Echocardiography “systematically underestimates” pulmonary hypertension (PH) severity in those who also have severe tricuspid regurgitation, but an “extreme gradient algorithm” used in a new study could plug knowledge gaps. This is according to a research team led by University Bochum’s Vera Fortmeier, MD, and Mark Lachmann, MD, from the Technical University of Munich and the German Center for Cardiovascular Research, working on retrospective data in 116 patients with severe tricuspid regurgitation undergoing transcatheter tricuspid valve intervention at 2 tertiary centers in Germany. The findings were published Monday online ahead of the Feb. 28 issue of JACC: Cardiovascular Interventions. PH is a key driver for functional tricuspid regurgitation. Because it is closely linked with pulmonary circulation, longstanding PH and subsequently increased right ventricular (RV) afterload cause progressive RV remodeling, eventually leading to functional tricuspid regurgitation in approximately 80% of cases. Yet echocardiographic assessment – “the cornerstone of diagnostics in valvular heart disease” – has significant limitations in assessing the true severity of PH in patients with severe defects of the tricuspid valve, said the researchers. This includes the fact that by focusing on systolic pulmonary artery pressure (sPAP) levels in echocardiography – as recommended in guidance for PH diagnosis – the true severity of PH may be underestimated in patients with impaired RV systolic function, as it is often encountered in advanced stages of cardiac failure. Echocardiographic measurement of peak tricuspid regurgitation velocity is also used to calculate the systolic pressure gradient between the RV and right atrium (RV-RA gradient), applying the modified Bernoulli’s equation. However, the modified equation cannot be properly applied to measure the peak tricuspid regurgitant velocity in patients with severe tricuspid regurgitation “because the equation presumes a unidirectional flow, and because a huge tricuspid valve effective regurgitant orifice area results in rapid pressure equalization between atrium and ventricle with an effective loss of RV-RA gradient,” the researchers stressed. In other words, the severity of PH can be underestimated in echocardiographic assessment, they said, particularly in very sick patients. Ultimately, this obscures the impact of PH on survival after transcatheter tricuspid valve intervention (TTVI), the researchers added, noting that the impact of PH on post-TTVI survival is, therefore, “controversially debated.” The study, therefore, set out to employ an extreme gradient boosting (XGB) algorithm as an emerging machine learning technique as a potential way to estimate the “true” severity of PH solely based on echocardiographic parameters and improve prognostic resolution in patients with severe tricuspid regurgitation undergoing TTVI. The patients underwent TTVI at the Heart Center at University of Cologne Hospital or the Heart and Diabetes Center North Rhine-Westphalia in Bad Oeynhausen, Germany, between January 2017 and December 2020. An external validation cohort of 142 equally treated patients was provided by the Heart Center at the University of Leipzig, Germany. Patients included had tricuspid regurgitation of at least III/V with high symptomatic burden despite optimal medical treatment and were deemed inoperable by the local heart team because of prohibitive perioperative risk. The XGB algorithm was trained using nine echocardiographic parameters as input variables: left ventricular ejection fraction, left ventricular end-systolic diameter, left atrial area, sPAP, basal RV diameter, tricuspid annular plane systolic excursion, tricuspid regurgitation vena contracta width, right atrial area, and inferior vena cava diameter. Across the data set, the researchers found that sPAP was consistently underestimated by echocardiography in comparison to right heart catheterization (40.3 ± 15.9 mmHg vs 44.1 ± 12.9 mmHg; P = 0.0066). “The assessment was most discrepant among patients with severe defects of the tricuspid valve and impaired right ventricular systolic function,” said the researchers. The XGB algorithm could reliably predict mean pulmonary artery pressure (mPAP) levels (R = 0.96, P < 2.2 x 10-16). The findings suggest that PH in patients with severe tricuspid regurgitation can be reliably assessed using echocardiographic parameters in conjunction with an XGB algorithm, the researchers said. Furthermore, elevations in predicted mPAP levels could show increased mortality risk after TTVI. Patients with elevations in predicted mPAP levels ≥ 29.9 mmHg showed significantly reduced 2-year survival after TTVI (58.3% [95% confidence interval (CI): 41.7%-81.6%] vs 78.8% [95% CI: 68.7%-90.5%]; P = 0.026). “Importantly, the poor prognosis associated with elevation in predicted mPAP levels was externally confirmed,” said the researchers (hazard ratio for 2-year mortality: 2.9 [95% CI: 1.5-5.7]; P = 0.002). Looking to the future, the researchers concluded that: “Patients presenting with tricuspid regurgitation subsequent to pulmonary hypertension are routinely excluded from surgical repair caused by distressing prognosis and prohibitively high perioperative risk. “Randomized controlled trials are therefore mandatory in order to investigate whether specifically patients with elevations of predicted mPAP still benefit from transcatheter interventions.” In an accompanying editorial, Nicole Karam, MD, PhD, from Université de Paris and European Hospital Georges Pompidou, France, and Jörg Hausleiter, MD, from Klinikum der Universität München and Munich Heart Alliance, Germany, welcomed the “new and interesting concept using artificial intelligence to improve the echocardiographic assessment of PAP.” They added that the study provides a potential alternative to systematic right heart catheterization in future patients with severe tricuspid regurgitation. “However, efforts are still needed not only to improve the PAP prediction model, but also to validate its performance in larger subgroups of patients with torrential or precapillary pulmonary hypertension, before this promising concept can be adopted in clinical practice,” the editorialists stressed. Sources: Fortmeier V, Lachmann M, Körber MI, et al. Solving the Pulmonary Hypertension Paradox in Patients With Severe Tricuspid Regurgitation by Employing Artificial Intelligence. JACC Cardiovasc Interv 2022;15:381-394. Karam N, Hausleiter J. Can Artificial Intelligence Solve Pulmonary Hypertension Paradox in Severe Tricuspid Regurgitation? JACC Cardiovasc Interv 2022;15:395-396. Image Credit: Yevhenii – stock.adobe.com