Early occurrence of heart failure hospitalization or ventricular arrhythmia re-define the long-term prognosis after CRT

Aims

Classifying patients as responders or non-responders to cardiac resynchronization therapy (CRT) has been recently challenged, suggesting that preventing heart failure (HF) progression may also provide survival benefits. We assessed a novel classification based on echocardiographic left ventricular (LV) reverse remodelling and the occurrence of acute HF hospitalization (HHF) or sustained ventricular arrhythmia (VA) within 1-year post-CRT.

Methods and results

Patients implanted with a CRT defibrillator (CRT-D) at our department between 2010 and 2020 were classified based on the change in LV ejection fraction (LVEF) within 1 year as responders (increase ≥ 10%), non-progressors (increase <10%) or progressors (decline). Patients in each category were further divided based on the occurrence or absence of an HHF/VA event within 1-year post-implantation. Long-term survival free of heart transplantation or LV assist device implantation was calculated for all six subgroups. Cohorts demonstrating no significant between-group differences were grouped together and reclassified as improved, stabilized or worsened. One hundred nineteen responders, 79 non-progressors and 69 progressors were identified based on the echocardiographic response. Long-term event-free survival was higher for responders as compared with non-progressors (hazard ratio [HR] 0.51, P = 0.002) or progressors (HR 0.34, P < 0.0001). Furthermore, non-progressors had better outcome than progressors (HR 0.63, P = 0.03). Long-term prognosis in patients was superior with versus without an HHF/VA event within each group of responders (HR 0.47, P = 0.03), non-progressors (HR 0.31, P = 0.0001) or progressors (HR 0.38, P = 0.0004). No survival difference was found between responders and non-progressors with no event (HR 0.69, P = 0.09), who were recategorized as improved. Long-term prognosis was also similar in responders with any event and in progressors with no event (HR 0.98, P = 0.88; stabilized), as well as in non-progressors and progressors with any event (HR 0.87, P = 0.63; worsened). Median survival rates demonstrated significant differences between the improved, stabilized and worsened groups (102.3, 62.0 and 24.4 months; HR 0.53, P = 0.006 between improved and stabilized; HR 0.41, P < 0.0001 between stabilized and worsened; HR 0.21, P < 0.0001 between improved and worsened cohorts, respectively).

Conclusions

Long-term survival can be predicted based on the change in LVEF and on the occurrence of an HHF/VA event within 1-year after CRT-D implantation. Stabilized patients have significantly better prognosis as compared with the worsened group. Patients with strikingly poor prognosis can be identified using this assessment method.