Carcinome hépatocellulaire : une incidence croissante, une prise en charge « optimisée »

Machine learning for malignant versus benign focal liver lesions on US and CEUS: a meta-analysis

OBJECTIVES

To perform a meta-analysis of the diagnostic performance of learning (ML) algorithms (conventional and deep learning algorithms) for the classification of malignant versus benign focal liver lesions (FLLs) on US and CEUS.

METHODS

Available databases were searched for relevant published studies through September 2022. Studies met eligibility criteria if they evaluate the diagnostic performance of ML for the classification of malignant and benign focal liver lesions on US and CEUS. The pooled per-lesion sensitivities and specificities for each modality with 95% confidence intervals were calculated.

RESULTS

A total of 8 studies on US, 11 on CEUS, and 1 study evaluating both methods met the inclusion criteria with a total of 34,245 FLLs evaluated. The pooled sensitivity and specificity of ML for the malignancy classification of FLLs were 81.7% (95% CI, 77.2-85.4%) and 84.8% (95% CI, 76.0-90.8%) for US, compared to 87.1% (95% CI, 81.8-91.0%) and 87.0% (95% CI, 83.1-90.1%) for CEUS. In the subgroup analysis of studies that evaluated deep learning algorithms, the sensitivity and specificity of CEUS (n = 4) increased to 92.4% (95% CI, 88.5-95.0%) and 88.2% (95% CI, 81.1-92.9%).

CONCLUSIONS

The diagnostic performance of ML algorithms for the malignant classification of FLLs was high for both US and CEUS with overall similar sensitivity and specificity. The similar performance of US may be related to the higher prevalence of DL models in that group.

Demographic, clinical and aetiological characteristics of patients with hepatocellular carcinoma followed between 2012 and 2017 at University Hospital Joseph Raseta Befelatanana, Antananarivo, Madagascar

Purpose

The aim of this study was to describe the demographic, clinical and aetiological characteristics of hepatocellular carcinoma (HCC) in a Malagasy population sample in view to defining an appropriate control program.

Methods

This was a retrospective, descriptive study conducted in the Gastroenterology Department, Joseph Raseta Befelatanana University Hospital, Antananarivo, over a period of 6 years (January 2012 to December 2017).

Results

A total of 42 patients were selected, 29 of whom were men (69.05%) and 13 women (30.95%) (sex ratio: 2.2). The mean age was 56.6 years with extremes of 21 and 82 years. Subjects aged 60-69 years were most affected (35.71%). Abdominal pain was the main revealing symptom (38.10%). The main aetiological factors were: hepatitis B virus (HBV) (42.86%), hepatitis C virus (19.05%) and chronic alcoholism (23.81%). All patients were cirrhotic, of which 23 patients (54.76%) had Child-Pugh B class and 15 (35.71%) Child-Pugh C. Twenty-six patients (61.90%) had α-foetoprotein level plus 500 ng/mL. Six patients (14.29%) had portal thrombosis at diagnosis. Twenty patients (47.62%) had advanced HCC (Barcelona Clinic Liver Cancer C (BCLC C)) and 21 (50%) had end-stage HCC (BCLC D). Management was palliative in 41/42 patients. The in-hospital death rate was 23.81%.

Conclusion

HCC are diagnosed at advanced stage in this study. The prognosis is poor for most patients. HBV infection is the main risk factor. An effort should be made for early diagnosis and prevention.

Deep Learning for the Detection, Localization, and Characterization of Focal Liver Lesions on Abdominal US Images

Purpose

To train and assess the performance of a deep learning-based network designed to detect, localize, and characterize focal liver lesions (FLLs) in the liver parenchyma on abdominal US images.

Materials and Methods

In this retrospective, multicenter, institutional review board-approved study, two object detectors, Faster region-based convolutional neural network (Faster R-CNN) and Detection Transformer (DETR), were fine-tuned on a dataset of 1026 patients (n = 2551 B-mode abdominal US images obtained between 2014 and 2018). Performance of the networks was analyzed on a test set of 48 additional patients (n = 155 B-mode abdominal US images obtained in 2019) and compared with the performance of three caregivers (one nonexpert and two experts) blinded to the clinical history. The sign test was used to compare accuracy, specificity, sensitivity, and positive predictive value among all raters.

Results

DETR achieved a specificity of 90% (95% CI: 75, 100) and a sensitivity of 97% (95% CI: 97, 97) for the detection of FLLs. The performance of DETR met or exceeded that of the three caregivers for this task. DETR correctly localized 80% of the lesions, and it achieved a specificity of 81% (95% CI: 67, 91) and a sensitivity of 82% (95% CI: 62, 100) for FLL characterization (benign vs malignant) among lesions localized by all raters. The performance of DETR met or exceeded that of two experts and Faster R-CNN for these tasks.

Conclusion

DETR demonstrated high specificity for detection, localization, and characterization of FLLs on abdominal US images. Supplemental material is available for this article.  RSNA, 2022Keywords: Computer-aided Diagnosis (CAD), Ultrasound, Abdomen/GI, Liver, Tissue Characterization, Supervised Learning, Transfer Learning, Convolutional Neural Network (CNN).

Fatty acid composition of the erythrocyte membrane and risk of hepatocellular carcinoma in cirrhotic patients

BACKGROUND

Disturbances in fatty acid (FA) metabolism have been reported in cirrhosis, but the role of FAs in the development of hepatocellular carcinoma (HCC) is still unclear. Biomarkers are a promising means to explore the associations between exogenous intake or endogenous production of FAs and cancer risk.

AIM

To estimate the relationship between fatty acid content in erythrocyte membranes and HCC risk in cirrhotic patients METHODS: The "CiRCE" case-control study recruited cirrhotic patients from six French hospitals between 2008 and 2012. Cases were cirrhotic patients with HCC (n = 349); controls were cirrhotic patients without HCC at inclusion (n = 550). FA composition of phospholipids in erythrocyte membranes was determined by high performance gas chromatography. Odds ratios for HCC risk according to FA concentrations were estimated with multivariable logistic regression.

RESULTS

HCC patients were older and more often men (P < 0.001). In both groups, saturated FAs represented more than 39% of all FAs in erythrocyte membranes, mono-unsaturated FAs around 14%, and polyunsaturated FAs around 46%. High levels of C15:0 + C17:0, C20:1 n-9, C18:2 n-6 and C20:2 n-6 were associated with higher risk of HCC. The levels of C18:0 and C20:4 n-6 were lower in HCC cases than in controls.

CONCLUSIONS

The FA composition of erythrocyte membranes differed according to the presence of HCC with higher levels of saturated FAs, linoleic and eicosadienoic acids, and lower levels of stearic and arachidonic acids. These alterations may reflect particular dietary patterns and/or altered FA metabolism. Further investigations are warranted.

[Stereotactic hypofractionated radiation therapy as a bridge to transplantation for hepatocellular carcinoma: Case report of a complete pathological response and review of the literature]

Patients with hepatocellular carcinoma who are on liver transplant waiting list usually require local treatment to limit any risk of tumour growth. Historically percutaneous radiofrequency ablation or transarterial chemoembolization represented the major therapeutic alternatives. Depending on the size, or the topography of the lesion these two techniques may not be feasible. Radiation therapy under stereotactic conditions has recently emerged in the management of localized hepatocellular carcinoma as an alternative to the focused therapies performed to date. We herein report the case of a 43-year-old patient harbouring a complete histological response on explant after liver stereotactic irradiation and discuss its role in the management of hepatocellular carcinoma before liver transplantation.

Hepatocellular carcinoma in 2014: current situation and future prospects

The leading causes of chronic liver disease associated with HCC are hepatitis B and C viruses throughout the world, and alcohol and NASH in France. After increasing for 20-30 years in France, the rise in the incidence of HCC appears to be slowing and the death rates appear to be falling. Screening for HCC by liver ultrasound is performed every 6 months. Assay of serum alpha-fetoprotein has no benefit. In developed countries, failure to identify patients with cirrhosis and inadequate adherence to guidelines greatly reduces the effectiveness of screening for HCC.

[The liver and Europe]

The past 30 years have been witnessing major progress in the knowledge and management of liver diseases. Approximately 29 million people in European Union suffer from chronic liver disease. However, liver diseases have been neglected by most European and national political and public health deciders. A number of measures should be implemented at the European and country levels: an efficient surveillance and data collection policy; broad information of the general population on the nature and importance of liver diseases and the means to diagnose and treat them; treatment and management policies; promotion of research on the causes, risk factors, prevention and management of liver diseases as well as on the most efficient interventions to reduce their incidence, prevalence, morbidity and mortality. The European Association for the Study of the Liver (EASL) has launched in 2006 an ambitious public policy and communication action in order to get recognition of liver diseases as a major cause of morbidity and mortality in Europe and push for appropriate measures. The future of hepatology as a specialty will depend on the success of this initiative.