P920The impact of tricuspid regurgitation on patient outcome in early and late heart failure

Abstract

Abstract Background Severe tricuspid regurgitation (TR) is associated with poor outcome, but TR remains poorly understood and under-treated. Purpose To examine the impact of TR at different stages of heart failure. Methods 3275 patients with outpatient echocardiogram done at our Hospital in 2013–15 with a mean follow-up of 1092 days were analyzed retrospectively. TR was graded by a semi-quantitative approach using jet-area on multiple views and inferior vena cava (IVC) flow pattern. Multivariate Cox proportional hazard model assessed for mortality, time-to-first heart failure hospitalization, and major adverse cardiovascular event in 3 years. Results were adjusted for age, sex, left ventricular ejection fraction (LVEF), left atrial enlargement, pre-existing cardiovascular, peripheral vascular and cerebrovascular disease, moderate-to-severe aortic or mitral valve disease, pulmonary hypertension, diabetes, chronic kidney disease, chronic obstructive pulmonary disease, malignancy, and heart failure stages (0=no heart failure, A=risk factor present, B=structural abnormality, C=symptomatic D=advanced). Subgroup analysis stratified by heart failure stage 0, stage A/B and stage C/D was done. Kaplan-Meier function, log-rank test, logistic regression, AURUC, and goodness-of-fit test were done. Results In patients with stage A-B heart failure, severe TR had a hazard ratio of 2.93 for death in 3 years compared to no TR (95% CI 1.11–7.73, p=0.03) and moderate TR had a hazard ratio of 2.35 (95% CI 1.28–4.31, P=0.006). In stage C/D, severe TR had a hazard ratio of 2.17 (95% CI 1.12–4.16, p=0.02) and moderate TR had no significant effect (hazard ratio 1.09, p=077). For heart failure hospitalization, severe TR had no significant association in stage A/B but had a hazard ratio of 3.74 in stage C/D (95% CI 1.81–7.7, p<0.001). TR had no impact on major adverse cardiovascular event in this model. No significant interaction was found between TR and heart failure stage, ejection fraction, and valvular heart disease. The model had C-statistics of 0.82 for 3-year mortality, 0.90 for heart failure hospitalization, and 0.81 for MACE, with insignificant Hosmer-Lemeshow goodness-of-fit test p for each, indicating good fit. Conclusion The association between TR and increased mortality in heart failure is apparent early and attenuated later, whereas that of TR and heart failure symptom decompensation appears late. Acknowledgement/Funding None