Objective(1) to test the association with prevalent and incident atrial fibrillation (AF), and prognosis of total N-terminal pro-B type natriuretic peptide (total NT-proBNP) and of a panel of biomarkers; (2) to assess iwhether the extent of glycosylation affects the relation of NT-proBNP with AF.MethodsIn a sub-study of the GISSI-AF trial on 382 patients in sinus rhythm with a history of AF, echocardiographic variables and eight circulating biomarkers were serially assayed over one year. The relations between circulating baseline biomarkers and AF and the risk of CV events, were assessed by Cox-analysis models adjusting the first by clinical variables, the second by clinical variables and the echocardiographic left-atrial-minimum-volume-index (LAVImin).ResultsOver a median follow-up of 365 days, 203/382 patients (53.1%) had at least one recurrence of AF and 16.3% were hospitalized for cardiovascular (CV) reasons. Total NT-proBNP, NT-proBNP, angiopoietin 2 (Ang2), myosin binding protein (MyBPC3) and bone morphogenic protein-10 (BMP-10) were strongly associated to ongoing AF. Natriuretic peptides and MyBPC3 predicted recurrent AF but this lost significance after adjustment for LAVImin. NT-proBNP and Ang2 predicted CV hospitalization even after adjustment for LAVImin, HR95%CI: 2.20 [1.02-4.80] and 5.26 [1.16-23.79].ConclusionsThe association of AF recurrence with the novel biomarker total NT-proBNP, is similar to that of NT-proBNP, suggesting no influence of glycosylation. Ang2, MyBPC3 and BMP10 were strongly associated with AF, indicating a possible role of extracellular matrix and myocardial injury. Abstract-words=233Key messagesWhat is already known on this subject?It is still complicated to predict the recurrence of AF in patients in sinus rhythm with a recent history of AF. Though several biomarkers have been associated with AF, few of them have proved to be independent predictors for recurrent AF or cardiovascular (CV) events. Their predictive sensitivity and specificity is modest at best. Previous studies showed that NT-proBNP was possibly the strongest predictor of recurrent AF and CV hospitalization. Natriuretic peptides circulate to a large extent as glycosylated molecules and a novel assay is now available to measure the glycosylated and non-glycosylated NT-proBNP in plasma, the total NT-proBNP. The extent of glycosylation varies in different diseases.What might this study add?No studies have assessed (a) the extent of NT-proBNP glycosylation in AF, or (b) the association and predictive value in patients with AF of total NT-proBNP. A multimarker approach, ratter than one based on a single biomarker, might predict AF better.The relation with AF of the novel biomarker, total NT-proBNP, is as strong as that of NT-proBNP, suggesting no-influence of glycosylation.Two biomarkers, MyBPC3, secreted few minutes after myocardial injury and Ang-2, involved in inflammation and coagulation, were strongly associated to AF.How might this impact on clinical practice?The identification of novel circulating biomarkers could have a direct impact on clinical practice when predicting the occurrence of AF, but unfortunately current data do not allow predictions based on biomarkers.The associations of different biomarkers with ongoing AF may cast light on the mechanisms of triggering and maintenance of AF.Strengths and limitations of this studyThe data came from to a multicenter randomized clinical trial with available concomitant serial echocardiographic and circulating biomarkers recorded and evaluated centrally, hence with minimal bias; AF recurrence during a 12-month follow up was checked weekly by trans-telephonic electrocardiographic monitoring, and with 12-lead ECG every six months.A comparative analysis of total NT-proBNP with other novel biomarkers and echocardiographic variables has never been done so far. The possible added value of total NT-proBNP to the benchmark biomarker NT-proBNP was assessed on the basis of different dimensions of performance, as recently proposed for new biomarkers. The main limitations are (1) the relatively small numbers of patients with AF during follow-up visits, (2) the very low prevalence of patients with other cardiac diseases such as coronary artery disease and heart failure, and (3) consequently, the low incidence of clinical events in one-year follow-up.