Funding Acknowledgements Type of funding sources: None. Introduction 3D left ventricular ejection fraction (LVEF) quantification methods are more accurate and reproducible than 2D echocardiography, however, conventional 3D is time consuming and requires extensive user expertise, thus hindering its routine implementation in busy echocardiography laboratories and its use by inexperienced physicians. HeartModel A.I. (HM) is a simple, fast, recently validated 3D automated analysis software that detects LV endocardial surfaces and calculates LVEF. The aim of this work is to evaluate the performance of HM with experienced and inexperienced physicians, its time saving potential and to assess whether this software can be a better alternative to 2D measurements in routine echocardiography. Methods Prospective analysis of 30 nonconsecutive patients referred for transthoracic echocardiogram in a university hospital echocardiography lab, from 1st February 2021 to 31st March 2021. 2D biplane LVEF was measured by an experienced and inexperienced physician (less than 250 echocardiograms performed), then the same physicians used the automated analysis software to assess LVEF (blinded for each other results). The time to make the measurements was registered. Comparisons of agreement between LVEF measurements (experienced versus inexperienced physicians) included linear regression with Pearson correlation coefficients and Bland-Altman analyses to assess the bias and limits of agreement (defined as 2SD around the mean). Results A total of 30 patients were included, mean age of 68.6 ± 20.1 years and 60% male. HM showed significantly lower acquisition times in both inexperienced (72±17s versus 173± 44s, P<0.01) and experienced (56±12s versus 126±29s, P<0.01) physicians. The difference in time of acquisition between 2D and HM was approximately 101s for inexperienced users and around 70s for experienced users. Regarding LVEF assessment, HM acquisitions compared to 2D measurements showed stronger correlations between experienced and inexperienced physicians (r= 0,98, P<0,01 versus r= 0,92, P<0,01) with minimal bias (−0,5 versus −0,6) and stronger agreement (HM limits of agreement: ± 5,8% versus 2D limits of agreement: ± 12,5%) Conclusion 3D LVEF assessment by HM significantly reduced acquisition times and exhibited higher interobserver agreement than 2D Simpson’s biplane method. These results suggest that automated 3D algorithms, such as HM, may play a key role in implementing 3D measurements in routine practice in busy echocardiography laboratories and allow the use of 3D echocardiography at early stages of physicians training.

Funding Acknowledgements Type of funding sources: None. Introduction 8-40% of patients with acute ST-elevation myocardial infarction (STEMI) present later than 12 hours after symptom onset. According to guidelines these late presenters maintain indication for primary percutaneous coronary intervention (PCI) when there are signs of ongoing ischemia. However, it prevails uncertainty in relation to the best approach in stable late presenters. Aims Describe the profile of stable late STEMI presenters and evaluate the trends of reperfusion decision in the Portuguese reality; compare early term outcomes between patients submitted to emergent primary PCI and those in which it was preferred an initial conservative approach. Methods Retrospective analysis of patients with STEMI presenting ≥12-48h hours after the beginning of the symptoms between October 2010 and December 2019 without evidence of ongoing ischemia, inserted in a national registry of acute coronary syndromes. Patients were dichotomized and compared according to whether or not were submitted to emergent reperfusion based on primary PCI. Results 274 patients were included (2.3% of all STEMI), predominantly men (67.5%), with a mean age of 68±13 years old. Emergent PCI was performed in a minority (15.7%; n=43); even so, coronarography ended up being executed in 61.3% of the admissions, with angioplasty performed in 47.1% of the cases. Right coronary artery was the most common intervened vessel (50.8%). Inotropes were necessary in 4.6% of the patients, with no reports of ventricular assistance device use. Mean ejection fraction was 51±12% with no differences between groups. Patients submitted to emergent PCI (15.7%) had a lower prevalence of atrial fibrillation (0 vs. 9.3%, p=0.04) and had more commonly electrocardiographic criteria for anterior STEMI (64.3% vs. 41.4%, p=0.006). Nitrates were significantly less prescribed at discharge in this subgroup (4.9% vs. 26.8%; p=0.002). Apart from aborted cardiac arrest, that was more prevalent in patients submitted to emergent reperfusion (4.8% vs. 0.9%, p=0.12), it was observed a tendency toward a lower percentage in this subgroup in all other early hard clinical outcomes such as re-infarction (0 vs. 0.4%, p=1.00), mechanical complications (0 vs. 2.2%; p=1.00), sustained ventricular tachycardia (0 vs. 0.9%, p=1.00) and in-hospital death (0% vs. 4.4%, p=0.37). However, none of the differences have reached statistical significance. Conclusion The study shows that, in the Portuguese reality, emergent reperfusion is adopted in only a minority of late stable STEMI patients, with a clear tendency to perform it more frequently in subacute anterior STEMI. Emergent PCI strategy did not show a clear benefit in terms of left ventricular function, risk of re-infarction, arrhythmic and mechanical complications, and in-hospital death. On the other hand, there was apparently a significant advantage of this strategy in ischemic symptom control.

Funding Acknowledgements Type of funding sources: None. Introduction Left ventricular systolic dysfunction (LVD) is a key concern in the context of cardio-oncology (CO). Usually, referral for suspected Cancer therapy-related cardiac dysfunction (CTRCD) is the main challenge, but heart failure with other more common causes, such as ischemic cardiomyopathy can also decompensate during cancer treatment or be diagnosed incidentally during cardiotoxicity echocardiographic (echo) surveillance. Multimodality imaging is essential in these patients in order to better establish aetiology and assure the most appropriate clinical management. Purpose evaluate clinical impact of multimodality imaging in the clinical management of CO patients. Methods retrospective study of a population followed in CO consultation. Statistical analysis of demographic, clinical, transthoracic echocardiography (TTE) and cardiac magnetic resonance (CMR) data was made. Results we included 115 pts, mean age 66.3 ± 10.2 years, 67,8% female, with mean follow-up of 16.1 ± 12.8 months. About half (56.5%) had breast cancer, followed by gastrointestinal tract (16.5%) and haematological (8,7%) malignancies, with a significant proportion (32,2%) with advanced disease. Prevalence of cardiovascular risk factors was high (hypertension in 74.8%, dyslipidaemia in 47%, type 2 diabetes mellitus in 17.4%), but also coronary artery disease (18,3%) and atrial fibrillation (18.3%). All of them were treated with different types of chemotherapy and 53,9% of pts with radiotherapy.  At baseline, 13% of pts had a left ventricular ejection fraction (LVEF) under 50% (LVD) assessed by TTE, which increased to 26,9% (n = 31) after oncological treatment initiation. Of these (n = 31), an ischemic aetiology was found in 32,3% and non-ischemic in 54,8%, which was significantly more frequent in patients with CTRCD (OR 2,7, p = 0,001). CMR was performed in 45,2%, mostly in CTRCD cases (p = 0,012, OR 8,4), which, apart from LVD, did not show any tissue changes in most patients (p = 0,026, OR 35). Only one patient with CTRCD (under treatment with trastuzumab and anthracyclines) had subepicardial late gadolinium enhancement, with wall motion abnormalities, suggesting a myocarditis-like mechanism for cardiotoxicity. Conclusion LVD has a major impact in patients" prognosis, particularly in CO context, where effective oncological treatments can be compromised due heart failure decompensation. Therefore, a thorough clinical evaluation should encompass etiological study in order to provide the most appropriate treatment strategies. Moreover, CTRCD can develop through different physiopathological mechanisms. Thus, multimodality imaging, particularly including CMR evaluation, can have a major role ensuring a good clinical outcome for these patients.

No abstract

Introduction Recent recommendations regarding myocardial infarction (MI) underline the adverse prognosis associated with right bundle branch block (RBBB), suggesting that, in some cases of non-ST-segment elevation MI (NSTEMI) with RBBB a primary percutaneous coronary intervention (PCI) strategy should be considered. However, it is unclear if this is due to a more difficult and late diagnosis or to the clinical severity inherent to these patients (pts). Purposes To characterize the NSTEMI with RBBB population and find predictors of worse prognosis. Methods Retrospective analysis of pts included in the Portuguese Registry of Acute Coronary Syndromes with NSTEMI, comparing pts with RBBB (group A) vs without RBBB (group B), regarding clinical and demographic variables, diagnostic and therapeutic approaches. Primary endpoint was heart failure, electrical and mechanical complications and death in the in-hospital period. Results We included 9375 pts, 686 in group A and 8689 in group B. Pts in group A were more likely to be male (p<0.001) and over 75 years old (p<0.001). Also, they were more prone to have cardiovascular risk factors (hypertension - p<0.001, diabetes – p<0.001) and history of coronary artery disease (stable angina p=0.007, previous MI p=0.002 and revascularization, either PCI – p=0.016 or surgery – p<0.001), stroke (p<0.001), chronic kidney disease (p<0.001) and cancer (p=0.025), comparing to pts in group B. There were no differences between time from onset of symptoms and first medical contact or hospital admission between groups. Upon admission, these pts presented more frequently with hypotension (p=0.026), Killip class>II (p<0.001) and atrial fibrillation (p<0.001) than pts in group B. There were statiscally significant differences between groups, regarding the use of inotropes (p<0.001), non-invasive (p=0.008) and invasive ventilation (p=0.018) and temporary pacing (p=0.001), all of them higher in group A. Pts with RBBB were less likely to undergo coronary angiography (CA) (p<0.001). However, among those who did, there were no differences in CA timing (p=0.091), but pts from group A had more frequently multivessel disease (p=0.044) and no revascularization was undertaken (p=0.012). About 16.64% of all pts reached the endpoint, but unfavourable in-hospital outcome was significantly more common in group A (p<0.001). RBBB remained an independent predictor of the endpoint (p=0.032) in a multivariate regression analysis, controlled for other variables (namely gender, age, cardiovascular risk factors, previous evidence of cardiovascular disease, and clinical and coronary anatomy data) – AUC of 0.833. Conclusion Although pts with NSTEMI and RBBB have a poorer in-hospital prognosis, partly due to their bigger clinical complexity (older age, multiple comorbidities and complex coronary anatomy), RBBB itself still remains an independent predictor of worse outcome. FUNDunding Acknowledgement Type of funding sources: None.