Search

Background and Aims: The natural history of hepatocellular adenomas (HCAs) remains to be better described, especially in nonresected patients. We aim to identify the predictive factors of HCA evolution after estrogen-based contraception discontinuation., Approach and Results: We retrospectively included patients with a histological diagnosis of HCA from three centers. Clinical, radiological, and pathological data were collected to identify predictive factors of radiological evolution per Response Evaluation Criteria in Solid Tumors, version 1.1, and occurrence of complications (bleeding, malignant transformation). We built a score using variables that modulate estrogen levels: body mass index and duration of estrogen-based contraception. An external cohort was used to validate this score. 183 patients were included in the cohort, including 161 women (89%) using estrogen-based contraception for a median of 12 years. Thirty percent of patients had at least one HNF1A -inactivated HCA, 45.5% at least one inflammatory HCA, and 11% at least one HCA with activation of β-catenin (bHCA). Twenty-one symptomatic bleedings (11%) and eleven malignant transformations (6%) occurred. Ages < 37 years old ( p = 0.004) and HCA > 5 cm at imaging were independently associated with symptomatic bleeding ( p = 0.003), whereas a bHCA was associated with malignant transformation ( p < 0.001). After a median follow-up of 5 years, radiological regression was observed in 31%, stabilization in 47%, and progression in 22% of patients. Weight loss was associated with regression ( p < 0.0001) and weight gain with progression ( p = 0.02). The estrogen exposure score predicted radiological regression (odds ratio, 2.33; confidence interval 95%, 1.29-4.19; p = 0.005) with a linear relationship between the rate of estrogen exposure and the probability of regression. This result was confirmed in an external cohort of 72 female patients ( p = 0.003)., Conclusion: Weight variation is strongly associated with radiological evolution after oral contraception discontinuation. A score of estrogen exposure, easily assessable in clinical practice at diagnosis, predicts regression of HCA., (Copyright © 2023 American Association for the Study of Liver Diseases.)

Background and Aims: Standardized and robust risk-stratification systems for patients with hepatocellular carcinoma (HCC) are required to improve therapeutic strategies and investigate the benefits of adjuvant systemic therapies after curative resection/ablation., Approach and Results: In this study, we used two deep-learning algorithms based on whole-slide digitized histological slides (whole-slide imaging; WSI) to build models for predicting survival of patients with HCC treated by surgical resection. Two independent series were investigated: a discovery set (Henri Mondor Hospital, n = 194) used to develop our algorithms and an independent validation set (The Cancer Genome Atlas [TCGA], n = 328). WSIs were first divided into small squares ("tiles"), and features were extracted with a pretrained convolutional neural network (preprocessing step). The first deep-learning-based algorithm ("SCHMOWDER") uses an attention mechanism on tumoral areas annotated by a pathologist whereas the second ("CHOWDER") does not require human expertise. In the discovery set, c-indices for survival prediction of SCHMOWDER and CHOWDER reached 0.78 and 0.75, respectively. Both models outperformed a composite score incorporating all baseline variables associated with survival. Prognostic value of the models was further validated in the TCGA data set, and, as observed in the discovery series, both models had a higher discriminatory power than a score combining all baseline variables associated with survival. Pathological review showed that the tumoral areas most predictive of poor survival were characterized by vascular spaces, the macrotrabecular architectural pattern, and a lack of immune infiltration., Conclusions: This study shows that artificial intelligence can help refine the prediction of HCC prognosis. It highlights the importance of pathologist/machine interactions for the construction of deep-learning algorithms that benefit from expert knowledge and allow a biological understanding of their output., (© 2020 by the American Association for the Study of Liver Diseases.)

To date, genomic analyses of hepatocellular carcinoma (HCC) have been limited to early stages obtained from liver resection. We aim to describe the genomic profiling of HCC from early to advanced stages. We analyzed 801 HCC from 720 patients (410 resections, 137 transplantations, 122 percutaneous ablations, and 52 noncurative) for 190 gene expressions and for 31 gene mutations. Forty-one advanced HCC and 156 whole exome of Barcelona Clinic Liver Cancer (BCLC) 0/A were analyzed by whole-exome sequencing. Genomic profiling was correlated with tumor stages, clinical features, and survival. Our cohort included patients classified in BCLC stage 0 (9.4%), A (59.5%), B (16.2%), and C (14.9%). Among the overall 801 HCC, the most frequently mutated genes were telomerase reverse transcriptase (TERT) (58.1%), catenin beta 1 (CTNNB1) (30.7%), tumor protein 53 (TP53; 18.7%), AT-rich interaction domain 1A (ARID1A) (13%), albumin (11.4%), apolipoprotein B (APOB) (9.4%), and AXIN1 (9.2%). Advanced-stage HCC (BCLC B/C) showed higher frequencies of splicing factor 3b subunit 1 (SF3B1) (P = 0.0003), TP53 (P = 0.0006), and RB Transcriptional Corepressor 1 mutations (P = 0.03). G1-G6 transcriptomic classification and the molecular prognostic 5-gene score showed different distributions according to the stage of the disease and the type of treatment with an enrichment of G3 (P < 0.0001), poor prognostic score (P < 0.0001), and increased proliferation and dedifferentiation at the transcriptomic level in advanced HCC. The 5-gene score predicted survival in patients treated by resection (P < 0.0001) and ablation (P = 0.01) and in advanced HCC (P = 0.04). Twenty-two percent of advanced HCC harbored potentially druggable genetic alterations, and MET amplification was associated with complete tumor response in patients with advanced HCC treated by a specific MET inhibitor. Conclusion: Genomic analysis across the different stages of HCC revealed the mechanisms of tumor progression and helped to identify biomarkers of response to targeted therapies., (© 2019 by the American Association for the Study of Liver Diseases.)

We recently identified a histological subtype of hepatocellular carcinoma (HCC), designated as "macrotrabecular-massive" (MTM-HCC) and associated with specific molecular features. In order to assess the clinical relevance of this variant, we investigated its prognostic value in two large series of patients with HCC treated by either surgical resection or radiofrequency ablation (RFA). We retrospectively included 237 HCC surgical samples and 284 HCC liver biopsies from patients treated by surgical resection and RFA, respectively. Histological slides were reviewed by pathologists specialized in liver disease, and the MTM-HCC subtype was defined by the presence of a predominant (>50%) macrotrabecular architecture (more than six cells thick). The main clinical and biological features were recorded at baseline. Clinical endpoints were early and overall recurrence. The MTM-HCC subtype was identified in 12% of the whole cohort (16% of surgically resected samples, 8.5% of liver biopsy samples). It was associated at baseline with known poor prognostic factors (tumor size, alpha-fetoprotein level, satellite nodules, and vascular invasion). Multivariate analysis showed that MTM-HCC subtype was an independent predictor of early and overall recurrence (surgical series: hazard ratio, 3.03; 95% confidence interval, 1.38-6.65; P = 0.006; and 2.76; 1.63-4.67; P < 0.001; RFA series: 2.37; 1.36-4.13; P = 0.002; and 2.19; 1.35-3.54; P = 0.001, respectively). Its prognostic value was retained even after patient stratification according to common clinical, biological, and pathological features of aggressiveness. No other baseline parameter was independently associated with recurrence in the RFA series., Conclusion: The MTM-HCC subtype, reliably observed in 12% of patients eligible for curative treatment, represents an aggressive form of HCC that may require more specific therapeutic strategies. (Hepatology 2018;68:103-112)., (© 2017 by the American Association for the Study of Liver Diseases.)

The prognosis of hepatocellular carcinoma (HCC) remains poor, with only one third of patients eligible for curative treatments and very limited survival benefits with the use of sorafenib, the current standard of care for advanced disease. Recently, agents targeting the programmed death ligand 1 (PD-L1)/programmed death receptor 1 (PD-1) immune checkpoint were shown to display impressive antitumor activity in various solid or hematological malignancies, including HCC. PD-L1 immunohistochemical expression is thought to represent a biomarker predictive of drug sensitivity. Here, we investigated PD-L1 expression in a series of 217 HCCs and correlated our results with clinical and histological features and immunohistochemical markers (PD-1, cytokeratin 19, glutamine synthetase, and β-catenin expression). PD-L1 expression by neoplastic cells was significantly associated with common markers of tumor aggressiveness (high serum alpha-fetoprotein levels, P = 0.038; satellite nodules, P < 0.001; macrovascular invasion, P < 0.001; microvascular invasion, P < 0.001; poor differentiation, P < 0.001) and with the progenitor subtype of HCC (cytokeratin 19 expression, P = 0.031). High PD-L1 expression by inflammatory cells from the tumor microenvironment also correlated with high serum alpha-fetoprotein levels (P < 0.001), macrovascular invasion (P = 0.001), poor differentiation (P = 0.001), high PD-1 expression (P < 0.001), and the so-called lymphoepithelioma-like histological subtype of HCC (P = 0.003)., Conclusion: PD-L1 expression by either neoplastic or intratumoral inflammatory cells is related to tumor aggressiveness and suggests that the response to treatments targeting the PD-L1/PD-1 immune checkpoint could be restricted to particular HCC variants; thus, enrichment of these tumor subtypes in future clinical trials should be considered. (Hepatology 2016;64:2038-2046)., (© 2016 by the American Association for the Study of Liver Diseases.)

Unlabelled: To evaluate whether hepatitis B virus (HBV) preS/S gene variability has any impact on serum hepatitis B surface antigen (HBsAg) levels and to analyze the replication capacity of naturally occurring preS/S variants, sera from 40 untreated patients with HBV-related chronic liver disease (hepatitis B e antigen [HBeAg]-positive, n = 11; HBeAg-negative, n = 29) were virologically characterized. Additionally, phenotypic analysis of three different preS/S variant isolates (carrying a 183-nucleotide deletion within the preS1 region, the deletion of preS2 start codon, and a stop signal at codon 182 within the S gene, respectively) was performed. HBV infecting 14 (35%) patients had single or multiple preS/S genomic mutations (i.e., preS1 and/or preS2 deletions, preS2 start codon mutations, C-terminally truncated and/or "a" determinant mutated S protein). Presence of preS/S variants negatively correlated with HBsAg titers (r = -0.431; P = 0.005) and its prevalence did not significantly differ between HBeAg-positive and HBeAg-negative patients. No correlation was found between HBsAg and HBV DNA levels in patients infected with preS/S mutants, whereas a significant correlation was found between HBsAg and viremia levels (r = 0.607; P = 0.001) in patients infected with wild-type HBV strains. HepG2 cells replicating the above-mentioned three preS/S variants showed significant reduction of HBsAg secretion, retention of envelope proteins in the endoplasmic reticulum, less efficient virion secretion and nuclear accumulation of significantly higher amounts of covalently closed circular DNA compared with wild-type HBV replicating cells., Conclusion: In patients infected with preS/S variants, HBV DNA replication and HBsAg synthesis/secretion appear to be dissociated. Therefore, the use of HBsAg titer as diagnostic/prognostic tool has to take into account the frequent emergence of preS/S variants in chronic HBV infection., (Copyright © 2012 American Association for the Study of Liver Diseases.)