ROC Curve Research Articles

The interplay between arterial stiffness and left ventricular diastolic dysfunction in systemic lupus erythematosus

Abstract Introduction Cardiovascular disease (CVD) is a leading cause of morbidity and mortality in systemic lupus erythematosus (SLE) patients. Both arterial stiffness and left ventricular diastolic dysfunction have predictive value and have been reported as early markers of CVD in SLE. The link between arterial stiffness and diastolic dysfunction has been described in patients with established CVD and heart failure with preserved ejection fraction. Yet, it is underexplored in the setting of SLE. Purpose The study aims to investigate the relationship between arterial rigidity measured by cardio-ankle vascular index (CAVI) and left ventricular diastolic dysfunction in patients with SLE. Methods This is a prospective cross-sectional study of adult SLE patients from a single center. Clinical, biological, electrocardiographic, transthoracic echocardiographic and CAVI data were collected for each participant. CAVI and vascular age were calculated using a Fukuda Denshi device. Diastolic dysfunction was defined according to the 2016 ASE/EACVI guidelines. The ROC analysis used the Youden index-associated criterion to determine cut-off values. Results Our cohort included 70 patients, 97.1% female, with a mean age of 53.6±11.7 years and a disease duration of 15.7±10.9 years. All patients had preserved ejection fraction; 50% had diastolic dysfunction. CAVI and vascular age were inversely correlated with the E/A ratio, septal and lateral e’ and positively correlated with the E/e’ ratio, isovolumetric relaxation time and indexed left atrial volume. CAVI (AUC 0.815, 95%CI 0.714-0.915), vascular age (AUC 0.827, 95%CI 0.728-0.927), age (AUC 0.791, 95%CI 0.690-0.892), disease duration (AUC 0.749, 95%CI 0.633-0.864) and NTproBNP (AUC 0.794, 95%CI 0.676-0.912) were associated with diastolic dysfunction. In the univariate analysis the presence of traditional risk factors (any of the following: obesity, hypertension, dyslipidemia, diabetes, smoking), ischemic heart disease, peripheral arterial disease and valvular heart disease were also associated with diastolic dysfunction. In the multivariable regression analysis disease duration (HR 1.1, 95%CI 1.01-1.19), vascular age (HR 1.1, 95%CI 1.04-1.18) and CAVI (HR 3.79, 95%CI 1.78-8.10) were independently correlated with diastolic dysfunction. Patients with CAVI> 8.1 and those with a vascular age> 59 had a RR of diastolic dysfunction of 4.1 (95%CI 1.8-9.1) and 3.2 (95%CI 1.7-6.4), respectively. Conclusions Arterial stiffness, evaluated by CAVI and vascular age, was independently associated with diastolic dysfunction in a cohort of adults with SLE, potentially useful for early detection of functional cardiovascular alterations.

Effective left ventricular ejection fraction by three-dimensional echocardiography is associated with survival and hospitalizations in ventricular secondary mitral regurgitation

Abstract Introduction Left ventricular (LV) systolic function is crucial for determining prognosis of patients with ventricular secondary mitral regurgitation (VsMR). However, both the LV ejection fraction (LVEF) and global longitudinal strain are influenced by the volume overload associated with MR. Effective LVEF (eLVEF) is a novel parameter that accounts for mitral regurgitant volumes and overcomes the load-dependency of LVEF. Pupose We sought to explore the association of eLVEF measured by three-dimensional (3D) echocardiography with mortality and heart failure (HF) hospitalizations in patients with VsMR. Methods We enrolled 212 VsMR patients (mean age 70±13 years, 26% female) from January 2015 to March 2022. 3D-eLVEF was calculated by dividing forward LV stroke volume (SV), measured by 2D-Doppler echocardiography by the LV end-diastolic volume measured by 3DE. The primary outcomes were all-cause mortality and HF hospitalization. Results During a mean follow-up of 22±11 months, 22% VsMR patients died and 28% had HF hospitalizations. In multivariable Cox regression, 3D-eLVEF was independently inversely associated with all-cause mortality (HR 0.93, 95%CI 0.89-0.97, p<0.001) and HF hospitalization (HR 0.95, 95%CI 0.92 – 0.99, p<0.001), outperforming 3D-LVEF and VsMR severity parameters. In ROC analysis, cut-off values associated with all-cause mortality were 3D-eLVEF≤24% (77% sensitivity, 61% specificity, AUC 0.74, 95%CI 0.67-0.79, p<0.0001), 3D-LVEF≤40% (87% sensitivity, 44% specificity, AUC 0.69, 95%CI 0.62-0.75, p<0.0001). Kaplan-Meier analysis demonstrated increasing event rate with 3D-eLVEF≤24% independent of LVEF higher or lower than 40% (log-rank, p<0.001). Patients with 3D-eLVEF≤24% and 3D-LVEF >40% had a worse prognosis than patients with 3D-LVEF≤40% and 3D-eLVEF>24% (log-rank, p<0.001). Those with 3D-eLVEF≤24% had the most HF hospitalizations, irrespective of 3D-LVEF (log-rank, p=0.002)(Figure 1). Conclusions In patients with VsMR, 3D-eLVEF was independently associated with all-cause mortality and HF hospitalization. 3D-eLVEF≤24% had a stronger association with both outcomes compared to 3D-LVEF≤40% and refined the risk stratification.

Performance of effective regurgitant orifice and 3D vena contracta area for the diagnosis of severe mitral regurgitation in patients with localized and diffuse mitral valve prolapse.

Abstract Background Mitral valve prolapse (MVP) is heterogenous, ranging from localized disease to diffuse bi-leaflet involvement in Barlow myxoid degeneration. Mitral regurgitation (MR) may be localized or diffuse, holosystolic or end-systolic, depending on the valve morphology. We aimed to compare the diagnostic accuracy of effective regurgitant orifice by conventional 2D PISA (ERO) on transthoracic echocardiography (TTE) with 3D vena contracta area (3D-VCA) on transoesophageal echocardiography (TOE) in patients with localized and diffuse MVP. Methods Patients with MVP assessed with TTE, 3D-TOE and cardiac magnetic resonance (CMR) for moderate to severe MR were included. MR was considered severe when the regurgitant fraction (RF) by CMR was ≥40%.1 3D-VCA was acquired in multi-beat mode during breath-hold and 3D-VCA was measured at end-systole. The diagnostic accuracy of 3D-VCA and ERO were compared using ROC analysis in patients with diffuse (6 prolapsing scallops) versus localized prolapse Results 77 patients had undergone all three exams. Diffuse bi-leaflet prolapse was present in 24 patients (31%, Barlow). In the group with localized prolapse (69%, non-Barlow), 90% had a single scallop prolapse, and 85% had a flail leaflet. 53 of the patients (70%) had severe MR by CMR, and severe MR was less frequent in Barlow (46%, versus 79% in non-Barlow). ERO could not be calculated accurately by TTE in 10 patients (8 with non-Barlow, 2 with Barlow). Hence, 67 patients were included in the comparative analysis. On average, 3D-VCA overestimated ERO by 20mm2 in non-Barlow and by 24mm2 in Barlow (Figure 1). Diagnostic accuracy of 3D-VCA and ERO was comparable in non-Barlow (AUC 0.79 vs 0.84, p=0.45), but 3D-VCA tended to perform poorly in Barlow (AUC 0.66 vs 0.84, p=0.2) (Figure 2). Conclusion 3D-VCA measured in end-systole consistently overestimated ERO as determined by the conventional 2D PISA method on TTE. While its diagnostic performance was comparable to ERO in the case of localized prolapse, it seems to demonstrate low performance in the context of Barlow disease. Figure 1 Figure 2

Left atrial stiffness index in patients with cryptogenic ischemic stroke: a clue of atrial fibrillation?

Abstract Introduction About 25% of cryptogenic ischemic stroke are due to undiagnosed atrial fibrillation. New echocardiographic markers derived from left atrial strain analysis could be useful to early identify patients candidates to oral anticoagulation. Purpose Aim of our study was to investigate the relationship between left atrial strain parameters assessed by speckle tracking echocardiography (STE) and the new documentation of atrial fibrillation at follow-up in patients admitted for acute cryptogenic ischemic stroke. Methods From 2020 to 2023 consecutive patients admitted to Stroke Unit of our Hospital for cryptogenic ischemic stroke undergoing loop-recorder implantation were retrospectively enrolled in our study. Main exclusion criteria were previous diagnosis of atrial fibrillation and poor image quality. All patients underwent full echocardiographic examination, including left atrial strain analysis with STE by dedicated software. Results Sixty-nine patients were finally selected, divided in patients with and without atrial fibrillation at follow-up (26 and 43 pts, respectively). No clinically significant differences were found between the two groups at baseline. Analysing echocardiographic parameters, significant differences were found in patients with and without diagnosis of atrial fibrillation at follow-up, respectively, with regard to left atrial volume index (LAVi, 44.26 ± 10.58 vs 34.54 ± 11.03 ml/m2, p 0.005), peak atrial longitudinal strain of reservoir phase (PALSr 19.50 ± 8.48 vs 26.22 ± 9.11%, p 0.01) and contraction phase (PALSct 5.70 (0; 20.92) vs 11.23 (5.98; 16), p 0.017). Left atrial stiffness index was defined as the ratio between E/E’ (expression of left ventricular filling pressures) and PALSr. Using a cut-point of 0.55, identified with ROC analysis, a significant difference was found between the two groups (10/26 (38.4%) vs 4/43 (9.3%), p 0.002). Among all these parameters, at the multivariate logistic regression analysis only left atrial stiffness index showed a trend to the significance (p 0.06). Conclusions Left atrial strain analysis can be useful to better select patients with cryptogenic stroke who need loop-recorder implantation and a closer follow-up to prompt recognize atrial fibrillation. In particular, Left atrial stiffness index seems to be the most useful parameter to stratify patients at higher risk of atrial fibrillation.

3D is superior to 2D echocardiography in identifying high risk patients with heart failure with reduced ejection fraction and indication for device therapy

Abstract Background In patients with heart failure with reduced ejection fraction (HFrEF), a left ventricular ejection fraction (LVEF) less than 35% has been identified as a significant risk factor for increased mortality. As a result, all studies assessing the indications and outcomes of implantable cardioverter-defibrillators and cardiac resynchronization therapy in HFrEF used this cut-off, measured by 2-dimensional echocardiography (2DE). However, many patients with HFrEF and a LVEF >35% by 2DE, who do not qualify for device therapy, still experience significant mortality. Purpose To investigate whether, in patients hospitalized with both ischemic and non-ischemic HFrEF, the LVEF measured by 3-dimensional echocardiography (3DE) could identify better patients at high risk of death, thereby indicating the need for device therapy. Methods 322 consecutive patients (60±11 years, 72% men), hospitalized for HFrEF or HFmrEF, were assessed using both 2DE and 3DE, including full-volume acquisitions of the LV. From 2D datasets, LV end-systolic (LVESV) and LV end-diastolic (LVEDV) volumes, and LVEF were estimated. From the 3D volumes, the same parameters were measured. Patients were discharged on optimal HF treatment, according to current ESC guidelines, and were followed for 24±8 months after the initial event. The primary outcome was cardiovascular death (CD). Results At admission, 43% patients were in NYHA class II, 25% in NYHA class III, and 33% in NYHA class IV. There was a similar burden of ischemic and non-ischemic etiology (52% vs 48%). Mean 2D LVEF was 31±11%, while mean 3D LVEF was 30±10% (p=NS). There were 62 CD (19%) during follow-up. Binary logistic regression for the 2D and 3DE parameters showed that only 3D LVEF was a significant predictor of death (β=-0.04, OR 0.96, 95%CI 0.93-0.98, p=0.003). By 2DE, 206 patients (64%) had LVEF <35%, of whom 47 patients died (23%) and 159 patients survived (77%), whereas by 3DE, 215 patients (67%) had LVEF <35%, of whom 60 patients died (28%) and 155 patients survived (72%). Consequently, 2D LVEF <35% identified 47 out of 62 deaths (76%), whereas 3D LVEF <35% identified 60 out of 62 deaths (97%). By ROC Analysis, 3DE was superior to 2DE for prediction of survival: AUC 0.62, p=0.03 for 3D LVEF vs. AUC 0.56, p=0.09 for 2D LVEF. Kaplan-Meier analysis (Figure) showed that 2D LVEF <35% discriminated between survivors and non-survivors with a Chi-square of 6.8, p=0.01, whereas 3D LVEF <35% discriminated between survivors and non-survivors with a Chi-square of 30.3, p<0.001. Conclusion 3DE assessment of LVEF and its cut-off of 35% enhances the identification of patients with HFrEF at high risk of death. This suggests that 3DE should be utilized for risk stratification in HFrEF, potentially leading to more accurate indications for device therapy. Survival according to 2D vs 3D LVEF

Apical rocking and septal flash revealed by an electrocardiographic parameter in the left bundle branch block

Abstract Background Assessing the relationship between electrical and mechanical dyssynchrony in patients with left bundle branch block (LBBB) remains a significant challenge for clinicians. Visual echocardiographic parameters, apical rocking (ApR), and septal flash (SF) are accurate markers of the contraction-and-stretch pattern in the septal and lateral wall in LBBB patients. Electrical activity of the LV apical septum and the LV lateral wall is ECG recorded by the aVL and aVF leads. An ECG parameter that integrates the electrical activation of these two contralateral LV walls could reveal subtle electrical features related to the uncoordinated contraction pattern in LBBB. To the best of our knowledge, no studies have examined the relationship between electrical and mechanical uncoordinated activation of the apical septum and lateral LV wall. Purpose This study aimed to assess the relationship between visual echocardiographic parameters of mechanical uncoordinated contraction pattern (ApR and SF) and ECG parameters of electrical intraventricular asynchrony in patients with LBBB. Methods We retrospectively included 157 consecutive patients with LBBB, in sinus rhythm, without significant coronary artery disease, and no more than mild organic valvular heart disease. The ApR and SF were echocardiographic visually assessed as a typical sequence of the septal-to-lateral apex motion for the ApR and the short leftward movement of the septum in early systole for the SF. We measured the following ECG parameters: LV activation times (LVAT), RR1 interval (as the longest interval from the first R wave to the last R1 wave) in lateral leads, intrinsicoid deflection (ID) in lateral and aVF leads, and a newly defined marker – the difference (Diff) between aVL and aVF ID’s (Diff_IDaVL_aVF). Based on the presence and absence of both ApR and SF (ApR_SF), the patients were divided into two groups: patients with (68.2%) and without (31.8%) ApR_SF. Results In the group of patients with ApR_SF, the LVAT, ID, and RR1 measured in lateral leads, and the Diff_IDaVL_aVF were significantly longer, but IDaVF was significantly shorter compared to the group of patients without ApR_SF (p<0.05) (Table 1). In multivariable analysis, the Diff_IDaVL_aVF (p < 0.001) was the only parameter independently correlated with the presence of ApR_SF (Table 2). On ROC analysis, the best cut-off of DIFF_IDaVL_IDaVF for predicting the presence of ApR_SF in LBBB patients was 46 ms (AUC of 0.92, sensitivity 86%, and specificity 80%) (Figure 2). Conclusions In LBBB, a simple ECG parameter incorporating the electrical activation of contralateral LV myocardial segments can provide insights into the myocardial contraction pattern in LBBB patients. A Diff_IDaVL_aVF longer than 46 msec identified LBBB patients with specific contraction patterns (ApR_SF). Considering ECG is the screening tool for LBBB diagnosis, these patients might benefit from early treatment to avoid a worse prognosis.

P-1579. Prediction of Prognosis of Visceral and Subcutaneous Fat Areas in Pyelonephritis

Abstract Background Visceral fat accumulation has been regarded as a poor prognostic factor for coronary artery disease. Moderate overweight and mild obesity are associated with a lower mortality in patients with several infectious diseases on intensive care units, and this phenomenon is referred to as the obesity paradox. Pyelonephritis is one of the common infectious diseases, and although computed tomography (CT) imaging is frequently used in its localization, its prognostic value is limited. In this study, we analyze the relationship between the visceral and subcutaneous fat areas in CT scans and prognosis of patients with pyelonephritis.Fig. 1ROC curves using subcutaneous and visceral fat area to discriminating surviving pyelonephritis patients from deceased pyelonephritis patientsThe black line shows the curve using subcutaneous fat area and the gray line shows the curve using visceral fat area. Methods We included hospitalized patients with pyelonephritis who underwent computed tomography of the abdomen. Clinical data were obtained from the medical records. Subcutaneous and visceral fat aeras were analyzed at the umbilical level using the abdominal analysis (2D) function of the 3D image analysis system volume analyzer. Results 505 patients were included in the study. There were 497 survivors and 8 non-survivors, with both subcutaneous and visceral fat areas being lower in the non-survivors than in the survivors (subcutaneous fat area 49.3 ± 73.9 cm2 vs 123.7 ± 85.3 cm2, p=0.012; visceral fat area 42.2 ± 29.2 cm2 vs 98.4 ± 69.1 cm2, p< 0.01, respectively. ROC analysis using subcutaneous and visceral fat areas to discriminate survival and death showed that the areas under the curve of subcutaneous and visceral fat areas were 0.794 (p=0.002) and 0.770 (p< 0.001), respectively (Fig.1). Conclusion Patients with poor prognosis of pyelonephritis tended to have lower subcutaneous and visceral fat area at onset than patients with improved prognosis. The results suggest that both subcutaneous and visceral fat area using CT may have prognostic value for pyelonephritis. Disclosures All Authors: No reported disclosures

Could Urinary Amino Acids be as New Biomarkers for Detection of Sarcopenia?

Abstract Background Sarcopenia is a syndrome that cause falls, fractures, and morbidity in geriatric patients. Early diagnosis of sarcopenia is important as it is known that muscle functions can improve with early intervention. We aimed to investigate whether urinary amino acid levels could serve as biomarkers of sarcopenia. Methods The study included ninety-one patients aged 45–65 who visited to our outpatient clinic. The patients underwent physical examinations, blood tests and sixteen distinct urinary amino acid levels were analyzed. Anthropometric measurements were performed. Physical performances were evaluated. Muscle strengths were measured. Muscle masses were analyzed. Patients were divided into 4 groups: pre-sarcopenic, sarcopenic, severe sarcopenic and non-sarcopenic. Statistical significance level was determined as p < 0.05. Results A total of ninety-one patients, 53 females and 38 males, were included in the study. Three patients had pre-sarcopenia, eleven had sarcopenia, and two met the criteria for severe sarcopenia, while seventy-five patients were non sarcopenic. A significant difference was found between the sarcopenia and non-sarcopenia groups in terms of glutamine and valine levels (p < 0.001 for both). In the ROC analysis, the cut-off value of glutamine and valine levels in detecting sarcopenia was determined as 492 µmol/L and 209 µmol/L (AUC: 0.875 and 0.968). Correlation analysis revealed a negative relationship between urinary leucine levels and both muscle strength and muscle mass Conclusions We found that patients with sarcopenia had high urinary glutamine and valine levels and that urinary leucine levels were associated with both muscle strength and muscle mass.

Prognostic predictors in constrictive pericarditis, 19-year experience at a tertiary care hospital

Abstract Introduction Constrictive pericarditis is a chronic inflammatory entity that may progress with thickening and calcification in the pericardial tissue, resulting in decreased pericardial elasticity and subsequent increase in ventricular filling pressures1,2. This rare treatable cause of HFpEF, is associated with various aetiologies, depending on the development level of the countries. Several algorithms have been developed for the diagnosis, whereas studies evaluating the prognostic factors are limited3–6. Our study aimed to examine the clinical, laboratory, electrocardiographic, imaging and catheterization parameters that determine the prognosis of constrictive pericarditis. Methods 60 patients diagnosed with constrictive pericarditis based on echocardiography, computed tomography and right heart catheterization parameters were included retrospectively in the study. Clinical presentation, laboratory, and electrocardiographic findings were assessed. Volumetric measurement, tissue Doppler, mitral inflow velocity analysis, and pericardial evaluation were made via 2D echocardiography. Pericardial thickness, effusion, and calcium score were obtained via thorax CT, while right-sided pressure measurements were acquired via catheterization. The primary clinical outcome was determined as mortality, and the effect of these parameters was evaluated by Cox regression analysis. Results Median follow-up was 3.5 years (0-19) and mean age was 52.7 ±17.5 years. 38 of the patients (63.3%) were male, the most common aetiologies were malignancy (21.4%) and other systemic diseases (21.4%). 7 patients were lost to follow-up and 18 (31.6%) patients died during the process. When different models were evaluated with Cox regression analysis, ascites, hemoglobin and elevated proBNP values were prognostic predictors (HR 10.991 p-value 0.006, HR 0.513 p-value 0.010 and HR 3.904 p-value 0.045, respectively). ROC analysis indicated the highest AUC value for the model that included hemoglobin (AUC 0.953). Conclusion Constrictive pericarditis is one of the treatable causes of HFpEF. While imaging and catheterization parameters play a crucial role in the diagnosis, clinical and laboratory parameters affect the prognosis.

Left ventricular global longitudinal strain (LV-GLS) by speckle-tracking echocardiography (STE) predicts appropriate defibrillator therapy in patients with dilated cardiomyopathy (DCM)

Abstract Background Patients with dilatied cardiomyopathy (DCM) have an increased risk for sudden cardiac death (SCD) which may be prevented by an implantable cardioverter-defibrillator (ICD). Current guidelines set the indication to implant an ICD in DCM patients based on a left ventricular ejection fraction (LVEF) of ≤ 35 %. However, this approach has been questioned by recent research in non-ischemic DCM. Left ventricular global longitudinal strain (LV-GLS) as assessed by speckle-tracking-echocardiography (STE), directly measuring myocardial deformation in the longitudinal dimension, is a new method to characterize global LV function. Data is scarce regarding the diagnostic value of LV-GLS in predicting sudden cardiac death (SCD) events. Purpose The aim of this study was to assess whether LV-GLS measurements predict the rate of appropriate ICD therapies in DCM patients better than LVEF. Methods Patients with DCM who underwent primary prophylactic ICD implantation from 2011-2016 at our hospital, who had baseline echocardiographic data adequate for STE analysis, and a follow-up duration of > 6 months were included. ICD interrogation data were retrieved from hospital electronic records. Patients with secondary preventive ICD indication were excluded. Results 115 patients were included (26% females, age: 61.3 ± 12.4 years, LVEF 27.9 ± 5.4%, NYHA functional class 2.7 ± 0.5). During a follow-up time of 35.2 ± 12.3 months 30 patients had at least one appropriate ICD therapy (25 patients had anti-tachycardia pacing (ATP), 8 patients had ICD-shocks). Mean LV-GLS in patients with ICD therapy was significantly worse than in those without ICD therapy (-6.4% vs. -7.8%, p=0.018). Multivariate logistic regression analysis showed that LV-GLS (odds ratio 1.31, 95% confidence interval: 1.07-1.60, p= 0.01) but not LVEF (OR 0.97, p= 0.47) was an independent predictors of appropriate ICD therapy. Using ROC analysis, we found a cut-off value of -9,7% for LV-GLS to best predict ICD therapy over follow-up with a sensitivity of 93.3% sensitivity (c-statistic 0.64). Conclusion Our findings indicate that LV-GLS is a useful marker to risk stratify patients with DCM, outperforming the current standard LVEF. LV-GLS should be implemented in the standard echocardiographic assessment of patients with cardiomyopathy.

Noninvasive left atrial stiffness is associated with atrial fibrillation recurrence after catheter ablation

Abstract Background Catheter ablation for atrial fibrillation (AF) improves quality of life and lowers cardiovascular adverse events. However, AF recurrence rates remain high - thus the need for better or improved predictors of AF recurrence. The role of left atrial (LA) stiffness remains unclear, as prior findings were controversial. Purpose To explore the relationship between a noninvasive LA stiffness index (LAstiffn) and AF recurrence after catheter ablation. Methods We analyzed a prospectively recruited cohort of patients who underwent catheter ablation for AF, were in sinus rhythm at 3 months post-ablation, had no significant (< moderate) valvular disease, and had LA volumes and LA diastolic function prospectively measured. LAStiffn was calculated as (E/e’)/LA emptying fraction (LAEF) ratio (abnormal defined as >15/45≈0.3). Outcome endpoint was the first documented AF/atrial flutter episode, lasting >30s, beyond 3 months post ablation (end of blanking period). Results A total of 589 patients were studied (mean age 60±10 years; 75% men; 50% with persistent AF; Table 1). At 3 months post-ablation, mean LAVI(Left atrial volume index) was 36±10 mL/m2 and LAstiffn 0.32±0.22. LAstiffn tended to be higher in older patients, in females, and patients with persistent AF and comorbidities (hypertension, congestive heart failure, coronary artery disease, CHA2DS2-VASc score ≥2, etc). At multivariable regression analysis, LAstiffn correlated with LV size and mass, degree of mitral/tricuspid regurgitation, and right ventricular systolic pressure. At 1 year, 3 years, and 8 years follow up, AF recurred in 118, 191, and 347 patients, respectively. At univariate Cox proportional hazards regression analysis, LAstiffn index >0.3 was significantly associated with AF recurrence at 1 year (log-rank P=.0005), 3 years (P=.0002), and 8 years follow up (P<.0001; Figure). In multivariate analysis including all clinical and laboratory factors with P<.10 in univariate analysis, LAstiffn >0.3 remained an independent factor for AF recurrence only at 3-years (adjusted HR 1.40 [1.03-1.91], P=.03) and 8 years follow up (adjusted HR 1.65 [1.29-2.13], P<.0001) after adjustment for age, sex, type of AF (persistent/paroxysmal), LA volume index (LAVi), obstructive sleep apnea, and prior pacemaker/ICD implantation (best independent factors); P=.14 at 1 year follow up. In ROC analysis, LAstiffn (AUC 0.61, P=.001) performed slightly better than LA deformation LAEF (AUC 0.59, P=.002) or E/e’ (AUC 0.58, P=.03), though with modest sensitivity/specificity (58%/62%). Conclusions LAstiffn may reflect LA myopathy that is likely due to the cumulative effects of comorbidities, diastolic dysfunction, and intrinsic LA tissue remodeling. LA stiffening may be a substrate facilitating redevelopment of AF, thus patients with increased LAstiffn may need more frequent monitoring to prevent/treat AF recurrence earlier.

P-1824. AI-Powered Differential Diagnosis: Distinguishing MIS-C from Kawasaki Disease

Abstract Background Distinguishing Multisystem Inflammatory Syndrome in Children from COVID-19 (MIS-C) from Kawasaki Disease (KD) can be challenging due to several overlapping clinical presentations. Clinical management of these conditions can vary. Hence, timely and accurate diagnosis is essential. We evaluated the potential of laboratory findings in the differentiation using machine learning (ML) models. Methods Patient demographics (Age, Sex) and laboratory data were compared between MIS-C and KD. Simultaneous CBC and routine biochemistry tests were evaluated. ML models were created using PyTorch, scikit-learn, Pandas, NumPy, and Matplotlib libraries. Specifically, Naive Bayes Classifier, KNN, Logistic Regression, SVC, Random Forest Classifier, XGBoost, GBM and LightGBM models were created and the model performances of the classifiers were evaluated with accuracy percentages. The confusion matrix of every learning model was also generated. K-fold CrossValidation was also performed to evaluate the predictive models. ROC and feature importance analysis were also performed. Identification of the disease as MIS-C or KD based on both regressive (Eg: Total Protein) and binary features (Eg: Sex) is posed as a binary classification ML problem. Train and test datasets were created using a couple of train-test splits (0.2 and 0.25). Results 209 paediatric patients with MIS-C (n=140), and KD (n=69) were included. • The accuracy performances of the models in the test set were determined as 67.8, 73.8, 75.9, 76.8, 79.9, 80.4, 81.0, 81.1%, respectively. • The area under curve (AUC) values of the LightGBM, GBM, XGBoost, were determined as 79.1, 79.5 and 78%, respectively. • Permutation importance analysis of the LightGBM model showed that elevated platelets, neutrophil count, SGPT, WBC levels, Creatinine, and NT ProBNP had influenced the model’s explainability. Conclusion Although four years have passed since the onset of COVID-19, the early diagnosis of MIS-C is still crucial. In addition to clinical findings, laboratory tests may contribute to the fight against the disease by providing early diagnosis of MIS-C and KD. Explainable ML models can help determine the appropriate markers for this purpose. Disclosures Aditya Basu, Bachelor of Technology, Triomics Healthcare: Stocks/Bonds (Private Company)

P-2241. Monocyte Distribution Width (MDW) as a Biomarker to Distinguish Septic Shock with Bacteremia from Cardiogenic Shock

Abstract Background Shock is associated with high mortality in intensive care units with resultant urgency for rapid diagnosis. The monocyte distribution width (MDW) is a biomarker that is elevated in sepsis and septic shock and is readily obtained from routine leukocyte differential testing. It is unknown whether MDW is also elevated in other kinds of shock. The purpose of this study was to evaluate the potential of MDW to discriminate between cardiogenic and septic shock. Boxplot of the MDW values by those with cardiogenic shock vs those with septic shock (A). The receiver operator curves for the prediction of septic shock (B) and cardiogenic shock (C) using MDW cut-off values, resulting in an AUC of 0.892. Methods This study was a single-center retrospective analysis of adult patients admitted between 1/31/2023 and 12/27/2023 with either cardiogenic shock requiring inotropic or mechanical support or septic shock with bacteremia requiring vasopressor support derived from EHR database search for diagnosis codes. All patients required hemodynamic support and were excluded if any confounding immunosuppressive medications or conditions (e.g., advanced HIV infection, transplant, active malignancy on treatment) were present. Patients with cardiogenic shock were excluded if concomitant infection at the time of admission was identified. MDW was collected at the time of initial presentation. Study population demographics and results by shock category. Results 171 patients met inclusion criteria: 81 with septic shock and 90 with cardiogenic shock. 56.7% of patients with cardiogenic shock received empiric antibiotics on admission. The MDW was significantly higher in the septic shock group. Using estimated cut-off values of MDW > 21.53 for septic shock and MDW < 28.99 for cardiogenic shock each produced an AUC (area under the curve) of 0.892. Elevated SOFA score, elevated WBC, and female sex were associated with increased odds of septic shock. Adding the WBC count to MDW in the logistic regression for septic shock only minimally improved the AUC. Conclusion MDW may be a useful tool in stratifying the risk of septic or cardiogenic shock. MDW above 28.99 supports a diagnosis of septic shock with bacteremia while MDW below 21.53 supports a diagnosis of non-septic shock; values from 21.53-28.99 are less predictive. A low MDW in patients with shock may exclude the need for empiric antibiotics and improve stewardship. Further work is needed to better characterize the performance of this assay in other populations and shock phenotypes. Disclosures Ryan C. Maves, MD, AiCuris: Grant/Research Support|Biotest: Grant/Research Support|GeoVax: Grant/Research Support|Shionogi: Advisor/Consultant|Shionogi: Honoraria|Sound Pharmaceuticals: Grant/Research Support

Pressure-volume analysis by three-dimensional echocardiography: a novel and superior method to predict response following cardiac resynchronization therapy

Abstract Introduction Long-term response to cardiac resynchronization therapy (CRT) is primarily assessed by 2D echocardiography. The combination of 3D echocardiography and left ventricular (LV) pressure curve estimation allows for pressure-volume (PV) analysis and LV work efficiency (LVWE) estimation which may be superior to today’s methods. Aim To investigate if PV-analysis by 3D echocardiography is superior to 2D volumes at predicting long-term survival in CRT. Methods In a recent prospective multicenter study, non-invasive PV-analysis (n=106) and 2D volume analysis by biplane Simpson’s method (n=195) were performed, before and 6 months after CRT. The LV pressure curve was estimated by aligning a reference pressure curve with mitral and aortic valve events obtained from echocardiography. The amplitude of the curve was adjusted to brachial systolic cuff pressure and combined with 3D volumes for PV-analysis. As shown in figure 1, LVWE was calculated as the ratio of estimated stroke work (red area) to total mechanical energy (red + blue areas). The cutoff value for change in LV end-systolic volume (ESV) was set at 15% and at >35% for LVWE using Receiver Operator Curve analysis. Volumetric non-responders were subdivided into response based on LVWE. Results After 6±2 years follow-up, 47 (23%) patients had died. Relative change in 3D and 2D LV ESV showed similar ability to predict long-term all-cause mortality (log rank p<0.001 for both) (Figure 2). However, LVWE identified patients with a positive effect of CRT who were volumetric non-responders. These patients had similar outcomes as volumetric responders (HR 1.91, 95% CI: 0.54-6.81) (Figure 3). LVWE also identified non-responders with particularly adverse outcomes (HR 3.46, 95% CI: 1.32-9.10 vs volumetric non-response and LVWE >35%). Conclusions PV-analysis by 3D-echoardiography identifies a new group of patients who respond to CRT without a 15% reduction in ESV who have similar outcomes to volumetric responders. In our data, 3D echocardiography was superior to 2D volumes at predicting long-term all-cause mortality after CRT. LV work efficiency before and after CRT LV work efficiency and survival

Regional left ventricle strain outperforms global longitudinal strain in predicting late gadolinium enhancement in arrhythmogenic cardiomyopathy

Abstract Background Arrhythmogenic cardiomyopathy (AC) is a genetically determined heart muscle disease characterized by myocyte necrosis and consequent fibro-fatty replacement. The recognition of fibrosis is a relevant finding with important diagnostic and prognostic implications. Speckle tracking echocardiography (STE) may be more accurate than conventional echocardiography in detecting early myocardial alterations that need to be confirmed with the gold standard technique, cardiac magnetic resonance (CMR). Purpose The aim of our study is to evaluate if STE-derived strain analysis can predict the presence of fibrosis on CMR in the corresponding segment. Methods 86 patients with AC (including right/left dominant and biventricular forms) underwent advanced echocardiography with STE-derived strain analysis and CMR. Using a 17-segment Bull's-eye model, the cut-off value that best predicts the presence of gadolinium enhancement on CMR was determined for each wall segment using ROC curve analysis. Results The medium age was 45 (32-58) years, male gender was prevalent 71%. Among our patients, 44.1% had a biventricular form, 26.7% had a right-dominant form, 26.7% had a left-dominant form, and 2.3% had a borderline diagnosis. Regarding cardiovascular risk factors, 11.6% of patients had arterial hypertension, 15.1% had dyslipidemia, 2% had diabetes, 1% were smokers and 1% had coronary artery disease with previous percutaneous revascularization. The mean left ventricle (LV) ejection fraction (EF) was preserved 59% (54-64), while the medium LV global longitudinal strain (GLS) was slightly reduced, -18.2% (-20.1 to -15.2). A detailed analysis of regional wall strain was performed, and ROC curve analysis was used to determine a cut-off value for each segment to predict the presence of fibrosis on CMR in the corresponding segment, as shown in Table 1. Lower-than-average values demonstrated good accuracy in predicting the presence of fibrosis on CMR. Conclusions STE-strain analysis is a clinically accessible tool that enhances the sensitivity of echocardiography in detecting myocardial tissue alterations in AC, which would otherwise be identifiable only through CMR. Notably, even when GLS is within normal limits, regional strain sub-analysis can additionally reveal myocardial involvement. Table 1

Pulmonary artery acceleration time: a valid alternative when tricuspid regurgitation is missing

Abstract Background the latest guidelines from the European Society of Cardiology (ESC) on the diagnosis of pulmonary hypertension recommend the use of peak tricuspid regurgitation velocity (TRV) as an initial measure. However, in a significant proportion of patients this measurement is not feasible due to the absence of regurgitation or inadequate quality of the Doppler signal curve. Pulmonary artery acceleration time (PAAT), on the other hand, is a measurement that can be easily obtained in nearly all patients. Purpose the aim of our study is to analyse the relationship between PAAT and TRV and to evaluate its diagnostic accuracy for the TRV values established in the latest ESC guidelines (2.8m/s and 3.4m/s), determining appropriate cut-off points for this purpose. Methods a retrospective analysis was conducted on 1,255 consecutive transthoracic echocardiograms in which both TRV and PAAT were measured. Patients with pulmonary stenosis (peak velocity ≥ 2m/s) were excluded. The relationship between both variables was analyzed using different linear and non-linear models, choosing the one with the lowest degree of residuals. Diagnostic accuracy was assessed using the area under the curve (AUC) and sensitivity (Se) and specificity (Sp) values. Optimal cut-off points (OCP) were derived from ROC curve analysis. Results the mean age of the sample was 72.5±13.8 years, with 54% (683) being women and 63% (795) in sinus rhythm. Eighty six patients (7%) exhibited severe, massive or torrential tricuspid regurgitation. The mean TRV was 2.77±0.47m/s and the mean PAAT was 100.50±29.40ms. The mean TAPSE was 21.34±4.70mm, with 196 patients (16%) having systolic dysfunction (defined as TAPSE < 17mm). The relationship between TRV and PAAT fits a quadratic model (p for the quadratic term <0.001). The diagnostic accuracy of PAAT for both TRV cut-off points was good: TRV > 2.8m/s: AUC 0.78 [95% CI 0.76-0.80], OCP 93ms [Se 72.4%, Sp 71.8%]; TRV > 3.4m/s: AUC 0.80 [95% CI 0.78-0.82], OCP 85ms [Se 77.5%, Sp 70.6%]. Conclusions the relationship between TRV and PAAT follows a non-linear model. Cut-off points of 93ms and 85ms demonstrate good discriminatory power for identifying patients with TRV superior to 2.8m/s and 3.4m/s, respectively. Therefore, PAAT constitutes a reliable alternative for the initial echocardiographic assessment of the pulmonary hypertension probability in patients in whom TRV cannot be determined. Relationship PAAT/TRV

P-1844. T cell Subtypes and DNA Methylome Dynamics Predict Infection after Kidney Transplantation

Abstract Background Infectious complications continue to cause significant morbidity and mortality after kidney transplantation (KTx), especially in older patients where immune senescence increases impact of immunosuppression. Given the importance of T cell function and of epigenetic regulation in infection vulnerability, we sought to evaluate whether the DNA methylome and T cell phenotypes could predict development of infection after KTx. Principal component analysis (PCA) of CD4 and CD8 subsets associated with PBMC collected pre (green) compared with post (blue) transplantation, as well as T cell subsets associated with development of infection (red) versus freedom from infection (orange). Methods We evaluated PBMC from 91 KTx patients with samples collected before and then 3 months after transplantation. Deep flow cytometry was performed to delineate naïve, central memory (CM) effector memory (EM), and senescence subtypes including TIM3 and TIGIT in CD4 and CD8 T cells. Targeted bisulfite sequencing (TBS-seq) was applied to PBMC, utilizing probes directed at regions important for T cell maturation, activation, and control of viral infections. Multivariate multiple linear regression model. A Spearman correlation matrix of actual-prediction traits. B Distribution of predicted values of ATG (left) and infection risk (right). C ROC curves of ATG (left) and infection risk (right). Results Transplantation had significant impact on T cell phenotypes, with increase in CD4 memory T cell subtypes CM TIM3+ (p< 0.001) and EM CD127+ (p=0.001) and decrease in CD4 naïve T cells, especially after receipt of antithymocyte globulin (ATG) induction (p=0.008). CD4 T cell subtypes measured pre-transplant were also significantly associated with infection after KTx, including CM (p=0.014) and EM TIGIT (p< 0.001). PCA analysis incorporating multiple maturation subtypes demonstrated ability to separate patients by transplant status and development of infection (Figure 1). Differential DNA methylation similarly demonstrated ability to predict infection: Multivariate linear regression identified significant association with infection post KTx with correction for age, sex, CMV serostatus, and ATG induction (Figure 2A). An integrated epigenetic risk score was associated with ATG induction and development of infection (Figure 2B), with AUC of 0.90 and 0.79, respectively (Figure 2C). Conclusion Immune markers based on T cell phenotype and differential methylation demonstrate the ability to predict infection after KTx. Pre- and post-transplant immune monitoring has the potential for patient risk stratification and individualization of immunosuppression to improve clinical outcomes by decreasing infection risk after transplantation. Disclosures Joanna M. Schaenman, MD, PhD, FAST, Eurofins Viracor: Honoraria|F2G: Grant/Research Support|MedCure: Advisor/Consultant|Moderna: Clinical trial support to institution|OneLegacy: Advisor/Consultant

Evaluation of the presence and extension valvular-related cardiac damage in patients with moderate aortic stenosis and its impact on outcomes

Abstract Introduction and objective Echocardiographic measurements form the basis for diagnosing and determining the indication for therapy in aortic stenosis (AS). These measurements may vary depending on different patient-related and operator-related conditions. Recently, staging cardiac damage has emerged as a prognostic tool in asymptomatic and moderate to severe AS. We aim to evaluate the presence and severity of cardiac damage (1,2), and its impact on outcomes, in patients with a first diagnosis of moderate and severe AS. Methods A total of 119 patients undergoing transcatheter aortic valve replacement (TAVR) from November 2019 to January 2021 were included in a single-center retrospective registry. In each patient, the first echocardiogram establishing the diagnosis of moderate and severe AS was analysed. Results 119 patients with a mean age of 82.8±5.33 years (51% men) were included. At the first echo study with the diagnosis of moderate AS, mean pressure gradient was 23±9 mmHg, maximum pressure gradient was 42±15 mmHg, and mean aortic valve area was 1.13±0.41 cm2 (0.64±0.23 cm2/m2). Regarding cardiac damage, mean left ventricular ejection fraction was 65.4±10.4%, mean longitudinal global strain was -18.5±5.4%, and mean left atrial volume index was 41.6±31.2 mL/m2. 16% of patients have significant mitral regurgitation, and 10% right ventricular dysfunction. The mean progression time between moderate and severe diagnoses was 36 (2.4-175.2) months. Figure 1 shows the distribution of patients in different staging classification systems (1,2) at the time of diagnosis of moderate and severe AS. According to one of these staging classifications (1), 75.7% of patients had left-sided cardiac damage (Stages 1 and 2) and 16.8% had right-sided damage (Stages 3 and 4), without differences compared to the same patients when the stenosis became severe (p=0.194). On the contrary, following the most recent second classification (2), 63% had no significant cardiac damage (Stage 0), 31.1% of patients had left-sided cardiac damage (Stages 1 and 2) and 5.9% had right-sided damage (Stage 3), and this system showed better ability to discriminate between moderate and severe AS related-damage (p=0.002). The area under the ROC curve for one-year mortality after TAVR for the 2 staging systems applied at the moment of diagnosis of moderate and severe AS is presented in Figure 2. When used in moderate AS, it showed fair discrimination. However, the best model for predicting one-year mortality among those used, is the one used in the second classification (2) when the stenosis became severe. Conclusions AS-related cardiac damage may be present in early stages of AS and may help to guide the timing of intervention in selected patients. Staging classification of cardiac damage AUROC curve for one-year mortality

From motion to emotion: exploring challenging behaviors in autism spectrum disorder through analysis of wearable physiology and movement

Objective.This study aims to evaluate the efficacy of wearable physiology and movement sensors in identifying a spectrum of challenging behaviors, including self-injurious behavior, in children and teenagers with autism spectrum disorder (ASD) in real-world settings.Approach.We utilized a long-short-term memory network with features derived using the wavelet scatter transform to analyze physiological biosignals, including electrodermal activity and skin temperature, alongside three-dimensional movement data captured via accelerometers. The study was conducted in naturalistic environments, focusing on participants' daily activities.Main results.Our findings indicate that the best performance in detecting challenging behaviors was achieved using movement data. The results showed a sensitivity of 0.62, specificity of 0.71, F1-score of 0.36, and an area under the ROC curve of 0.71. These results are particularly significant given the study's focus on real-world scenarios and the limited existing research in this area.Significance.This study demonstrates that using wearable technology to record physiological and movement signals can detect challenging behaviors in children with ASD in real-world settings. This methodology has the potential to greatly improve the management of these behaviors, thereby enhancing the quality of life for children with ASD and their caregivers. This approach marks a significant step forward in applying the outcome of ASD research in practical, everyday environments.

Atrial strain and strain rate in a general population: do these measures improve the assessment of elevated NT-proBNP levels?

Abstract Background and Purpose Noninvasive assessment of elevated filling pressure in the left ventricle (LV) remains an unresolved problem. Of the many echocardiographic parameters used to evaluate diastolic pressure, the left atrial strain and strain rate (LA S/SR) have shown promise in clinical settings. However, only a few previous studies have evaluated LA S/SR in larger populations. Methods A total of 2033 participants from Norwegian and Russian (Know Your Heart) population studies, equally distributed by age and sex, underwent echocardiography, including atrial and ventricular S/SR and NT-proBNP measurements. Of these, 1069 were identified as healthy (without hypertension (HT), atrial fibrillation (AF), or structural cardiac disease) and were used to define the age- and sex-adjusted normal ranges of LA S/SR. Furthermore, the total study population was divided into groups according to ejection fraction (EF) ≥50%, EF <50%, and AF. In each group, uni- and multiple regression and receiver operating characteristic curve analyses were performed to test LA and LV functional parameters as potential indicators of NT-proBNP levels above 250 ng/ml. Results The mean LA S/SR values in this study were higher than those in previous large studies, whereas the lower references were comparable. In normal hearts, atrial total strain (ATS) and mitral valve E deceleration time (MV DT) were independent factors indicating elevated NT-proBNP levels, whereas in hearts with reduced EFs, the independent indicators were peak atrial contraction strain (PACS) and LV stroke volume. The areas under the curve for these significant indicators to discriminate elevated NT-proBNP levels were 0.639 (95% confidence interval (CI): 0.577-0.701) for normal EF and 0.805 (CI: 0.675-0.935) for reduced EF. Conclusion The results confirm good intra-study reproducibility, with mean values in the upper range of previous meta-analyses. In the future automated border-detection algorithms may be able to generate highly reproducible normal values. Furthermore, the study showed atrial S/SR as an additional indicator of elevated NT-proBNP levels in the general population, demonstrating the incremental value of both ATS and PACS in addition to conventional and ventricular strain echocardiography. Thus, the LA S/SR may be regarded as an important addition to the multiparametric approach used for evaluating LV filling. Normal Ranges ROC curve analysis