Long-term imaging analysis of a myocarditis case: Utilizing strain with echocardiography and cardiovascular magnetic resonance findings.

Herein, we report a case of myocarditis in a 27-year-old male with long-term follow-up using longitudinal peak systolic strain (LPSS) measurements with transthoracic echocardiography (TTE) and late gadolinium enhancement (LGE) in cardiovascular magnetic resonance imaging (CMR). On admission, a predominant decrease was observed in the LPSS in the posterolateral segments of the TTE. After a period of two weeks, the values of the LPSS observed in the posterolateral segments were still slightly reduced, which is consistent with the LGE results in CMR. After a duration of 16 months, an improvement was noted in the LPSS and LGE results in all the segments. Moreover, a time-phase discrepancy was observed in the segmental longitudinal strain curve for a period of two weeks from the onset of myocarditis. However, an improvement in the discrepancy was detected after 16 months. Learning objective: Longitudinal peak systolic strain (LPSS) on transthoracic echocardiography (TTE) has predominantly focused on diagnosing the acute phase of myocarditis. Herein, LPSS was evaluated not only in the acute phase but also in the chronic phase. Furthermore, the relationship between the results of segmental LPSS and late gadolinium enhancement was documented. We would like to emphasize the usefulness of LPSS on TTE both for identifying myocarditis and as a tool for the long-term follow-up of patients.

Risk of Mortality Prediction Involving Time-Varying Covariates for Patients with Heart Failure Using Deep Learning.

Heart failure (HF) is challenging public medical and healthcare systems. This study aimed to develop and validate a novel deep learning-based prognostic model to predict the risk of all-cause mortality for patients with HF. We also compared the performance of the proposed model with those of classical deep learning- and traditional statistical-based models. The present study enrolled 730 patients with HF hospitalized at Toho University Ohashi Medical Center between April 2016 and March 2020. A recurrent neural network-based model (RNNSurv) involving time-varying covariates was developed and validated. The proposed RNNSurv showed better prediction performance than those of a deep feed-forward neural network-based model (referred as "DeepSurv") and a multivariate Cox proportional hazard model in view of discrimination (C-index: 0.839 vs. 0.755 vs. 0.762, respectively), calibration (better fit with a 45-degree line), and ability of risk stratification, especially identifying patients with high risk of mortality. The proposed RNNSurv demonstrated an improved prediction performance in consideration of temporal information from time-varying covariates that could assist clinical decision-making. Additionally, this study found that significant risk and protective factors of mortality were specific to risk levels, highlighting the demand for an individual-specific clinical strategy instead of a uniform one for all patients.

Exploring and Identifying Prognostic Phenotypes of Patients with Heart Failure Guided by Explainable Machine Learning.

Identifying patient prognostic phenotypes facilitates precision medicine. This study aimed to explore phenotypes of patients with heart failure (HF) corresponding to prognostic condition (risk of mortality) and identify the phenotype of new patients by machine learning (ML). A unsupervised ML was applied to explore phenotypes of patients in a derivation dataset (n = 562) based on their medical records. Thereafter, supervised ML models were trained on the derivation dataset to classify these identified phenotypes. Then, the trained classifiers were further validated on an independent validation dataset (n = 168). Finally, Shapley additive explanations were used to interpret decision making of phenotype classification. Three patient phenotypes corresponding to stratified mortality risk (high, low, and intermediate) were identified. Kaplan−Meier survival curves among the three phenotypes had significant difference (pairwise comparison p < 0.05). Hazard ratio of all-cause mortality between patients in phenotype 1 (n = 91; high risk) and phenotype 3 (n = 329; intermediate risk) was 2.08 (95%CI 1.29−3.37, p = 0.003), and 0.26 (95%CI 0.11−0.61, p = 0.002) between phenotype 2 (n = 142; low risk) and phenotype 3. For phenotypes classification by random forest, AUCs of phenotypes 1, 2, and 3 were 0.736 ± 0.038, 0.815 ± 0.035, and 0.721 ± 0.03, respectively, slightly better than the decision tree. Then, the classifier effectively identified the phenotypes for new patients in the validation dataset with significant difference on survival curves and hazard ratios. Finally, age and creatinine clearance rate were identified as the top two most important predictors. ML could effectively identify patient prognostic phenotypes, facilitating reasonable management and treatment considering prognostic condition.

Deep Learning-Based Recurrence Prediction of Atrial Fibrillation After Catheter Ablation.

BACKGROUND: Radiofrequency catheter ablation (RFCA) is an effective therapy for atrial fibrillation (AF). However, it the problem of AF recurrence remains. This study investigates whether a deep convolutional neural network (CNN) can accurately predict AF recurrence in patients with AF who underwent RFCA, and compares CNN with conventional statistical analysis.Methods and Results:Three-hundred and ten patients with AF after RFCA treatment, including 94 patients with AF recurrence, were enrolled. Nine variables are identified as candidate predictors by univariate Cox proportional hazards regression (CPH). A CNNSurv model for AF recurrence prediction was proposed. The model's discrimination ability is validated by a 10-fold cross validation method and measured by C-index. After back elimination, 4 predictors are used for model development, they are N-terminal pro-BNP (NT-proBNP), paroxysmal AF (PAF), left atrial appendage volume (LAAV) and left atrial volume (LAV). The average testing C-index is 0.76 (0.72-0.79). The corresponding calibration plot appears to fit well to a diagonal, and the P value of the Hosmer-Lemeshow test also indicates the proposed model has good calibration ability. The proposed model has superior performance compared with the DeepSurv and multivariate CPH. The result of risk stratification indicates that patients with non-PAF, higher NT-proBNP, larger LAAV and LAV would have higher risks of AF recurrence. CONCLUSIONS: The proposed CNNSurv model has better performance than conventional statistical analysis, which may provide valuable guidance for clinical practice.

Safety and feasibility of implanting a transvenous implantable cardioverter defibrillator (TV-ICD) in the left axilla.

BACKGROUND: Transvenous implantable cardioverter defibrillator (TV-ICD) systems are commonly implanted in the left anterior chest because of an easier implantation and better defibrillation threshold. This study aimed to evaluate the safety and feasibility of left axillary implantations of TV-ICD systems. METHODS: We performed left axillary TV-ICD implantations and compared that to the major complication rate and operation time of the conventional TV-ICD implantation site (left anterior chest). The electrical parameter trends were also assessed in the left axilla group. RESULTS: Seventy-six consecutive patients were evaluated for the analysis. Thirty-one patients had their system implanted in the left axilla and the reasons for the implantations included 29 patients for cosmetic reasons and two for post-infection conditions. The operation time and major complication rate were similar between the two groups (left anterior chest vs. left axilla: 134±62.4 min vs. 114±33.5 min, p = .11, 1/45 patient, 2.2% [pocket hematoma] vs. 1/31 patient, 3.2% [lead dislodgement], p = .77). During the follow up period (4.9±2.3years), no lead interruptions were observed in either group. The electrical lead parameters at the time of the implantation and follow up were similar in the study group (R wave sensing 20.8±33.4 vs. 11.2±7.42 mv, p = .34; lead impedance 464±64.7 vs. 418±135ohm, p = .22; pacing threshold [at 0.4 ms] 1.0±0.76 vs. 1.21±0.93V, p = .49). CONCLUSION: TV-ICD implantations in the left axilla were performed safely without increasing the operation time as compared to the conventional ICD implantation site. ICD implantations in the left axilla are an alternative in those not suitable for implanting TV-ICDs in the conventional implantation site.

Impact of right atrial structural remodeling on recurrence after ablation for atrial fibrillation.

BACKGROUND: Recurrence of atrial fibrillation (AF) after pulmonary vein isolation (PVI) is associated with left atrial (LA) remodeling; however, its association with right atrial (RA) remodeling remains unclear. OBJECTIVE: This study aimed to identify whether RA structural remodeling could predict recurrence of AF after PVI. METHODS: This study prospectively analyzed 245 patients with AF who had undergone PVI. RA and LA volumes were determined by contrast-enhanced computed tomography. Atrial structural remodeling was defined as an atrial volume of ≥110 mL according to previous reports and receiver operating characteristic curve analysis. RESULTS: After excluding 32 patients, 213 patients were analyzed. During a follow-up period of 12 months, 41 patients (19%) demonstrated atrial arrhythmia recurrence after PVI. With the Cox proportional-hazards model, RA structural remodeling was the only predictor of arrhythmia recurrence (hazard ratio, 1.012; 95% confidence interval 1.003-1.021; P = .009). Kaplan-Meier analysis showed that arrhythmia recurrence was more frequent in the RA structural remodeling group compared with the group without RA remodeling (log-rank, P < .001), and the arrhythmia-free survival rates in these groups at 12 months were 68.0% and 91.4%, respectively. Additionally, there was a significant difference in recurrence-free survival after RA structural remodeling in each type of AF (log-rank, P < .001). CONCLUSIONS: RA structural remodeling is a useful predictor of clinical outcome after PVI regardless of the type of AF. Our results suggest that patients without RA structural remodeling may be good candidates for successful ablation with PVI.

Impact of atrial mitral and tricuspid regurgitation on atrial fibrillation recurrence after ablation.

BACKGROUND: Atrial fibrillation (AF) induces functional mitral regurgitation (FMR) and tricuspid regurgitation (FTR) during atrial remodeling. FMR and FTR are associated with AF prognosis, but the effects for AF recurrence after ablation have not been determined conclusively. METHODS: Two hundred thirty nine patients who underwent AF ablation were enrolled. Forty five patients were excluded. In total, 194 patients were analyzed. FMR and FTR were assessed by echocardiography. The left atrial volume index (LAVI) was evaluated by contrast-enhanced computed tomography. RESULTS: Significant FMR and moderate FTR were observed in 15 (7.7%) and in 25 (12.9%) patients, respectively. The severity of tricuspid regurgitation (TR) significant correlated with age, NT-proBNP, and LAVI. During a 13.4 month follow-up period of, 39 patients (20.1%) demonstrated AF recurrence. In the Cox proportional-hazards model, E/e', FTR, and LAVI, were termed as predictor factors of AF recurrence (E/e'. hazard ratio [HR] = 1.117; P = 0.019, significant FTR. HR = 4.679; P = 0.041, LAVI. HR = 1.057; P = 0.003). Kaplan-Meier analysis showed that AF recurrence was more frequent in FTR compared with the nonsignificant FTR cases (log-rank, P = 0.001). Although survival analysis showed no difference with or without FMR, the presence of FMR and FTR was strongly associated with high-AF recurrence (log-rank, P = 0.004). CONCLUSIONS: AF recurrence was associated with E/e', LAVI, and extensive FTR. Specifically, the combination of FTR and FMR markedly worsens the AF prognosis.

Lesion size and adjacent tissue damage assessment with high power and short duration radiofrequency ablation: comparison to conventional radiofrequency ablation power setting.

There is increased interest in creating high-power short duration (HPSD) ablation lesions in the field of atrial fibrillation (AF) radiofrequency ablation (RFA). We evaluated the lesion characteristics and collateral damage using two separate RFA protocols setting (HPSD: 50 W and 7 s vs control: 25 W and 30 s) in vitro model. Sixteen freshly killed porcine hearts were obtained, and the atrium and ventricle slabs were harvested for ablation. The each slabs were placed in a tissue bath with circulating 0.9% NaCl at maintained temperature 37 °C. RFA was performed with 4 mm tip irrigated force sensing catheter. All lesions were ablated under recording the electrical parameters using with Ensite Navx system (St. Jude Medical, St. Paul, Minnesota). After RFA, lesion characteristics were assessed for each lesion. Thirty-five lesions were made for each ablation protocol (total 70 lesions for analysis). Ablation parameters were similar between two groups (HPSD vs control; impedance drop (Ω): 34.2 ± 13.1 vs 36.1 ± 8.65 P = 0.49, contact force (g): 13.9 ± 4.37 vs 14.6 ± 5.09, P = 0.51, lesion size index: 4.8 ± 0.52 vs 4.73 ± 0.59, P = 0.62). Although the lesion volume was similar, the HPSD ablation creates wider but more shallower lesions compared to control group (HPSD vs control; lesion volume: 29.6 ± 18.1 mm3 vs 35.5 ± 17.1 mm3 P = 0.16, lesion diameter: 4.98 ± 0.91 mm vs 4.45 ± 0.74 mm P = 0.0095, lesion depth: 2.2 ± 0.76 mm vs 2.8 ± 1.56 mm P = 0.046). Of these, 38 lesions were assessed for adjacent tissue damage and adjacent tissue damages were more frequent seen in control group (HPSD vs control; 1/19 (5.26%) vs 6/19 (31.5%), P = 0.036). Effective lesions were made with HPSD, thereby reducing RFA procedure time. Although the lesion volume was similar between two groups, collateral damage was less seen in HPSD group attributed by lesion characteristics.

The Association of Cardio-Ankle Vascular Index (CAVI) with Biatrial Remodeling in Atrial Fibrillation.

AIM: Arterial stiffness results in elevated left ventricular filling pressure and can promote atrial remodeling due to chronic pressure overload. However, the impact of arterial stiffness on the process of atrial remodeling in association with atrial fibrillation (AF) has not been fully evaluated. METHODS: We enrolled 237 consecutive patients diagnosed with AF who had undergone ablation; data from 213 patients were analyzed. Cardio-ankle vascular index (CAVI) was used as a marker of arterial stiffness. The left atrial (LA) and right atrial (RA) volumes were determined by computed tomography imaging; atrial conduction and voltage amplitude were evaluated using a three-dimensional electromapping system used to guide the ablation procedure. RESULT: In univariate analysis, CAVI significantly correlated with atrial structural and electrical remodeling (LA volume index, r=0.297, P=0.001; RA volume index, r=0.252, P=0.004; LA conduction velocity, r=0.254, P= 0.003; LA mean voltage, r=-0.343, P=0.001, RA mean voltage; r=-0.245, P=0.015). Multivariate regression analysis revealed that CAVI and plasma levels of N-terminal B-type natriuretic peptide were independent determinants of LA and RA remodeling, respectively. On the other hand, age and LA conduction velocity were independent variables with respect to CAVI. Age-adjusted CAVI was highest in long-standing persistent AF when compared with measures of persistent or paroxysmal AF. CONCLUSION: CAVI was closely associated with biatrial remodeling in patients diagnosed with AF. These results suggest that arterial stiffness may play a significant role with respect to disease progression.

Decreased Defibrillation Threshold and Minimized Myocardial Damage With Left Axilla Implantable Cardioverter Defibrillator Implantation.

BACKGROUND: To reduce myocardial damage caused by implantable cardioverter defibrillator (ICD) shock, the left axilla was studied as an alternative pulse generator implantation site, and compared with the traditional implantation site, the left anterior chest. METHODS AND RESULTS: Computer simulation was used to study the defibrillation conduction pattern and estimate the simulated defibrillation threshold (DFT) and myocardial damage when pulse generators were placed in the left axilla and left anterior chest, respectively; pulse generators were also newly implanted in the left axilla (n=30) and anterior chest (n=40) to compare the corresponding DFT. On simulation, when ICD generators were implanted in the left axilla, compared with the left anterior chest, the whole heart may be defibrillated with a lower defibrillation energy (left axilla 6.4 J vs. left anterior chest 12.0 J) and thus the proportion of cardiac myocardial damage may be reduced (2.1 vs. 4.2%). Clinically, ventricular fibrillation was successfully terminated with a defibrillation output ≤5 J in 86.7% (26/30) of the left axillary group, and in 27.5% (11/40) of the left anterior group (P<0.001). CONCLUSIONS: Clinically and theoretically, the left axilla was shown to be an improved ICD implantation site that may reduce DFT and lessen myocardial damage due to shock. Lower DFT also facilitates less myocardial damage, as a result of the lower shock required.

Efficacy and Myocardial Injury With Subcutaneous Implantable Cardioverter Defibrillators　- Computer Simulation of Defibrillation Shock Conduction.

BACKGROUND: Subcutaneous implantable cardiac defibrillator (S-ICD) systems have a lower invasiveness than traditional ICD systems, and expand the indications of ICD implantations. The S-ICD standard defibrillation shock output energy, however, is approximately 4 times that of the traditional ICD system. This raises concern about the efficacy of the defibrillation and myocardial injury. In this study, we investigated the defibrillation efficacy and myocardial injury with S-ICD systems based on computer simulations. METHODS AND RESULTS: First, computer simulations were performed based on the S-ICD system configurations proposed in a previous study. Furthermore, simulations were performed by placing the lead at the left or right parasternal margin and the pulse generator in the superior and inferior positions (0-10 cm) of the recommended site. The simulated defibrillation threshold (DFT) for the 4 S-ICD system configurations were 30.1, 41.6, 40.6, and 32.8 J, which were generally similar to the corresponding clinical results of 33.5, 40.4, 40.1, and 34.3 J. CONCLUSIONS: The simulated DFT were generally similar to their clinical counterparts. In the simulation, the S-ICD system had a higher DFT but relatively less severe myocardial injury compared with the traditional ICD system. Further, the lead at the right parasternal margin may correspond to a lower DFT and cause less myocardial injury.