Innovative approaches to risk stratification in interventional cardiology

For the podcast associated with this article, please visit https://academic.oup.com/eurheartj/pages/Podcasts. This Focus Issue on interventional cardiology and cardiac and vascular surgery contains a State of the Art Review article entitled ‘Aspirin-free antiplatelet strategies after percutaneous coronary interventions’ by Piera Capranzano from the University of Catania in Italy, and colleagues.1 The authors note that dual-antiplatelet therapy (DAPT) with aspirin and a platelet P2Y12 receptor inhibitor is the standard antithrombotic treatment after percutaneous coronary interventions (PCIs). Several trials have challenged guideline-recommended DAPT after PCI by testing the clinical effects of an aspirin-free antiplatelet approach—consisting of P2Y12 inhibitor monotherapy after a short course (mostly 1–3 months) of DAPT—among patients undergoing PCI without a concomitant indication for oral anticoagulation (OAC). Overall, these studies have shown P2Y12 inhibitor monotherapy after short DAPT to be associated with a significant reduction in the risk of bleeding without an increase in thrombotic or ischaemic events compared with continued DAPT. Yet a recent study reported unfavourable efficacy results associated with the aspirin-free approach compared with conventional DAPT. A P2Y12 inhibitor alone has also been compared with aspirin alone as chronic therapy after DAPT discontinuation, thus challenging the historical role of aspirin as a standard of care for secondary prevention following PCI. A thorough understanding of study designs, populations, treatments, results, and limitations of trials testing P2Y12 inhibitor monotherapy vs. DAPT or vs. aspirin is required to consider adopting this treatment in clinical practice. This review addresses the use of aspirin-free antiplatelet strategies among patients undergoing PCI without a concomitant indication for OAC, providing an overview of clinical evidence, guideline indications, practical implications, ongoing issues, and future perspectives. There is growing interest in the mechanisms and treatment of tricuspid regurgitation (TR).2–7 In a Fast Track Clinical Research contribution entitled ‘TRI-SCORE and benefit of intervention in patients with severe tricuspid regurgitation’, Julien Dreyfus from the Centre Cardiologique du Nord in Saint-Denis, France, and colleagues note that the benefits of TR correction and timing of intervention are unclear.8 This study aimed to compare survival rates after surgical or transcatheter intervention with conservative management according to a TR clinical stage as assessed using the TRI-SCORE. A total of 2413 patients with severe isolated functional TR were enrolled in TRIGISTRY (1217 conservatively managed, 551 isolated tricuspid valve surgery, and 645 transcatheter valve repair). The primary endpoint was survival at 2 years. The TRI-SCORE was low (≤3) in 32%, intermediate (4–5) in 33%, and high (≥6) in 35%. A successful correction was achieved in 97% and 65% of patients in the surgical and transcatheter groups, respectively. Survival rates decreased with increasing values of the TRI-SCORE in the three treatment groups (all P < .0001). In the low TRI-SCORE category, survival rates were higher in the surgical and transcatheter groups than in the conservative management group (93, 87, and 79%, respectively, P = .0002). In the intermediate category, no significant difference between groups was observed overall (80, 71, and 71%, respectively, P = .13), but benefit of the intervention became significant when the analysis was restricted to patients with successful correction (80, 81, and 71%, respectively, P = .009). In the high TRI-SCORE category, survival was not different from conservative management in the surgical and successful repair group (61% and 68% vs. 58%, P = .26 and P = .18 respectively) (Figure 1). Comparison of the survival rates at 2 years between the different treatment modalities according to the TRI-SCORE category in patients with severe isolated functional tricuspid regurgitation.8 The authors conclude that survival progressively decreases with the TRI-SCORE irrespective of treatment modality. Compared with conservative management, an early and successful surgical or transcatheter intervention improves 2-year survival in patients with a low and, to a lesser extent, intermediate TRI-SCORE, while no benefit is observed in the high TRI-SCORE subset. The contribution is accompanied by an Editorial by Andreas Baumbach and Kush P. Patel from the Queen Mary University of London in the UK.9 The authors conclude that TR’s reputation as a ‘bystander, benign condition’ has shifted to a clinically significant and, importantly, treatable condition. Refinement in risk stratification, development of transcatheter technologies, and awareness of TR can substantially improve patient outcomes, making the future of TR both interesting and promising. Predicting personalized risk for adverse events following PCI remains critical in weighing treatment options, employing risk mitigation strategies, and enhancing shared decision-making.1,10–15 In a Clinical Research manuscript entitled ‘Merging machine learning and patient preference: a novel tool for risk prediction of percutaneous coronary interventions’, David Hamilton from the University of Michigan in Ann Arbor, MI, USA aimed to employ machine learning models using pre-procedural variables to accurately predict common post-PCI complications.16 A group of 66 adults underwent a semi-quantitative survey assessing a preferred list of outcomes and model display. The machine learning cohort included 107 793 patients undergoing PCI procedures performed at 48 hospitals in Michigan between 1 April 2018 and 31 December 2021 in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) registry separated into training and validation cohorts. External validation was conducted in the Cardiac Care Outcomes Assessment Program database of 56 583 procedures in 33 hospitals in Washington. Overall, the rate of in-hospital mortality was 1.85%, acute kidney injury 2.51%, new-onset dialysis 0.44%, stroke 0.41%, major bleeding 0.89%, and transfusion 2.41%. The model demonstrated robust discrimination and calibration for mortality [area under the receiver operating characteristic curve (AUC) 0.930], acute kidney injury (AUC 0.893), dialysis (AUC 0.951), stroke (AUC 0.751), transfusion (AUC 0.917), and major bleeding (AUC 0.887). Similar discrimination was noted in the external validation population. Survey subjects preferred a comprehensive list of individually reported post-procedure outcomes (Figure 2). (Top) Graphical representation of the overall study goal combining machine learning and patient feedback to create a patient-centred personalized risk prediction tool. (Bottom) Area under the receiver operating characteristic curves for XGBoost model performance for in-hospital mortality, acute kidney injury (AKI), new requirement for dialysis, stroke, transfusion, and major bleeding.16 The authors conclude that using common pre-procedural risk factors, the BMC2 machine learning models accurately predict post-PCI outcomes. Utilizing patient feedback, the BMC2 models employ a patient-centred tool to clearly display risks to patients and providers (https://shiny.bmc2.org/pci-prediction/). Enhanced risk prediction prior to PCI could help inform treatment selection and shared decision-making discussions. The contribution is accompanied by an Editorial by Patrick Serruys, Yoshinobu Onuma, and Shigetaka Kageyama from the University of Galway in Ireland.17 The authors highlight that the systematic screening approach employing machine learning has the potential to unravel novel and non-traditional critical prognostic factors in the context of randomized trials or registries. Utilization of ‘big data’ for individual decision-making undoubtedly offers substantial benefits to patients. However, it is important to acknowledge that the machine learning process can be challenging to visualize the decision process and weight of each factor, and, while it consistently demonstrates high performance, it requires ongoing scrutiny and validation by clinicians to ensure its accuracy. The fusion of advanced technology with clinical expertise holds promise in advancing the field of personalized medicine. Left ventricular assist devices (LVADs) play a growing role in the management of patients with end-stage heart failure.18,19 Left assist HeartMate 3 (HM3) is a fully magnetically levitated continuous flow LVAD, which received CE marking in 2015. In a Clinical Research article entitled ‘Fully magnetically centrifugal left ventricular assist device and long-term outcomes: the ELEVATE registry’, Jan Schmitto from Hannover Medical School in Germany, and colleagues indicate that the ELEVATE registry was initiated to collect real-world outcomes in patients treated with HM3 post-CE mark approval.20 A total of 540 subjects implanted at 26 centres between March 2015 and February 2017 were included in this registry. Of these, 463 received the device as a primary implant, 19 as a pump exchange, and in 58 patients only anonymized survival data were collected. Survival, adverse events, quality of life (QoL) (EuroQoL-5 Dimensions-5 Levels visual analogue scale), and functional capacity (6-min walk distance) were assessed. Primary implant cohort patients had a mean age of 56 years and were predominantly male (89%) with 48% ischaemic aetiology. The majority of subjects was designated bridge to transplant (66%) and had INTERMACS profiles 1–3 (70%). At baseline, the subjects had poor functional capacity (104 ± 140 m) and impaired QoL (35 ± 19 points). The overall survival rate was 63% and survival free of stroke was 58% at 5 years. Significant improvements in functional capacity and QoL were observed and maintained for 5 years. The authors conclude that real-world data from the ELEVATE registry demonstrate an overall high survival rate for primary implants. Improved QoL and functional capacity are observed at 5 years in this patient population with advanced heart failure. The contribution is accompanied by an Editorial by Mandeep Mehra and Francesco Castagna from the Brigham and Women’s Hospital and Harvard Medical School in Boston, MA, USA, and Javed Butler from the University of Mississippi Medical Center in Jackson, MS, USA.21 The authors conclude that the next paradigm shift will be defined with more natural and pulsatile flow LVADs that are physiologically responsive (smart pumps), more biocompatible, and without onerous external peripheral components. Even as we look towards a more evolved future with support technology, we must not ignore the vividly present opportunity to relieve the ‘therapeutic orphan’ status of advanced heart failure and reduce such suffering. In a Rapid Communications article entitled ‘Mechanisms of rosuvastatin-related acute kidney injury following cardiac surgery: the STICS trial’, Rohan Wijesurendra from the University of Oxford in the UK, and colleagues note that The Statin Therapy In Cardiac Surgery (STICS) trial was a randomized, double-blind, placebo-controlled trial investigating the effects of peri-operative rosuvastatin on post-operative atrial fibrillation and cardiac injury in patients undergoing cardiac surgery.22 While rosuvastatin did not significantly affect either outcome, acute kidney injury (AKI) was unexpectedly and significantly more common in rosuvastatin-allocated patients. To understand the mechanism of post-operative AKI, the authors undertook further analysis of the STICS samples, including measurement of cystatin C, which may have advantages over creatinine for diagnosis of AKI in this context, and several other biomarkers relevant to inflammation and AKI. Creatinine and cystatin C levels in both treatment groups increased from baseline to 48 h after surgery; however, baseline-adjusted creatinine and cystatin C levels at 48 h were significantly higher in rosuvastatin-allocated than in placebo-allocated patients. Concentrations of growth differentiation factor 15 (GDF-15), interleukin-6 (IL-6), procalcitonin (PCT), placental growth factor (PLGF), kidney injury molecule-1 (KIM-1), and neutrophil gelatinase-associated lipocalin (NGAL) were all substantially higher after surgery than at baseline (all P < .001), whereas there was no significant difference between the groups in the rise for GDF-15, IL-6, PCT, PLGF, and NGAL. The authors conclude that that the adverse effect of peri-operative rosuvastatin on renal function may be independent of systemic inflammation and renal epithelial tissue injury. The issue is also complemented by two Discussion Forum contributions. In a commentary entitled ‘There is insufficient evidence to lower the threshold for prophylactic aortic surgery’, Olivier Milleron from the Hopital Bichat-Claude-Bernard in Paris, France and colleagues comment on the recent publication ‘Fate of the unoperated ascending thoracic aortic aneurysm: three-decade experience from the Aortic Institute at Yale University’ by Jinlin Wu from Yale University School of Medicine in New Haven, CT, USA.23 Wu et al. respond in a separate comment.24 The editors hope that this issue of the European Heart Journal will be of interest to its readers. Dr. Crea reports speaker fees from Abbott, Amgen, Astra Zeneca, BMS, Chiesi, Daiichi Sankyo, Menarini outside the submitted work. With thanks to Amelia Meier-Batschelet, Johanna Huggler, and Martin Meyer for help with compilation of this article.