Choi, Sang Hyun; Kim, So Yeon; Park, Seong Ho; Kim, Kyung Won; Lee, Ja Youn; Lee, Seung Soo; Lee, Moon-Gyu
Journal of magnetic resonance imaging : JMRI
2018May ; 47 ( 5 ) :1237-1250.
PMID : 28901685
ÀúÀÚ »ó¼¼Á¤º¸
Choi, Sang Hyun - Department of Radiology and the Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea.
Kim, So Yeon - Department of Radiology and the Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea.
Park, Seong Ho - Department of Radiology and the Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea.
Kim, Kyung Won - Department of Radiology and the Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea.
Lee, Ja Youn - National Evidence-based Healthcare Collaborating Agency, Seoul, South Korea.
Lee, Seung Soo - Department of Radiology and the Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea.
Lee, Moon-Gyu - Department of Radiology and the Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea.
ABSTRACT
BACKGROUND: Imaging studies, including computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET), have an essential role in the detection and localization of colorectal liver metastasis (CRLM). PURPOSE: To systematically determine the diagnostic accuracy of multidetector row CT (MDCT), gadoxetate disodium-enhanced MRI, and PET/CT for diagnosing CRLM and the sources of heterogeneity between the reported results. STUDY TYPE: Systematic review and meta-analysis. SUBJECTS: In all, 2151 lesions in CT studies, 2301 lesions in MRI studies, 1846 lesions in PET/CT studies, FIELD STRENGTH: 1.5T and 3.0T. ASSESSMENT: We identified research studies that investigated MDCT, gadoxetate disodium-enhanced MRI, and PET/CT to diagnose CRLM by performing a systematic search of PubMed MEDLINE and EMBASE. Study quality was assessed using Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2). STATISTICAL TESTS: According to the types of imaging tests, study heterogeneity and the threshold effect were analyzed and the meta-analytic summary of sensitivity and specificity were estimated. Meta-regression analysis was performed to further investigate study heterogeneity.
RESULTS: Of the 860 articles screened, we found 36 studies from 24 articles reporting a diagnosis of CRLM (11 CT studies, 12 MRI studies, and 13 PET/CT studies). The meta-analytic summary sensitivity for CT, MRI, and PET/CT were 82.1% (95% confidence interval [CI], 74.0-88.1%), 93.1% (95% CI, 88.4-96.0%), and 74.1% (95% CI, 62.1-83.3%), respectively. The meta-analytic summary specificity for CT, MRI, and PET/CT were 73.5% (95% CI, 53.7-86.9%), 87.3% (95% CI, 77.5-93.2%), and 93.9% (95% CI, 83.9-97.8%), respectively. There was no threshold effect in any of the imaging tests. Neoadjuvant chemotherapy significantly decreased the sensitivity of CT and MRI (P < 0.01), although it did not significantly affect the sensitivity of PET/CT. The study design, type of reference standard, and study quality also affected the diagnostic performances of imaging studies. DATA CONCLUSION: Despite the heterogeneous accuracy between studies, gadoxetate disodium-enhanced MRI showed the highest sensitivity, and gadoxetate disodium-enhanced MRI and PET/CT had similar specificities for diagnosing CRLM. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:1237-1250. CI - ??2017 International Society for Magnetic Resonance in Medicine.
keyword
colon; computed tomography; magnetic resonance imaging; metastasis; positron emission tomography
¸µÅ©