Gadolinium and NSF
What is fact and what is theory?

Adrian Holden
Medical Director Europe/International,
GE Healthcare Medical Diagnostics

Address for correspondence:
Dr Adrian Holden, MB, BS, MFPM
GE Healthcare Medical Diagnostics, Amersham Place,
Little Chalfont, Bucks, HP7 9NA, UK
Tel: +44-(0)1494-542069 Fax: +44-(0)1494-542600
Email: adrian.holden@ge.com

Abstract
Since the publication by Grobner and colleagues in 2006 that first suggested the association between gadolinium-based contrast agents (GBCA) and nephrogenic systemic fibrosis (NSF), there have been many articles published on the subject.

However, there is a lack of clinical evidence to support a number of the theories being put forward regarding differential risk between GBCA in those patients at risk of developing the condition. To date, there is no definitive evidence demonstrating that any gadolinium product is safer than any other with respect to NSF. Nor is there any evidence that the gadolinium complex stability, especially when measured in vitro in a nonphysiological environment, is relevant to NSF causality.

While NSF is a serious condition that needs to be approached with the utmost care, radiologists need also to consider the overall safety profile and effectiveness of agents for that majority of patients where there is no evidence of a risk of NSF. Consideration of adverse events and general patient tolerability for various agents, as well as differences in licensed indications, need to be fully understood so that a balanced decision is made and the appropriate agent for the appropriate patient group is selected.

This article addresses a number of the theories that have been proposed with respect to NSF, covers some of the data that has been collected through GE Healthcare's pharmacovigilance reporting systems and asks where we, as a medical community, are with respect to our understanding of this increasingly rare condition.

Introduction
Since the association between gadolinium-based contrast agents (GBCA) and nephrogenic systemic fibrosis (NSF) was first suggested in 2006,[1] a number of papers have been published on the subject. Awareness has increased, but definitive clinical evidence about this association is still lacking and much is still unknown about NSF and the role of GBCA in its pathophysiology.

A great deal of the discussion in the published case reports, as well as the justification for much of the guidance around this subject, remains theory and speculation. Other than avoiding the use of gadolinium agents - especially at high dose - in patients at risk of the condition, there is uncertainty about what can be done to prevent this potentially serious condition. The rarity of the disorder, the retrospective nature of most case reports and the relatively short time since an association with gadolinium has been proposed all contribute to the challenges the medical community faces when trying to understand this condition.

Any discussions on the subject of NSF should take place against the backdrop of the excellent safety record of linear GBCA in more than 130 million administrations over a period of more than 20 years and the fact that they have contributed to the early diagnosis and subsequent treatment of a vast number of patients.

The facts
It seems clear that NSF has only been documented in patients with acute or severe renal impairment, particularly those with end-stage renal disease on dialysis. It also seems to be clear that repeated or higher than recommended doses of a GBCA may increase the risk for NSF[2].

In Europe, both gadodiamide (Omniscan®) and gadopentetate dimeglumine (Magnevist®) are contraindicated in patients with severe renal insufficiency and there are now warnings and precautions in place for all available GBCA about use in patients with renal impairment.

However, no causal link between gadolinium and NSF has been proven and there is no clear evidence that any of the available GBCAs are safe in the at-risk population.

This has led the Food and Drug Administration (FDA)[3] to advise clinicians to:

• Avoid use of GBCAs unless the diagnostic information is essential and not available with non-contrast enhanced magnetic resonance imaging (MRI).
• Screen all patients for renal dysfunction by obtaining a history and/or laboratory tests.
• Avoid the use of doses of GBCA exceeding the dose recommended in product labelling. Allow sufficient time for elimination of the GBCA prior to any re-administration.

The apparent effect of changing protocols
It is always challenging to draw conclusions from data collected through a voluntary spontaneous reporting system. However, taking these considerations into account, it appears that the FDA guidance has resulted in a decrease in the number of new cases of NSF being reported since the association was first suggested. Figure 1 shows the numbers of cases of NSF that have been reported to or identified by GE Healthcare through literature and other sources as associated with gadodiamide by year of diagnosis (excluding patients for whom the date of diagnosis is not available). The decrease in the number of cases from 2006 to 2007 is greater than can be explained by any overall decrease in the use of gadodiamide in the same period. As of the end of January 2009, there have been no reports of new cases associated with Omniscan with a documented date of onset after July 2007, a period of over 18 months.

A similar trend is seen in the number of NSF cases associated with gadopentetate dimeglumine[4], supporting the view that the decrease in the number of new cases is the result of changes in clinical practice.

Potential flaws in the foundation of some current theories
Much of the currently published advice is based on differences in the incidence of spontaneous reports of NSF associated with different GBCA and attempts have been made to show that the risk of NSF varies between products. Some authors have chosen to rely only on published data, arguing that these data are of higher quality for comparison purposes than reports to regulatory authorities such as the FDA[5]. However, if a physician reports a case, it is because they believe that it has occurred, regardless of whether it is subsequently published or not. The FDA clearly states that reports have been received of NSF associated with all GBCA available in the USA during the period when most cases were reported.

Analysis of spontaneous reports in isolation can lead to misleading conclusions and the numbers of cases reported should be considered in relation to the market share and usage of the various products over the relevant period. For example, gadopentetate dimeglumine and gadodiamide were the most commonly used agents in Europe and the USA during the time in which NSF cases have occurred. Therefore, it is not surprising that the majority of cases have been linked to these products. It is important to consider the patterns of use of the various products in those patients at greatest risk of NSF, and the doses used:

• How long has a product been available, and was it widely used between 2002 and 2006, when most cases of NSF occurred?
• Was the product available in those markets where most cases were reported?
• Is the product licensed for angiography or whole body imaging, the procedures most likely to be undertaken in patients with renal insufficiency and to involve the use of GBCA at higher doses?
• Does the product have a pre-existing contraindication in patients with severe renal insufficiency in any of the major markets, thus potentially limiting any historical exposure to patients at greatest risk of NSF?

Figure 2. The geographical divide in reported cases (as of 31 July 2008).

As an example, Figure 2 shows that as of the end of July 2008, the majority of cases associated with gadodiamide, as reported to GE Healthcare, have occurred in the USA. Under these circumstances, it might be expected that GBCA that are not licensed or used in the USA, such as gadoteric acid (Dotarem), gadofosveset trisodium (Vasovist) and gadobutrol (Gadovist) would be associated with fewer cases.

It should also be recognised that in Germany, the largest user of GBCA in Europe, gadoteric acid was contraindicated in patients with severe renal impairment, i.e. was not approved for use in patients at greatest risk of NSF, prior to 2008 when the vast majority of cases occurred.

The majority of cases of NSF have occurred in clusters from a relatively small number of healthcare institutions. For example, data from the Danish Medicines Agency in 2008 showed that in Denmark, which has reported the second largest number of cases after the USA, 33 of the 36 cases that the agency were then aware of had been reported from a single centre (Herlev Hospital).[6] In January 2009, further reports of NSF cases in Danish institutions, related to a variety of GBCA, have been received, although the details of these cases are still being clarified.

In the USA, GE Healthcare has received 356 reports of NSF; 196 of these have been received from law firms and some of these appear to duplicate reports from other sources. Of the healthcare professional and literature reports for which the US administering facility is known, a total of 156 cases of Omniscan-associated NSF have been received from 45 imaging centres since NSF was first linked to gadolinium-based contrast agents in 2006. Of these, 9 of 45 (20%) centres that each reported more than 5 cases account for 98/156 (63%) cases; of the approximate 2100 US imaging centres that used Omniscan during the period in which NSF emerged, 98% have not reported any cases of NSF.

The relevance of this clustering has yet to be adequately explained, but serves as an example of the lack of understanding that exists about the occurrence of NSF in patients following GBCA. Only a minority of sites that have used gadodiamide historically have reported any cases of NSF associated with it.

Potential for reporting bias
The relatively high number of reports of NSF associated with gadodiamide may, in part, be due to reporting bias due to greater physician awareness of the concerns around this product. This may stem from the fact that GE Healthcare was the first GBCA manufacturer to inform healthcare professionals about the association between NSF and gadodiamide and to actively solicit reports of cases. This could potentially have led to 'hindsight bias', where medical notes of patients who had received gadodiamide but not other products were selectively reviewed following a "Dear Doctor" letter or media coverage, leading to the reporting of an adverse event.

In order to estimate the true incidence of NSF associated with different GBCAs, it would be necessary to know the actual number of NSF cases associated with each agent, as well as the number of patients at risk for NSF who were exposed to each specific GBCA; these figures are not known for any GBCA.

Of note is the fact that the American College of Radiology (ACR) have recently updated their guidelines with respect to NSF[7] and state that "it must be emphasised that the frequency with which NSF has been associated with different gadolinium-based contrast media (GBCM) may reflect a combination of differences in agent toxicity and market share. In addition, reported frequency is also likely further confounded by the fact that some agents, particularly gadodiamide, may have been used disproportionately more frequently in patients receiving high doses of GBCM for MRA."

The stability hypothesis
In attempting to explain the difference in reported numbers between the various GBCA, many authors have noted that the majority of cases are with gadodiamide or gadopentetate dimeglumine, which are both linear molecules. Some have extrapolated from this to argue that the risk of NSF in patients with severe renal impairment is greater with linear GBCA than with macrocyclic agents. The rationale for this argument is largely based on molecular stability, leading to the hypothesis that the stability of gadolinium agents is a factor in the development of NSF, a theory as yet not supported by any clinical evidence.

If gadodiamide were prone to dechelation in vivo, it would be expected that the resulting metabolites would be detectable, and that the distribution and retention kinetics would be similar to free gadolinium metal. However, several studies have shown that there is no detectable metabolism of the injected chelates, even in patients with prolonged retention due to renal impairment[8,9].

Table 1. The thermodynamic stability constant of several GBCAs.

Theories regarding stability are largely based on comparison of the thermodynamic stability constant (K_therm) of the compounds, which is measured at pH 11, and does not reflect physiological conditions.

Although there are differences in (K_therm) (as shown in Table 1), the scale is logarithmic and it can be concluded that all GBCA are extremely stable even at pH 11. Hence it is inaccurate to describe any of the GBCA as "unstable", a description that has been used by some authors in recent publications.

Table 2. The conditional stability constant of several GBCAs.

As shown recently by Ersoy[10], the stability constants for different GBCA become increasingly similar when measured under more physiological conditions. The conditional stability constants (K_cond), which reflect the calculated stability at pH 7.4, are shown in Table 2.

It is worth noting that the drop in stability from thermodynamic to conditional constants is much greater for the macrocyclic GBCA (gadoterate meglumine and gadoteridol) than for gadodiamide. The stability of gadoterate meglumine and gadoteridol, for example, decreases approximately 10⁶ times compared to 10² times for gadodiamide. It is unknown whether this trend continues as conditions become more physiological, for example, where other cations are present, such as sodium, calcium, potassium, zinc, etc.

As no comparative in-vivo stability constants are available, it is impossible to know how the products compare in the truly physiological setting. This is one of a number of areas of research that GE Healthcare is pursuing to better understand the real differences between the available GBCA in clinical practice and to gain further insight into the association with NSF.

Judging by the conditional stability constant, gadopentetate dimeglumine - a linear GBCA for which 314 cases of NSF have been reported according to the manufacturer's website (as of 31 July 2008) - is more stable than gadoteridol, a macrocyclic GBCA. This raises questions about the simplistic message that macrocyclic GBCA are safer than linear ones with respect to NSF risk due to their greater stability, especially as there is no evidence that stability is a factor in NSF. Any discussion regarding risk associated with possible instability should also address a number of as yet unanswered questions:

• How can we explain the small, but increasing, number of reports of cases of NSF with no prior exposure to GBCA?
• Why do approximately 97% of patients with endstage renal disease not develop NSF when exposed to gadolinium agents, even in those centres with the highest incidence of NSF?
• Why are there no reports of symptoms of general gadolinium toxicity, as might be expected if this toxic metal was being released in the bodies of patients developing this condition?
• How can we explain the confirmed and non-confounded cases of NSF associated with gadoteridol, which is a macrocyclic compound and therefore not thought to be associated with the release of free gadolinium?

It is also important that we do not dismiss confounded cases as irrelevant. For example, in one of the eight confounded cases reported as associated with gadoterate meglumine on the manufacturer's website, the patient developed NSF in 2005, the same year that they received two macrocyclic agents, gadoterate meglumine and gadobutrol. However, the only exposure listed to a linear product, gadopentetate dimeglumine, was in 1999 and 2000. If it is suggested that the stability of GBCA plays a role in the development of NSF, then it is also important to consider the potential role of the various GBCA administered over the various timescales involved.

New evidence
New evidence is emerging all the time and recent work by both Edward[11] and Wermuth[12] raises further questions about the role of GBCA stability in NSF. The pathophysiology of NSF is not yet understood, but one theory is that the condition results from effects on circulating fibrocytes[13]. Edward and colleagues showed that gadolinium chloride (free gadolinium) did not stimulate circulating fibrocytes, which would lead to fibrosis, whereas the chelated form of a GBCA did. The results of further similar studies with a range of GBCA are awaited with interest. In addition to Edward's work, Jimenez and colleagues have shown that gadodiamide and gadopentetate dimeglumine both stimulated the release of profibrotic mediators (IL-6, IL-13, TGF- and VEGF) from human monocytes, whereas free gadolinium (GdCl3) only stimulated TGF-ß and to a lesser effect than gadopentetate dimeglumine or gadodiamide.

A recent paper by Prince and colleagues[14] examined data from 31 confirmed cases of NSF diagnosed at 2 major institutions in the USA. They found no statistically significant difference in the incidence of cases between the various GBCA used at the institutions and went on to present a number of "advances in knowledge" that include:

• More patients with NSF had pro-inflammatory events and, compared with patients who did not have NSF, the NSF group had lower pH, younger age, lower eGFR, elevated serum phosphorus levels, and a longer delay between GBCA injection and haemodialysis.
• For patients with an eGFR lower than 15 mL/min, haemodialysis after GBCA administration was associated with a significant reduction (p<0.05) in the incidence of NSF; no patient who underwent haemodialysis the same day as the injection developed NSF.
• There were no cases of NSF with a standard dose (0.1 mmol/kg) of GBCA in 74,124 patients examined without being screened for renal function and when gadodiamide was used in 86% of patients.
• There was a lower incidence of NSF (0 in 5725 patients) with multiple 0.1 mmol/kg doses of GBCA administered across multiple examinations compared with the incidence with a single high dose (9 in 5119) patients.

The study therefore raises questions about the differential risk between products and the effects of cumulative dose versus high single dose.

Gadolinium retention
Another area where opinions have been expressed without strong supporting clinical evidence is the differential retention of GBCA in the tissues of the body, with speculation about a possible cumulative dose effect.

Gibby[15] and White[16] reported that more gadolinium was detectable in human bone following gadodiamide administration compared to gadoteridol - although both agents resulted in measurable levels of bone retention - and interpreted this as evidence of in-vivo dechelation of gadodiamide. However, no adjustment was made for the different time intervals between administration of the contrast agents and analyses. The mean interval between administration and analysis was shorter for the gadodiamide group which could result in bias. Furthermore, the analytical method used detects only the gadolinium ion, and cannot distinguish between the intact gadolinium complex and uncomplexed, 'free' gadolinium. Hence, it cannot be concluded that this is evidence of gadolinium release from the chelate. Finally, the relative amounts of retained gadolinium do not correlate with stability constants, or with the numbers of reported cases of NSF. Consequently, the results of such studies, and any conclusions drawn from them, should be treated with caution.

The research by Gibby[15] did not investigate long-term retention of gadolinium in the tissues. If gadolinium retention is a causal factor in NSF, it may be important to collect clinical information over longer time periods than in this study. Since gadolinium was detected in patients that received a macrocyclic agent, this is not a discussion that should be limited by the structure of the compound.

Evidence-based medicine
Definitive clinical evidence is lacking to support any of the theories being proposed to explain a possible differential risk between GBCA in patients at risk. As in all fields of medicine, it is important that clinical decisions are based on sound clinical evidence. To date, there is no evidence to show that any gadolinium product is safer than any other with respect to NSF. Nor is there any evidence that the stability of the gadolinium complex, especially when measured in vitro under non-physiological conditions, is relevant to the pathogenesis of NSF. However, there is no doubt that the patients with acute renal failure or severe renal impairment, especially end-stage renal disease, are at risk for NSF. The use of relatively large (>20 ml) volumes of contrast appears to increase the risk.

Future research
As a responsible manufacturer (and contrary to the suggestion in a previous publication[17]), GE Healthcare will continue to pursue pre-clinical and clinical studies with external collaborators to understand better the science behind NSF and to help ensure patient safety. As much information as possible has been collected about every case of NSF reported to the company's pharmacovigilance team, and discussions continue with experts to ensure that understanding within the company and within the medical community at large, is as complete and balanced as possible.

Since knowledge of NSF and its association with GBCA is far from complete, caution should be exercised in patients at risk of this rare, but potentially serious, condition. The significant decrease in new cases of NSF reported since the end of 2006 associated with either gadodiamide or gadopentetate dimeglumine suggests that a cautious approach can lead to a reduction in the risk of NSF.

Conclusions
It is recommended that gadodiamide should not be used in patients with GFR <30 ml/min/1.73m2 and there is no evidence to suggest that any GBCA is safe in this high risk group.

This reflects the latest guidelines provided by the ACR in September 2008[7]. Version 6 of the ACR Manual on Contrast Media states that, in patients with severe or end-stage renal failure, if MRI contrast administration is absolutely essential, judicious use of the lowest possible dose of selected GBCA is probably safest.

"There is no proof that any GBCM is completely safe in this patient group."

Reassuringly, these latest guidelines also state that with respect to patients with stage 1 or 2 renal failure, there is no evidence of an increased risk of developing NSF and that all GBCM can be used as long as no more than a maximum dose of 0.1 mmol/kg is used.

While NSF is a serious condition, radiologists must also consider the overall safety profile and effectiveness of GBCA for the large majority of patients in whom there is no evidence of a risk of NSF. The adverse events and general patient tolerability for various agents, as well as the differences in licensed indications, need to be fully understood so that a balanced decision is made and the appropriate agent is selected for each patient group.



References

1. Grobner T. Gadolinium - a specific trigger for the development of nephrogenic fibrosing dermopathy and nephrogenic systemic fibrosis? Nephrol Dial Transplant 2006;21:1104-8.
2. Wiginton CD, Kelly B, Oto A, et al. Gadolinium-based contrast exposure, nephrogenic systemic fibrosis, and gadolinium detection in tissue. AJR Am J Roentgenol 2008;190:1060-8.
3. FDA Alert. Gadolinium-based contrast agents for magnetic resonance imaging (marketed as Magnevist, MultiHance, Omniscan, OptiMARK, ProHance). Rockville, MD, USA: US Food and Drug Administration 2007 [cited 2008 October 9] Available from: http://www.fda.gov/Cder/Drug/ InfoSheets/HCP/gcca_200705.htm
4. Bayer Healthcare Pharmaceuticals. Magnevist Safety Information. 2008 April [cited 2008 October 24] Available from: http://www. imaging.bayerhealthcare.com/html/magnevist/nsf_nfd.html
5. Broome DR. Nephrogenic systemic fibrosis associated with gadolinium based contrast agents: A summary of the medical literature reporting. Eur J Radiol 2008; 66:230-4.
6. Lægemiddelstyrelsen. Redegørelse om kontrastmidlet Omniscan®. 2008 March [cited 2008 October 13] Available from: http://www. sum.dk/IMEVERest/Publications/imdk%20x2D%20dansk/Redeg_ omniscan/20080328103545/CurrentVersion/Redeg_%20om_%20 Omniscan.pdf
7. American College of Radiology. Manual on Contrast Media Version 6 2008. 2008 [cited 2008 October 13] Available from: http://www.acr.org/ SecondaryMainMenuCategories/quality_safety/contrast_manual.aspx
8. Joffe P, Thomsen HS, Meusel M. Pharmacokinetics of gadodiamide injection in patients with severe renal insufficiency and patients undergoing hemodialysis or continuous ambulatory peritoneal dialysis. Acad Radiol 1998;5:491-502.
9. Normann PT, Joffe P, Martinsen I, et al. Quantification of gadodiamide as Gd in serum, peritoneal dialysate and faeces by inductively coupled plasma atomic emission spectroscopy and comparative analysis by highperformance liquid chromatography. J Pharm Biomed Anal 2000;22:939-47.
10. Ersoy H, Rybicki FJ. Biochemical safety profiles of gadolinium-based extracellular contrast agents and nephrogenic systemic fibrosis. J Magn Reson Imaging 2007;26:1190-7.
11. Edward M, Quinn JA, Mukherjee S, et al. Gadodiamide contrast agent 'activates' fibroblasts: a possible cause of nephrogenic systemic fibrosis. J Pathol 2008;214:584-93.
12. Wermuth PJ, Del Galdo F, Jiménez SA Induction of expression of profibrotic/proinflammatory proteins in normal human peripheral blood mononuclear cells (PBMC). Role in pathogenesis of gadolinium-associated nephrogenic systemic fibrosis. Ann Rheum Dis 2008;67(Suppl II); 366.
13. Bucala R. Circulating fibrocytes: cellular basis for NSF. J Am Coll Radiol 2008;5:36-9.
14. Prince MR, Zhang H, Morris M, et al. Incidence of nephrogenic systemic fibrosis at two large medical centers. Radiology 2008;248: 807-16.
15. Gibby WA, Gibby KA, Gibby WA. Comparison of Gd DTPA-BMA (Omniscan) versus Gd HP-DO3A (ProHance) retention in human bone tissue by inductively coupled plasma atomic emission spectroscopy. Invest Radiol 2004;39:138-42.
16. White GW, Gibby WA, Tweedle MF. Comparison of Gd(DTPA-BMA) (Omniscan) versus Gd(HP-DO3A) (ProHance) relative to gadolinium retention in human bone tissue by inductively coupled plasma mass spectroscopy. Invest Radiol 2006;41:272-8.
17. Thomsen H, Marckmann P. Extracellular Gd-CA: Differences in prevalence of NSF. Eur J Radiol 2008;66:180-3.


01-2009 BUY1163892/JB3484/CMC 17th edition

 

Source: C2I2, Volume VI, Issue 2, 2008

Read also - related articles