Comments on the case report reported by Elmholdt et al.

Sir,

We read with interest the recent report by Elmholdt et al. [1], documenting two cases of nephrogenic systemic fibrosis (NSF) following exposure to a large dose of the macrocyclic gadolinium (Gd) contrast-enhancing agent (GdCA), gadobutrol (Gadovist, Bayer-Schering, Germany). However, we have several reservations about this report including the following:

In Case 1, in which the patient developed a mild form of NSF, the authors have not provided information either about the blood chemistry of the patient (serum creatinine, phosphate, calcium and parathyroid hormone) or about medications the patient was receiving at the time of administration of the GdCA. The serum creatinine cited had been measured a month before CA administration when estimated glomerular filtration rate (eGFR) was 34 mL/min. It is not clear from the report whether the renal function was stable during this month or whether there was a progressive decline.

Case 2 was ‘confounded’ as the patient received both gadodiamide (Omniscan) (10 mL) in 2001 when renal function was normal (eGFR >60 mL/min) and gadobutrol (15 mL) in 2008 when the patient was suffering from end-stage renal disease with eGFR 11 mL/min. The patient also had an elevated serum phosphate and parathyroid hormone at the time of the last MRI examination. The patient also developed severe peritonitis a week after the gadobutrol injection. Six months later, the patient developed NSF affecting mainly his hands. Again, no information was provided about medications the patient was receiving during this period.

The authors concluded that these two cases were caused by gadobutrol but did not explore the possible role of a number of co-factors that may promote fibrosis; these include inflammation, hyperphosphataemia and medications such as erythropoietin.

In Case 2, the patient received gadodiamide 7 years prior to the administration of gadobutrol. It is feasible that some Gd had been deposited in the bone after this gadodiamide injection [2]. The delayed development of NSF could have been due to mobilization of the Gd from the bone induced by the development of hyperparathyroidism in 2008.

It would have been useful to examine the biopsy specimens for the presence of Gd to confirm that the observed NSF is due to Gd exposure.

The two patients received high doses of gadobutrol (17.5 mL in Case 1 and 15 mL in Case 2) which are equal to 35 and 30 mL, respectively, of a GdCA with a 0.5-mol/L concentration. The European Society of Uroradiology (ESUR) and regulatory authorities in Europe recommend the use of the lowest possible dose of the most stable GdCA in patients with advanced renal impairment in order to minimize the risk of NSF. This advice has not been adopted in the management of these reported cases. It may be noted that high-quality diagnostic MR angiography can now be achieved with ‘half’ of the traditional doses used in the past with modern MRI software and equipment.

It is well recognized that there are many pathways that may lead to fibrosis, the final outcome of chronic inflammatory insults of affected tissues. According to Wahba et al. [3], a mild form of NSF may develop in an absence of Gd exposure, caused by background pro-inflammatory and pro-fibrotic conditions. It was also recognized as far back as the early 1980s that lanthanides promote fibrillogenesis [4,5]. More recently, it has been shown experimentally that Gd can stimulate the proliferation of human fibroblasts and accumulation of collagen in vivo and in vitro [6,7]. This effect was evident with low-stability, non-ionic linear chelates but was absent in vivo with the stable macrocyclic agents [6] and required a massive dose of the non-ionic macrocyclic agent gadoteridol (ProHance, Bracco, Italy) in vitro to induce a stimulatory effect on fibroblasts [7].

The authors’ conclusion that macrocyclic agents have similar potential to induce NSF as the linear chelates is misleading and is not supported by a very large body of available clinical and experimental data [8]. The suggestion that the low prevalence of NSF with the macrocyclic agents is due to a lower share of the market is also misleading. In France, where the ionic macrocyclic agent gadoterate (Dotarem, Guerbet, France) has been the most commonly used GdCA for the last 20 years, they have not observed a single case of NSF in spite of using this agent in patients with end-stage renal disease [9]. In the USA, a centre, which used only the macrocyclic agent gadoteridol (ProHance, Bracco, Italy) in dialysis patients with ESRD, also did not find any cases of NSF [10].

In summary, the two cases reported by the authors do not constitute conclusive evidence that there is a significant risk of NSF after exposure to macrocyclic GdCA.

Conflict of interest statement. S.K.M. received a research grant from Guerbet, France and lecture fees from Guerbet, France and Bracco, Italy, and is an expert witness in NSF litigations in the USA. P.D. had nothing to declare.