By Dr. Oliver Jones
- Fabry disease is characterised by increased risk of cardiovascular events including myocardial infarction, heart failure, and sudden cardiac death, but evidence-based risk predictors are lacking
- T1 dispersion, a novel CMR biomarker, combined with age and left ventricular mass index, was shown to accurately predict risk of a composite cardiac endpoint on internal validation
- The proposed novel risk prediction tool could be easily integrated into current clinical pathways
Fabry disease and cardiac risk
Fabry disease, an X-linked inherited deficiency of the lysosomal enzyme alpha-galactosidase, leads to accumulation of globotriaosylceramide in a variety of tissues. Cardiac features include left (and right) concentric, asymmetric, and eccentric ventricular hypertrophy; coronary artery disease; aortic root dilatation; aortic and mitral regurgitation; and atrial and ventricular tachyarrhythmias and conduction system defects.
Fabry disease patients are also at increased risk for adverse cardiovascular events, including myocardial infarction, heart failure, and sudden cardiac death. While males are typically more severely affected than females, evidence to support individualised risk prediction is lacking.
T1 dispersion: a novel magnetic resonance imaging biomarker
BCS Conference 2022 Young Investigators Award Competition Winner, Dr Christopher Orsborne, presented exciting steps towards developing a novel risk prediction model for adverse cardiac outcomes in Fabry disease.
The proposed model comprised a function of age, left ventricular mass index, and a novel cardiac magnetic resonance (CMR) biomarker known as native myocardial T1 dispersion.
T1 dispersion was derived using a custom script written in Python, which extracted the T1 times from all voxels within a circumferential region of interest in the middle third of myocardium in the basal and mid ventricular short axis T1 maps. The degree of variation in these values determined the T1 dispersion, which was hypothesised to reflect disease severity in Fabry by reflecting the opposing effects on T1 time observed during early and late disease.
A novel risk prediction tool
In his study, 200 patients were selected from an initial screening cohort of 232 patients. These 200 patients underwent CMR and were subsequently followed-up over a median period of 4.5 years – none were lost to follow-up. A composite outcome combined first hospitalisation for heart failure, myocardial infarction, coronary revascularisation, clinically significant arrhythmias, and cardiovascular death. 43 of the 200 patients experienced the composite outcome, with 38 of these events represented by arrhythmia (24 instances of non-sustained ventricular tachycardia, eight of new atrial fibrillation, and six of bradycardia resulting in a pacemaker).
Candidate predictor variables were initially identified according to clinical practice and literature review. These were then selected using the 10 events per variable method, then the predictor variables were analysed by univariable then multivariable Cox-non-proportional hazards regression modelling, to generate two competing multivariable risk prediction models.
On internal validation, the better-performing of these models produced a Harrel’s C-statistic of 0.77 – a value of 0.5 indicates no discrimination, whilst a value of 1 indicates perfect discrimination. For reference, Dr Orsborne compared this to a value of 0.70 for the European Society of Cardiology 2014 hypertrophic cardiomyopathy sudden cardiac death risk calculator.
A Kaplan-Meier curve of patients stratified into three risk groups based on the risk prediction demonstrated excellent separation: event-free survival at 6 years was roughly >95%, 70%, and 50% respectively, with p <0.0001.
In his presentation, Dr Orsborne highlighted many potential benefits of the proposed tool, including the ease of integration into current clinical pathways, the comparable performance in both males and females, and the lack of requirement for contrast, which can lead to complications in the context of Fabry-related renal impairment.
Acknowledging limitations including the relatively small cohort size and the associated need for a composite outcome, Dr Orsborne pointed towards the need for further study including external validation in a separate cohort.
- Germain DP. Fabry disease. Orphanet journal of rare diseases. 2010 Dec;5(1):1-49.
- O’Mahony C, Jichi F, Pavlou M, Monserrat L, Anastasakis A, Rapezzi C, Biagini E, Gimeno JR, Limongelli G, McKenna WJ, Omar RZ. A novel clinical risk prediction model for sudden cardiac death in hypertrophic cardiomyopathy (HCM risk-SCD). European heart journal. 2014 Aug 7;35(30):2010-20.