The risk of heart failure may persist after TAVI due to myocardial changes induced by prolonged pressure overload on the left ventricle related to aortic stenosis. This remains a significant burden after TAVR, noted Bassim El-Sabawi from Vanderbilt University Medical Center in Nashville (Tennessee) and colleagues. Identifying patients at risk, who may benefit from closer surveillance and adjunctive medical therapy to reduce the risk of heart failure, remains an unmet need, they added.
Using data from the Society of Thoracic Surgeons/American College of Cardiology TAVI Registry, the investigators developed and internally validated a clinical prediction model to determine the one-year risk of death and hospitalisation for heart failure following TAVI.
The study included 78,384 patients who underwent TAVI and survived the index hospitalisation between 2016 and 2019. Among them, the one-year rate of all-cause death or hospitalisation for heart failure was 17.4%, including 10.9% deaths, 1.6% of patients rehospitalised at least twice, and 4.9% rehospitalised once.
The predictive model was developed in 62,536 patients from this cohort, based on 39 covariates such as age, sex, race and ethnicity, smoking status, use of certain medications, hemoglobin level, gait speed, among others. It was internally validated in the remaining 15,848 patients.
Analyses showed good predictive performance of the model for the composite endpoint of death and rehospitalisation for heart failure at one year, with a C-index of 0.753 in the development cohort and 0.747 in the validation cohort. The C-index measures a model’s ability to rank patients by risk from highest to lowest and generally ranges from 0.5 (random concordance) to 1 (perfect concordance).
Among low-risk patients, the model also demonstrated good performance, with a C-index of 0.772.
The investigators subsequently developed a simplified model including readily available clinical variables, selecting the 12 main variables associated with the composite endpoint: discharge to home after hospitalisation, KCCQ-OS score, hemoglobin, atrial fibrillation, home oxygen therapy, left ventricular ejection fraction, stage of chronic kidney disease, age above or below 75 years, tricuspid regurgitation, insulin-dependent or non–insulin-dependent diabetes, body mass index, and female sex.
This simplified model demonstrated comparable performance for prediction of the primary composite endpoint, with a C-index of 0.745, and 0.758 in the subgroup of patients at low surgical risk. Its performance for predicting rehospitalisation was similar, with a C-index of 0.744.
“The proposed model identifies that approximately 60% of patients treated with TAVI have an estimated risk of death or rehospitalisation for heart failure of at least 10% within one year following TAVI,” the authors commented.
“To reduce the residual risk related to heart failure after TAVI, randomised clinical trials are needed to rigorously test the efficacy of novel therapies and existing medications whose benefit has already been demonstrated in populations without aortic stenosis,” they added.
In the meantime, in the absence of established medical therapy, identification of at-risk patients may improve outcomes by enabling closer clinical surveillance in high-risk patients. “Indeed, routine post-TAVI follow-up, typically performed at one month and then annually, may be insufficient for high-risk individuals,” in whom medical therapies could be optimised through closer monitoring.
(Circulation: Cardiovascular Interventions)
Source:
