Colorectal Cancer Liver Metastases: Diagnostic Performance and Prognostic Value of PET/MR Imaging.

Purpose To evaluate the diagnostic performance of combined positron emission tomography (PET) and magnetic resonace (MR) imaging (hereafter, PET/MR imaging) in the detection of liver metastases and to assess its prognostic value in patients with colorectal cancer liver metastases (CRLMs). Materials and Methods Institutional review board approval was obtained for this retrospective study, with waiver of informed consent. A total of 55 patients with 98 CRLMs who underwent PET/MR imaging and multidetector computed tomography (CT) between January 2013 and June 2014 comprised the study population. Of these patients, 34 underwent hepatic resection, 18 of whom also underwent neoadjuvant chemotherapy (NAC). Two board-certificated radiologists independently assessed the four image sets (ie, multidetector CT, whole-body PET, MR imaging with a liver-specific contrast agent [hereafter, EOB MR imaging], and PET/MR imaging). To compare the diagnostic performance of each imaging modality, jackknife alternative free-response receiver operating characteristic and generalized estimating equations were used. To assess prognostic value, recurrence-free survival of the 18 patients who underwent NAC followed by hepatic resection was analyzed by using the Kaplan-Meier method and log-rank test. Results The reader-averaged figure of merit of PET/MR imaging was significantly higher than that of either multidetector CT (P = .003) or PET (P = .020) in the detection of CRLMs. However, no significant difference was observed between figure of merit for PET/MR imaging and that for EOB MR imaging (P = .231). After NAC, six of the 18 patients had isometabolic CRLMs on PET images, and 12 patients had hypermetabolic CRLMs. The 1-year recurrence-free survival rate was 80% in patients with isometabolic CRLMs and 14% in patients with hypermetabolic CRLMs, showing a significant difference (P = .026). Conclusion PET/MR imaging can yield significantly higher diagnostic performance in the detection of CRLMs when compared with the performance of multidetector CT or PET. However, no significant difference in diagnostic performance was observed between PET/MR imaging and EOB MR imaging. In addition, the degree of fluorodeoxyglucose uptake after NAC may have the potential to yield prognostic information for recurrence-free survival after hepatic resection. (©) RSNA, 2016 Online supplemental material is available for this article.