LI-RADS Criteria Research Articles

Abstract No. 194 Tumor response after ablative Y-90 transarterial radioembolization for hepatocellular carcinoma based on post-hoc voxel-based dosimetry

Dosimetry for glass trans-arterial radioembolization (TARE) is commonly calculated via a uniform activity distribution model. Recent studies have demonstrated improved tumor response with increased dose to perfused area though no tumor dose threshold is established. This study evaluates dose to tumor and treatment response (mRECIST and LI-RADS criteria) after TARE for early-stage hepatocellular carcinoma (HCC) to determine a threshold tumor dose to predict a complete imaging response.

Automated machine learning for differentiation of hepatocellular carcinoma from intrahepatic cholangiocarcinoma on multiphasic MRI

With modern management of primary liver cancer shifting towards non-invasive diagnostics, accurate tumor classification on medical imaging is increasingly critical for disease surveillance and appropriate targeting of therapy. Recent advancements in machine learning raise the possibility of automated tools that can accelerate workflow, enhance performance, and increase the accessibility of artificial intelligence to clinical researchers. We explore the use of an automated Tree-Based Optimization Tool that leverages a genetic programming algorithm for differentiation of the two common primary liver cancers on multiphasic MRI. Manual and automated analyses were performed to select an optimal machine learning model, with an accuracy of 73–75% (95% CI 0.59–0.85), sensitivity of 70–75% (95% CI 0.48–0.89), and specificity of 71–79% (95% CI 0.52–0.90) on manual optimization, and an accuracy of 73–75% (95% CI 0.59–0.85), sensitivity of 65–75% (95% CI 0.43–0.89) and specificity of 75–79% (95% CI 0.56–0.90) for automated machine learning. We found that automated machine learning performance was similar to that of manual optimization, and it could classify hepatocellular carcinoma and intrahepatic cholangiocarcinoma with an sensitivity and specificity comparable to that of radiologists. However, automated machine learning performance was poor on a subset of scans that met LI-RADS criteria for LR-M. Exploration of additional feature selection and classifier methods with automated machine learning to improve performance on LR-M cases as well as prospective validation in the clinical setting are needed prior to implementation.

Hepatobiliary and Pancreatic: Liver lesions in cirrhosis: When to biopsy?

A 70-year-old man originally from Poland was referred to the hepatoma service for investigation of two liver lesions found during routine hepatocellular carcinoma (HCC) screening. Active medical history included both castrate sensitive metastatic prostate cancer and Child–Pugh A cirrhosis secondary to type 2 diabetes mellitus and previous heavy alcohol use. For his prostate cancer, he was treated with androgen deprivation therapy (ADT) with goserelin and received upfront docetaxel chemotherapy with prednisolone, before being enrolled on a clinical trial of enzalutamide. Prostate specific antigen (PSA) level remained suppressed at <0.01 μg/L on therapy. The patient was asymptomatic with no features of hepatic decompensation. His Child–Pugh status was A6. Biochemistry revealed preserved synthetic liver function; alpha fetoprotein (AFP) level was 5 μg/L. Computed tomography (CT) chest, abdomen, and pelvis demonstrated two 10-mm arterially enhancing hepatic nodules in segments 7/8 and 4a (Fig. 1a) with no portal venous washout (Fig. 1b). On magnetic resonance imaging (MRI) with Primovist, both lesions retained Primovist and hence both were indeterminate (Liver Imaging Reporting and Data System [LI-RADS]-3). Differentials at discussion at the Hepatoma Multidisciplinary meeting included regenerative or dysplastic nodules, given metastatic prostate cancer typically shows hypodensity on contrast-enhanced imaging. The nodules were followed up in tandem with the patient's prostate cancer at 6 monthly intervals and were stable on contrast-enhanced CT (LI-RADS 3) at 12 and 18 months (Fig. 1c) (LI-RADS 3) and therefore did not meet criteria for biopsy. In addition, there were numerous new pulmonary nodules. At 24 months however, CT revealed new lesions concerning for multifocal HCC (Fig. 1d), increase in the size of the segment 7 and 8 lesions with a new 1.5-cm lesion in segment 2 with washout (LI-RADS 5). The patient remained clinically stable and had a suppressed PSA level. He underwent an ultrasound guided biopsy of the segment 2 lesion. Histopathology was consistent with a diagnosis of metastatic small cell neuroendocrine carcinoma. The patient was commenced on single agent Carboplatin to treat his now de-differentiated small cell prostate cancer, to which he has responded well with interval response demonstrated on serial CT of pulmonary and liver lesions in addition to the bony disease on CT SPECT. The diagnosis of HCC is typically made on cross-sectional radiological findings without the need for confirmatory tissue in cirrhotics. LI-RADS scoring system has high sensitivity and specificity in diagnosing HCC in cirrhotic patients and is applicable to lesions ≥ 1 cm. Thus, biopsy is rarely required to diagnose HCC. Current guidelines support the use of the LI-RADS as a classification system for reporting liver lesions. The decision to biopsy is a contentious clinical question that is not definitively answered by current guidelines. Biopsy carries known risks of bleeding, tumor seeding, and failure to obtain adequate tissue for sampling. There are several mimics of HCC on imaging. Intrahepatic cholangiocarcinomas are known to mimic HCC. Tumors such as neuroendocrine tumors, squamous cell carcinoma metastases, and lung cancer metastases require pathological differentiation from HCC. Prostate cancer metastasis to the liver is heralded by hypoenhancement on contrast-enhanced imaging, whereas small cell prostate cancer itself is ill defined in imaging and there is currently a paucity of literature describing its imaging characteristics. It is aggressive in nature and 40–50% of cases are reported to occur in men with pre-existing prostate adenocarcinoma. The PSA is highly sensitive to progression of metastatic prostate adenocarcinoma but not to de-differentiated carcinoma, and it can therefore be difficult to differentiate de-differentiated prostate cancer from HCC in the cirrhotic patient. This case demonstrates that although current clinical practice guidelines deliver a safe approach to the diagnosis of HCC, alternative diagnoses can be missed if they carry similar characteristics on cross-sectional imaging, especially in cirrhotic patients and particularly in those with a cancer history. The role of biopsy in indeterminate liver lesions is often a discussion point and requires multidisciplinary team discussion, and the LI-RADS criteria should always be interpreted in the clinical context.

Validation trial to diagnose liver tumors by LI-RADS criteria

Background and Aims LI-RADS® was created to standardize the reporting and data collection of CT and MR imaging for hepatocellular carcinoma (HCC). Recently, LI-RADS was gained attention because LI-RADS are consistent with and integrated into the American Association for the Study of Liver Diseases (AASLD) 2018 hepatocellular carcinoma (HCC) clinical practice guidance. However, there are few reports that have been verified LI-RADS. We aimed to assess diagnose with LI-RADS from image findings and examine whether the diagnosis made is appropriate. Methods The number of cases was 159 nodules in 159 cases in 3 facilities. There were 81 cases (50.9%) of the high risk group (HBV carrier or cirrhosis case) in LI-RADS. In addition, we examined cases other than the high risk group in LI-RADS in this study. We categorized into US-1 (Negative), US-2 (Subthreshold), US-3 (Positive) with ultrasound (US) LI-RADS. Furthermore, we analyzed cases diagnosed as malignant tumors by CT/MRI but no malignant findings (LR-4 or 5) in US LI-RADS. Conversely, the proportion of cases with benign tumors diagnosed as malignant in US LI-RADS was calculated. Results There were no cases equivalent to US-1 in this study. There were 4 cases of US-2, of which 2 non-malignant tumors, 1 hepatocellular carcinoma (highly differentiated) and 1 metastatic liver tumor were included. Both hepatocellular carcinoma and metastatic liver tumor were LR-3 on the basis of CT/MRI, and the diagnosis was different from that of US diagnosis (1.3%). Conversely, two non-malignant tumors were CT/MRI findings of LR-5 (1.3%), malignant tumors could not be denied and one case had hepatectomy. The cases of US-3 were 121 cases of hepatocellular carcinoma, 19 cases of intrahepatic bile duct cancer, 2 cases of mixed type liver cancer, 10 cases of metastatic liver tumor, and 3 cases of other malignant tumors. 10 cases of hepatocellular carcinoma (6.3%) and 4 cases of metastatic liver tumors (2.5%) were judged as LR-3 by CT/MRI, and diagnosis was different between US and CT/MRI. Conclusion Except for a few cases, US LI - RADS generally performed adequate diagnosis. Further upgrading to contrast enhanced US LI-RADS which corresponds to perflubutane will further improve diagnostic accuracy.

The effects of a transjugular intrahepatic portosystemic shunt on the diagnosis of hepatocellular cancer.

Background and aimsTransjugular intrahepatic portosystemic shunt (TIPS) may be placed to treat complications of portal hypertension by creating a conduit between the hepatic and portal vein. The diagnosis of hepatocellular carcinoma (HCC) is typically made by multiphasic imaging studies demonstrating arterial enhancement with washout on arterial, portal venous, and delayed phase imaging. The aim of our study was to determine how the presence of TIPS would affect the imaging diagnosis of HCC.MethodsThis was a single-center electronic database review of all patients who underwent multiphasic imaging with MRI or CT scan for HCC screening between January 2000 and July 2017 and who were subsequently diagnosed with HCC. Data collected included patient demographics, liver disease characteristics including CPT score, MELD-Na, AFP, type of imaging, tumor stage, and lab values at the time of HCC diagnosis. The diagnosis of HCC was made using LI-RADS criteria on contrast-enhanced CT or MR imaging and confirmed by chart abstraction as documented by the treating clinician. Demographic and imaging characteristics for HCC patients with and without TIPS were compared.ResultsA total of 279 patients met eligibility criteria for the study, 37 (13.2%) of whom had TIPS placed prior to diagnosis of HCC. There was no significant difference in demographics or liver disease characteristics between patients with and without TIPS. Compared to cirrhotic patients with no TIPS prior to HCC diagnosis, patients with TIPS had significantly more scans with a longer duration of surveillance until HCC diagnosis. However, LI-RADS criteria and stage of HCC at diagnosis were not significantly different between both groups. There were no differences in outcomes including liver transplant and survival.ConclusionThe presence of TIPS does not lead to a delayed diagnosis of HCC. It is associated, however, with greater duration of time from first scan to diagnosis of HCC.

Comparison of the accuracy of AASLD and LI-RADS criteria for the non-invasive diagnosis of HCC smaller than 3 cm

Non-invasive imaging is crucial for the early diagnosis and successful treatment of hepatocellular carcinoma (HCC). Terminology and criteria for interpreting and reporting imaging results must be standardized to optimize diagnosis. The aim of this study was to prospectively compare the diagnostic accuracy of the American Association for the Study of Liver Diseases (AASLD) and the 2014 version of Liver Imaging Reporting and Data System (LI-RADS®) criteria for the non-invasive diagnosis of small HCC, and to evaluate the diagnostic value of ancillary features used in the LI-RADS criteria. Between April 2009 and April 2012, patients with cirrhosis and one to three 10-30 mm nodules were enrolled and underwent computed tomography (CT) and magnetic resonance (MR) imaging. The diagnostic accuracy of both the AASLD and the LI-RADS criteria were determined based on their sensitivity, specificity, positive (PPV) and negative predictive values (NPV). A total of 595 nodules were included (559 [341 HCC, 61%] with MR imaging and 529 [332 HCC, 63%] with CT). Overall, no (0%) LR-1 and LR-2, 44 (33%) and 47 (41%) LR-3, 50 (53%) and 54 (55%) LR-4, 244 (94%) and 222 (91%) LR-5 and 4 (67%) and 9 (82%) LR-5V were HCC on MR imaging and CT, respectively. The sensitivity, specificity, PPV/NPV of the AASLD score was 72.5%, 87.6%, 90.2%, and 66.9% for MR imaging, and 71.4%, 77.7%, 84.3%, 61.7% for CT, respectively. For the combination of LR-5V and LR-5 nodules these measures were 72.5%, 89.9%, 91.9% and 67.5% on MRI and 66.9%, 88.3%, 90.9% and 63.3% on CT, respectively. For the combination of LR-5V, LR-5 and LR-4 nodules they were 87.1%, 69.1%, 81.6% and 77.3% on MRI and 85.8%, 66%, 81% on 73.5% on CT, respectively. The 2014 version of the LI-RADS is no more accurate than the AASLD score for the non-invasive diagnosis of small HCC in high-risk patients, but it provides important and complementary information on the probability of having HCC in high-risk patients, allowing for possible changes in the management of these patients. The 2014 version of Liver Imaging Reporting and Data System criteria does not outperform the American Association for the Study of Liver Diseases criteria for the non-invasive diagnosis of hepatocellular carcinoma (HCC) smaller than 3 cm. Liver Imaging Reporting and Data System offers a nodule-based evaluation of the risk of HCC, allowing possible changes in management in these patients. The added value of ancillary features appears limited for the non-invasive diagnosis of small HCC.