The addition of genetic testing and cardiovascular magnetic resonance to routine clinical data for stratification of etiology in dilated cardiomyopathy

Abstract

Background: Guidelines recommend genetic testing and cardiovascular magnetic resonance (CMR) for the investigation of dilated cardiomyopathy (DCM). However, the incremental value is unclear. We assessed the impact of these investigations in determining etiology.

Methods: Sixty consecutive patients referred with DCM and recruited to our hospital biobank were selected. Six independent experts determined the etiology of each phenotype in a step-wise manner based on (1) routine clinical data, (2) clinical and genetic data and (3) clinical, genetic and CMR data. They indicated their confidence (1-3) in the classification and any changes to management at each step.

Results: Six physicians adjudicated 60 cases. The addition of genetics and CMR resulted in 57 (15.8%) and 26 (7.2%) changes in the classification of etiology, including an increased number of genetic diagnoses and a reduction in idiopathic diagnoses. Diagnostic confidence improved at each step (p < 0.0005). The number of diagnoses made with low confidence reduced from 105 (29.2%) with routine clinical data to 71 (19.7%) following the addition of genetics and 37 (10.3%) with the addition of CMR. The addition of genetics and CMR led to 101 (28.1%) and 112 (31.1%) proposed changes to management, respectively. Interobserver variability showed moderate agreement with clinical data (κ = 0.44) which improved following the addition of genetics (κ = 0.65) and CMR (κ = 0.68).

Conclusion: We demonstrate that genetics and CMR, frequently changed the classification of etiology in DCM, improved confidence and interobserver variability in determining the diagnosis and had an impact on proposed management.

Keywords: cardiac magnetic resonance imaging; dilated cardiomyopathy; etiology; genetic testing; precision stratification.

Figure 1

Derivation of cohort. A graphic highlighting the process used to derive the patient cohort used in the study. The 17 patients that were not screened for eligibility were only excluded because the 60 patient threshold desired for the patient cohort had already been met. Clinical information was deemed insufficient where more than one section of the basic investigations could not be completed such as no history of presenting complaint and no blood results.

Figure 2

Ring-like pattern of fibrosis on cardiovascular magnetic resonance. A patient with dilated cardiomyopathy secondary to a truncating variant in DSP, with ring-like late gadolinium enhancement on CMR.

Figure 3

Sankey diagram of change in etiology by investigation. A Sankey diagram illustrating the number of times etiology was changed when genetic testing and subsequently CMR results were added to the information provided by the basic investigations, where N = 360 each time an investigation was added. Etiologies of dilated cardiomyopathy were grouped into idiopathic, genetic/familial, toxic (comprised of iron overload, recreational drugs, anti-neoplastic drugs, other medications, and alcohol), inflammatory (comprised of viral myocarditis and autoimmune), peripartum, and other (comprised of other, nutritional deficiency, and hypothyroidism).

Figure 4

Comparison of confidence scores. Confidence scores were recorded each time the expert was asked to diagnose etiology, resulting in N = 360 confidence scores at each set of investigations. Numbers in each color-coded box reflect the percentage frequency of each score.