Clinical use of [15O]H2O/[18F]FDG viability positron emission tomography does not reliably predict left ventricular ejection fraction improvement or survival after revascularization.

Abstract

Aims: Previous observational studies suggest that preoperative imaging in patients with chronic ischaemic heart failure (iHF) may identify non-contractile, hypoperfused, yet metabolically viable (hibernating) myocardial segments that can regain function after coronary revascularization. Various imaging techniques, including positron emission tomography (PET) with retention tracers like 82Rb, have shown equivocal results. However, recent randomized studies have found limited value in these methods for predicting postoperative recovery and survival. This study, therefore, aims to assess whether PET viability imaging using the optimal perfusion tracer [15O]H2O, combined with [18F]FDG, provides better predictive accuracy.

Methods and results: Seventy-three patients with chronic iHF and reduced left ventricular ejection fraction (LVEF) (mean baseline LVEF 31 ± 9%) underwent [15O]H2O/[18F]FDG PET viability imaging before potential revascularization. The primary endpoint was a ≥5% absolute increase in LVEF from baseline to follow-up, assessed by echocardiography. In total, 31 of 73 (42%) patients were revascularized, with 16 of 31 (52%) experiencing a ≥5% LVEF improvement postoperatively. Baseline characteristics and revascularization type did not significantly differ between improvers and non-improvers. Receiver operating characteristic analysis of PET metrics to predict LVEF improvement yielded area under curve values ≤0.60, and no baseline characteristics or PET measures predicted survival in revascularized patients.

Conclusion: No [15O]H2O/[18F]FDG PET parameters predicted post-revascularization LVEF improvement or survival in patients with suspected chronic iHF. Thus, the clinical use of PET viability imaging still warrants reconsideration, particularly if non-ischaemic HF is present.