Background and Aims:
Tricuspid regurgitation (TR) is a highly prevalent disease with potentially dismal prognosis. Development of transcatheter tricuspid edge-to-edge repair (T-TEER) is a therapeutic milestone. However, a specific periprocedural risk assessment tool is lacking. Recently, a dedicated risk score for tricuspid valve surgery has been introduced: TRI-SCORE. In contrast to other risk scores used in clinical practice, such as the EuroSCORE II, this score includes echocardiographic parameters of right ventricular function and clinical signs of right ventricular failure. To date, use of TRI-SCORE in T-TEER patients has not been evaluated. The present study is the first to analyze the predictive performance of TRI-SCORE following T-TEER.

Methods:
180 patients who underwent T-TEER at Ulm University Hospital were consecutively included and were stratified into TRI-SCORE quartiles. Predictive performance of TRI-SCORE for adverse outcome was assessed throughout a follow-up period of 30 days and up to two years.

Results:
All patients had TR grade III – V and were highly symptomatic in terms of heart failure: 78.9% were in New York Heart Association (NYHA) functional class III or IV. Patients had severe comorbidities such as renal failure (median glomerular filtration rate (GFR) 40.0 ml/min (IQR 30.0 – 54.0)), atrial fibrillation (87.8%) and chronic lung disease (11.1%). Median LV-EF was 50.0% (IQR 40.0 – 56.8) and pulmonary hypertension was frequent (median mean pulmonary artery pressure (mPAP) 31.0 mmHg (IQR 25.0 – 37.5)). Median EuroSCORE II was 6.4% (IQR 3.8 – 10.1%), median TRI-SCORE was 6.0 (IQR 4.0 – 7.0). 15 patients (8.3%) were in TRI-SCORE quartile 1, 69 (38.3%) in quartile 2, 71 (39.4%) in quartile 3 and 25 (13.9%) in quartile 4. Procedural success rate was 97.8% and 74.5% of patients had TR grade II or lower at discharge (p < 0.001 compared to baseline). 30 day mortality was 0% in TRI-SCORE quartiles 1 and 2, 2.0% in quartile 3 and 17.0% in quartile 4 (p < 0.001). Early rehospitalisations due to heart failure (HFH) occurred in one patient in quartile 2 (2.0%), in four patients in quartile 3 (7.0%) and in five patients in quartile 4 (22.0%) (p = 0.003 by log-rank test). During a median follow-up period of 168 days mortality was 7.0%, 0%, 15.0% and 52.0%, respectively (p < 0.001). Long-term HFH occurred in two patients in quartile 1 (13.3%), seven in quartile 2 (11.1%), 17 in quartile 3 (27.4%) and 11 in quartile 4 (47.8%) (p < 0.001 by log-rank test). Predictive performance of TRI-SCORE for mortality was excellent (area under the receiver operating characteristics curve (AUROC) for 30 day mortality: 90.3% (95% confidence interval (CI) 81.7 – 98.9%), AUROC for long-term mortality: 85.0% (95% CI 75.4 – 94.7%)) and superior to EuroSCORE II (AUROC 56.6% (95% CI 32.5 – 80.7%) and 64.4% (95% CI 52.8 – 76.6%), respectively). TRI-SCORE also accurately predicted HFH at 30 days (AUROC 77.9% (95% CI 63.0 – 92.8%)) and up to two years (AUROC 70.8% (95% CI 61.1 – 80.5%)).

Conclusions:
TRI-SCORE is a valuable tool for prediction of 30 day and long-term mortality after T-TEER. Its performance is superior to EuroSCORE II. TRI-SCORE might be used for heart team decisions and patient counseling in the future.