Search

Supplemental Digital Content is available in the text.

Supplemental Digital Content is available in the text.

Introduction:The genetic basis of hypertrophic cardiomyopathy (HCM) is complex, and the genotype-phenotype correlations have not been completely resolved. Patients lacking a pathogenic variant are presumed to have a better prognosis than patients with sarcomeric pathogenic mutations.Aim:We sought to assess the relationship between the genotype status and the adverse clinical events.Methods:Consecutive patients with HCM underwent genetic testing focused on HCM-related genes. Patients with pathogenic or likely pathogenic variants were considered genotype positive (G+), and those without definite disease-causing mutation were considered genotype negative (G-). Patients were followed for 5.2 ±4.9 years for outcomes. Sudden death (SD) events included out-of-hospital cardiac arrests, appropriate ICD interventions, and HCM-related death.Results:Of 988 HCM patients, 212 (21%) were G+, while 776 (79%) were G -. As compared to G-, G+ HCM patients were younger (45 ± 14 vs. 53 ± 13 years), had more extensive LV hypertrophy (maximum wall thickness: 19.6 ± 5 mm vs. 18.4 ± 4 mm, p<0.001), and with no difference in the presence of LVOT obstruction (59% vs. 66%, p=0.06). Despite G- status, 3% of G- patients had extensive scarring on cardiac MRI (≥15% of LV mass; no different from G+, p=0.54). Over follow-up, 66 patients had SD events. While SD events were more common in G+ patients (11.3% vs. 5.4%, p=0.002), the majority of SD events occurred in G- patients (64%). Similarly, end-stage HCM developed 54 patients, more commonly in G+ (9.0% vs. 4.5%, p=0.01), but with majority of events in G- patients (65%).Conclusion:Contrary to certain prevailing views, in this large consecutive genotyped HCM cohort, sarcomere genotype-negative patients (representing a distinctive majority) did not express an entirely benign prognosis. In fact, nearly two-thirds of SD events and end-stage HF progression occurred in G- HCM patients, albeit at somewhat lower rates than in G+ patients. These data do not support that genotype should not be included as a risk marker of adverse arrhythmic and advanced heart failure events.

Introduction:In hypertrophic cardiomyopathy (HCM), the presence of left ventricular (LV) apical aneurysm warrants implantable cardioverter defibrillator (ICD) therapy consideration for primary prevention of sudden death (SD). However, a significant proportion of these patients may never experience an SD event. Thus, identifying parameters that modify SD risk in this population may be important clinically, but likely requires phenotyping methods that include robust patient-level information.Methods:Clinical, demographic and echocardiographic parameters were compiled for a large HCM referral population. Variables (N=52) were normalized to the cohort and the Pearson’s correlation was calculated between patients using r>0.6 (after Benjamini-Hochberg correction), selected to maximize the patient:edge ratio. The SD event/risk profile was defined as: SD event/≥1 major HCM SD risk factor.Results:From N=3,412 HCM patients (male, 62%; 49 17 years at diagnosis) in the network, the clinical profile included maximum LV thickness, 18 4 mm; left atrial diameter, 41 7 mm; LV outflow tract gradient 30 mmHg, and N=68 (3%) patients with LV apical aneurysm. We identified N=17 distinct patient modules (range, N=19-483 patients/module). Module #15 was highly enriched for LV apical aneurysm (N=56/56); however, the SD event rate was significantly less compared to LV apical aneurysm patients in all other modules (N=12; 4% vs. 75%; P<0.0001). Lower blood pressure and smaller left atrial size delineated LV apical aneurysm patient in module #15.Conclusions:This proof-of-concept patient-patient network analysis suggests that lower blood pressure and smaller left atrial size may contribute to a favorable SD risk profile among HCM LV apical aneurysm patients.

Background:In hypertrophic cardiomyopathy (HCM), late gadolinium enhancement (LGE) on cardiac magnetic resonance imaging (CMR) is an in vivo marker of replacement fibrosis with a continuous relationship between amount of LGE and both risk for sudden death and development of endstage disease. Cost and accessibility can limit routine utilization of CMR. In contrast, electrocardiogram (ECG) is an inexpensive readily available test and identifies patients with myocardial scarring in ischemic heart disease. Its ability to identify presence and extent of LGE in HCM is unknown.Methods:1983 consecutive HCM patients were included; 62% male, average wall thickness 18 ± 3mm, and 36% with resting LV outflow tract obstruction ≥ 30mmHg. ECGs were analyzed for LV hypertrophy, T wave abnormalities, ST-segment depression or elevations, abnormal Q waves, conduction disease, atrial enlargement and QTc prolongation. Extent of LGE was quantified and expressed as a proportion of total LV myocardium.Results:822 (41%) had no LGE, while 1161 (59%) had LGE, including 687 (35%) with <5%, 299 (15%) with 5-9%, 92 (5%) with 10-14%, and 83 (4%) with ≥ 15%. Compared to no LGE, Patients with LGE were more likely to have pathologic T wave inversions (55% vs. 36%, p<0.01), positive T waves in aVR (34% vs. 20%, p<0.01), and ST-depressions (29% vs. 21%, p<0.01), with no difference in LV hypertrophy, bundle branch block, pathologic Q waves, first degree AV block, QTc prolongation ≥ 480ms, or ST-elevations (p>0.05). Notably, as extent of LGE increased there was a corresponding progressive increase in the prevalence of both positive T waves in aVR (20% in no LGE to 42% in ≥ 15% LGE, p<0.01) and pathologic T wave inversions (36% in no LGE to 61% in ≥ 15% LGE, p<0.01) but with no incremental differences in ST depressions (p>0.05). In contrast, 43% of patients with LGE did not have either positive T waves in aVR or pathologic T wave inversions, including 34% with ≥ 15% LGE.Conclusion:Most ECG findings including pathologic Q waves have no relation to presence or extent of LGE. While positive T waves in aVR and pathologic T wave inversions are associated with LGE, they have limited sensitivity. These finding support that routinely evaluated ECG abnormalities do not predict presence or extent of scarring in HCM.

Introduction:Subcutaneous implantable cardioverter defibrillators (SICD) are an attractive option for sudden death (SCD) prevention in younger hypertrophic cardiomyopathy (HCM) patients. Conversely, SICDs have higher rate of inappropriate shock (IAS) when compared to transvenous devices.Objective:We characterize incidence of appropriate shock (AS) and IAS and analyze predictors of IAS in HCM SICD patients.Methods:Data was collected from HCM SICD patients from 2013 to 2021. We used multivariable logistic regression to assess for predictors of IAS in patients with > 6 mo follow up.Results:94 HCM patients (age 47 ± 15 years) underwent SICD implant with mean follow up of 3.7 ± 2.0 year. Maximal LV thickness 20.5 ± 5.8 mm with massive hypertrophy (> 30 mm) in 10 patients (11.8%). Initial DFT with 65J was successful in 88 patients, with 5 more successful after device adjustment. 5 patients (5.9%) had 10 AS (3.2 AS per 100 pt-years). 10 patients (11.8%) had 19 IAS (6.0 IAS per 100 pt-years) due to T wave oversensing (n = 13), P wave oversensing (n = 2), atrial arrhythmia (n=2), and external noise (n = 1). IAS rate decreased over time, with IAS occurring in 8 patients of the initial half of the cohort and in only 2 of the second half. Time to IAS from implant was 11 ± 10.8 mo. QRS duration (OR 1.025, 95% CI 0.997-1.053; P = 0.084) showed trend to prediction of IAS, but no characteristic proved independently significant.Conclusions:This data adds to increasing evidence that in high-risk HCM patients, SICD represents a reliable treatment option for SCD prevention. IAS mainly due to cardiac oversensing, was seen in 11.8% of patients within this HCM cohort. IAS decreased over time, possibly due to improved patient selection, implant technique and device programming (SMART pass filter). A larger dataset is likely necessary to better understand this trend.

Introduction:Myocardium scar volume in late gadolinium enhancement (LGE) MRI is associated with cardiac adverse outcomes in hypertrophic cardiomyopathy (HCM). However, scar quantification is time-consuming and challenging. Also, LGE contains information about scar heterogeneity, location, and shape not readily quantified by current LGE analyses.Hypothesis:Automated deep phenotyping of LGE images can extract novel imaging markers of heart failure (HF) progression in HCM.Methods:We developed a deep convolutional neural network (CNN) (Fig 1) to extract LGE image features associated with progressive HF in HCM. We developed the CNN using data for 572 HCM patients (183 F; 51±16 years) evaluated at Tufts medical center. The CNN input was 5 LGE images and the output was the probability of developing a HF outcome (defined as a change of New York Heart Association class I/II at baseline to class III/IV during follow-up). The dataset was split into training, 50%, validation, 25%, and independent testing, 25%. We trained the CNN for 50 epochs (learn-rate = 0.001, batch size = 12, class-weighted cross-entropy loss). To avoid overfitting, the model that best predicted HF outcomes in the validation cohort was used as the final model. We compared the CNN model to a single-variable logistic regression (LR) model (L2 regularization, class-weighted loss) that predicts HF outcomes using scar volume (manually quantified by an expert reader).Results:Progressive HF occurred in 96 (17%) patients during follow-up (4±2 years). In the testing cohort (n = 138), CNN correctly predicted HF in 60% of patients with positive HF outcomes (n = 22; specificity 77%, accuracy 73%, AUC = 0.73). Using the same dataset, the LR classifier showed much lower HF prediction power with AUC = 0.54 (accuracy 38%, specificity 30%).Conclusions:Deep phenotyping of LGE allows accurate prediction of progressive HF in HCM patients and outperforms classical LR models that use scar volume as a sole predictor of HF.

Introduction:LBBB results in impaired transseptal conduction and delayed posterolateral LV activation. Criteria for left bundle branch block (LBBB) vary across scientific organizations. Septal myectomy for hypertrophic cardiomyopathy (HCM) could serve as an iatrogenic anatomic model to characterize LBBB.Objective:To describe ECG features of LBBB after septal myectomy for HCM.Methods:ECG data were analyzed for 377 HCM patients (204 male) who developed post-operative LBBB after extended septal myectomy between 2004-2018.Results:Average age of the cohort at myectomy was 53 ± 14 years. Baseline QRS duration (QRSd) was 94 ± 10ms. The post myectomy QRSd was 152 ± 15ms, consistent with an average ΔQRSd of 58 ± 13ms. There was positive correlation between pre and post myectomy QRSd (r = 0.485; p < 0.0001). The average LV end-diastolic diameter (LVEDd) pre and 2 months post myectomy was 40 ± 5.6mm and 46.2 ± 6.5mm, respectively. Positive correlation between post myectomy QRSd and post myectomy LVEDd was also observed (r = 0.340; p < 0.0001). Females and males had pre myectomy QRSd of 93 ± 10ms and 95 ± 10ms respectively (p = 0.007) and post myectomy QRSd of 147 ± 13ms (120-184ms) and 157 ± 14ms (126-209ms) respectively (p < 0.0001). ΔQRSd was less in females than males (54 ± 13ms vs. 62 ± 11ms; p<0.0001). Only 13 females and 3 males had post myectomy QRSd < 130ms. Only 23 males had QRSd < 140ms.Conclusions:Surgical myectomy serves as a model to characterize LBBB. Following myectomy, QRSd correlated with LVEDd as well as pre-myectomy QRSd, and was longer in males due to delayed depolarization of larger hearts in men. Notably only 3 males and a minority of females had QRSd < 130ms. These data provide important insights for the validation of electrocardiographic LBBB criteria.

Introduction:Subcutaneous implantable cardioverter-defibrillators (SICD) are effective for sudden death (SCD) prevention in patients (pts) with hypertrophic cardiomyopathy (HCM). Pre-implant ECG screening to evaluate R and T wave sensing is done to decrease the risk of inappropriate shocks (IAS). Pts who undergo septal myectomy are at high risk of developing left bundle branch block (LBBB) that could potentially result in SICD T wave oversensing (TWOS).Objective:We describe 6 cases of pts with SICD and septal myectomy to understand the risks and benefits of maintaining SICD therapy in this cohort.Methods:We performed a retrospective cohort study of all pts with HCM who underwent SICD implantation and septal myectomy from 2013 to 2021.Results:Six pts with HCM underwent both SICD implantation and myectomy. Post myectomy ECGs for all pts show LBBB with QRS between 137 to 166 ms. Four of these pts were implanted prior to myectomy. The average time to myectomy after SICD was 1.3 years. Three out of the four pts have had no IAS or TWOS since myectomy. One patient experienced TWOS with 2 IAS 4 months after myectomy. The sensing vector was changed from primary to alternate with no further TWOS for the last 6 years. The remaining two pts underwent SICD implantation after myectomy, both of whom screened “OK” using the manual screening tool prior to implant. Neither patient has experienced IAS. One patient displayed episodes of TWOS and underwent vector change with no TWOS for the last 5 years. At the time of recent SICD generator replacement, he was screened using the automated programmer software and failed screening in all 3 vectors (Figure 1).Conclusions:This series demonstrates that maintenance of SICD is a viable option for SCD prevention in young pts who develop LBBB after septal myectomy. With close monitoring and vector adjustment, IAS and TWOS can be avoided. Further larger studies are necessary to assess the long-term risk of TWOS and IAS in this unique population.

Introduction:Safety of vigorous exercise for individuals with appropriately-treated HCM remains debated. Physical activity practices in this population have not been described. The objective of this report is to describe physical activity pattern in HCM patients. Long term follow up which is ongoing will determine safety of vigorous exercise.Methods:The NIH-funded LIVE-HCM study prospectively enrolled individuals age 8-60 years with clinically overt HCM or patients with a positive genetic test for sarcomeric HCM. Participants (or parents, for children) answered surveys describing activity patterns. Vigorous exercise was defined as > 6 METS for > 60 hours per year. Clinical and demographic data were derived from medical record review.Results:Among 1798 participants, 745 (41%) reported exercising vigorously, including 296 competitive athletes, with 59 at high-school or college varsity level, 770 (43%) were exercising at moderate, and 283 (16%) at low levels. Of those working, 6% described jobs more active than walking, including 16 in protective services, and 11 in sports/coaching. Clinical/demographic data including imaging characteristics and treatments, are shown in Table, and were mostly similar amongst the exercise-groups.Conclusions:Many patients with HCM are engaged in vigorous exercise including competitive athletics. Prospective follow-up of this cohort is ongoing. Comparison of arrhythmic outcomes outcomes in vigorous exercisers vs moderate/low level exercisers will help determine the safety of exercise in HCM patients.

Introduction:Defibrillation threshold testing (DFT) is generally reserved for populations with perceived concern for defibrillation failure, including in hypertrophic cardiomyopathy (HCM). Higher shock impedance has been associated with risk of DFT failure with subcutaneous defibrillators (SICD). The PRAETORIAN score is a chest radiograph based tool that evaluates device positioning to predict risk of DFT failure.Objective:We assessed predictors of impedance during DFT in a single center HCM SICD cohort.Methods:Data was collected from HCM patients who underwent SICD implantation from 2013 to 2021. Predictors of impedance were evaluated with linear and logistic regression.Results:Impedance was reported in 75 HCM SICD patients during DFT with 65J. DFT was successful in 74 patients with mean impedance of 72.7 ohm and septal thickness of 20.6 mm. Ten patients had massive HCM (> 30 mm) with mean impedance of 66.4 ohm. Septal thickness was not predictive of impedance (β 0.37, 95% CI 1.24-0.50; P = .40). Independent predictors of impedance were BMI (β 1.41, 95% CI 0.61-2.21; P = .0008) and PRAETORIAN score (β 0.14, 95% CI 0.05-0.22; P = .0015). Multivariable regression revealed BMI no longer significant and PRAETORIAN score remaining significant (β 0.10, 95% CI 0.01-1.66; P = .038).Conclusion:In this HCM SICD cohort BMI and PRAETORIAN score were predictive of impedance. However, in a multivariable model, BMI was no longer predictive, highlighting importance of implant technique to improve impedance and DFT. Additionally, patients with massive hypertrophy all had successful DFT and septal thickness was not predictive of impedance. HCM has historically been portrayed as a high DFT population. Further prospective trials are necessary to establish whether standard DFT practice can be applied in HCM.

Background:Atrial fibrillation (AF) is the most common sustained arrhythmia in hypertrophic cardiomyopathy (HCM), with patients having a 4 to 6 fold greater likelihood of developing AF as compared to the general population, and is an important cause of morbidity, impaired quality of life and embolic stroke in HCM. The impact of outflow obstruction and the influence of surgical septal myectomy on the development of new-onset AF has not been well described.Aim:To characterize the risk of new-onset AF in nonobstructive and obstructive HCM and compare to AF risk in patients after myectomy.Methods:1767 consecutive HCM patients without prior AF history, seen at the Tufts HCM Institute between 2004 and 2019, including 712 with obstruction (outflow gradients ≥ 30mmHg at rest or following provocation) who did not undergo invasive septal reduction therapy (SRT), 552 with nonobstructive HCM (outflow gradients < 30mmHg), and 503 who underwent surgical myectomy were followed for an average of 4.3 ±4.2 years for new-onset symptomatic AF. Isolated post-operative episodes of AF, i.e., occurring within the first 3 months after myectomy, were excluded from analysis.Results:Patients with obstructive HCM who did not undergo SRT had a 1.5-fold increased risk for new-onset AF as compared to nonobstructive HCM (rate onset of 2.8%/year vs.1.7%/year, p=0.04), with freedom from AF at 10-years of 72% in obstructive patients. Patients undergoing myectomy were older (52 ± 15 vs. 50 ± 15 years; p<0.01), had larger LA dimension (42 ± 7 vs 40 ± 6mm; p<0.01), and higher HCM-AF risk scores (22 ± 3.5 vs. 19 ± 3.8; p<0.01) as compared to nonobstructive HCM. However, despite a higher predicted risk for new-onset AF at time of myectomy (2.7%/year vs. 1.5%/year), after myectomy the risk of new onset AF was no different from patients with nonobstructive HCM (hazard ratio 0.9; rate of onset of 1.8%/year vs. 1.7%/year; p=.73), with freedom from AF at 10-years of 83% post-myectomy.Conclusions:Patients with obstructive HCM have a higher risk of AF as compared to nonobstructive HCM. After myectomy risk of new-onset AF is substantially reduced, such that risk is no different from patients with nonobstructive HCM. These data demonstrate efficacy of myectomy in decreasing long term risk of AF in obstructive HCM.

Introduction:The Hypertrophic Cardiomyopathy Registry (HCMR) is an NHLBI-funded registry of 2755 patients with HCM recruited from 44 sites in 6 countries, designed to improve risk prediction by incorporating cardiovascular magnetic resonance (CMR), genetic, and biomarker data.Hypothesis:Baseline characteristics could differentiate specific subgroups in HCM.Methods:Demographic and echocardiographic data were collected. Patients underwent CMR including cine imaging, late gadolinium enhancement imaging (LGE) and T1 mapping for measurement of extracellular volume (ECV) s. Blood was drawn for biomarker (N-terminal pro brain natriuretic peptide and high sensitivity troponin T) and genetic analysis.Results:Of 2755 patients studied, age (mean?S.D.) was 49?11 years, 71% were male, and 17% non-white. ESC risk score was 2.48?0.56. Twenty four % had resting left ventricular outflow tract gradient ?30mmHg. Thirty six percent had a sarcomere mutation identified. Fifty percent had any LGE, which was more common and extensive in patients who were sarcomere mutation (+), had reverse septal curvature morphology and no significant resting LVOT obstruction. Those without LGE were more likely sarcomere mutation (-), had isolated basal septal hypertrophy, and more LVOT obstruction. Elevated ECV was noted in segments both with and without LGE. Serum biomarker levels correlated with increasing LGE and ECV in a graded fashion.Conclusions:The HCMR population has characteristics of mild-to-moderate risk HCM. Baseline data separated patients broadly into 2 categories. One group was sarcomere mutation (+), more likely had reverse septal curvature morphology, more fibrosis, but less resting obstruction whereas the other was sarcomere mutation (-), more likely had isolated basal septal hypertrophy with obstruction, but less fibrosis. Further follow-up will allow better understanding of these subgroups and development of an improved risk prediction model.

Introduction:The implantable cardioverter-defibrillator (ICD) has been proven effective for preventing sudden death in patients with hypertrophic cardiomyopathy (HCM). However, performance and complications of ICDs over particularly long time periods in this unpredictable disease is unresolved.Methods:The study cohort comprises 984 consecutive HCM patients implanted with primary or secondary prevention ICDs, 235 of whom had ?10 years of follow-up (mean 12 ? 4; range to 33).Results:Patients were 38 ? 17 years at implant and 53 of the 235 (22%) experienced appropriate interventions terminating VT/VF. Time from device implant to first appropriate intervention was 6 ? 5 years but varied from 1 month to 16 years, with 14 (26%) initial interventions ?5 years after implant and 17 (32%) ?10 years after implant. Patients with early interventions (<1 year) were more likely to experience future interventions and the likelihood of an appropriate intervention increased in a linear fashion with respect to time from implant. Seventy-two significant device complications including inappropriate shocks (n=43 [18%]) and device and/or lead complications requiring intervention (n=29 [12%]) occurred in 69 patients (29%) 7 ? 6 years after implant, with inappropriate therapy most common in the first 5 years after implant (1.9%/year), but decreasing in frequency thereafter to 0.4%/year. Of the 69 patients experiencing complications, 15 (22%) also experienced appropriate interventions.Conclusions:ICDs are effective in terminating ventricular tachyarrhythmias in young HCM patients with appropriate interventions increasing in frequency with greater time delays after implant. One third of first ICD interventions occurred 10 or more years after implant. Notably, long-term ICD use in HCM is also associated with risk of device and/or lead-related complications, suggesting consideration of subcutaneous devices in young high-risk HCM patients.

Background:Despite recognition that hypertrophic cardiomyopathy (HCM) can be compatible with extended longevity in the modern treatment era, the perception remains that children with HCM remain at high risk for mortality, with risk stratification for sudden death (SD) unreliable for identification of at-risk children, and surgical myectomy only palliative for treatment of drug-refractory heart failure (HF) symptoms in children with obstructive HCM. Whether these perceptions are true is unresolved in the contemporary treatment era for HCM.Methods:A retrospective review of 152 consecutive children and adolescents between 5 and 19 years of age (16 ? 3 years), initially evaluated at the Tufts HCM Institute between 2002 and 2018, and followed for 6 ? 4 years for clinical outcomes.Results:Of the 152 patients, 60 (39%) underwent primary prevention implantable cardioverter-defibrillator (ICD) placement due to the presence of ? 1 major risk marker (family history of SD, nonsustained ventricular tachycardia on event monitor, syncope, and extreme left ventricular wall thickness). Eight (13%) of the 60 patients experienced ICD therapy terminating ventricular arrhythmia episodes. An additional 7 patients (5%) had ICD implanted for secondary prevention of SD after resuscitated cardiac arrest, each occurring prior to initial visit at our center, including 5 of 7 with major risk markers and 4 of 7 with subsequent ICD interventions. None of the 152 patients experienced arrhythmic SD. In addition, 27 patients (18%) required surgical myectomy for progressive HF due to left ventricular outflow obstruction, with improvement in symptoms to New York Heart Association class I/II in 24 (89%) at 5 ? 4 years, while one additional non-obstructive patient underwent heart transplant due to end-stage disease. Each of the 153 patients have survived to 22 ? 5 years (61% to >20 years of age), with no HCM-related mortality or all-cause mortality.Conclusions:HCM in children is compatible with extended longevity into adulthood with modern treatment strategies, including ICD for prevention of sudden death and surgical myectomy for reversal of HF secondary to obstruction. Current risk stratification strategies are effective in identifying children at risk for SD events.

Introduction:The transvenous (TV) implantable cardioverter defibrillator (ICD) is effective for preventing sudden cardiac death (SCD) in patients with hypertrophic cardiomyopathy (HCM). The efficacy of the newer subcutaneous ICD (S-ICD) in real world HCM practice is not well defined.Hypothesis:To report S-ICD performance in a large HCM institute study cohort.Methods:This observational, longitudinal study cohort comprises 123 consecutive HCM patients with primary or secondary prevention S-ICDs (n=46) or TV-ICDs (n=77) followed at Tufts Medical Center after November 2012, with longitudinal follow-up of at least 6 months.Results:HCM patients who received an S-ICD were 39 ? 15 years of age with max LV wall thickness (MLVWT) of 19 ? 7 mm (n=10 [22%] ?30 mm); 36 (78%) were male. Of the 44 patients who underwent defibrillation threshold testing (DFT), 43 (98%) patients were successfully defibrillated. Appropriate therapy for ventricular tachycardia (VT) and ventricular fibrillation (VF) was delivered in 3 (7%) S-ICD patients over 2 ? 1.7 years of follow-up and in 5 (7%) TV-ICD patients over 2.8 ? 1.7 years of follow-up. In S-ICD patients, time from implant to first appropriate shock was 1.9 ? 1.4 years. The 3 S-ICD patients were aged 16, 35, and 42 years with MLVWT of 30, 23, and 17 mm respectively. Appropriate shocks were delivered for monomorphic or polymorphic VT that deteriorated to VF and rapid atrial fibrillation (AF) that deteriorated to VT. S-ICD patients who received an appropriate shock had 1 or 2 major SCD risk factors including massive LV hypertrophy, extensive late gadolinium enhancement, NSVT on ambulatory monitor, and family history of SCD. There were no deaths due to ventricular arrhythmia in patients with S-ICDs. Inappropriate shocks were experienced by 9 (19%) S-ICD patients mostly due to T-wave oversensing and 2 (3%) TV-ICD patients due to rapid AF.Conclusions:The S-ICD is effective in terminating life threatening ventricular arrhythmias in a heterogenous cohort of young high-risk HCM patients in a real-world HCM practice. These data suggest that the S-ICD is effective for SCD prevention in HCM, although inappropriate shocks might be more common than in transvenous devices.

Background:Women with hypertrophic cardiomyopathy (HCM) have been described as having worse long-term survival, with adverse events occurring later in life as compared to men. Sudden death (SD) is the most visible and devastating disease complication in HCM, however it remains unknown whether women are at increased risk for SD events. Therefore, we investigated the difference in SD risk, SD risk markers, and timing of SD events in women vs. men.Methods:Retrospective assessment of SD events including sudden cardiac death, resuscitated cardiac arrest and appropriate ICD interventions for ventricular tachycardia (VT) / ventricular fibrillation (VF), in HCM patients followed in Tufts HCM Institute from 2001 to 2016. Of the 2123 patients included in this analysis, 794 were women, and 1329 were men.Results:SD events occurred in 40 (5%) women vs. 84 (6%) men (p=0.25); most commonly appropriate ICD intervention for VT/VF (25 women vs. 57 men), while sudden cardiac death was uncommon in both sexes (1 woman and 4 men). There was no difference in event rates; 0.9% of women per year vs. 0.8% of men per year [HR 0.92 (95% CI:0.56,1.5) p= 0.73].Age at SD events was similar in women and men, occurring at 41 ? 17 years in women vs. 46 ? 17 years in men (p=0.17). While there was no difference in proportion of male and female HCM patients with SD events in individuals < 30 years (13% of women vs. 10% of men; p = 0.65), or 30-59 years (7% of women vs. 6% of men; p = 0.54), unexpectedly, a larger proportion of men had SD events at >60 years of age than women (2% of women vs. 6% of men; p = 0.007).SD risk factors were similar in women and men with SD events including unexplained syncope (25% of women vs. 31% men; p = 0.63), NSVT on ambulatory monitoring (13% vs. 23%; p = 0.22), family history of SD (28% vs. 23%; p = 0.65), massive LV hypertrophy (wall thickness ? 30mm [13% vs. 23%; p = 0.22]), and LV apical aneurysms (10% vs. 15% men; p = 0.58).Conclusion:Sudden death events do not differ in men and women with HCM. Events occur at young ages in women, similar to men, with no evidence that SD events occur at more advanced ages in women. The presence of risk markers in women and men at increased SD risk is similar.

Background:In hypertrophic cardiomyopathy (HCM), late-gadolinium enhancement (LGE) is routinely performed for evaluation of left ventricular (LV) scar. However, 50% of HCM patients do not have scar yet these patients are exposed to the potential risk of gadolinium administration. Fibrosis influences regional myocardial T1which manifests as signal variation on cine steady-state free-precession (SSFP) images. We hypothesize that radiomic analysis of cine-SSFP images including established morphological variables as well as novel imaging markers of myocardial texture can identify patients without scar.Methods:We accessed a multicenter, international cohort dataset of 1.5T cardiac MRI from 1152 HCM patients. Cine images were acquired using SSFP prior to gadolinium administration in the short-axis orientation. The entire stack of LV cine images was segmented to contour endo- and epicardial borders. A total of 2747 radiomic features (including LV volumes, mass, wall thickness, myocardial signal intensity, texture, shape and wavelet) were extracted using PyRadiomics. LGE assessment was visually performed. A random forest algorithm was used to select the discriminative features and generate a radiomics signature of LGE+ myocardium. A gradient boosting algorithm was, subsequently, used to build a classification model in the training dataset. The receiver operating characteristics curve was used for model performance evaluation in the testing dataset.Results:LGE was present in 49% of patients. Radiomics signature of LGE+ consisted of measures of thickness and motion of the anterior and septal walls and textural features describing myocardial heterogeneity. The model yielded values for the area under the curve of 0.78 [95% confidence interval (CI): 0.72-0.84], sensitivity of 68% (95% CI: 60-84), precision of 68% (95% CI: 60-84), and accuracy of 71% (95% CI: 63-78). When selecting a threshold with no preference for sensitivity, the overall accuracy was 68% (95% CI: 60-84), with a sensitivity of 94% (95% CI: 91-100), and a negative predictive value of 88% (95% CI: 80-96).Conclusion:In HCM, radiomic analysis of non-contrast cine-SSFP images has the potential to identify patients without scar in whom gadolinium administration can be safely avoided.

Introduction:hypertrophic cardiomyopathy (HCM) is a leading cause of sudden cardiac death (SCD). Treatment planning requires identifying patients at high risk of SCD. We sought to develop a novel deep learning model for risk stratification of SCD in HCM patients.Methods:Data from patients followed at Tufts HCM Institute were interrogated. Primary outcome of SCD was defined as appropriate ICD shock, cardiac arrest, or sudden death. A deep learning SCD prediction model with 3 dense layers (nodes = 32/4/1) was trained and tested. Batch normalization and dropouts were used to enhance training and avoid overfitting. Weighted loss was used to compensate small number of SCD outcomes. Training/testing was repeated 10 times with exclusive subsets (i.e. 10-fold cross-validation). We built 12 risk prediction models using different combinations of SCD risk factors (e.g. demographics, family history of SCD, rest/stress echocardiography measurements). The selected model was determined as one with highest c-statistics and lowest standard deviation over the different experiments.Results:2574 patients with complete risk factor data and follow up were used to build/test the model. The follow-up time was 4.4?3.6 yr (median 3.4 yr). Primary outcomes occurred in 87 (3.4%) patients. The model was trained for 300 epochs (batch size=100), and the model with lowest error during training was used to predict SCD in the testing cases. A combination of age, history of syncope and non-sustained ventricular tachycardia, max wall thickness and presence of apical aneurysm (assessed on echocardiography) resulted in best SCD prediction accuracy (c-statistics = 0.76?0.04) (Fig. 1). Our model correctly identified 90% of patients who met the primary outcome with specificity of 50%.Conclusions:The proposed deep learning model can accurately identify patients at high risk of SCD. Further studies in independent cohorts with longer follow-up time are warranted to evaluate the proposed model.