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

Improving the detection of hypo-vascular liver metastases in multiphase contrast-enhanced CT with slice thickness less than 5 mm using DenseNet

INTRODUCTION

Thinner slices are more susceptible in detecting small lesions but suffer from higher statistical fluctuation. This work aimed to reduce image noise in multiphase contrast-enhanced CT reconstructed with slice thickness thinner than the clinical setting (i.e., 5 mm) using convolutional neural network (CNN) for enabling better detection of hypo-vascular liver metastasis.

METHODS

A DenseNet model was used to generate noise map for multiphase CT reconstructed with slice thickness of 2.5 mm and 1.25 mm. Image denoising was conducted by subtracting the CNN-generated noise map from CT images with reduced photon flux due to thinner slice thickness. The performance of DenseNet was evaluated on CT scans of electron density phantoms and patients with hypovascular liver metastases less than 1.5 cm in terms of Hounsfield Unit (HU) variation, statistical fluctuation, and contrast-to-noise ratio (CNR).

RESULTS

The phantom study demonstrated that the CNN-based denoising method was able to reduce statistical fluctuation in CT images reconstructed with slice thickness of 2.5 mm and 1.25 mm without causing significant edge blurring or variation in HU values. With regards to patient study, it was found that the denoised 2.5-mm and 1.25-mm slices had higher CNR than the conventional 5-mm slices for hypo-vascular liver metastases in all 4 phases of multiphase CT.

CONCLUSION

Our results demonstrated that the detection of hypo-vascular liver metastases in multiphase contrast-enhanced CT with slice thickness less than 5 mm could be improved by using the CNN-based denoising method.

IMPLICATIONS FOR PRACTICE

Reconstruction slice thickness has a strong influence on the image quality of CT imaging. A CNN-based denoising method was used in this work to reduce the image noise in multiphase contrast-enhanced CT reconstructed with slice thickness less than 5 mm.

Assessment of postoperative circulating tumour DNA to predict early recurrence in patients with stage I-III right-sided colon cancer: prospective observational study

BACKGROUND

Right-sided colon cancer (RCC) differs in mutation profile and risk of recurrence compared to distal colon cancer. Circulating tumour DNA (ctDNA) present after surgery can identify patients with residual disease after curative surgery and predict risk of early recurrence.

METHODS

This is a prospective observational biomarker trial with exploration of ctDNA in 50 non-metastatic RCC patients for which oncological right-sided colectomy was performed. Blood samples were collected preoperatively, within 1 month post surgery, 3 months (not mandatory), 6 months and every 6 months thereafter. Plasma cell free DNA and/or tumour was investigated for cancer-related mutations by the next-generation sequencing (NGS) panel AVENIO surveillance specifically designed for ctDNA analysis. Detected mutations were quantified using digital droplet PCR (ddPCR) for follow-up. Recurrence-free survival was explored.

RESULTS

50 patients were recruited. Somatic cancer-related mutations were detected in 47/50 patients. ddPCR validated results from NGS for 27/34 (plasma) and 72/72 samples (tumour). Preoperative ctDNA was detected in 31/47 of the stage I/III patients and the majority of ctDNA positive patients showed reduction of ctDNA after surgery (27/31). ctDNA-positive patients at first postoperative sample had high recurrence risk compared to patients without measurable ctDNA (adjusted hazard ratio: 172.91; 95% c.i.: 8.70 to 3437.24; P: 0.001).

CONCLUSION

ctDNA was detectable in most patients with non-metastatic RCC before surgery. Positive postoperative ctDNA was strongly associated with early recurrence. Detectable postoperative ctDNA is a prognostic factor with high (100%) positive predictive value for recurrence in this cohort of non-metastatic RCC.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov ID: NCT03776591.

Artificial intelligence-powered software detected more than half of the liver metastases overlooked by radiologists on contrast-enhanced CT

PURPOSE

To evaluate the sensitivity of artificial intelligence (AI)-powered software in detecting liver metastases, especially those overlooked by radiologists.

METHODS

Records of 746 patients diagnosed with liver metastases (November 2010-September 2017) were reviewed. Images from when radiologists first diagnosed liver metastases were reviewed, and prior contrast-enhanced CT (CECT) images were checked for availability. Two abdominal radiologists classified the lesions into overlooked lesions (all metastases missed by radiologists on prior CECT) and detected lesions (all metastases if any of them were correctly identified and invisible on prior CECT or those with no prior CECT). Finally, images from 137 patients were identified, 68 of which were classified as "overlooked cases." The same radiologists created the ground truth for these lesions and compared them with the software's output at 2-month intervals. The primary endpoint was the sensitivity in detecting all liver lesion types, liver metastases, and liver metastases overlooked by radiologists.

RESULTS

The software successfully processed images from 135 patients. The per-lesion sensitivity for all liver lesion types, liver metastases, and liver metastases overlooked by radiologists was 70.1%, 70.8%, and 55.0%, respectively. The software detected liver metastases in 92.7% and 53.7% of patients in detected and overlooked cases, respectively. The average number of false positives was 0.48 per patient.

CONCLUSION

The AI-powered software detected more than half of liver metastases overlooked by radiologists while maintaining a relatively low number of false positives. Our results suggest the potential of AI-powered software in reducing the frequency of overlooked liver metastases when used in conjunction with the radiologists' clinical interpretation.

Imaging evaluation of the liver in oncology patients: A comparison of techniques

The liver is commonly affected by metastatic disease. Therefore, it is essential to detect and characterize liver metastases, assuming that patient management and prognosis rely on it. The imaging techniques that allow non-invasive assessment of liver metastases include ultrasonography, computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET)/CT, and PET/MRI. In this paper, we review the imaging findings of liver metastases, focusing on each imaging modality’s advantages and potential limitations. We also assess the importance of different imaging modalities for the management, follow-up, and therapy response of liver metastases. To date, both CT and MRI are the most appropriate imaging methods for initial lesion detection, follow-up, and assessment of treatment response. Multiparametric MRI is frequently used as a problem-solving technique for liver lesions and has evolved substantially over the past decade, including hardware and software developments and specific intravenous contrast agents. Several studies have shown that MRI performs better in small-sized metastases and moderate to severe liver steatosis cases. Although state-of-the-art MRI shows a greater sensitivity for detecting and characterizing liver metastases, CT remains the chosen method. We also present the controversial subject of the "economic implication" to use CT over MRI.

Clinical value of contrast-enhanced computed tomography (CECT) combined with contrast-enhanced ultrasound (CEUS) for characterization and diagnosis of small nodular lesions in liver

Objectives

To explore the clinical value of contrast-enhanced computed tomography (CECT) combined with contrast-enhanced ultrasound (CEUS) for characterization and diagnosis of small nodular lesions in the liver and investigate the association between such small nodular lesions and the degree of tumor differentiation.

Methods

Combined imaging modalities were performed on 120 patients who were admitted by Linyi Maternal and Child Health hospital from December 2018 to December 2020 and diagnosed with hepatic nodular lesions. The CT scans were interpreted by two senior imageologists while the ultrasound scans were analyzed by two senior sonographers. A comparative analysis was carried out on different scan modes and the postoperative or post-puncture pathological results using the t-test, the χ2 test, and the Pearson's correlation analysis.

Results

Compared to the pathological results, definite diagnoses of 55 malignant cases were made using CECT alone, with the coincidence rate of 78.6%; CECT combined with CEUS formed correct diagnoses in 64 cases, and the coincidence rate was up to 91.4%. The difference between the two scan modes was statistically significant (p= 0.03). Based on pathological diagnosis, seventy out of the 120 cases of small nodular lesions were identified as malignant, while the other 50 cases were benign. The single imaging modality diagnosed 63 malignant and 57 benign nodules, whereas the combined modalities identified 68 malignancies and 52 benign conditions. Compared to CECT as a single imaging modality, the combined modalities showed a higher degree of sensitivity and accuracy, and the difference was statistically significant (sensitivity: p= 0.03; accuracy: p= 0.02); in the malignant cases, the magnitudes of contrast enhancement of CT and ultrasound imaging decreased with an increase in the degree of differentiation, indicating a negative correlation between these factors.

Conclusions

CECT combined with CEUS has a higher coincidence rate, greater sensitivity, and better diagnostic accuracy when being used for characterization and diagnosis of small nodular lesions in the liver. A higher degree of tumor differentiation means a decreased magnitude of contrast enhancement and a blurrier boundary, which indicates that CECT and CEUS are complementary to each other in classifying malignant liver nodules. The use of the combined imaging modalities shows clinical value for characterizing small liver nodules and predicting the degree of malignancy.

Management of indeterminate hepatic nodules and evaluation of factors predicting their malignant potential in patients with colorectal cancer

Some liver nodules remain indeterminate despite hepatocyte-specific contrast MRI in patients with colorectal liver metastasis (CRLM). Our objective was to study the natural course and evaluate possible treatment strategies for indeterminate nodules. We retrospectively evaluated patients in whom MRI revealed 'indeterminate' or 'equivocal' nodules between January 2008 and October 2018. Patients were followed up until October 2019 or until death (median, 18 months; (1-130 months)). The incidence of patients with indeterminate nodules on MRI was 15.4% (60 of 389). The sensitivity and specificity of intraoperative ultrasound for detecting indeterminate nodules were 73.68% and 93.75%, respectively, with a positive predictive value of 96.6%. Over half of the patients followed up had benign nodules (58.8%). By comparing characteristics of patients with benign or malignant nodules in the follow up group, the ratio of positive lymph nodes to total number of lymph nodes resected (pLNR) was significantly greater in patients with malignant nodules (P = 0.006). Intraoperative ultrasound could be considered as an adjunct to MRI in patients with indeterminate nodules owing to its high positive predictive value. The pLNR could be used to help select which patients can undergo conservative therapy, at least in metachronous CRLM.

Contrast-enhanced ultrasound can guide the therapeutic strategy by improving the detection of colorectal liver metastases

BACKGROUND & AIMS

CT may miss up to 30% of cases of colorectal liver metastases (CRLMs). We assessed the impact of contrast-enhanced ultrasound (CEUS) on the detection of CRLMs and on changes to the therapeutic strategy; additionally, we assessed the accuracy of CEUS in differentiating unclear focal liver lesions (FLLs) compared to staging-CT.

METHODS

We prospectively analyzed all patients with newly diagnosed and histologically confirmed colorectal cancer (CRC) at our tertiary gastroenterological center between December 2015 and May 2019. CEUS was performed in a total of 296 patients without CRLMs after staging-CT using the contrast agent (SonoVue®). Standard of reference was obtained by MRI or histology to diagnose CRLMs missed by CT. Benign FLLs were confirmed by MRI or follow-up CT (mean follow-up interval: 18 months).

RESULTS

Eight additional CRLMs were detected by CEUS (overall 2.7%; sensitivity 88.9%, specificity 99.0%, positive predictive value 100%, negative predictive value 99.6%). All patients with CRLMs detected only by CEUS were in tumor stage T3/T4 (4.0% additionally detected CRLMs). The number needed to screen to detect 1 additional CRLM by CEUS was 37 in all patients and 24.5 in T3/T4-patients. When results were reviewed by a board-certified radiologist and oncologist, the therapeutic strategy changed in 6 of these 8 patients. Among the 62 patients (20.9%) with unclear FLLs after staging-CT, CEUS determined the dignity (malignant vs. benign) of 98.4% of the FLLs.

CONCLUSION

Overall, CEUS detected 2.7% additional CRLMs (including 4.0% in tumor stage T3/T4) with a significant impact on the oncological therapeutic strategy for 75% of these patients. Patients with tumor stage T3/T4 would particularly benefit from CEUS. We propose CEUS as the first imaging modality for CT-detected lesions of unknown dignity.

LAY SUMMARY

In patients with newly diagnosed colorectal cancer, contrast-enhanced ultrasound (CEUS) detected additional liver metastases after computed tomography (CT). In the majority of these patients, the oncological therapy was changed after obtaining the CEUS results. After staging-CT, 21% of hepatic lesions remained unclear. In these cases, CEUS was accurate to either reveal or exclude liver metastasis in nearly all patients and could reduce costs (e.g., number of MRI scans).

Multimodality Approach in Detection and Characterization of Hepatic Metastases

Early detection of liver metastases is important in patients with known primary malignancies. This plays an important role in treatment planning and impacts on further management of certain primary malignancies.

Magnetic resonance imaging (MRI), computed tomography (CT), and positron emission tomography-computed tomography scans are reported to have high accuracy in the diagnosis of intrahepatic lesions. MRI in particular has the advantages of its high tissue sensitivity and its multiparametric approach.

Hepatic metastatic lesions have considerable overlap in their radiological appearance, and in this article the imaging appearance of various hepatic metastasis and approach is described.

Deep Learning for Automated Segmentation of Liver Lesions at CT in Patients with Colorectal Cancer Liver Metastases

Purpose

To evaluate the performance, agreement, and efficiency of a fully convolutional network (FCN) for liver lesion detection and segmentation at CT examinations in patients with colorectal liver metastases (CLMs).

Materials and Methods

This retrospective study evaluated an automated method using an FCN that was trained, validated, and tested with 115, 15, and 26 contrast material-enhanced CT examinations containing 261, 22, and 105 lesions, respectively. Manual detection and segmentation by a radiologist was the reference standard. Performance of fully automated and user-corrected segmentations was compared with that of manual segmentations. The interuser agreement and interaction time of manual and user-corrected segmentations were assessed. Analyses included sensitivity and positive predictive value of detection, segmentation accuracy, Cohen κ, Bland-Altman analyses, and analysis of variance.

Results

In the test cohort, for lesion size smaller than 10 mm (n = 30), 10-20 mm (n = 35), and larger than 20 mm (n = 40), the detection sensitivity of the automated method was 10%, 71%, and 85%; positive predictive value was 25%, 83%, and 94%; Dice similarity coefficient was 0.14, 0.53, and 0.68; maximum symmetric surface distance was 5.2, 6.0, and 10.4 mm; and average symmetric surface distance was 2.7, 1.7, and 2.8 mm, respectively. For manual and user-corrected segmentation, κ values were 0.42 (95% confidence interval: 0.24, 0.63) and 0.52 (95% confidence interval: 0.36, 0.72); normalized interreader agreement for lesion volume was -0.10 ± 0.07 (95% confidence interval) and -0.10 ± 0.08; and mean interaction time was 7.7 minutes ± 2.4 (standard deviation) and 4.8 minutes ± 2.1 (P < .001), respectively.

Conclusion

Automated detection and segmentation of CLM by using deep learning with convolutional neural networks, when manually corrected, improved efficiency but did not substantially change agreement on volumetric measurements.© RSNA, 2019Supplemental material is available for this article.

Current status of surgical treatment of colorectal liver metastases

Liver metastasis (LM) is one of the major causes of death in patients with colorectal cancer (CRC). Approximately 60% of CRC patients develop LM during the course of their illness. About 85% of these patients have unresectable disease at the time of presentation. Surgical resection is currently the only curative treatment for patients with colorectal LM (CRLM). In recent years, with the help of modern multimodality therapy including systemic chemotherapy, radiation therapy, and surgery, the outcomes of CRLM treatment have significantly improved. This article summarizes the current status of surgical treatment of CRLM including evaluation of resectability, treatment for resectable LM, conversion therapy and liver transplantation for unresectable cases, liver resection for recurrent CRLM and elderly patients, and surgery for concomitant hepatic and extra-hepatic metastatic disease (EHMD). We believe that with the help of modern multimodality therapy, an aggressive oncosurgical approach should be implemented as it has the possibility of achieving a cure, even when EHMD is present in patients with CRLM.