Revisiting the risks of MRI with Gadolinium based contrast agents—review of literature and guidelines

In the year 2000, 15 patients with chronic kidney disease were identified presenting with scleromyxoedema-like cutaneous manifestations yet having significant clinical and histo-pathological differences; the term nephrogenic fibrosing dermopathy was initially proposed. These clinical findings are now recognized as characteristic of NSF [4]. Following this, case reports of similar findings and also significant systemic involvement found on autopsy were reported [5, 6]. Patients with end-stage renal disease were reported to develop symptoms as early as two to four weeks after exposure to GBCAs for MRI [1]. Exact pathogenesis remains unclear; however, postulation of likely early dermal manifestation of this gadolinium toxicity is proposed [7]. A strong association is observed in the presence of both acute renal impairment and chronic dialysis dependent renal insufficiency and other influencing co-factors that may play a role, such as a background inflammatory process [8, 9]. As the evidence in published literature increased, the United States Food and Drug Administration (FDA), followed by the European Medicines Agency (EMA) issued an alert on the use of GBCAs in patients with renal insufficiency [10, 11]. Since then, public health and practice guideline bodies have published recommendations on its use [12‐14]. A 2008 multi-centre retrospective review reported 15 cases of NSF in a total population of 83,121 (0.02 % incidence), all of whom received at least one administration of a GBCA. All 15 of these cases of NSF were found in patients who had received a higher than standard dose, increasing the incidence to 0.17 % (15 of 8997 patients). A higher than normal dose was described as approximately between 0.2 to 0.4 mmol per kilogram body weight. In the entire cohort, 265 patients were on haemodialysis, but only one of them was reported to have developed NSF (incidence 0.4 %) [3]. Another publication retrospectively collating data from four centres set to determine the benchmark incidence of NSF related to the confirmed use of two GBCAs [15]. They reported an overall incidence of 0.04 % at two centres that used Gadodiamide (32 cases in 82,260 patients—administered total dose range 1 to 9.5 mmol), as compared to the 0.02 % from the previously publish study. The other two centres that used Gadopentate dimeglumin reported an incidence of 0.003 % (four cases in 135,347 patients) with an administered dose ranging between 2.5 and 8.5.

As reported in the literature, specific cutaneous findings on clinical examination with relevant past history of GBCA exposure trigger a differential of NSF, but require histological confirmation [16‐18]. It has been postulated that the deposition of disassociated free gadolinium causes this fibrous connective tissue formation [5]. Patients may present with firm, erythematous and indurated plaques of skin associated with subcutaneous oedema. The presentation may range from hyperpigmentation, yellow papules or plaques, blistering or even ulceration [1, 2, 8]. Resultant manifestations include pain, severe pruritus, paraesthesia and flexion contractures that can begin on the hands or feet and extend proximally. Cutaneous calcifications maybe noted on a plain film radiograph and confirmed on biopsy [19]. Lesions are frequently symmetrical, often located on the lower limbs, followed by the forearms. Idiopathic, rapid onset, unstable hypertension has been described prior to onset of skin lesions. Its systemic involvement of lungs, heart, diaphragm, liver or kidneys can vary. The international centre for research on NSF, led by Prof. Dr. S.E. Cowper, states that approximately 5 % are reported to have a fulminant course [20]. The Girardi Score (Fig. 1, Table 1) was proposed in 2011 based on reported clinical presentations and expert consensus, as no single laboratory test could be used as a gold standard to diagnose NSF. This encompassed identification of major and minor criteria on clinical findings, coupled with histological findings.

Recent theory suggests that administration of large amounts of GBCAs, (solely excreted via kidneys for earlier used formulations and again mainly excreted by kidneys in the remainder) persist in the body and may dissociate from their carrier ligands/chelates [16, 21, 22]. These may then bind with readily available phosphates, carbonates or citrates, and form insoluble molecules. Several authors mention histological findings of increased dermal collagen bundles, CD34+ fibroblast like cells, macrophages, mucin and transforming growth factor beta (TGFβ) in this cohort of patients [2, 4, 5, 16, 23‐25]. Pre-existing renal disease has been the most prevalent patient characteristic.

Although the disassociated gadolinium theory has been the widely acknowledged trigger of NSF, questions of chelated gadolinium in combination with pre-existing cofactors have been raised [26]. Cofactors such as high dose erythropoietin treatment, pro-inflammatory state, high serum phosphate and calcium, and absence of ace inhibitor treatment, have also been linked to the appearance of NSF [27‐30]. Chelated gadolinium such as gadodiamide and gadopentetate have been shown to directly stimulate macrophages and monocytes in vitro to release profibrotic cytokines and growth factors capable of initiating and supporting the characteristic tissue fibrosis [23, 31‐33].

As evidence in published literature increased following the initial June 2006 alert, manufacturers of GBCAs were ordered by the FDA to add a “black box warning” the following year [34]. This elaborated on the alert to extend caution of GBCA use in any acute or chronic renal insufficiency patient (GFR < 30 mL/min/1.73 m2), or acute renal insufficiency of any severity due to the hepato-renal syndrome or in the perioperative liver transplantation period. A 2009 systematic review of NSF in liver disease patients found no compelling evidence to suggest liver disease in itself as being a risk factor for developing NSF. The authors concluded that NSF developed only in the setting of pre-existing severe renal insufficiency, irrespective of liver disease status [35].

To date, very little data is available regarding GBCA administration in pregnancy. Most of the published studies are either animal studies or patient cohorts that are understandably historic, and have small numbers and limited follow-up periods [36‐42]. As evidence of gadolinium retention after exposure continues to pile and taking into consideration immature foetal renal function, recommendations are much more restrictive. Guidelines recommend that GBCAs should only be given to pregnant women when there is a very strong clinical indication. Breast feeding should be stopped for at least 24 hours. One of the more stable, macrocyclic gadolinium agents (gadoterate meglumine, gadoteridol, or gadobutrol) should be used in the lowest dose consistent with a diagnostic result [43].

The vast majority of GBCA preparations contain excess chelates to reduce or ensure absence of free gadolinium in the solution, and some studies have suggested the possibility of excess chelate to inhibit the collagenolytic properties of matrix metalloproteinase 1. The addition of excess chelate to non-ionic linear chelate dramatically reduces the acute toxicity [9, 37, 43]. Table 2 summates the commonly used GBCAs, their elimination pathway, reports of NSF and the amount of excess chelate within the preparations.