Diagnostic performance of diffusion MRI for pancreatic ductal adenocarcinoma characterisation: A meta-analysis. Issue 139 (June 2021)

• Record Type: Journal Article
• Authors: Chen, Jing; Liu, Shuxue; Tang, Yude; Zhang, Xiongbiao; Cao, Mingming; Xiao, Zheng; Ren, Mingda; Chen, Xianteng
• Is Part Of: European journal of radiology. Issue 139(2021)
• Publication Date: 2021-06
• Subjects: Pancreatic ductal adenocarcinoma -- Magnetic resonance imaging -- Diffusion weighted imaging -- Intravoxel incoherent motion; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109672
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: DWI and IVIM have comparable diagnostic power and high diagnostic accuracy for PDAC. DWI and IVIM can be used to differentiate between PDAC and other non-PDAC diseases. An IVIM model may serve as a non-invasive tool to evaluate tumour vascularity. Abstract: Purpose: To assess the diagnostic performance of intravoxel incoherent motion (IVIM) and diffusion-weighted imaging (DWI) for characterising pancreatic ductal adenocarcinoma (PDAC). Method: A literature search was performed through PubMed, Web of Science, the Cochrane Library, and Embase databases. The search date was updated to extend until 28 October 2020, with no starting time limitation. The pooled sensitivity and specificity were calculated using a bivariate random effects model. Summary receiver operating characteristic curves were constructed, and area under the curve (AUC) of each diffusion parameter was calculated. Subgroup and meta-regression analyses were performed to assess for heterogeneity. Study quality was assessed. Results: Twenty-nine studies involving 1579 participants were included, of which 26 evaluated the apparent diffusion coefficient (ADC) and eight evaluated IVIM, with five evaluating both ADC and IVIM. Pooled sensitivity and specificity of ADC were 83 % (95 % CI, 76 %–88 %, I 2 = 86 %) and 85 % (95 % CI, 79 %–90 %, I 2 = 77 %), respectively, and AUC was 0.91 (95 % CI, 0.88−0.93). The perfusion fraction had the highest diagnostic accuracy in the IVIM model; the pooled sensitivity, specificity, and AUC were 87 % (95 % CI, 81 %–92 %, I 2 = 45 %), 88 % (95 % CI, 77 %–94 %, I 2 = 57 %), and 0.93 (95 % CI, 0.91−0.95), respectively. The pooled sensitivity, specificity and AUC for the tissue diffusion coefficient were 74 % (95 % CI, 55 %–87 %, I 2 = 87 %), 69 % (95 % CI, 52 %–82 %, I 2 = 73 %), and 0.77 (95 % CI, 0.73−0.81), respectively. And the pooled sensitivity, specificity, and AUC for the pseudodiffusion coefficient were 89 % (95 % CI, 77 %–96 %, I 2 = 79 %), 74 % (95 % CI, 60 %–84 %, I 2 = 78 %), and 0.88(95 %CI, 0.85−0.91), respectively. Meta-regression analyses revealed that study design (specificity, P＜0.01), region-of-interest delineation (sensitivity, P = 0.02;specificity, P = 0.03), field strength (sensitivity, P＜0.01), and thickness (sensitivity, P＜0.01; specificity, P = 0.01) were sources of ADC heterogeneity. Conclusions: DWI and IVIM have comparable diagnostic power and good diagnostic performance for characterising PDAC.