be a logical approach. Unfortunately, there are no data to support that dialysis will prevent development of NSF, making avoidance of exposure more critical for the time being. The recommendations for gadolinium use in patients with AKI and stages 4 and 5 CKD (including patients with chronic allograft nephropathy and those awaiting pre-emptive renal transplantation) who are not on dialysis are similar. One possible exception is the use of hemodialysis following gadolinium exposure in patients who are not on dialysis and would not ordinarily have vascular access for this procedure. This is a difficult decision as there are no data to guide therapy in this large group of patients. Physicians will have to assess risk and make avoidance and dialysis recommendations on a case-by-case basis.

An effective therapy for NSF is not currently available. Aggressive physical therapy and pain management help reduce joint contractures and maintain mobility in some patients. Restoration of renal function with kidney transplantation may reverse early NSF or stabilize chronic disease if excellent transplant renal function is achieved. Short of renal transplantation, numerous case reports note some benefit with a variety of treatments, which include pentoxyfilline, steroids, plasmapheresis, high dose intravenous immunoglobulin and extracorporeal photopheresis.³ None are, however, documented conclusively as a proven therapy that will reverse or stabilize disease. Cautious optimism for intravenous sodium thiosulfate (STS) as a therapy is garnered from a single case report that describes an excellent response to this medication.⁶ Anecdotal experience with intravenous STS therapy for NSF at our institution has been generally positive, but not all patients respond dramatically. Because STS is both a chelator and an antioxidant, it may work in NSF by chelating Gd3+ from tissues and modifying the “endothelial dysfunction” that is present in patients with underlying kidney disease. The best therapy for NSF is prevention; limiting gadolinium exposure in at risk patients is a logical goal for the present time. This may not always be an acceptable option, and thus we suggest that those involved with the use of gadolinium-based contrast in high-risk patients work in close conjunction with nephrologists.

• Email this page