Background

In the last decade, cardiovascular magnetic resonance (CMR) has grown dramatically and in amyloidosis has become the gold standard for non invasive diagnosis of cardiac involvement with the characteristic subendocardial late gadolinium enhancement (LGE) [1]. The number of new referrals for cardiac amyloidosis has significantly increased in the last few years, with the diagnosis being based on CMR and the characteristic LGE. New contrast agents favoured for their high relaxivity related to blood protein binding, [2] have become available but no one has ever tested the diagnostic performance of these new agents in infiltrative disease.

Case presentation

A 77-year old man presented with breathlessness and weight loss. Endoscopy and duodenal biopsy showed amyloid deposits, subtyped as transthyretin type (TTR). Investigations excluded a plasma cell dyscrasia, and showed atrial fibrillation on ECG (Fig. 1), and moderately increased wall thickness with moderate to severe diastolic dysfunction on echocardiography (Fig. 2, videos available as Additional files 1 and 2). TTR gene sequencing showed no mutation so the diagnosis was that of wild-type ATTR amyloidosis (senile systemic amyloidosis). Cardiac involvement, though likely, was not unequivocal as age/hypertension can confound. Technetium-labelled bone scintigraphy and CMR were performed. Scintigraphy demonstrated grade 2 myocardial uptake, highly specific for cardiac amyloidosis. Cine CMR imaging supported the echocardiographic and scintigraphy findings. Tissue characterization showed clear amyloidosis with elevated native myocardial T1 (ShMOLLI, 1072 ms, normal 960 ms+/-30 ms) and, after contrast (Gadoterate meglumine, 0.1 mmol/kg, gadolinium-DOTA, Dotarem® Guerbet S.A. France), abnormal gadolinium kinetics on TI scout (nulling of the myocardium before the blood) and diffuse biventricular subendocardial LGE (Fig. 2) [1]. The extracellular volume fraction (ECV) at 15 min was 0.53.

Fig. 1
figure 1

Twelve lead ECG at presentation showing atrial fibrillation

Fig. 2
figure 2

Parasternal long axis (left) and apical four-chamber (right) views echocardiography images. Videos are available as Additional files 1 and 2

Two weeks later, the CMR was repeated (same scanner and team with no changes in the patient regarding treatment or disease status) within the screening tests of a clinical trial, this time with a different contrast agent, gadobenate dimeglumine (0.1 mmol/kg, gadolinium-BOPTA, MultiHance®, Bracco, Milan, Italy). The cines and native T1 findings were the same (Fig. 3). The LGE however was markedly different – the subendocardial LGE so characteristic of amyloidosis was not visible, even with phase-sensitive inversion recovery (PSIR). There was just non-specific right ventricle insertion point LGE (Fig. 4). The ECV at 15 min obtained using Gd-BOPTA (MultiHance®) was now lower – 0.45 compared to 0.53 with gadolinium-DOTA (Dotarem®) (same acquisition time after contrast administration). This difference is higher than the expected variability of the technique. Prior publications demonstrate the ECV to be stable to a standard deviation of 0.02 (0.03 for amyloid); [3] here a 0.08 difference with contrast agent is recorded.

Fig. 3
figure 3

Short axis ShMOLLI T1 mapping images pre-contrast (top) and post-contrast (bottom) with gadolinium-DOTA, Dotarem® (a) and with gadolinium-BOPTA, MultiHance® (b). Myocardial T1 post-contrast = 486 ms and blood T1 post-contrast =506 ms 15 min after gadolinium-DOTA, Dotarem®; and myocardial T1 post-contrast = 344 ms and blood T1 post-contrast = 287 ms 15 min after gadolinium-BOPTA, MultiHance®

Fig. 4
figure 4

CMR end-diastolic cine (a); late gadolinium images with Magnitude-FLASH (b) and late gadolinium images with PSIR-FLASH (c) after gadolinium-DOTA, Dotarem® showing the classic amyloid global, subendocardial late gadolinium enhancement pattern; late gadolinium images with Magnitude-FLASH (d) and late gadolinium images with PSIR-FLASH (e) after gadolinium-BOPTA MultiHance®, showing no features of amyloid

Discussion

Gd-BOPTA (MultiHance®) is an ionic linear chelate favoured for its high relaxivity related to blood protein binding, [2] which also makes it a partially “intra-vascular” contrast agent. Here, paired scanning performed serendipitously for a clinical trial failed to demonstrate the characteristic pattern of myocardial amyloidosis, and performed poorly compared to gadoterate meglumine. MultiHance® is widely used for CMR. This effect was unexpected – but theoretical concerns had been raised for ECV mapping with protein bound contrast agents [4]. Due to difference in relaxivity between the contrast agents, some Authors proposed the use of a lower dose of Gd-BOPTA for reaching similar T1 and ECV values. In this case the same dose was used for both contrasts and this could explain in part the differences observed. In this case, the use of MultiHance® did not affect the diagnosis, but reporting the MultiHance® CMR in isolation without the previous work-up by the National Amyloid Centre could have conceivably led to a misdiagnosis.

Conclusion

We suggest that protein bound agents for interstitial enhancement of non-ischaemic cardiomyopathy (myocardium, ECV measurement) should be used with caution whilst further work is undertaken as the diagnostic performance may not be the same as non-protein bound variants.