Diagnostic Performance of CMR, SPECT, and PET Imagings for the Detection of Cardiac Amyloidosis: A Meta-Analysis


 Background: Noninvasive myocardial imaging modalities, such as cardiac magnetic resonance (CMR), single photon emission computed tomography (SPECT), and Positron emission tomography (PET), are well-established and extensively used to detect cardiac amyloid (CA). The purpose of this study is to directly compare CMR, SPECT, and PET scans in the diagnosis of CA, and to provide evidence for further scientific research and clinical decision-making.Methods: PubMed, Embase, and Cochrane Library were searched. Studies used CMR, SPECT and/or PET for the diagnosis of CA were included. Pooled sensitivity, specificity, positive and negative likelihood ratio (LR), diagnostic odds ratio (DOR), their respective 95% confidence intervals (CIs) and the area under the summary receiver operating characteristic (SROC) curve (AUC) were calculated. Quality assessment of included studies was conducted.Results: A total of 31 articles were identified for inclusion in this meta-analysis. The pooled sensitivities of CMR, SPECT and PET was 0.84, 0.98 and 0.78, respectively. Their respective overall specificities were 0.87, 0.92 and 0.83. Subgroup analysis demonstrated that 99mTc-HMDP manifested the highest sensitivity (0.99). 99mTc-PYP had the highest specificity (0.95). The AUC values of 99mTc-DPD, 99mTc-PYP, 99mTc-HMDP were 0.89, 0.99, and 0.99, respectively. PET scan with 11C-PIB demonstrated a pooled sensitivity of 0.91 and specificity of 0.97 with an AUC value of 0.98.Conclusion: Our meta-analysis reveals that SEPCT scans present better diagnostic performance for the identification of CA as compared with other two modalities.

Data extraction and quality assessments Two reviewers independently performed the screening of types of articles, titles and abstracts according to the protocol of study selection, hereafter the fulltext reading of the articles was conducted for the nal inclusion. The following information was retrieved from each study included: name of rst author, year of publication, number of patients analyzed, reference standard, type of detection modalities and type of radiopharmaceuticals used in the study, absolute number of participants with TP, TN, FP and FN results. Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) criteria was used to assess the quality of each included studies, this quality scale includes components in terms of participant selection, index test, reference standard, as well as ow and timing [25]. Any disagreements occurred in the process of data extraction and quality assessments were resolved by consensus.

Statistical analysis
Data were analyzed employing the Stata version 15.0 software and Review Manager version 5.3 software at the study level. A p value <0.05 was considered to be statistically signi cant. We calculated pooled sensitivity, speci city, positive and negative likelihood ratio (LR), diagnostic odds ratio (DOR), and their respective 95% con dence intervals (CIs) and the area under the summary receiver operating characteristic (SROC) curve (AUC). The Cochran Q and the I 2 statistics were introduced to assess the heterogeneity of studies included on qualitative and quantitative basis. I 2 values within 0-25%, 25-50%, 50-75%, and 75-100% manifested insigni cant, low, moderate, and high heterogeneity, respectively [26]. Funnel plots were conducted to qualitatively assess potential bias of publication, A Deeks' method was used to statistically test the asymmetry of the funnel plots and detect publication bias. Moreover, we used sensitivity analysis to evaluate the impacts of each single study on the pooled outcomes.

Study selection and characteristics
A total of 367 articles were identi ed from the databases searched. Among them, 51 duplicates were removed and 254 studies were excluded through an initial screening. After a full text assessment for eligibility of the remaining 62 articles, 31 articles with 37 studies, 2577 patients with con rmed or suspected CA were identi ed for inclusion in this meta-analysis. No additional studies were found through reference screening of the included papers. Figure 1 shows the ow of the database search and literature selection process. The quality of the included studies was regarded as high according to the QUADAS-2 scale ( Figure 2). Table 1

Subgroup analysis of SPECT tracers
The numbers of studies using 99m Tc-DPD, 99m Tc-PYP, 99m Tc-HMDP, and 99m Tc-aprotinin for SPECT radiotracers were 5, 8, 5, and 2, respectively. Studies using 99m Tc-aprotinin were not enrolled in pooled analysis for the inadequate number of studies. Overall results demonstrated that 99m Tc-HMDP manifested the

Heterogeneity and publication bias
The I 2 values for meta-analysis of CMR were 64 (pooled sensitivity) and 61 (pooled speci city). The respective I 2 static for SPECT were 94 and 93. As for PET, the I 2 values for pooled analysis of sensitivity and pooled speci city were 84 and 0. Deek's funnel plot asymmetry tests for publication bias yielded p values of 0.89, 0.88, and 0.06 for CMR, SPECT and PET, which revealed that there may be no potential publication bias in the study.

Sensitivity Analysis
Sensitivity analysis was conducted to assess the potential in uence of single study on the overall results. After omitting each study one by one, the pooled results of CMR, SPECT, PET and the corresponding subgroup analysis remained robust.

Discussion
Cardiac amyloidosis is part of systemic amyloidosis, it's characterized by the abnormal accumulation of amyloid brils within the extracellular of the myocardial tissue [27]. Accurate and timely con rmation of CA is of particular importance because cardiac involvement usually can be lethal [28]. Endomyocardial biopsy remains the gold standard for the detection and evaluation of prognosis of CA [29]. However, it's an invasive method and introduces potential damage to human body [30,31]. Among those noninvasive modalities, cardiac ultrasound is widely used, but the diagnostic accuracy is relatively low [32]. It is reported that CMR manifested favorable sensitivity and speci city in the identi cation of CA regardless of its low cost-effectiveness [10,33].
Previous meta-analysis commonly focused on single detection tool of CA [19,[21][22][23]. We conducted a meta-analysis to directly compare the performance of CMR, SPECT and PET for the diagnosis of CA. The analysis was on the updated articles with respect to study design, type of radiotracers in SPECT and PET scans. This is one of the strengths of this study. It is worth noting 20 of the total 31 articles included in this meta-analysis were published in the years of 2019 and 2020, which indicated that noninvasive diagnostic modalities especially SPECT and PET scans have been extensively investigated. In general, results of this meta-analysis revealed that CMR, SPECT, and PET presented high sensitivity and speci city for the diagnosis of CA. and 0.99 [0.98, 1.00], respectively. As for PET scans, PET studies using 11 C-PIB was included in pooled analysis, both the pooled sensitivity and speci city reached more than 0.90, the AUC value of was surprisingly 0.98. One study reported that the sensitivity and speci city of 18 F-orbetaben PET for the detection of CA were 100%, the level of evidence in this study was relatively lower than a meta-analysis, and therefore a possibly pooled analysis of PET scans using 18 F-orbetaben is recommended in the future.
In this meta-analysis, we comprehensively searched the online database to enhance the possibility of retrieving as more eligible studies as we could. Two researchers independently performed the whole process of information extraction under the guidance of the study protocol. Moreover, the heterogeneity between the studies included were assessed using statistical methods. In general, there existed signi cant heterogeneities among studies. The sources of heterogeneity may be attributed to difference in the year of publication, study design (as mentioned above), and patient characteristics. We indented to conduct meta-regression to explore the possible origins of heterogeneity, unfortunately, the numbers of PET and CMR studies was insu cient to complete meta-regression. The underlying sources of heterogeneity would be investigated in further studies. Moreover, results of sensitivity analysis claimed that after omitting individual study one after another, the pooled indicators were robust in this study. The Deek's funnel plot asymmetry tests for publication bias revealed that there may not be publication bias in the meta-analysis. Despite the existence of heterogeneity, we may conclude based on the pooled results that this analysis could provide evidence-based information for scienti c research and practical applications in the process of CA diagnosis. As far as scienti c research is concerned, prospective studies and PET radiotracers with higher spatial resolution need to be further investigated on the basis of results of this meta-analysis. Meta-analysis with larger sample-sized and amount of studies are recommended. With regard to applications in clinical settings, decisionmaking of practitioners in the diagnosis of CA should be made according to technical merit, consideration of cost-effectiveness, and the availability of speci c modalities. In order to enhance diagnostic accuracy of CA, if possible, the combination of different diagnostic tools is recommended.

Availability of data and materials
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
There is no fund support for this study.

Authors' Contributions
ZW conceived and designed this study. ZW and CY were responsible for the collection, extraction, and analysis of the data. ZW was responsible for data analysis and writing the paper. CY performed the quality evaluation of the writing and polished the English language. All authors reviewed the paper and reached an agreement to approve the nal manuscript.