Left atrial dysfunction predicts left ventricular remodelling in patients with preserved ejection fraction after acute ST-elevation myocardial infarction

Left ventricular remodelling (LVR) potentially occurs even in patients with preserved ejection fraction (EF) after ST-elevation myocardial infarction (STEMI), where diastolic dysfunction may develop or worsen despite a small infarct size, as a result of myocyte necrosis and reduced left ventricular (LV) compliance.1 The left atrium (LA) is susceptible to changes in shape and haemodynamics that occur in the downstream LV. Left atrial dysfunction is emerging as a bellwether and prognosticator in ischaemic heart disease2,3 and in heart failure with preserved ejection fraction (HFpEF).4 Cardiovascular magnetic resonance (CMR) is the gold standard technique for quantifying LV volumes and their temporal variations. Additionally, it allows the assessment of LA function using the feature-tracking algorithm (CMR-FT)5 applied on routinely available cineSSFP sequences. The present research sought to explore the potential role of LA strain as an early predictor of evolution towards LVR in patients with preserved EF after STEMI. The study population derives from the prospective, observational, and multicentre VavirimS trial6 (Valore predittivo della vitalità miocardica nell’identificazione del rimodellamento ventricolare sinistro 6 mesi dopo uno STEMI; ClinicalTrials.gov ID: NCT04699565). One hundred and four patients with preserved EF after the first episode of STEMI, treated with successful primary PCI, underwent CMR at 1 and 6 months after the index event. The CMR protocol included cineSSFP, T2-weighted, and late gadolinium enhancement (LGE) sequences. All images were transferred to the Core-Lab (Multimodality Cardiac Imaging Section at IRCCS Policlinico San Donato, San Donato Milanese, Milan, Italy) for centralized analysis by two experienced operators with Level 3 EACVI CMR certification. Cardiovascular magnetic resonance analysis was performed using dedicated software (Medis Medical Imaging Systems, Leiden, The Netherlands) and included the quantification of biventricular volumes and LA parameters using a semi-automated threshold-based method and the CMR-FT algorithm, respectively. Twenty-four subjects matched for age, sex, and major cardiovascular risk factors affecting LA function (i.e. arterial hypertension and diabetes mellitus) were selected as the control group for LA evaluation (see Supplementary material online, Figure S1). The study outcome was LVR, defined as an increase of ≥12% in both left ventricular end-diastolic volume indexed (LV-EDV-i) and left ventricular end-sistolic volume indexed (LV-ESV-i) at 6 months CMR after STEMI. According to this endpoint, the cohort was divided into two groups: (i) LVR (n = 11) and (ii) no-LVR (n = 93). The two groups did not differ for baseline demographic, clinical, and angiographic parameters, nor for blood biomarkers (nt-proBNP, troponin), CMR parameters [infarct size, wall motion score index, LV global longitudinal strain (GLS), total LV mass, LV viable mass, presence of haemorrhage, and microvascular obstruction], or medical therapy (see Supplementary material online, Table S1). In the no-LVR group, LVEF increased at 6 months [ejection fraction (EF) 59% vs. 64%, P-value 0.032], while no substantial differences occurred in the LV group (EF 62% vs. 61%, P-value 0.853). The whole STEMI population showed LA dysfunction (i.e. reduced LA EF and LA strain), at 1 and 6 months after the index event (see Supplementary material online, Table S2). At 1-month CMR, comparing LVR and no-LVR groups, the former showed smaller LA volumes and greater impairment of LA reservoir and conduit functions. At 6-month CMR, the differences in atrial volumes and function between the two groups were no longer evident, as the LVR group showed improvement in LA functions along with an increase in atrial volumes (Table 1; Supplementary material online, Table S3). At multivariate analysis, LA reservoir [P-value 0.027; odds ration 1.17 (95% confidential interval 1.35–1.02)] emerged as the only independent predictor of LVR in patients with preserved EF after STEMI. Youden’s index was maximized to determine a LA reservoir function of 17% as the best cut-off value for predicting LVR (Figure 1). Example of left atrial dysfunction and left ventricular remodeling after STEMI. Left atrial cardiovascular magnetic resonance feature-tracking analysis: comparison between left ventricular remodelling and no-left ventricular remodelling ST-elevation myocardial infarction patients EF, ejection fraction; LA, left atrial; LAV max, maximum LA volume; LAV min, minimum LA volume; LAV preA, LA volume pre-atrial contraction. Left atrial cardiovascular magnetic resonance feature-tracking analysis: comparison between left ventricular remodelling and no-left ventricular remodelling ST-elevation myocardial infarction patients EF, ejection fraction; LA, left atrial; LAV max, maximum LA volume; LAV min, minimum LA volume; LAV preA, LA volume pre-atrial contraction. This is the first CMR study evaluating LA dysfunction and LVR in patients with preserved EF after STEMI. The main findings are as follows: (i) patients with preserved EF after STEMI show LA dysfunction at 1 month that remains evident even 6 months after the index event; (ii) ∼11% experience LVR; (iii) LVR patients exhibit greater impairment of atrial function 1 month after STEMI; and (iv) LA reservoir emerged as an independent predictor of LVR. Although patients with preserved EF after STEMI have been traditionally considered at low risk, recent evidence shows that a non-negligible percentage of them may experience adverse outcomes.7 The search for prognostic imaging markers in this population has mainly focused on ventricular imaging, including tissue characterization and functional analysis. However, traditional markers of adverse outcomes, such as LVEF and infarct size, have demonstrated poor predictive power in patients with limited myocardial damage.8 Consistently, also in our study, neither LGE nor LV functional alterations (LVEF, LV GLS) proved to distinguish LVR patients. In recent years, the LA has been gaining increasing relevance, not only as the mere counterpart upstream of the LV but also as an independent and vulnerable cardiac chamber that shows early involvement in the event of myocardial disease. The LA reservoir appears to be the most promising parameter for many reasons. First, from a diagnostic standpoint, it provides a more accurate categorization of diastolic dysfunction compared with traditional echocardiographic parameters such as E/eʹ, LA volume, and LV GLS.9 Second, from a clinical perspective, it has proven to be an early diagnostic and prognostic marker across most cardiac diseases.10 Left atrial strain parameters assessed by CMR are easy to obtain, as no specific sequences are needed. Various scientific studies have reported good reliability not only in terms of intra- and inter-operator variability but also in inter-vendor comparisons.5,11 Since reservoir function has been validated against invasive measures9 and a meta-analysis defined a value of LA reservoir ≤19%12 for identifying an increase in LV filling pressures, we can hypothesize that such reservoir impairment in LVR patients at 1 month after STEMI may testify to the development of diastolic dysfunction despite limited tissue alterations and preserved LVEF. The atrial dysfunction observed in the LVR group could be a useful telltale indicator of patients predisposed to develop diastolic dysfunction and volumetric remodelling following myocardial damage, albeit to a limited extent. These preliminary data necessitate validation in more extensive cohorts, integration with other potential imaging markers, and follow-up data, to define the clinical impact of LVR in patients with preserved EF after STEMI. Moreover, given the emerging role of the LA as a prognosticator, the use of dedicated CMR long-axis sequences may provide a more accurate assessment of LA remodelling and dysfunction.13 Evaluating LA strain may serve as an easy and reliable prognostic tool for screening patients with preserved EF, to identify those at a higher risk of developing LVR and heart failure. This is particularly relevant today, considering the availability of new therapeutic options for HFpEF such as SGLT2 inhibitors. Supplementary material is available at European Journal of Preventive Cardiology. We acknowledge Consorzio di Bioingegneria e Informatica Medica–CBIM for methodological support and technical assistance with eCRF (electronic case report form) design, development, and management. The present study is dedicated to Silvia Pica, without whom this work would not have been possible. A.A. and L.T. conceived the study, performed the formal analysis, and contributed to writing the manuscript. M.P. and S.C. contributed to the refinement of the final manuscript. M.L. and P.G.C. provided supervision for the findings of this work. All authors actively participated in the discussion of results and contributed to the finalization of the manuscript. The study was funded by a grant to the Cardiology IRCCS network from the Italian Ministry of Health (RCR-2019-23669118_003). The data underlying this article will be shared upon reasonable request to the corresponding author.