All-cause mortality in post–myocardial infarction patients with reduced LVEF with and without ICD therapy: insights from the PROFID study
T Chiba, A Sepehri Shamloo, G Hindricks, N DagresAbstract
Background
Previous randomized trials have demonstrated a survival benefit of implantable cardioverter-defibrillator (ICD) therapy in patients with reduced left ventricular ejection fraction (LVEF). However, advances in heart failure therapy and post–myocardial infarction (MI) management over the past decade may have modified contemporary outcome patterns.
Purpose
This study aimed to evaluate the association between ICD implantation and all-cause mortality (ACM) in post-MI patients with LVEF ≤35%, and to assess whether this association differs according to era of inclusion
Methods
We analyzed 140,204 post-MI patients in the PROFID study cohort, including 7,543 ICD patients and 25,058 non-ICD patients. The primary endpoint was ACM. Kaplan–Meier methods were used to estimate mortality. Missing data were handled using multiple imputation. Baseline differences between ICD and non-ICD patients with LVEF ≤35% were adjusted using propensity score overlap weighting. Weighted Cox proportional hazards models were used to evaluate the association between ICD therapy and mortality, including an interaction term between ICD status and era (pre-2016 vs ≥2016). Eras were defined by the year of inclusion or ICD implantation, with 2016 as the cut-off to reflect broader temporal changes in management patterns in this population.
Results
During a median follow-up of 54.5±38.3 months, 33,699 deaths occurred in the overall cohort. Crude mortality proportions were 17.8% in ICD patients, 43.5% in non-ICD patients with LVEF ≤35%. In the propensity score–weighted cohort, before 2016, ACM was lower in ICD patients than in non-ICD patients with LVEF ≤35%, whereas in the post-2016 era (2016 and later), no survival advantage in ICD patients was observed (Figure). Before 2016, ICD implantation was associated with a substantially lower risk of mortality compared with no ICD (HR 0.58, 95% CI 0.54-0.62, P < 0.01). In contrast, in the post-2016 era, the association was attenuated and not statistically significant (HR 1.23, 95% CI 0.93-1.64, P = 0.14). The association between ICD implantation and mortality differed significantly between eras (multiplicative interaction HR 2.13, 95% CI 1.59-2.85, P < 0.01), indicating a marked attenuation of the previously observed association in the contemporary era.
Conclusion
In the contemporary era (2016 and later), ACM in ICD patients did not differ significantly from that in matched non-ICD patients with LVEF ≤35%. These findings suggest evolving outcome patterns in the modern therapeutic era and warrant further investigation.Figure