Resting myocardial thallium washout rate of the right ventricle predicts 1-year cardiovascular events in heart failure with reduced ejection fraction

Abstract Background Nuclear imaging is widely utilized for prognostic assessment in heart failure (HF), and iodine-123 metaiodobenzylguanidine (MIBG) washout rate (WR) is a well-established parameter for evaluating cardiac sympathetic activity in patients with HF. Although WR of the nuclear imaging primarily assesses the whole heart, no standardized method has been established for the separate assessment of the left ventricle (LV) and right ventricle (RV). Furthermore, perfusion imaging with thallium-201 (Tl) is often performed simultaneously with MIBG imaging to assess the Tl-MIBG mismatch in the heart. While stress-induced Tl WR has been applied in clinical practice, the clinical significance of resting Tl WR remains unclear. Purpose The aim is to evaluate the utility of resting Tl and MIBG WR for predicting HF-related adverse events (HFAEs) in patients with HF with reduced ejection fraction (HFrEF) by assessing the LV and RV separately. Methods In this single center study, 38 consecutive patients with HFrEF (30 males, 53 ± 16 years old, and left ventricular ejection fraction of 24.1 ± 7.3%) who underwent Tl and MIBG dual isotope myocardial scintigraphy from January 2021 to September 2024 were analyzed retrospectively. To assess the LV and RV separately, the collection of planar images was performed in the left anterior oblique position, which was optimized using single-photon emission computed tomography (SPECT) transverse imaging to separate the LV from the RV (Figure A). We evaluated the resting WR of the whole heart, LV, and RV respectively for each radiotracer. The primary endpoint was HFAEs, the composite of HF hospitalizations, cardiovascular deaths, or left ventricular assist device (LVAD) implantations within 1 year. Results Among all 38 patients, 6 combined events (3 HF hospitalizations and 3 LVAD implantations) were observed within 1 year. There were no significant differences between patients with and without HFAEs for resting Tl WR of the whole heart and LV, as well as resting MIBG WR of the whole heart, LV, and RV. In contrast, resting Tl WR of the RV was significantly higher in patients with HFAEs than those without HFAEs (26.2 ± 6.2% vs 20.9 ± 4.2%, P = 0.016). In the Cox proportional hazards model, resting Tl WR of the RV (Hazard ratio 1.30, 95% confidence interval (CI) 1.06-­1.58, P = 0.010) was a significant predictor of HFAEs. The optimal cut­off value of resting Tl WR of the RV for prediction of HFAEs within 1 year was 27.77%, and the sensitivity and specificity were 67% and 97%, respectively. The area under the curve on the receiver operating characteristics curve analysis was 0.74 (95% CI 0.46-1.00) (Figure B). Patients with resting Tl WR of the RV ≥ 27.77% (n = 5) had significantly higher HFAEs compared to those with Tl WR of the RV 27.77% as shown by the Kaplan­Meier analysis (P 0.001) (Figure C). Conclusion In patients with HFrEF, resting Tl WR of the RV is useful for prediction of HFAEs within 1 year.