Implementing Machine Learning in Interventional Cardiology: The Benefits Are Worth the Trouble.

Driven by recent innovations and technological progress, the increasing quality and amount of biomedical data coupled with the advances in computing power allowed for much progress in artificial intelligence (AI) approaches for health and biomedical research. In interventional cardiology, the hope is for AI to provide automated analysis and deeper interpretation of data from electrocardiography, computed tomography, magnetic resonance imaging, and electronic health records, among others. Furthermore, high-performance predictive models supporting decision-making hold the potential to improve safety, diagnostic and prognostic prediction in patients undergoing interventional cardiology procedures. These applications include robotic-assisted percutaneous coronary intervention procedures and automatic assessment of coronary stenosis during diagnostic coronary angiograms. Machine learning (ML) has been used in these innovations that have improved the field of interventional cardiology, and more recently, deep Learning (DL) has emerged as one of the most successful branches of ML in many applications. It remains to be seen if DL approaches will have a major impact on current and future practice. DL-based predictive systems also have several limitations, including lack of interpretability and lack of generalizability due to cohort heterogeneity and low sample sizes. There are also challenges for the clinical implementation of these systems, such as ethical limits and data privacy. This review is intended to bring the attention of health practitioners and interventional cardiologists to the broad and helpful applications of ML and DL algorithms to date in the field. Their implementation challenges in daily practice and future applications in the field of interventional cardiology are also discussed.

Deep learning-based prediction of early cerebrovascular events after transcatheter aortic valve replacement.

Cerebrovascular events (CVE) are among the most feared complications of transcatheter aortic valve replacement (TAVR). CVE appear difficult to predict due to their multifactorial origin incompletely explained by clinical predictors. We aimed to build a deep learning-based predictive tool for TAVR-related CVE. Integrated clinical and imaging characteristics from consecutive patients enrolled into a prospective TAVR registry were analysed. CVE comprised any strokes and transient ischemic attacks. Predictive variables were selected by recursive feature reduction to train an autoencoder predictive model. Area under the curve (AUC) represented the model's performance to predict 30-day CVE. Among 2279 patients included between 2007 and 2019, both clinical and imaging data were available in 1492 patients. Median age was 83 years and STS score was 4.6%. Acute (< 24 h) and subacute (day 2-30) CVE occurred in 19 (1.3%) and 36 (2.4%) patients, respectively. The occurrence of CVE was associated with an increased risk of death (HR [95% CI] 2.62 [1.82-3.78]). The constructed predictive model uses less than 107 clinical and imaging variables and has an AUC of 0.79 (0.65-0.93). TAVR-related CVE can be predicted using a deep learning-based predictive algorithm. The model is implemented online for broad usage.

Double S-Curve Versus Cusp-Overlap Technique: Defining the Optimal Fluoroscopic Projection for TAVR With a Self-Expanding Device.

OBJECTIVES: The purpose of this study was to assess the concordance between transcatheter aortic valve implantation angles generated by the "double S-curve" and "cusp-overlap" techniques. BACKGROUND: The "double S-curve" and "cusp-overlap" methods aim to define optimal fluoroscopic projections for transcatheter aortic valve replacement (TAVR) with a self-expandable device. METHODS: The study included 100 consecutive patients undergoing TAVR with self-expanding device planned by multidetector computed tomography. TAVR was performed using the double S-curve model, as a view in which both the aortic valve annulus and delivery catheter planes are displayed perpendicularly on fluoroscopy. Optimal projection according to the cusp-overlap technique was retrospectively generated by overlapping the right and left cups on the multidetector computed tomography annular plane. The angular difference between methods was assessed in spherical 3 dimensions and on the left and right anterior oblique (RAO) and cranial and caudal (CAU) axes. RESULTS: The double S-curve and cusp-overlap methods provided views located in the same quadrant, mostly the RAO and CAU, in 92% of patients with a median 3-dimensional angular difference of 10.0° (interquartile range: 5.5° to 17.9°). The 3-dimensional deviation between the average angulation obtained by each method was not statistically significant (1.49°; p = 0.349). No significant differences in average coordinates were noted between the double S-curve and cusp-overlap methods (RAO: 14.7 ± 15.2 vs. 12.9 ± 12.5; p = 0.36; and CAU: 27.0 ± 9.4 vs. 26.9 ± 10.4; p = 0.90). TAVR using the double S-curve was associated with 98% device success, low complication rate, and absence of moderate-to-severe paravalvular leak. CONCLUSIONS: The double S-curve and cusp-overlap methods provide comparable TAVR projections, mostly RAO and CAU. TAVR using the double S-curve model is associated with a high rate of device success and low rate of procedural complications.

Predictive Value of the Residual SYNTAX Score in Patients With Cardiogenic Shock.

BACKGROUND: In hemodynamically stable patients, complete revascularization (CR) following percutaneous coronary intervention (PCI) is associated with a better prognosis in chronic and acute coronary syndromes. OBJECTIVES: This study sought to assess the extent, severity, and prognostic value of remaining coronary stenoses following PCI, by using the residual SYNTAX score (rSS), in patients with cardiogenic shock (CS) related to myocardial infarction (MI). METHODS: The CULPRIT-SHOCK (Culprit Lesion Only Percutaneous Coronary Intervention [PCI] Versus Multivessel PCI in Cardiogenic Shock) trial compared a multivessel PCI (MV-PCI) strategy with a culprit lesion-only PCI (CLO-PCI) strategy in patients with multivessel coronary artery disease who presented with MI-related CS. The rSS was assessed by a central core laboratory. The study group was divided in 4 subgroups according to tertiles of rSS of the participants, thereby isolating patients with an rSS of 0 (CR). The predictive value of rSS for the 30-day primary endpoint (mortality or severe renal failure) and for 30-day and 1-year mortality was assessed using multivariate logistic regression. RESULTS: Among the 587 patients with an rSS available, the median rSS was 9.0 (interquartile range: 3.0 to 17.0); 102 (17.4%), 100 (17.0%), 196 (33.4%), and 189 (32.2%) patients had rSS = 0, 0 < rSS ≤5, 5 < rSS ≤14, and rSS >14, respectively. CR was achieved in 75 (25.2%; 95% confidence interval [CI]: 20.3% to 30.5%) and 27 (9.3%; 95% CI: 6.2% to 13.3%) of patients treated using the MV-PCI and CLO-PCI strategies, respectively. After multiple adjustments, rSS was independently associated with 30-day mortality (adjusted odds ratio per 10 units: 1.49; 95% CI: 1.11 to 2.01) and 1-year mortality (adjusted odds ratio per 10 units: 1.52; 95% CI: 1.11 to 2.07). CONCLUSIONS: Among patients with multivessel disease and MI-related CS, CR is achieved only in one-fourth of the patients treated using an MV-PCI strategy. and the residual SYNTAX score is independently associated with early and late mortality.

Angiographic predictors of outcome in myocardial infarction patients presenting with cardiogenic shock: a CULPRIT-SHOCK angiographic substudy.

AIMS: The aim of this study was to determine the prognostic impact of pre- and post-PCI TIMI flow grade and TIMI myocardial perfusion grade (TMPG) in a well-defined group of patients with cardiogenic shock due to acute myocardial infarction. METHODS AND RESULTS: Patients with infarct-related cardiogenic shock randomised into the CULPRIT-SHOCK trial were included in the angiographic predictor analysis whenever their TIMI flow grade or TMPG was available in the core lab database (96.9% of cases). A multivariable logistic regression analysis, adjusted on non-angiographic covariates, was performed to investigate whether TIMI flow grade or TMPG was independently associated with all-cause mortality or renal replacement therapy up to one year. Pre-PCI TIMI flow grade and TMPG did not impact on mortality. When analysed in separate multivariable models, post-PCI TIMI 3 flow and TMPG grade 3 were both significantly associated with reduced risk of 30-day mortality: aOR 0.61 (95% CI: 0.38-0.97, p=0.037) and 0.46 (95% CI: 0.29-0.72, p<0.001), respectively. When considered in the same multivariable model, only TMPG was significantly associated with 30-day mortality (aOR 0.38 [0.20-0.71], p=0.002), the 30-day composite of all-cause mortality and renal replacement therapy (aOR 0.34 [0.18-0.66], p=0.001) and mortality at one-year follow-up (aOR 0.46 [0.24-0.88], p=0.02). CONCLUSIONS: Post-PCI TIMI flow grade and TMPG are associated with mortality after PCI. TMPG is a better discriminator, supporting microcirculation rather than epicardial reperfusion for prognosis estimation.

Impact of chronic total occlusion and revascularization strategy in patients with infarct-related cardiogenic shock: A subanalysis of the culprit-shock trial.

BACKGROUND: The impact of coronary artery chronic total occlusion (CTO) and its management with percutaneous coronary intervention (PCI) in the setting of myocardial infarction (MI) related cardiogenic shock (CS) remains unclear. METHODS: This is a pre-specified analysis from the culprit-lesion-only PCI vs multivessel PCI in CS (CULPRIT-SHOCK) trial which randomized patients presenting with MI and multivessel disease complicated by CS to a culprit-lesion-only or immediate multivessel PCI strategy. CTO was defined by central core-laboratory evaluation. The independent associations between the presence of CTO and adverse outcomes at 30 days and 1 year were assessed using multivariate logistics models. RESULTS: A noninfarct related CTO was present in 157 of 667 (23.5%) analyzed patients. Patients presenting with CTO had more frequent diabetes mellitus or prior PCI but less frequently presented with ST segment elevation MI as index event. The presence of CTO was associated with higher rate of death at 30 days (adjusted Odds ratio 1.63; 95% confidence interval [CI] 1.01-2.60). Rate of death at 1 year was also increased but did not reach statistical significance (adjusted Odds ratio 1.62; 95%CI 0.99-2.66). Compare to immediate multivessel PCI, a strategy of culprit-lesion-only PCI was associated with lower rates of death or renal replacement therapy at 30 days in patients with and without CTO (Odds ratio 0.79 95%CI 0.42-1.49 and Odds ratio 0.67 95%CI 0.48-0.96, respectively), without significant interaction (P = .68). CONCLUSIONS: In patients with MI-related CS and multivessel disease, the presence of CTO is associated with adverse outcomes while a strategy of culprit-lesion-only PCI seems beneficial regardless of the presence of CTO.

Transcatheter Treatment of Residual Significant Mitral Regurgitation Following TAVR: A Multicenter Registry.

OBJECTIVES: The aim of this study was to describe baseline characteristics, and periprocedural and mid-term outcomes of patients undergoing transcatheter mitral valve interventions post-transcatheter aortic valve replacement (TAVR) and examine their clinical benefit. BACKGROUND: The optimal management of residual mitral regurgitation (MR) post-TAVR is challenging. METHODS: This was an international registry of 23 TAVR centers. RESULTS: In total, 106 of 24,178 patients (0.43%) underwent mitral interventions post-TAVR (100 staged, 6 concomitant), most commonly percutaneous edge-to-edge mitral valve repair (PMVR). The median interval post-TAVR was 164 days. Mean age was 79.5 ± 7.2 years, MR was >moderate in 97.2%, technical success was 99.1%, and 30-day device success rate was 88.7%. There were 18 periprocedural complications (16.9%) including 4 deaths. During a median follow-up of 464 days, the cumulative risk for 3-year mortality was 29.0%. MR grade and New York Heart Association (NYHA) functional class improved dramatically; at 1 year, MR was moderate or less in 90.9% of patients (mild or less in 69.1%), and 85.9% of patients were in NYHA functional class I/II. Staged PMVR was associated with lower mortality versus medical treatment (57.5% vs. 30.8%) in a propensity-matched cohort (n = 156), but this was not statistically significant (hazard ratio: 1.75; p = 0.05). CONCLUSIONS: For patients who continue to have significant MR, remain symptomatic post-TAVR, and are anatomically suitable for transcatheter interventions, these interventions are feasible, safe, and associated with significant improvement in MR grade and NYHA functional class. These results apply mainly to PMVR. A staged PMVR strategy was associated with markedly lower mortality, but this was not statistically significant. (Transcatheter Treatment for Combined Aortic and Mitral Valve Disease. The Aortic+Mitral TRAnsCatheter Valve Registry [AMTRAC]; NCT04031274).

Optimal Fluoroscopic Projections of Coronary Ostia and Bifurcations Defined by Computed Tomographic Coronary Angiography.

OBJECTIVES: The aim of this study was to define the optimal fluoroscopic viewing angles of both coronary ostia and important coronary bifurcations by using 3-dimensional multislice computed tomographic data. BACKGROUND: Optimal fluoroscopic projections are crucial for coronary imaging and interventions. Historically, coronary fluoroscopic viewing angles were derived empirically from experienced operators. METHODS: In this analysis, 100 consecutive patients who underwent computed tomographic coronary angiography (CTCA) for suspected coronary artery disease were studied. A CTCA-based method is described to define optimal viewing angles of both coronary ostia and important coronary bifurcations to guide percutaneous coronary interventions. RESULTS: The average optimal viewing angle for ostial left main stenting was left anterior oblique (LAO) 37°, cranial (CRA) 22° (95% confidence interval [CI]: LAO 33° to 40°, CRA 19° to 25°) and for ostial right coronary stenting was LAO 79°, CRA 41° (95% CI: LAO 74° to 84°, CRA 37° to 45°). Estimated mean optimal viewing angles for bifurcation stenting were as follows: left main: LAO 0°, caudal (CAU) 49° (95% CI: right anterior oblique [RAO] 8° to LAO 8°, CAU 43° to 54°); left anterior descending with first diagonal branch: LAO 11°, CRA 71° (95% CI: RAO 6° to LAO 27°, CRA 66° to 77°); left circumflex bifurcation with first marginal branch: LAO 24°, CAU 33° (95% CI: LAO 15° to 33°, CAU 25° to 41°); and posterior descending artery and posterolateral branch: LAO 44°, CRA 34° (95% CI: LAO 35° to 52°, CRA 27° to 41°). CONCLUSIONS: CTCA can suggest optimal fluoroscopic viewing angles of coronary artery ostia and bifurcations. As the frequency of use of diagnostic CTCA increases in the future, it has the potential to provide additional information for planning and guiding percutaneous coronary intervention procedures.

A Case of TAVR Complicated by Severe Functional Mitral Regurgitation.

Mitral regurgitation (MR) is a known complication of transcatheter aortic valve replacement (TAVR). We report a case of a 90-year-old man with severe symptomatic aortic stenosis who underwent elective TAVR. The procedure was complicated by severe functional MR from left ventricular stunning and dilatation caused by hypotension throughout the procedure. An Impella CP (Abiomed, Inc, Danvers, MA) was inserted to unload the left ventricle and decrease its size with subsequent improvement in MR severity, which was sustained after Impella CP removal. In conclusion, we present the first case of successful management of post-TAVR severe functional MR with an Impella CP.

Utility of Three-Dimensional Transesophageal Echocardiography for Mitral Annular Sizing in Transcatheter Mitral Valve Replacement Procedures: A Cardiac Computed Tomographic Comparative Study.

BACKGROUND: Three-dimensional (3D) transesophageal echocardiographic (TEE) imaging is frequently used as an initial screening tool in the evaluation of patients who are candidates for transcatheter mitral valve replacement (TMVR). However, little is known about the imaging correlation with the gold standard, computed tomographic (CT) imaging. The aims of this study were to test the quantitative differences between these two modalities and to determine the best 3D TEE parameters for TMVR screening. METHODS: Fifty-seven patients referred to the heart valve clinic for TMVR with prostheses specifically designed for the mitral valve were included. Mitral annular (MA) analyses were performed using commercially available software on 3D TEE and CT imaging. RESULTS: Three-dimensional TEE imaging was feasible in 52 patients (91%). Although 3D TEE measurements were slightly lower than those obtained on CT imaging, measurements of both projected MA area and perimeter showed excellent correlations, with small differences between the two modalities (r = 0.88 and r = 0.92, respectively, P < .0001). Correlations were significant but lower for MA diameters (r = 0.68-0.72, P < .0001) and mitroaortic angle (r = 0.53, P = .0001). Receiver operating characteristic curve analyses showed that 3D TEE imaging had a good ability to predict TMVR screening success, defined by constructors on the basis of CT measurements, with ranges of 12.9 to 15 cm2 for MA area (area under the curve [AUC] = 0.88-0.91, P < .0001), 128 to 139 mm for MA perimeter (AUC = 0.85-0.91, P < .0001), 35 to 39 mm for anteroposterior diameter (AUC = 0.79-0.84, P < .0001), and 37 to 42 mm for posteromedial-anterolateral diameter (AUC = 0.81-0.89, P < .0001). CONCLUSIONS: Three-dimensional TEE measurements of MA dimensions display strong correlations with CT measurements in patients undergoing TMVR screening. Three-dimensional TEE imaging should be proposed as a reasonable alternative to CT imaging in this vulnerable population.

Prognostic Value of SYNTAX Score in Patients With Infarct-Related Cardiogenic Shock: Insights From the CULPRIT-SHOCK Trial.

OBJECTIVES: This study sought to evaluate the prognostic value of the SYNTAX (SYNergy between PCI with TAXUS and Cardiac Surgery) scores in patients undergoing percutaneous coronary intervention (PCI) for multivessel coronary disease with infarct-related cardiogenic shock (CS). BACKGROUND: The prognostic value of the SYNTAX score in this high-risk setting remains unclear. METHODS: The CULPRIT-SHOCK (Culprit Lesion Only PCI versus Multivessel PCI in Cardiogenic Shock) trial was an international, open-label trial, where patients presenting with infarct-related CS and multivessel disease were randomized to a culprit-lesion-only or an immediate multivessel PCI strategy. Baseline SYNTAX score was assessed by a central core laboratory and categorized as low SYNTAX score (SS ≤22), intermediate SYNTAX score (2232). Adjudicated endpoints of interest were the 30-day risk of death or renal replacement therapy (RRT) and 1-year death. Associations between baseline SYNTAX score and outcomes were assessed using multivariate logistic regression. RESULTS: Pre-PCI SYNTAX score was available in 624 patients, of whom 263 (42.1%), 207 (33.2%) and 154 (24.7%) presented with low, intermediate and high SYNTAX score, respectively. A stepwise increase in the incidence of adverse events was observed from low to intermediate and high SYNTAX score for the 30-day risk of death or RRT and the 1-year risk of death (p < 0.001, for all). After multiple adjustments, intermediate and high SYNTAX score remained strongly associated with 30-day risk of death or renal replacement therapy and 1-year risk of all-cause death. There was no significant interaction between SYNTAX score and the coronary revascularization strategy for any outcomes. CONCLUSIONS: In patients presenting with multivessel disease and infarct-related CS, the SYNTAX score was strongly associated with 30-day death or RRT and 1-year mortality.

Effect of oral anticoagulation on clinical outcomes and haemodynamic variables after successful transcatheter aortic valve implantation.

BACKGROUND: The effect of oral anticoagulation on clinical and haemodynamic outcomes following successful transcatheter aortic valve implantation is unclear. AIMS: To evaluate the effect of oral anticoagulation within the first year after transcatheter aortic valve implantation. METHODS: All patients undergoing transcatheter aortic valve implantation in two French tertiary centres from 2010 to 2016 were included prospectively. The composite outcome of death, stroke, readmission for heart failure or major/life-threatening bleeding according to Valve Academic Research Consortium 2 criteria within 1year was evaluated. Valvular haemodynamic deterioration was defined as mean transprosthetic gradient ≥20mmHg or an increase of ≥10mmHg during echocardiographic follow-up. RESULTS: Of the 1139 patients included, 400 (35.1%) were discharged on oral anticoagulation. The primary endpoint was more frequent in the group with versus without oral anticoagulation (29.4% vs. 17.3% 21.5%; hazard ratio 1.83, 95% confidence interval 1.42-2.35). Composite endpoint risk factors were chronic pulmonary and kidney diseases, previous atrial fibrillation, left ventricular ejection fraction ≤30% at discharge and no femoral vascular approach, but not oral anticoagulation prescription at discharge. Conversely, 58 patients were identified with valvular haemodynamic deterioration, including 11 (19%) in the group with oral anticoagulation and 47 (81%) in the group without oral anticoagulation. Valvular haemodynamic deterioration risk factors were absence of oral anticoagulation exposure, increased body mass index, use of a balloon-expandable bioprosthesis and use of a bioprosthesis with diameter ≤23mm. Antithrombotic treatment crossover (i.e. oral anticoagulation interruption or introduction during follow-up) occurred in 9.6% of patients, and was a risk factor for death (adjusted hazard ratio 3.39, 95% confidence interval 1.63-7.07). CONCLUSIONS: Baseline characteristics, rather than oral anticoagulation prescription at discharge, were associated with adverse outcomes following successful transcatheter aortic valve implantation. Conversely, oral anticoagulation was associated with reduced valvular haemodynamic deterioration.

Advances in transcatheter mitral and tricuspid therapies.

BACKGROUND: While rheumatic mitral stenosis has been effectively treated percutaneously for more than 20 years, mitral and tricuspid regurgitation treatment appear as a contemporary unmet need. The advent of transcatheter therapies offer new treatment options to often elderly and frail patients at high risk for open surgery. We aimed at providing an updated review of fast-growing domain of transcatheter mitral and tricuspid technology. MAIN BODY: We reviewed the existing literature on mitral and tricuspid transcatheter therapies. Mitraclip is becoming an established therapy for secondary mitral regurgitation in selected patients with disproportionately severe regurgitation associated with moderate left ventricle dysfunction. Evidence is less convincing for primary mitral regurgitation. Transcatheter mitral valve replacement is a promising emerging alternative to transcatheter repair, for secondary as well as primary mitral regurgitation. But further development is needed to improve delivery. Transcatheter tricuspid intervention arrives late after similar technologies have been developed for aortic and mitral valves and is currently at its infancy. This is likely due in part to previously under-recognized impact of tricuspid regurgitation on patient outcomes. Edge-to-edge repair is the most advanced transcatheter solution in development. Data on tricuspid annuloplasty and replacement is limited, and more research is warranted. CONCLUSION: The future appears bright for transcatheter mitral therapies, albeit their place in clinical practice is yet to be clearly defined. Tricuspid transcatheter therapies might address the unmet need of tricuspid regurgitation treatment.

Predictors of adverse outcomes after transcatheter mitral valve replacement.

Introduction: Transcatheter mitral valve replacement (TMVR) is still a recent technology with numerous unknowns but also great promises. The risk of complications reported in observational studies have limited its adoption by interventional cardiology and surgical communities. Areas covered: Some of the major setbacks of TMVR are complications related to the devices and those related to the pathway. Device-related complications include left ventricle outflow tract (LVOT) obstruction, transcatheter heart valve (THV) dislocation or embolization, thrombosis, and stroke. The transapical approach currently remains the main pathway for TMVR but is associated with high risk of major bleeding and residual apical myocardial scarring. Complication prediction and prevention seem possible. Device-related complication prediction is based on pre-operative imaging including multi-slice computed tomography with 3-dimensional reconstructions and echocardiography which allow LVOT obstruction prediction and appropriate sizing aiming at avoiding dislocation. Industry should aim at the development of transfemoral delivery systems. Nevertheless, several recent feasibility observational studies suggested acceptable safety and efficacy of transcatheter mitral valve replacement. Expert opinion: TMVR complications and transapical delivery are some of the main setbacks which need to be addressed for TMVR to be adopted for broad clinical use.