Diagnostic value of positron emission tomography using amyloid-binding radiotracers for cardiac amyloidosis: a systematic review and meta-analysis
TL;DR
Amyloid PET demonstrates high sensitivity and specificity for CA detection, even though the evidence base remains limited in terms of the number and scope of available studies.
Abstract
Abstract Aims Amyloid-binding positron emission tomography (PET) tracers have emerged as a promising non-invasive tool for the diagnosis of cardiac amyloidosis (CA). This study aimed to assess the diagnostic accuracy of amyloid PET for detecting CA through a meta-analysis of diagnostic test accuracy. Methods and results PubMed, Cochrane Library, Embase, and Web of Science were searched through 16 March 2026 for studies reporting the diagnostic performance of amyloid PET for CA. The primary analysis was a bivariate random-effects (Reitsma) model yielding summary sensitivity, specificity, the summary receiver operating characteristics curve with corresponding area under the curve (AUC), and positive and negative likelihood ratios. Subgroup analyses were performed by radiotracer and by amyloid subtype. Risk of bias was assessed using QUADAS-2 and publication bias by Deeks’ funnel-plot asymmetry test. Twenty studies comprising 577 patients (407 with CA, 170 controls) met inclusion criteria, covering five radiotracers (11C-PiB, 18F-florbetaben, 18F-florbetapir, 18F-flutemetamol, and 124I-evuzamitide). In the primary bivariate model (k = 17), amyloid PET demonstrated a summary sensitivity of 0.91 (95% CI 0.83–0.96), specificity of 0.90 (95% CI 0.79–0.96), and AUC of 0.96, with low between-study heterogeneity (sensitivity I2 = 13.4%; specificity I2 = 0%). Sensitivity was high and consistent across 11C-PiB, 18F-florbetaben, 18F-florbetapir, and 124I-evuzamitide (0.96–1.00), with the exception of 18F-flutemetamol, which showed lower and heterogeneous sensitivity (0.70). Conclusion Amyloid PET demonstrates high sensitivity and specificity for CA detection, even though the evidence base remains limited in terms of the number and scope of available studies. Prospective multicentre studies in consecutive patients with suspected CA are required to establish its clinical role and standardize acquisition protocols.