Mid-regional pro-adrenomedullin: a new tool in prognosticating asymptomatic severe aortic stenosis?

Aortic stenosis (AS) is characterised both by progressive valve narrowing and the remodelling response of the left ventricle (LV) that occurs secondary to an increased afterload. The latter is of particular importance when considering the development of patient symptoms, adverse clinical events and the need for aortic valve replacement (AVR). The hypertrophic response of the left ventricle is protective for many years, even decades, yet with time it decompensates and patients transition to heart failure. Current wisdom is that valve replacement should be performed in patients with severe stenosis just as that decompensation is starting to occur. Most commonly we use symptom development as our barometer of developing myocardial ill health, however there is increasing interest in more objective markers of LV dysfunction with which to optimise the timing of AVR. These include imaging markers of myocardial fibrosis and early systolic dysfunction as well as serum biomarkers such as highsensitivity troponin and Nterminalprobeta natriuretic peptide (NTproBNP)). 6 It is on this background that Dr Tan and colleagues report a multibiomarker study in 173 patients (55% male, age 69±11 years) with moderatetosevere AS, preserved LV function and New York Heart Association (NYHA) class III symptoms. A range of widely used (highsensitivity troponin T, NTproBNP) and novel (growth differentiation factor, suppression of tumorigenicity2, midregional proadrenomedullin (MRproADM), and MRproatrial natriuretic peptide) biomarkers were obtained from each participant, who were then followed up for a median of 2.7 years. The primary outcome was a composite of allcause mortality, progression to NYHA class IIIIV, and heart failure hospitalisation with the secondary outcome also incorporating syncope and acute coronary syndromes. Impressively, all but one participant had followup data available until either (1) aortic valve replacement was performed, (2) the first outcome of interest was reached, or (3) the date of final followup in those without events. Fiftynine participants fulfilled criteria for the primary outcome (34%) and 66 the secondary outcome (38%). Thirtyfour patients died (20%) with causes being cardiovascular (n=18), respiratory (n=4), sepsis (n=1) and unknown (n=15). There were additionally 28 (16%) heart failure hospitalisations, 10 (6%) syncopal events, 22 (13%) episodes of the acute coronary syndrome and 37 (21%) episodes of symtom progression to NYHA class IIIIV. Across all outcomes, MRproADM emerged as the biomarker with the best prognostic potential, being associated with HRs of 11 and 13 for the primary and secondary outcomes, respectively. The authors also looked at combining multiple biomarkers. Although the strongest dualbiomarker combination was NTproBNP combined with MRproADM, MRproADM remained stronger alone than in any combination. Receiver operating characteristic curve analysis suggested a cutoff level of 0.645 nmol/L was optimal for prediction of the primary outcome. MRproADM is the midregional precursor of ADM, a regulatory peptide first isolated from patients with phaeochromocytomas. ADM is a potent autocrine and paracrine hormone with vasodilatory and natriuretic actions, allowing it to contribute to fluid and electrolyte homoeostasis among other roles (figure 1). Widely expressed throughout the body, elevated plasma ADM is seen in disorders including hypertension, chronic kidney disease, myocardial infarction and finally heart failure, where elevated plasma ADM predicts those most likely to undergo cardiac transplantation or suffer a cardiovascular death. 9 Elevated ADM therefore highlights patients with a state of adverse neurohormonal activation. The results from this initial study suggest MRproADM is an exciting novel biomarker for patients with AS. Its ability to predict adverse events and symptom development prior to AVR indicates that MRproADM may help identify patients in whom prompt valve replacement is required. Plasma biomarkers are particularly attractive for this purpose given that they are minimally invasive, easy to access and relatively cheap when compared with more expensive imaging assessments. However, further research is required before this biomarker can be recommended for clinical use. In particular external validation of these findings and the proposed thresholds are required in much larger multicentre cohorts. These should also include less stable patient cohorts, investigate sex differences in MRproADM expression and assess its ability to predict AVR as well as adverse events after AVR. Ideally these studies should also include detailed imaging assessments so we can understand the