Predicting Cardiac Structural And Functional Improvement On Left Ventricular Assist Device Support: The Externally Validated UCAR Score

Abstract

Introduction Although significant cardiac structural and functional improvement is a prerequisite for an LVAD-supported patient to be considered for device weaning, multiple factors including patient willingness, physician expertise, and center experience, weigh in on this complex decision. Existing predictive models defining cardiac recovery as device removal might underestimate the reverse remodeling rates. Objective To develop a predictive tool to identify patients prone to improve cardiac structure and function, independent of the complex decision to deactivate a durable, surgically deployed device. Methods Advanced chronic HF patients (N=703) requiring continuous-flow LVADs were evaluated. After excluding patients with acute HF etiologies and/or inadequate post-LVAD follow-up (<3 months), 3 US programs contributed 390 patients (derivation cohort), while 5 US programs formed the validation cohort (n=245). We defined responders as patients with an LVEF ≥40% and LVEDD ≤6.0 cm within the first year of LVAD support. Multivariable logistic regression on the derivation set yielded a predictive model, which was externally tested; a predictive score was developed. Results Overall, 18.5% of the patients were responders. Compared to the validation, patients in the derivation cohort were more likely white (83 vs 69%; p<0.001), with ischemic HF (42 vs 33%; p=0.04), and longer HF duration (88 vs 66 months; p<0.001). Multivariable predictors of responders and the derived scoring system are shown in the Figure (c-statistic 0.74 [95% CI: 0.66-0.83] and 0.71 [95% CI: 0.59-0.77] in the derivation and validation set). Conclusion Focused on patients experiencing significant reverse remodeling post-LVAD support we developed a predictive score achieving good discriminative performance in distinct, heterogeneous, contemporary cohorts. This tool can be useful in selecting patients to implement diagnostic and therapeutic protocols to promote reverse remodeling and myocardial recovery. Although significant cardiac structural and functional improvement is a prerequisite for an LVAD-supported patient to be considered for device weaning, multiple factors including patient willingness, physician expertise, and center experience, weigh in on this complex decision. Existing predictive models defining cardiac recovery as device removal might underestimate the reverse remodeling rates. To develop a predictive tool to identify patients prone to improve cardiac structure and function, independent of the complex decision to deactivate a durable, surgically deployed device. Advanced chronic HF patients (N=703) requiring continuous-flow LVADs were evaluated. After excluding patients with acute HF etiologies and/or inadequate post-LVAD follow-up (<3 months), 3 US programs contributed 390 patients (derivation cohort), while 5 US programs formed the validation cohort (n=245). We defined responders as patients with an LVEF ≥40% and LVEDD ≤6.0 cm within the first year of LVAD support. Multivariable logistic regression on the derivation set yielded a predictive model, which was externally tested; a predictive score was developed. Overall, 18.5% of the patients were responders. Compared to the validation, patients in the derivation cohort were more likely white (83 vs 69%; p<0.001), with ischemic HF (42 vs 33%; p=0.04), and longer HF duration (88 vs 66 months; p<0.001). Multivariable predictors of responders and the derived scoring system are shown in the Figure (c-statistic 0.74 [95% CI: 0.66-0.83] and 0.71 [95% CI: 0.59-0.77] in the derivation and validation set). Focused on patients experiencing significant reverse remodeling post-LVAD support we developed a predictive score achieving good discriminative performance in distinct, heterogeneous, contemporary cohorts. This tool can be useful in selecting patients to implement diagnostic and therapeutic protocols to promote reverse remodeling and myocardial recovery.