Feasibility of quantitative 99mTc-DPD scintigraphy SPECT/CT for assessing the burden of ATTR cardiac amyloidosis

Abstract: Background: Amyloid transthyretin (ATTR) cardiomyopathy is caused by the deposition of misfolded proteins, known as amyloid fibrils, in the myocardium. Quantitative Single Photon Emission Computed Tomography (QSPECT) utilizing 99mTc-DPD scintigraphy has the potential to assess ATTR-suspected cardiac amyloidosis (CA). This method could offer improved risk stratification and therapy response monitoring compared to the established Perugini score system. The primary objective of this study is to evaluate the feasibility of employing a quantitative approach and correlating various parameters with LVMI (left ventricular mass index). Method: Initially, planar and volumetric sensitivity measurements were conducted, followed by verification of accuracy measurements. Several torso phantom acquisitions were then performed to evaluate the accuracy and repeatability in terms of recovery coefficient (RC) and repeatability deviation (RD). This served as the foundation for the subsequent in vivo quantification. In this retrospective study, 10 patients underwent 99mTc-DPD scintigraphy, including SPECT/CT of the thorax and echocardiography examinations, as part of a clinical routine for suspected CA. The myocardial SUVmax was determined using a semi-automatic segmentation of the entire heart, excluding the descending and ascending aortas. The bone uptake was also quantified using the SUVmean parameter within the automatically delineated volume of all bones. This enabled the determination of the normalized uptake value nSUVmax, (SUVmax to SUVmean bone). Moreover, an attempt was made to apply an automatic segmentation of the myocardium based on 26% and 36% thresholds, which were developed from the torso phantom acquisitions. This approach allowed for the utilization of the injected dose (ID). Results: The planar calibration factor (CF) exceeded the volumetric cross-calibration factor (CCF) by 3.4%. The anthropomorphic phantom exhibited an underestimation of approximately 50% in the myocardium and around 21% in the kidneys. The average RD in the myocardium and kidneys was 2.3% and 1.6%, respectively. A significant quantitative separation was observed between the ATTR and control groups, comprising 6 and 4 patients, respectively (p<0.01 for SUVmax and nSUVmax, and p<0.02 for SUVmean bone). Correlation analysis revealed a weak positive correlation between SUVmax and LVMI; however, the correlation was not statistically significant (ρ = 0.31, p = 0.39). The ID in the 26% and 36% threshold-based segmented myocardium showed a relatively lower negative correlation with LVMI; although this was observed in the small ATTR cohort and was not statistically significant. The latter outcome resulted from the automatic-delineation method, which was unable to segment grade 0 and 1 patients. Conclusion: This study suggests that both CFs can be used in QSPECT. The phantom measurements indicate good repeatability and a significant quantitative underestimation, primarily due to the partial volume effect (PVE). Size and shape-specific PVE corrections, encompassing various activity concentrations of myocardium-to-blood pool ratios, are essential in QSPECT of the myocardium. Moreover, quantitative SPECT/CT utilizing 99mTc-DPD scintigraphy can effectively distinguish between patients with grade 0/1 and those with grade 2/3. This approach has the potential to enhance diagnostic accuracy and improve risk stratification.