Predicting Survival Beyond 5 Years On Continuous-Flow LVAD: A Landmark Analysis From The STELLAR Registry

Abstract Background Five-year survival after continuous-flow left ventricular assist device (LVAD) implantation has improved substantially, resulting in a growing population of ultra-long-term LVAD patients, defined as five years and longer survival on LVAD. Remarkably, determinants of mortality beyond the five-year landmark remain poorly characterized, and ultra-long-term outcomes are rarely investigated. Methods We performed a large bicenter observational landmark analysis including patients who survived at least five years on LVAD support. Implanted devices included HeartMate II, HeartWare (HVAD), and HeartMate 3. Clinical, laboratory, echocardiographic, and hemodynamic parameters at the five-year time point were evaluated to represent predictors of mortality. Cox proportional hazards regression was used to identify independent predictors of mortality from the five-year landmark. Model discrimination was assessed using Harrell’s C-index and time-dependent area under the curve (AUC). Results Among 201 patients surviving ≥5 years on LVAD support, 189 had complete data for multivariable analysis. Median age at implantation was 59.9 years (IQR 52.6-65.8), and 86% were male. During post-landmark follow-up, 77 deaths occurred. Age (HR 1.04 per year, 95% CI 1.02-1.07; p=0.002), hemoglobin (HR 0.84 per mmol/L, 95% CI 0.73-0.97; p=0.016), MELD-XI score (HR 1.10 per unit, 95% CI 1.06-1.14; p<0.001), and ICD presence (HR 1.60, 95% CI 1.00-2.54; p=0.049) were independently associated with mortality beyond the five-year landmark. The model showed moderate discrimination (Harrell’s C-index approximately 0.74). LVAD model was not independently associated with post-landmark mortality (likelihood-ratio p=0.63). Conclusions Among patients who have already survived five years on continuous-flow LVAD support, subsequent mortality was associated with routinely available markers of physiological and end-organ reserve. These findings characterize risk specifically in prevalent long-term LVAD recipients and warrant external validation.