Predictors of left ventricular ejection fraction improvement following conduction system and biventricular resynchronization: a subanalysis of the PhysioSync-HF trial
I Bitencourt, A Zimerman, A Vigo, G Gazzoni, F Donner Alves, A dal Forno, L Zimerman, A d'Avila, C A. Polanczyk, L E RohdeAbstract
Background
The degree of improvement in left ventricular ejection fraction following cardiac resynchronization therapy (CRT) is associated with baseline characteristics and post-procedural electrical parameters. However, the relative contribution of these factors in patients undergoing conduction system and biventricular resynchronization strategies remains uncertain.
Purpose
To explore baseline and post-procedural predictors of left ventricular ejection fraction (LVEF) change following conduction system and biventricular resynchronization.
Methods
PhysioSync-HF was a randomized, multicenter, investigator-led trial comparing conduction system pacing (CSP) with biventricular resynchronization (BiVP) in 173 adults with symptomatic heart failure, LVEF<35%, and left bundle branch block (QRS ≥130 ms). In this post hoc analysis, multivariable linear regression models were applied to assess predictors of change in LVEF (ΔLVEF) from baseline up to 12 months, incorporating baseline clinical variables (age, sex, BMI, LVEF strata, tertiles of BNP/NT-proBNP, non-ischemic etiology) and immediate post-procedural ΔQRS from baseline. A formal interaction analysis was conducted to assess whether the predictive effect of immediate post-procedural ΔQRS differed according to resynchronization strategy (CSP vs BiVP). Patients were analyzed according to randomization assignment (modified intention-to-treat).
Results
Among 173 randomized patients, 151 with complete baseline and follow-up data were included in the present analysis (51% female; median LVEF 27.1%; median QRS 180 ms). Across both CSP and BiVP arms, median 12-month ΔLVEF was +10.1% (IQR, 1 to 17.6%) and median post-procedural ΔQRS was -60 ms (IQR, -40 to -75.5 ms). In the multivariable model, ΔQRS was significantly associated with LVEF improvement (+1.2% per 10 ms reduction; 95% CI 0.33–2.1; p=0.008) (Table). Lower baseline LVEF, lower natriuretic peptides, and randomization to BiVP were also independently associated with LVEF improvement. No significant interaction was observed between resynchronization strategy and ΔQRS (p=0.51) (Figure).
Conclusion
In patients with HFrEF and LBBB undergoing CSP and BiVP resynchronization, lower baseline LVEF, lower values of natriuretic peptides, and greater post-procedural QRS narrowing were associated with a greater degree of LVEF improvement at 12 months, with no heterogeneity between resynchronization strategies. While no formal interaction was observed, the trend toward differential slopes across strategies highlights the need for further evaluation of treatment-effect heterogeneity.Table:Predictors of LVEF ChangeImpact of QRS narrowing on LVEF Recovery