Interplay between guideline-directed medical therapy and QRS duration in patients with heart failure and reduced ejection fraction: data from the Swedish Heart Failure registry
R Laborante, A Abdin, L Benson, V Valente, B N Beer, C Basile, A Villaschi, D D'Amario, C C Cabrera, R Scorza, G SavareseAbstract
Background
The interplay between QRS duration/morphology and effectiveness of contemporary guideline-directed medical therapy(GDMT) is debated, which might affect the implementation of cardiac resynchronization therapy (CRT) and GDMT itself in patients with heart failure with reduced ejection fraction(HFrEF) and dyssynchrony.
Purpose
In patients with HFrEF, to assess: a)whether the association between GDMT and outcomes varies according to QRS duration and morphology; b)whether QRS duration remains an independent prognostic predictor in the context of contemporary HFrEF pharmacotherapy care.
Methods
Patients from the Swedish HF Registry with an EF <40% and receiving optimal medical therapy between 2013 and 2024 were included. Patients with CRT or pacemaker rhythm on the baseline electrocardiogram were excluded. QRS duration was modelled both as a categorical variable with three levels [QRS <120ms; QRS ≥120ms without left bundle branch block(LBBB); LBBB] and as a continuous variable using restricted cubic splines. GDMT was considered both as a three-level categorical variable (≤2, 3, and 4 components) and at the level of individual components(i.e., beta-blocker, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker/angiotensin receptor–neprilysin inhibitor, mineralocorticoid receptor antagonist, and sodium–glucose cotransporter 2 inhibitor). The primary endpoint was a composite of time to first HF hospitalisation or cardiovascular death, with censoring at 3 years or CRT implantation, whichever occurred first. Secondary endpoints were the components of the primary endpoint individually, and time to all-cause death. Multivariable Cox regression models were fitted adjusting for clinical, demographic, and treatment covariates, including also an interaction term between GDMT and QRS morphology/duration.
Results
A total of 16,596 patients were included (median age 74 years; 27% female; 44%, 42%, and 14% receiving ≤2, 3, and 4 components, respectively). Of these, 55% had QRS <120ms, 20% had QRS ≥120ms without LBBB, and 25% had LBBB. Treatment with both three and four components was associated with a lower risk across all outcomes as compared with ≤2 components, irrespective of QRS duration(Figure 1). Consistent findings were observed when each component was analysed separately. Restricted cubic spline modelling, using 120ms of QRS duration as the reference, demonstrated that longer QRS durations were independently associated with a progressively increasing risk of all the outcomes(Figure 2).
Conclusions
GDMT was associated with better prognosis irrespective of QRS duration/morphology. Despite therapeutic advances, QRS prolongation remained progressively and independently associated with worse outcomes, underscoring the persistence of substantial residual risk which can be further reduced by CRT.Figure 1Figure 2