Cardiac sympathetic innervation and mortality risk scores in patients with heart failure

In the risk stratification and selection of patients with heart failure (HF) eligible for implantable cardioverter-defibrillator (ICD) therapy, ¹²³ I-meta-IodineBenzylGuanidine (¹²³ I-mIBG) scintigraphy has emerged as an effective non-invasive method to assess cardiac adrenergic innervation. Similarly, clinical risk scores have been proposed to identify patients with HF at risk of all-cause mortality, for whom the net clinical benefit of device implantation would presumably be lower. Nevertheless, the association between the two classes of tools, one suggestive of arrhythmic risk, the other of all-cause mortality, needs further investigation.

To test the relationship between the risk scores for predicting mortality and cardiac sympathetic innervation, assessed through myocardial ¹²³ I-mIBG imaging, in a population of patients with HF.

In HF patients undergoing ¹²³ I-mIBG scintigraphy, eight risk stratification models were assessed: AAACC, FADES, MADIT, MADIT-ICD non-arrhythmic mortality score, PACE, Parkash, SHOCKED and Sjoblom. Cardiac adrenergic impairment was assessed by late heart-to-mediastinum ratio (H/M) <1.6.

Among 269 patients suffering from HF, late H/M showed significant negative correlation with all the predicting models, although generally weak, ranging from -0.15 (p = .013) for PACE to -0.32 (p < .001) for FADES. The scores showed poor discrimination for cardiac innervation, with areas under the curve (AUC) ranging from 0.546 for Parkash to 0.621 for FADES.

A weak association emerged among mortality risk scores and cardiac innervation, suggesting to integrate in clinical practice tools indicative of both arrhythmic and general mortality risks, when evaluating patients affected by HF eligible for device implantation.