Reliability and feasibility of visual ranking systems and quantitative indices on [99mTc]Tc-DPD imaging for cardiac amyloidosis
In this study, we assessed visual grades and quantitative indices of bone scan and SPECT/CT for CA, in terms of inter-observer agreement and correlation with other indices. Both well-known visual scoring systems showed high inter-observer agreement, and DS corresponded better to H/CL and H/WB than to PS. All SPECT/CT indices tested showed better inter-observer agreement than planar indices. In particular, SUVmax showed not only the highest inter-observer agreement but also significant correlations with other indices.
Bone scan using [99mTc]Tc-DPD or [99mTc]Tc-PYP is an accurate and efficient imaging for ATTR CA, although the exact mechanism of uptake is still unclear. Currently, bone scan is included in the standard diagnostic processes recommended by major cardiology societies12. Because bone scan uptake is stronger in ATTR than in AL, a visual classification of uptake has been proposed and grades 2 or higher are used for ATTR diagnostic criteria. CA in recommendations. In the previously proposed system, PS, attenuation of bone fixation is assessed in addition to myocardial fixation, whereas only myocardial versus rib fixation is considered in DS. Although these visual scoring systems are well established and have been used for clinical practice and numerous studies, they are dependent on the expertise of readers and are vulnerable to inter-observer variability. Thus, quantitative indices have been proposed as complementary methods, and H/CL and H/WB are frequently used on planar scanning.13. Since these indices represent the degree of myocardial absorption, they are expected to correspond better to DS than to PS. In our study, H/CL and H/WB showed clear differences between grades 2 and 3 in DS, but not in PS.
In recent years, quantitative SPECT/CT has been used in clinical practice based on advances in CT scans and imaging technology. The heart and reference organs can be precisely delineated on CT images, and the SUV of a VOI can be measured by counting corrections for attenuation, scatter, sensitivity, and others. Several quantitative indices have been proposed for bone SPECT/CT in AC6.7. In the present study, all SPECT/CT indices tested showed higher inter-observer agreement than planar image indices, likely due to correct delineation of the heart based on CT images. Among them, SUVmax showed almost perfect agreement between the two readers, except for only two grade 1 cases (Fig. 2a). This is a reasonable result because SUVmax is not affected by a minor difference in segmentation if the myocardial uptake is greater than that of adjacent tissues. The two cases that had discordant SUVmax between the two readers had slight uptake in the myocardium, and spillover activity from adjacent ribs may have affected the measurements. Unlike the SUVmax, the C-index showed relatively low inter-observer agreement and low correlations with the planar indices. As spinal absorption is used for C-index calculation, it can be affected by abnormal spinal absorption such as degenerative osteophytes.
In addition to inter-observer agreement, SUVmax showed high correlations with planar image indices of H/WB and H/CL. SUVmean and TMU had slightly higher correlation coefficients than SUVmax, possibly due to similarity in measurement methods. However, SUVmax was more effective in discriminating visual grades, especially low grades 1 vs 2 (Fig. 3) while the other indices were not. Both SUVmean and TMU are assumed to underestimate myocardial uptake because they use the average SUV for the entire VG VOI. Additionally, spillover activity from the ribs to the LV cavity can affect mean absorption, especially in low-quality absorption.
In ROC curve analyses, SUVmax showed excellent diagnostic performance (AUC 0.98–1.00) for discriminating visual grades. The optimal cutoff values of SUVmax to distinguish DS grades 0 versus 1, 1 versus 2, and 2 versus 3 were 2.14, 3.76, and 7.77, respectively. In a recent study, SUVmax 6.0 was reported as an optimal cut-off value for distinguishing grades 0/1 versus 2/314, which is different from ours. Thus, further studies are needed to standardize protocols for [99mTc]Acquisition and measurement Tc-DPD SPECT/CT.
SUVmax also showed significant differences by pathologic diagnosis, AL, and ATTR, despite the small number of EMB cases. Although H/CL and H/WB were also significantly different by pathology, SUVmax showed lower variations in each group, compared to H/CL and H/WB (Fig. 5). SUVmax showed only a weak correlation with E/A, and E/e’ did not show a significant correlation with any of the tests tested. [99mTc]Tc-DPD Image Index. In this study, many patients were negative for CA, and it is assumed that many patients with restrictive cardiomyopathy from other causes such as cardiac sarcoidosis and pulmonary hypertension were included in the study.15.16. However, ECV on cardiac MRI showed a significant correlation with SUVmax, which is consistent with previous reports.7.8.
This study has certain limitations. First, [99mTc]Tc-DPD imaging is used as a screening test in our institution, the number of positive cases was low. Additionally, other cardiac MRI and EMB studies have been performed in a limited number of cases. However, it was not meant to be a critical limitation to assess inter-observer agreement and the efficacy of SUVmax as a quantitative index for [99mTc]Tc-DPD SPECT/CT. The sample size of 18 is acceptable in case the ICC is 0.90 to 0.97, and the statistical power and significance level are 0.8 and 0.05, respectively17. The variable time intervals (2 to 3 h) between [99mTc]Injection of Tc-DPD and image acquisition would be another limitation. However, a previous study reported that the interobserver concordance of visual scores and H/CL index was comparable between 1 and 3 h after injection.18.