Diagnostic Accuracy of Preoperative Gadoxetic Acid–enhanced 3-T MR Imaging for Malignant Liver Lesions by Using Ex Vivo MR Imaging–matched Pathologic Findings as the Reference Standard

Limited detection of small (≤ 10 mm) colorectal liver metastasis at preoperative CT in patients undergoing liver resection

“…Although our results rather showed no notable difference between the two subgroups in the sensitivity and overall diagnostic yield (i.e., nodule number) according to size category, the comparability between the two subgroups is limited due to the confounders associated with study period. As suggested by previous studies [ 9 , 10 ] on malignant hepatic nodules (not limited to CRLM), gadoxetic acid-enhanced MR imaging appears to detect additional nodules of 6–10 mm. However, this improvement in sensitivity is likely limited for larger nodules because the sensitivity of CT for larger lesions is already high.…”

“…Our first-step matching with predefined criteria yielded only a small number of true-positive (n = 37) and false-negative (n = 33) nodules, and left many nodules uncategorized (n = 391). This finding clearly shows that individual nodule matching is frequently not straightforward, and therefore inevitably involves investigators’ subjective opinion to some extent [ 9 , 25 ], as in our second-step adjudication. Surprisingly, previous studies on CRLM have rarely addressed the difficulties in nodule matching, the associated study limitations, or the measures required to overcome those limitations.…”

“…In the investigations on malignant hepatic nodules (not limited to CRLM), several researchers have attempted rigorous nodule matching by using advanced techniques. These techniques included using ex vivo multiplanar reformation MR images as a link between preoperative cross-sectional images and pathological sections [ 9 ], using infusion fixation of liver specimens [ 9 , 26 ], cutting liver specimens in the approximate plane of preoperative imaging [ 27 ], and cutting liver specimens into very thin (e.g., 3 mm) slices [ 9 , 26 ]. As expected, these studies reported low sensitivities of preoperative imaging for the detection of malignant hepatic nodules.…”

“…While CT is the primary imaging modality for preoperative staging and post-treatment surveillance, meta-analyses [ 6 – 8 ] have reported that magnetic resonance (MR) imaging is more sensitive than CT in the detection of CRLM. In detecting small (≤ 10 mm) malignant hepatic nodules, promising results have been reported particularly by using a hepatocyte-specific MR contrast agent (gadoxetic acid disodium; Primovist, Bayer Healthcare, Germany) [ 9 , 10 ]. Such results have led radiology communities to suggest consensus guidelines recommending gadoxetic acid-enhanced MR imaging for preoperative evaluation of CRLM [ 11 – 13 ].…”

LI-RADS for MR Imaging Diagnosis of Hepatocellular Carcinoma: Performance of Major and Ancillary Features

“…Finally, we used a composite reference standard for determination of either malignancy or benignity. Application of a strict requirement of pathologic proof for all observations would have introduced a selection bias by restricting our study to explant specimens (32). Because we also accepted hepatectomy and biopsy specimens, imaging as part of…”