Grown-up congenital heart disease: building evidence where it is badly needed

For the podcast associated with this article, please visit https://academic.oup.com/eurheartj/pages/Podcasts. Nearly one in a hundred babies are born with a congenital heart defect. While in the past their outcome was dismal, thanks to dedicated cardiologists1 and modern surgical, interventional, and medical management, an increasing number of such patients reach adulthood and some even enjoy an almost normal life span.2 However, many have to undergo multiple surgical and interventional procedures3,4 and require lifelong care by specialized physicians working in a team.5 Like other cardiac patients, grown-ups with congenital heart disease also are increasingly managed with catheter-based rather than only surgical interventions. Thus, the ESC Working Group Position Paper ‘Transcatheter adult congenital heart disease interventions: organization of care—recommendations from a Joint Working Group of the European Society of Cardiology (ESC), European Association of Pediatric and Congenital Cardiology (AEPC), and the European Association of Percutaneous Cardiac Intervention (EAPCI)’ is a very timely document.5 Indeed, catheter-based interventions, as either stand-alone or hybrid procedures, continue to expand rapidly in this patient population. This document of the ESC Working Group on Grown-up Congenital Heart Disease therefore not only provides recommendations for the standards of care for these adult patients6 and for the organization of such units6 as in a previous document, but also includes recommendations on transcatheter interventions. These recommendations are important to optimize indications and avoid complications,6 and are based both on published evidence and on a consensus of experienced physicians in the care of adult congenital heart disease along with paediatric and adult interventionists. Unlike coronary artery disease7,8 and heart failure,9 the evidence of medical therapy in adults with congenital heart disease is much less solid. Nevertheless, it is of great importance for such patients, as outlined in a review entitled ‘Pharmacological therapy in adult congenital heart disease: growing need, yet limited evidence’ by Michael Gatzoulis and colleagues from the Royal Brompton Hospital in London, UK.10 They remind us that congenital heart disease is the most common inborn defect. In spite of all the advances in paediatric care, surgical, and catheter procedures, most of these patients have residua from their original operation(s) and require lifelong care. Thus, they also require effective drug therapy for late complications such as heart failure, arrhythmias, pulmonary and systemic hypertension, and thrombo-embolic events. Unlike other cardiovascular areas, drug therapy in adult congenital heart disease is largely empiric. In this review, the rationale, limited evidence, and knowledge gaps of drug therapy in this growing cardiovascular field are discussed specifically for conditions such as heart failure, arrhythmias, thrombosis, pulmonary arterial hypertension, contraception, and pregnancy. The Fontan procedure is a palliative surgical intervention used in children with univentricular hearts. Technically the procedure diverts the venous circulation from the superior and inferior vena cava directly to the pulmonary circulation.11 While effective initially, it is associated with a series of complications with impaired outcomes, among them liver fibrosis and cirrhosis. It is unclear whether non-invasive assessment of liver pathology is helpful to provide clinically relevant information. In their paper ‘The clinical spectrum of Fontan-associated liver disease: results from a prospective multimodality screening cohort’ Eric T.T.L. Tjwa and colleagues from Radboudumc in Nijmegen, The Netherlands comprehensively assessed this issue in 38 consecutive patients.12 Hepatic screening consisted of a blood panel, ultrasonography, elastography, contrast-enhanced magnetic resonance imaging (MRI) or computed tomography (CT), and liver biopsy. Fibrosis parameters, varices, ascites, and splenomegaly were measured on imaging. Liver fibrosis on biopsy was present in all patients and classified as severe in 68%. Collagen proportionate area was 22.5% and correlated with individual fibrosis scores. ELF® and liver stiffness were elevated, but MELD-XI scores were low in all patients. Fibrosis severity correlated neither to ELF® and liver stiffness nor to fibrosis parameters on MRI or CT. Varices were present in 50% and hyperenhancing nodules in 25% of patients, both independent of fibrosis stage, but varices were associated with a higher collagen proportionate area. Thus, the Fontan-associated liver disease spectrum includes both hepatic congestion and severe fibrosis, with signs of portal hypertension and hyperenhancing nodules as significant manifestations. Routine imaging, transient elastography, and serum biomarkers are unable to assess the severity of liver fibrosis accurately in this cohort. Future research should focus on validating new diagnostic tools with biopsy as the reference standard. Although individual biomarkers provide diagnostic information also in patients with congenital heart disease,13 their precision is far from perfect. Thus, machine learning has been increasingly employed to allow for more precise diagnostics and complex prognostic scores.14,15 This has been applied for the first time in this patient population in the study described in the article ‘Machine learning algorithms estimating prognosis and guiding therapy in adult congenital heart disease: data from a single tertiary centre including 10 019 patients’ by Gerhard-Paul Diller and colleagues from the University University Hospital Münster in Germany and the Royal Brompton Hospital in London, UK.16 The authors included 10 019 adult patients and used clinical and demographic data, ECG parameters, cardiopulmonary exercise testing, and selected laboratory markers in deep learning algorithms. Overall, the deep learning algorithms based on >44 000 medical records categorized diagnosis, disease complexity, and NYHA class with an accuracy of 91.1, 97.0, and 90.6%, respectively (Figure 1). Similarly, patient presentation at HeartTeam meetings was predicted with a test sample accuracy of 90.2%. Thus, machine learning algorithms trained on large data sets to estimate prognosis and potentially to guide therapy in adult congenital heart disease appear to provide useful information. These novel findings are put into context in an Editorial by Folkert Meijboom from the Dijkzigt Hospital in Utrecht, The Netherlands.17 Model overview illustrating the combination of deep learning architecture and semiparametric survival model. The deep learning network accepts raw text input and predicts main diagnostic group, disease complexity, and New York Heart Association class. In addition, a disease severity score modelled on 5-year mortality is provided that is combined with additional variables in a multivariate Cox model to provide prognostic information. The speech bubbles illustrate the type of input accepted by the deep learning network. Input examples shown are from different patients and slightly modified to avoid patient confidentiality issues (from Diller G-P, Kempny A, Babu-Narayan SV, Henrichs M, Brida M, Uebing A, Lammers AE, Baumgartner H, Li W, Wort SJ, Dimopoulos K, Gatzoulis MA. Machine learning algorithms estimating prognosis and guiding therapy in adult congenital heart disease: data from a single tertiary centre including 10 019 patients. See pages 1069–1077). Patients with congenital heart disease often have to undergo several surgical procedures during their lifetime. Thus effective transfusion strategies are important in this patient population. In their meta-analysis entitled ‘Restrictive compared with liberal red cell transfusion strategies in cardiac surgery: a meta-analysis’, C. David Mazer and colleagues from the University of Toronto in Canada sought to determine whether a restrictive strategy of red blood cell transfusion at lower haemoglobin concentrations is inferior to a liberal strategy of red blood cell transfusion at higher haemoglobin concentrations in patients undergoing cardiac surgery.18 They searched the electronic databases MEDLINE, EMBASE, and the Cochrane Library up to 17 November 2017 and included 13 trials. The risk ratio of mortality derived from 4545 patients assigned to a restrictive strategy as compared with 4547 transfused according to a liberal strategy was 0.96 [95% confidence interval (CI) 0.76–1.21, I2 = 0] (Figure 2). A restrictive strategy did not have a statistically significant effect on the risk of myocardial infarction, renal failure, or infection. Thus, current evidence does not support the notion that restrictive red blood cell transfusion strategies are inferior to liberal red blood cell strategies in patients undergoing cardiac surgery. These findings are further discussed in an Editorial by Nishith Patel from the Papworth Hospital NHS Foundation Trust in Cambridge (UK) and Gavin J. Murphy from the University of Leicester.19 The current evidence suggests restrictive transfusion strategies are not inferior to liberal transfusion strategies in adult and pediatric patients undergoing cardiac surgery. RBC indicates red blood cell; ICU indicates intensive care unit. ICU stay and hospital stay are reported as days (from Shehata N, Mistry N, da Costa BR, Pereira TV, Whitlock R, Curley GF, Scott DA, Hare GMT, Jüni P, Mazer CD. Restrictive compared with liberal red cell transfusion strategies in cardiac surgery: a meta-analysis. See pages 1081–1088). The editors hope that readers of this issue of the European Heart Journal will find it of interest. With thanks to Amelia Meier-Batschelet for help with compilation of this article.