A rigorous standardized uptake value–based framework for monitoring tafamidis response in transthyretin cardiac amyloidosis with bone-avid single-photon emission computed tomography/computed tomography

BACKGROUND: Bone scintigraphy with technetium-labeled tracers has become a cornerstone in diagnosing transthyretin cardiac amyloidosis (ATTR-CM). While visual grading systems like the Perugini Score are widely used, they lack sensitivity for monitoring therapeutic response. Quantitative single-photon emission computed tomography (SPECT)/computed tomography (CT) offers potential for objective assessment of myocardial tracer uptake and disease activity. METHODS: We retrospectively analyzed 25 wild-type ATTR-CM patients who underwent sequential 99mTc-HMDP whole-body scintigraphy and thoracic SPECT/CT before and after tafamidis therapy. SPECT images were reconstructed quantitatively, and myocardial volumes were manually segmented on fused SPECT/CT images. Standardized uptake values (SUVmean, SUVmax, SUVpeak) were extracted and normalized to a reference vertebral volume. Nonparametric tests were used for paired comparisons and correlations with the Perugini Score. RESULTS: All myocardial SUV parameters (SUVmean, SUVmax, SUVpeak) decreased significantly after tafamidis (Wilcoxon signed-rank, two-sided P < 0.0001 for all), while vertebral SUVmean remained stable (P = 0.5539). Myocardial-to-bone SUV ratios also decreased significantly. At baseline, SUVpeak showed a moderate positive association with the Perugini Score (Kendall's τ = 0.476, P = 0.0045); a monotonic increasing trend across categories was confirmed (Jonckheere-Terpstra z = 2.93, two-sided P = 0.0034; Kruskal-Wallis H = 7.75, P = 0.0207). Effect sizes for paired changes (Hodges-Lehmann and rank-biserial r) indicated large and concordant within-patient reductions for SUVmax and SUVpeak. CONCLUSION: Quantitative SPECT/CT using SUV-based metrics provides a reproducible and sensitive method for assessing tafamidis response in ATTR-CM. Compared with existing approaches, our method - based on SUVpeak and bone normalization - offers robustness and clinical relevance, supporting its integration into routine practice.