Magnetic Resonance Imaging for Colorectal Cancer Metastasis to the Liver: Comparative Effectiveness Research for the Choice of Contrast Agents

Clinical performance of a simulated abbreviated liver magnetic resonance imaging in combination with contrast-enhanced computed tomography for the baseline evaluation of the liver in patients with colorectal cancer

AIM

To assess the diagnostic accuracy and inter-reader agreement of a simulated abbreviated gadoxetate liver magnetic resonance imaging (MRI) protocol together with contrast-enhanced computed tomography (CE-CT) against a standard gadoxetate MRI for the detection of colorectal liver metastases at baseline.

MATERIALS AND METHODS

Three readers independently evaluated two sets of images per patient, recording number and location of metastases and benign lesions. Set 1 comprised T1w, T2w, DWI, multiphase CE-T1w, and hepatobiliary phase (HBP) images (standard). Set 2 included T2w, HBP, DWI (from Set 1) and CE-CT (simulated abbreviated). Diagnostic performance was compared using McNemar's test. The level of agreement between sets 1 and 2 was determined with Cohen kappa. For agreement in the number of benign lesions and metastases, we applied intraclass correlation coefficient (ICC).

RESULTS

Seventy-five patients (245 metastases, 122 benign lesions) were evaluated. There was no significant difference in diagnostic accuracy between set 1 and 2 for each reader (mean P = 0.74). The total number of metastases and benign lesions showed high agreement between reading set 1 and 2 (κ = 0.81, 0.78). The total number of metastases showed substantial agreement between readers for set 1 and 2 (ICC = 0.99, 0.99). Good agreement was seen for metastatic segmental involvement (κ = 0.84-0.99).

CONCLUSION

At baseline, using a simulated abbreviated liver MRI together with CE-CT showed excellent agreement with standard MRI protocol for liver metastasis detection.

Diagnostic Performance of Deep Learning-Based Lesion Detection Algorithm in CT for Detecting Hepatic Metastasis from Colorectal Cancer

OBJECTIVE

To compare the performance of the deep learning-based lesion detection algorithm (DLLD) in detecting liver metastasis with that of radiologists.

MATERIALS AND METHODS

This clinical retrospective study used 4386-slice computed tomography (CT) images and labels from a training cohort (502 patients with colorectal cancer [CRC] from November 2005 to December 2010) to train the DLLD for detecting liver metastasis, and used CT images of a validation cohort (40 patients with 99 liver metastatic lesions and 45 patients without liver metastasis from January 2011 to December 2011) for comparing the performance of the DLLD with that of readers (three abdominal radiologists and three radiology residents). For per-lesion binary classification, the sensitivity and false positives per patient were measured.

RESULTS

A total of 85 patients with CRC were included in the validation cohort. In the comparison based on per-lesion binary classification, the sensitivity of DLLD (81.82%, [81/99]) was comparable to that of abdominal radiologists (80.81%, p = 0.80) and radiology residents (79.46%, p = 0.57). However, the false positives per patient with DLLD (1.330) was higher than that of abdominal radiologists (0.357, p < 0.001) and radiology residents (0.667, p < 0.001).

CONCLUSION

DLLD showed a sensitivity comparable to that of radiologists when detecting liver metastasis in patients initially diagnosed with CRC. However, the false positives of DLLD were higher than those of radiologists. Therefore, DLLD could serve as an assistant tool for detecting liver metastasis instead of a standalone diagnostic tool.

Development and assessment of an individualized nomogram to predict colorectal cancer liver metastases

Background

This article aims to develop and assess the radiomics paradigm for predicting colorectal cancer liver metastasis (CRLM) from the primary tumor.

Methods

This retrospective study included 100 patients from the First Hospital of Jilin University from June 2017 to December 2017. The 100 patients comprised 50 patients with and 50 without CRLM. The maximum-level enhanced computed tomography (CT) image of primary cancer in the portal venous phase of each patient was selected as the original image data. To automatically implement radiomics-related paradigms, we developed a toolkit called Radiomics Intelligent Analysis Toolkit (RIAT).

Results

With RIAT, the model based on logistic regression (LR) using both the radiomics and clinical information signatures showed the maximum net benefit. The area under the curve (AUC) value was 0.90±0.02 (sensitivity =0.85±0.02, specificity =0.79±0.04) for the training set, 0.86±0.11 (sensitivity =0.85±0.09, specificity =0.75±0.19) for the verification set, 0.906 (95% CI, 0.840-0.971; sensitivity =0.81, specificity =0.84) for the cross-validation set, and 0.899 (95% CI, 0.761-1.000; sensitivity =0.78, specificity =0.91) for the test set.

Conclusions

The radiomics nomogram-based LR with clinical risk and radiomics features allows for a more accurate classification of CRLM using CT images with RIAT.

Intraoperative Ultrasound Staging for Colorectal Liver Metastases in the Era of Liver-Specific Magnetic Resonance Imaging: Is It Still Worthwhile?

Background

To assess the efficacy of intraoperative ultrasound (IOUS) compared with liver-specific magnetic resonance imaging (MRI) in patients with colorectal liver metastases (CRLMs).

Methods

From January 2010 to December 2017, 721 patients underwent MRI as a part of preoperative workup within 1 month before hepatectomy and were considered for the study. Early intrahepatic recurrence (relapse at cut surface excluded) was assessed 6 months after the resection and was considered as residual disease undetected by IOUS and/or MRI. IOUS and MRI performance was compared on a patient-by-patient basis. Long-term results were also studied.

Results

A total of 2845 CRLMs were detected by MRI, and the median number of CRLMs per patient was 2 (1–31). Preoperative chemotherapy was administered in 489 patients (67.8%). In 177 patients, 379 new nodules were intraoperatively found and resected. Among 379 newly identified nodules, 317 were histologically proven CRLMs (11.1% of entire series). The median size of new CRLMs was 6 ± 2.5 mm. Relationships between intrahepatic vessels and tumors differed between IOUS and MRI in 128 patients (17.7%). The preoperative surgical plan was intraoperatively changed for 171 patients (23.7%). Overall, early intrahepatic recurrence occurred in 8.7% of cases. To assess the diagnostic performance, 24 (3.3%) recurrences at the cut surface were excluded; thus, 5.4% of early relapses were considered for analysis. The sensitivity of IOUS was superior to MRI (94.5% vs 75.1%), while the specificity was similar (95.7% vs 95.9%). Multivariate analysis at the hepatic dome or subglissonian and mucinous histology revealed predictive factors of metastases missing at MRI. The 5-year OS (52.1% vs 37.8%, p=0.006) and DF survival (45.1% vs 33%, p=0.002) were significantly worse among patients with new CRLMs than without.

Conclusions

IOUS improves staging in patients undergoing resection for CRLMs even in the era of liver-specific MRI. Intraoperative detection of new CRLMs negatively affects oncologic outcomes.

Abbreviated Gadoxetic Acid-enhanced MRI with Second-Shot Arterial Phase Imaging for Liver Metastasis Evaluation

PURPOSE

To evaluate the feasibility of an abbreviated gadoxetic acid-enhanced MRI protocol including second-shot arterial phase (SSAP) imaging for liver metastasis evaluation.

MATERIALS AND METHODS

For this retrospective study, a total of 197 patients with cancer (117 men and 80 women; mean age, 62.9 years) were included who underwent gadoxetic acid-enhanced MRI performed by using a modified injection protocol for liver metastasis evaluation from July to August 2017. The modified injection protocol included routine dynamic imaging after a first injection of 6 mL and SSAP imaging after a second injection of 4 mL. Image set 1 was obtained with the full original protocol. Image set 2 consisted of T2-weighted, diffusion-weighted, hepatobiliary phase, and SSAP images (the simulated abbreviated protocol). Acquisition time was measured in each image set. The diagnostic performance of each image set was compared by using a jackknife alternative free-response receiver operating characteristic analysis. Image quality evaluation and visual assessment of vascularity were performed on the original arterial phase images, the SSAP images, and their subtraction images.

RESULTS

The acquisition time was significantly shorter in image set 2 than in image set 1 (18.6 vs 6.2 minutes, P <.0001). The reader-averaged figure-of-merit was not significantly different between image sets 1 and 2 (P = .197). The mean motion artifact score was significantly lower for the SSAP images than for the original arterial phase images (P <.001). All hypervascular metastases (n = 72) showed hyperintensity on the SSAP and/or the second subtraction images.

CONCLUSION

An abbreviated MRI protocol including SSAP is feasible for liver metastasis evaluation, providing faster image acquisition while preserving diagnostic performance, image quality, and visual vascularity.Keywords: Abdomen/GI, Comparative Studies, Liver, MR-Imaging, Metastases© RSNA, 2019Supplemental material is available for this article.

Incremental Role of Pancreatic Magnetic Resonance Imaging after Staging Computed Tomography to Evaluate Patients with Pancreatic Ductal Adenocarcinoma

Purpose

The purpose of this study was to investigate the impact of contrast enhanced pancreatic magnetic resonance imaging (MRI) in resectability and prognosis evaluation after staging computed tomography (CT) in patients with pancreatic ductal adenocarcinoma (PDA).

Materials and Methods

From January 2005 to December 2012, 298 patients were diagnosed to have potentially resectable stage PDA on CT. Patients were divided into CT+MR (patients underwent both CT and MRI; n=216) and CT only groups (n=82). Changes in resectability staging in the CT+MR group were evaluated. The overall survival was compared between the two groups. The recurrence-free survival and median time to liver metastasis after curative surgery were compared between the two groups.

Results

Staging was changed from resectable on CT to unresectable state on MRI in 14.4% of (31 of 216 patients) patients of the CT+MR group. The overall survival and recurrence-free survival rates were not significantly different between the two groups (p=0.162 and p=0.721, respectively). The median time to liver metastases after curative surgery in the CT+MR group (9.9 months) was significantly longer than that in the CT group (4.2 months) (p=0.011).

Conclusion

Additional MRI resulted in changes of resectability and treatment modifications in a significant proportion of patients who have potentially resectable state at CT and in prolonged time to liver metastases in patients after curative surgery. Additional MRI to standard staging CT can be recommended for surgical candidates of PDA.