Cardiac output in severe tricuspid regurgitation: when more is less

This article refers to ‘Cardiac output states in patients with severe functional tricuspid regurgitation: impact on treatment success and prognosis’ by M. Unterhuber et al., published in this issue on pages 1784–1794. Severe tricuspid regurgitation (TR), with its backward haemodynamic consequences on the kidney, liver and gut, carries an indisputable signature of adverse clinical outcome.1, 2 Advanced right heart failure concurs to the unfavourable outcome, though TR maintains an independent strong prognostic linear trajectory over time starting at earlier mild to moderate degrees.3, 4 Tricuspid regurgitation develops as a consequence of a number of disease states of different aetiological types, broadly consisting of primary or secondary valve involvement. Secondary or functional TR is the most common and worrisome condition, which is sustained by pulmonary hypertension due to left heart disease, valve disease and/or pulmonary vascular disease. Remarkably, once asymptomatic TR of mild to moderate entity is detected, the progression to advanced stages with right ventricular (RV) remodelling may be quite rapid over a period from 2 to 4 years.5 Medical therapies are limited and/or ineffective in these patients and surgical correction of severe and symptomatic isolated TR is burdened by high in-hospital mortality rates.6 In the last few years, novel techniques of transcatheter tricuspid valve repair (TTVR) or replacement have rapidly gained consensus because of lower procedural risk and potential impact on the unrelenting haemodynamic evolution to congestion and multiorgan failure.7, 8 However, interventional strategies are still pending on evidence-based indications, correct patient selection and timing of the procedure. Many determinants of post-procedural success have been identified with a highly informative profile for pulmonary and systemic haemodynamic data, in addition to clinical phenotypes.9, 10 Moreover, markers of advanced RV failure, such as renal injury, liver dysfunction and nutritional impairment, were found to be associated with an increased risk of clinical events after TTVR.11 Remarkably, post-procedural success is one of the strongest predictors of mid-term mortality,12, 13 and propensity matched analyses have shown an impressive prognostic benefit in patients receiving TTVR compared to those treated conservatively, especially in case of heart failure with preserved ejection fraction.2, 14 In the present issue of the Journal, Unterhuber et al.15 report a single centre, single arm study including 114 patients with severe or massive TR, in New York Heart Association functional class III to IV undergoing TTVR. The aetiological background was functional in all cases and the primary study endpoint was all-cause mortality at 1 year. Patients were investigated by a parallel right heart catheterization and echo-derived analysis before and during the periprocedural phase and were divided into three groups according to the echo-estimated cardiac index (CI) with the help of a computer-generated clustering: 33 patients pertaining to the low CI tertile (median CI 1.6 L/min/m2); 77 to the intermediate CI group (median CI 2.2 L/min/m2) and 26 to the high CI cluster (median CI 3.4 L/min/m2). Interestingly, the group of subjects with the highest median CI exhibited the most pronounced comorbid state with renal dysfunction, liver and visceral congestion. However, no intergroup differences in pulmonary haemodynamics, RV dimensions and function were observed under similar therapeutic regimens. The only haemodynamic difference between the high CI group and the others was a significant reduction in systemic vascular resistance. Patients of this group behaved as non-responders showing no changes in periprocedural cardiac output and presenting with the worse outcome (Figure 1). These findings are challenging on previous observations that show an expected linear decline in cardiac output with increasing TR severity16 and recall the attention toward the cardiac output pattern, a basic but often forgotten way to phenotype cardiac failure out of left ventricular ejection fraction, especially useful when signs, symptoms and degree of congestion overlap.17 Of note, all groups showed a similar post-procedural reduction in TR degree but the key point in interpreting the study findings is to prove why subjects with a high CI behaved as clinical non-responders as those in the low CI range. The high CI phenotype was the less represented and what is sure is that, despite comparable intergroup TR grade, pulmonary haemodynamics and RV to pulmonary circulation uncoupling, patients suffered of more advanced right heart failure and hepato-visceral congestion. If reduced systemic vascular resistance is the mediator of a high CI or just an epiphenomenon is not easy to be established because the study lacks mechanistic insights. Nonetheless, findings move into the direction that high CI is secondary to reduced afterload. A series of interesting generating hypotheses which involve the liver–renal–gut axis have been put forth. In particular, a major putative role has been raised for the inflammatory pathways related to visceral congestion and intestinal microenvironment disruption. Of note, chronic hepatic congestion may result in liver fibrosis or cardiac cirrhosis. The difference between the two conditions has not been addressed directly in the study but the higher levels of gamma-glutamyl transpeptidase and alkaline phosphatase in the presence of normal serum aminotransferases in the high CI group portend for a high rate of cardiac chirrosis.18 This would explain per se the worse prognosis regardless of haemodynamic conditions. At variance with very recent findings by the same group focusing on different patterns of longitudinal vs. meridional RV function and outcome before TTVR,19 the present data point against a main role in outcome difference of the right heart and pulmonary vascular disease given a comparable RV to pulmonary circulation uncoupling as assessed by th tricuspid annular plane systolic excursion/pulmonary artery systolic pressure ratio. This finding is quite surprising and remains of not easy interpretation. Nonetheless, the degree of RV to left ventricular interaction, derived by the difference between pulmonary artery wedge pressure and right atrial pressure, as an estimate of cardiac transmural pressure, points toward a higher level of unfavourable interaction for the high CI group, likely reflecting more advanced alterations in RV size, geometry and filling. Despite the combined secondary endpoint of all-cause mortality and heart failure hospitalization, the analysis was not extended to cardiac-related causes of deaths likely because of the limited number of events. This weakens the study findings and leaves unanswered how much cardiac output vs. multiorgan system failure influenced the clinical course. Another potential weakness is the lack of post-TTVR CI monitoring over time. It is unknown whether in the high CI subgroup, the observed steadiness in right atrial pressure and RV dimensions in the early post-procedural setting was followed by a subsequent progressive RV mismatch and unfavourable RV to left ventricular interaction favouring a decline in cardiac output. This is a possibility worth to be tested in future studies. Also, the study does not provide comparative information on TTVR treatment vs. medical therapy which would have helped to clarify whether in the high CI phenotype the interventional procedure was neutral or potentially contributed to accelerate right heart failure and the congestive state. In conclusion, the interest in severe TR and its treatment is quickly expanding and findings by Unterhuber et al.15 are a step forward. They leave us with a paradox, i.e. more (CI) is less (prognosis), and with the conviction that a comprehensive haemodynamic phenotyping incorporating CI assessment in severe TR is central for staging and predicting success after its correction. Conflict of interest: none declared.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.