Artificial intelligence models in prediction of response to cardiac resynchronization therapy: a systematic review.

The aim of the presented review is to summarize the literature data on the accuracy and clinical applicability of artificial intelligence (AI) models as a valuable alternative to the current guidelines in predicting cardiac resynchronization therapy (CRT) response and phenotyping of patients eligible for CRT implantation. This systematic review was performed according to the PRISMA guidelines. After a search of Scopus, PubMed, Cochrane Library, and Embase databases, 675 records were identified. Twenty supervised (prediction of CRT response) and 9 unsupervised (clustering and phenotyping) AI models were analyzed qualitatively (22 studies, 14,258 patients). Fifty-five percent of AI models were based on retrospective studies. Unsupervised AI models were able to identify clusters of patients with significantly different rates of primary outcome events (death, heart failure event). In comparison to the guideline-based CRT response prediction accuracy of 70%, supervised AI models trained on cohorts with > 100 patients achieved up to 85% accuracy and an AUC of 0.86 in their prediction of response to CRT for echocardiographic and clinical outcomes, respectively. AI models seem to be an accurate and clinically applicable tool in phenotyping of patients eligible for CRT implantation and predicting potential responders. In the future, AI may help to increase CRT response rates to over 80% and improve clinical decision-making and prognosis of the patients, including reduction of mortality rates. However, these findings must be validated in randomized controlled trials.

Therapeutic difficulties in a patient with Ehlers-Danlos syndrome and numerous symptomatic premature ventricular contractions-case report and literature review.

A 28-year-old female patient diagnosed with Ehlers-Danlos syndrome type III (hypermobile EDS, hEDS) was admitted to the cardiology clinic due to a 3-year history of symptomatic ventricular arrhythmia in the form of multiple premature ventricular contractions (PVCs). Attempts at antiarrhythmic treatment with beta-blockers, propafenone, and verapamil were unsuccessful. Due to the diagnosis of hEDS and the high risk of vascular complications related to the ablation procedure, invasive treatment was abandoned, and it was decided to implement flecainide. After the flecainide treatment initiation, a spectacular improvement in the number of ventricular arrhythmias was observed, along with the disappearance of the complaints previously reported by the patient. To the best of our knowledge, this is the first described case of spectacular flecainide antiarrhythmic effect in a patient with numerous PVCs also diagnosed with EDS. Flecainide treatment in the EDS group could be a successful alternative to ablation, which can lead to serious vascular and even life-threatening complications, especially after the failure of propafenone and beta-blockers treatment.

Left Ventricle Segmental Longitudinal Strain and Regional Myocardial Work Index Could Help Determine Mitral Valve Prolapse Patients with Increased Risk of Ventricular Arrhythmias.

Mitral valve prolapse (MVP) could associate with malignant ventricular arrhythmias (VAs). Mitral annular disjunction, a putative mechanism for an arrhythmic substrate, leads to excessive mobility, stretch, and damage of some segments. Speckle tracking echocardiography (STE), with particular attention to the segmental longitudinal strain and myocardial work index (MWI), could be an indicator of the segments we aimed to check. Seventy-two MVP patients and twenty controls underwent echocardiography. Complex VAs documented prospectively after the enrollment was qualified as the primary endpoint, which was noticed in 29 (40%) patients. Pre-specified cut-off values for peak segmental longitudinal strain (PSS) and segmental MWI for basal lateral (-25%, 2200 mmHg%), mid-lateral (-25%, 2500 mmHg%), mid-posterior (-25%, 2400 mmHg%), and mid-inferior (-23%, 2400 mmHg%) segments were accurate predictors of complex VAs. A combination of PSS and MWI increased the probability of the endpoint, reaching the highest predictive value for the basal lateral segment: odds ratio 32.15 (3.78-273.8), p < 0.001 for PSS ≥ -25% and MWI ≥ 2200 mmHg%. STE may be a valuable tool for assessing the arrhythmic risk in MVP patients. Excessively increased segmental longitudinal strain with an augmented regional myocardial work index identifies patients with the highest risk of complex VAs.

Echocardiographic predictors of thrombus in left atrial appendage-The role of novel transthoracic parameters.

Introduction: The left atrium appendage thrombus (LAAT) formation is a complex process. A CHA2DS2-VASc scale is an established tool for determining the thromboembolic risk and initiation of anticoagulation treatment in patients with atrial fibrillation or flutter (AF/AFL). We aimed to identify whether any transthoracic echocardiography (TTE) parameters could have an additional impact on LAAT detection. Methods: That is a sub-study of multicenter, prospective, observational study LATTEE (NCT03591627), which enrolled 3,109 consecutive patients with AF/AFL referred for transesophageal echocardiography (TEE) before cardioversion or ablation. Results: LAAT was diagnosed in 8.0% of patients. The univariate logistic regression analysis [based on pre-specified in the receiver operating characteristic (ROC) analysis cut-off values with AUC ≥ 0.7] identified left ventricular ejection fraction (LVEF) ≤ 48% and novel TTE parameters i.e., the ratios of LVEF and left atrial diameter (LAD) ≤ 1.1 (AUC 0.75; OR 5.64; 95% CI 4.03-7.9; p < 0.001), LVEF to left atrial area (LAA) ≤ 1.7 (AUC 0.75; OR 5.64; 95% CI 4.02-7.9; p < 0.001), and LVEF to indexed left atrial volume (LAVI) ≤ 1.1 (AUC 0.75, OR 6.77; 95% CI 4.25-10.8; p < 0.001) as significant predictors of LAAT. In a multivariate logistic regression analysis, LVEF/LAVI and LVEF/LAA maintained statistical significance. Calculating the accuracy of the abovementioned ratios according to the CHA2DS2-VASc scale values revealed their highest predictive power for LAAT in a setting with low thromboembolic risk. Conclusion: Novel TTE indices could help identify patients with increased probability of the LAAT, with particular applicability for patients at low thromboembolic risk.

Baroreflex sensitivity but not microvolt T-wave alternans can predict major adverse cardiac events in ischemic heart failure.

BACKGROUND: Major adverse cardiovascular events (MACE) constitutes the main cause of morbidity and mortality in ischemic heart failure (HF) patients. The prognostic value of the autonomic nervous system parameters and microvolt T-wave alternans (MTWA) in this issue has not been identified to date. The aim herein, was to assess the usefulness of the abovementioned parameters in the prediction of MACE in HF patients with left ventricular systolic dysfunction of ischemic origin. METHODS: Baroreflex sensitivity (BRS), heart rate variability (HRV), MTWA and other well-known clinical parameters were analyzed in 188 ischemic HF outpatients with left ventricular ejection fraction (LVEF) ≤ 50%. During 34 (14-71) months of follow-up, 56 (30%) endpoints were noted. RESULTS: Univariate Cox analyses revealed BRS (but not HRV), MTWA, age, New York Heart Association functional class III, LVEF, implantable cardioverter-defibrillator presence, use of diuretics and antiarrhythmic drugs, diabetes, and kidney insufficiency were defined as significant predictors of MACE. Pre-specified cut-off values for MACE occurrence for the aforementioned continuous parameters (age, LVEF, and BRS) were: ≥ 72 years, ≤ 33%, and ≤ 3 ms/mmHg, respectively. In a multivariate Cox analysis only BRS (HR 2.97, 95% CI 1.35-6.36, p < 0.006), and LVEF (HR 1.98, 95% CI 0.61-4.52, p < 0.038) maintained statistical significance in the prediction of MACE. CONCLUSIONS: Baroreflex sensitivity and LVEF are independent of other well-known clinical parameters in the prediction of MACE in patients with HF of ischemic origin and LVEF up to 50%. BRS ≤ 3 ms/mmHg and LVEF ≤ 33% identified individuals with the highest probability of MACE during the follow-up period.

Comprehensive Echocardiography of Left Atrium and Left Ventricle Using Modern Techniques Helps in Better Revealing Atrial Fibrillation in Patients with Hypertrophic Cardiomyopathy.

Atrial fibrillation (AF) is an important arrhythmia in hypertrophic cardiomyopathy (HCM). We aimed to explore whether a complex evaluation of the left ventricle (LV) using modern echocardiography techniques, additionally to the left atrium (LA) boosts the probability of AF diagnosis. Standard echocardiography, 2D and 3D speckle tracking, were performed for LA and LV evaluation in HCM patients and healthy volunteers. Of 128 initially qualified HCM patients, 60 fulfilled included criteria, from which 43 had a history of AF, and 17 were without AF. LA volume index and peak strain, LV ejection fraction, and strains were significant predictors of AF. In addition, 2D global longitudinal strain (GLS) for LV at cut off -16% turned out to be the most accurate predictor of AF (OR 48.00 [95% CI 2.68-859.36], p = 0.001), whereas the combination of LA peak strain ≤ 22% and LV GLS ≥ -16% had the highest discriminatory power (OR 76.36 [95% CI 4.13-1411.36], p = 0.001). AF in HCM patients seems to be LA as well as LV disease. Revealing lower strain for LV, in addition to lower LA strain, may have an important impact on accurate characteristics of HCM patients with AF history.

Segmental longitudinal strain as the most accurate predictor of the patchy pattern late gadolinium enhancement in hypertrophic cardiomyopathy.

BACKGROUND: The prognostic value of myocardial fibrosis in patients with hypertrophic cardiomyopathy (HCM) has been well-established. Although cardiac magnetic resonance (CMR) is the method of choice in its revealing as the presence of late gadolinium enhancement (LGE), this technique still has limited availability in daily clinical practice. Two-dimensional speckle tracking echocardiography (2D STE) seems to be helpful in verification which HCM patient has the highest probability of LGE presence and hence needs to be qualified to CMR. While the majority of HCM patients have a patchy pattern of myocardial fibrosis, the aim of this study was to evaluate whether segmental rather than global longitudinal strain is more accurate in the identification of the presence of LGE. METHODS: Forty-six HCM patients had transthoracic echocardiography and CMR imaging performed. Each patient had global longitudinal strain and rotation parameters calculated, as well as segmental analyses for wall thickness, longitudinal strain, and LGE presence based on 736 segments of the left ventricle (LV). The presence of LGE in CMR was confirmed on a per-segment basis, which was similar to LV segments in the echocardiographic examination. All patients were divided into two groups according to the CMR result: LGE (+) and LGE (-). RESULTS: Receiver-operating characteristic analyses identified peak global longitudinal strain and peak twisting velocity with the cut-off values -14.4% and 116°/s respectively as the accurate predictors of LGE presence in CMR, whereas segmental longitudinal strain of -12.5% cut-off value had the highest area under the curve value (87.4%, confidence interval 84.5-90.3%), with 93.7% sensitivity, 86.5% negative predictive value, and 55% specificity. CONCLUSIONS: Segmental longitudinal strain with the cut-off value of -12.5% has the highest discriminatory power for LGE presence and seems to be more adequate than global speckle tracking parameters in identification of HCM patients with strong indications for CMR for more accurate risk stratification.

Electrocardiographic abnormalities in amateur male marathon runners.

BACKGROUND: Sports activity has become extremely popular among amateurs. Electrocardiography is a useful tool in screening for cardiac pathologies in athletes; however, there is little data on electrocardiographic abnormalities in the group of amateur athletes. OBJECTIVES: The aim of this study was to analyze the abnormalities in resting and exercise electrocardiograms (ECGs) in a group of amateur athletes, and try to determine whether the criteria applied for the general population or for athletes' ECGs should be implemented in this group. MATERIAL AND METHODS: In 40 amateur male marathon runners, 3 consecutive 12-lead ECGs were performed: 2-3 weeks before (stage 1), just after the run (stage 2) and 2-3 weeks after the marathon (stage 3). Resting (stage 1) and exercise (stage 2) ECGs were analyzed following the refined criteria for the assessment of athlete's ECG (changes classified as training-related, borderline or training-unrelated). RESULTS: In resting ECGs, at least 1 abnormality was found in 92.5% of the subjects and the most common was sinus bradycardia (62.5%). In post-exercise ECGs, at least 1 abnormality was present in 77.5% of the subjects and the most common was right atrium enlargement (RAE) (42.5%). Training-related ECG variants were more frequent at rest (82.5% vs 42.5%; p = 0.0008), while borderline variants - after the run (22.5% vs 57.5%; p = 0.0004). Training-unrelated abnormalities were found in 15% and 10% of the subjects, respectively (p-value - nonsignificant), and the most common was T-wave inversion. CONCLUSIONS: Even if the refined criteria rather than the criteria used for normal sedentary population were applied, the vast majority of amateur runners showed at least 1 abnormality in resting ECGs, which were mainly training-related variants. However, at rest, in 15% of the subjects, pathologic training-unrelated abnormalities were found. The most frequent post-exercise abnormality was right atrial enlargement. General electrocardiographic screening in amateur athletes should be taken into consideration.