A multicenter clinical AI system study for detection and diagnosis of focal liver lesions.

Early and accurate diagnosis of focal liver lesions is crucial for effective treatment and prognosis. We developed and validated a fully automated diagnostic system named Liver Artificial Intelligence Diagnosis System (LiAIDS) based on a diverse sample of 12,610 patients from 18 hospitals, both retrospectively and prospectively. In this study, LiAIDS achieved an F1-score of 0.940 for benign and 0.692 for malignant lesions, outperforming junior radiologists (benign: 0.830-0.890, malignant: 0.230-0.360) and being on par with senior radiologists (benign: 0.920-0.950, malignant: 0.550-0.650). Furthermore, with the assistance of LiAIDS, the diagnostic accuracy of all radiologists improved. For benign and malignant lesions, junior radiologists' F1-scores improved to 0.936-0.946 and 0.667-0.680 respectively, while seniors improved to 0.950-0.961 and 0.679-0.753. Additionally, in a triage study of 13,192 consecutive patients, LiAIDS automatically classified 76.46% of patients as low risk with a high NPV of 99.0%. The evidence suggests that LiAIDS can serve as a routine diagnostic tool and enhance the diagnostic capabilities of radiologists for liver lesions.

A deep learning model for prediction of post hepatectomy liver failure after hemihepatectomy using preoperative contrast-enhanced computed tomography: a retrospective study.

Objective: Post-hepatectomy liver failure (PHLF) remains clinical challenges after major hepatectomy. The aim of this study was to establish and validate a deep learning model to predict PHLF after hemihepatectomy using preoperative contrast-enhancedcomputed tomography with three phases (Non-contrast, arterial phase and venous phase). Methods: 265 patients undergoing hemihepatectomy in Sir Run Run Shaw Hospital were enrolled in this study. The primary endpoint was PHLF, according to the International Study Group of Liver Surgery's definition. In this study, to evaluate the proposed method, 5-fold cross-validation technique was used. The dataset was split into 5 folds of equal size, and each fold was used as a test set once, while the other folds were temporarily combined to form a training set. Performance metrics on the test set were then calculated and stored. At the end of the 5-fold cross-validation run, the accuracy, precision, sensitivity and specificity for predicting PHLF with the deep learning model and the area under receiver operating characteristic curve (AUC) were calculated. Results: Of the 265 patients, 170 patients with left liver resection and 95 patients with right liver resection. The diagnosis had 6 types: hepatocellular carcinoma, intrahepatic cholangiocarcinoma, liver metastases, benign tumor, hepatolithiasis, and other liver diseases. Laparoscopic liver resection was performed in 187 patients. The accuracy of prediction was 84.15%. The AUC was 0.7927. In 170 left hemihepatectomy cases, the accuracy was 89.41% (152/170), and the AUC was 82.72%. The accuracy was 77.47% (141/182) with liver mass, 78.33% (47/60) with liver cirrhosis and 80.46% (70/87) with viral hepatitis. Conclusion: The deep learning model showed excellent performance in prediction of PHLF and could be useful for identifying high-risk patients to modify the treatment planning.

Safety and feasibility of laparoscopic liver resection for intrahepatic cholangiocarcinoma: a propensity score-matched study.

BACKGROUND: Laparoscopic liver resection (LLR) is controversial in treating intrahepatic cholangiocarcinoma (ICC). Therefore, this study aimed to evaluate the safety and feasibility of LLR for the treatment of ICC and explored the independent factors affecting the long-term prognosis of ICC. METHODS: We included 170 patients undergoing hepatectomy for ICC from December 2010 to December 2021 and divided them into LLR group and open liver resection (OLR) group. We used propensity score matching (PSM) analysis to reduce the impact of data bias and confounding variables and then compared the short-term and long-term prognosis of LLR and OLR in treating ICC; Cox proportional hazards regression model was adopted to explore the independent factors affecting the long-term prognosis of ICC. RESULTS: A total of 105 patients (70 in the LLR group and 35 in the OLR group) were included after 2:1 PSM analysis. There was no difference in demographic characteristics and preoperative indexes between the two groups. The perioperative results of the OLR group were worse than those of the LLR group, that is, the intraoperative blood transfusion rate (24 (68.6) vs 21 (30.0)), blood loss (500 (200-1500) vs 200 (100-525)), and the morbidity of major postoperative complications (9 (25.7) vs 6 (8.5)) in the OLR group were worse than those in LLR group. LLR could enable patients to obtain an equivalent long-term prognosis compared to OLR. The Cox proportional hazards regression model exhibited that no matter before or after PSM, preoperative serum CA12-5 and postoperative hospital stay were independent factors affecting overall survival, while only lymph node metastasis independently influenced recurrence-free survival. CONCLUSIONS: Compared with ICC treated by OLR, the LLR group obtained superior perioperative period outcomes. In the long run, LLR could enable ICC patients to receive an equivalent long-term prognosis compared to OLR. In addition, ICC patients with preoperative abnormal CA12-5, lymph node metastasis, and more extended postoperative hospital stay might suffer from a worse long-term prognosis. However, these conclusions still need multicenter extensive sample prospective research to demonstrate.

A Knowledge-Guided Framework for Fine-Grained Classification of Liver Lesions Based on Multi-Phase CT Images.

Automatic and accurate differentiation of liver lesions from multi-phase computed tomography imaging is critical for the early detection of liver cancer. Multi-phase data can provide more diagnostic information than single-phase data, and the effective use of multi-phase data can significantly improve diagnostic accuracy. Current fusion methods usually fuse multi-phase information at the image level or feature level, ignoring the specificity of each modality, therefore, the information integration capacity is always limited. In this paper, we propose a Knowledge-guided framework, named MCCNet, which adaptively integrates multi-phase liver lesion information from three different stages to fully utilize and fuse multi-phase liver information. Specifically, 1) a multi-phase self-attention module was designed to adaptively combine and integrate complementary information from three phases using multi-level phase features; 2) a cross-feature interaction module was proposed to further integrate multi-phase fine-grained features from a global perspective; 3) a cross-lesion correlation module was proposed for the first time to imitate the clinical diagnosis process by exploiting inter-lesion correlation in the same patient. By integrating the above three modules into a 3D backbone, we constructed a lesion classification network. The proposed lesion classification network was validated on an in-house dataset containing 3,683 lesions from 2,333 patients in 9 hospitals. Extensive experimental results and evaluations on real-world clinical applications demonstrate the effectiveness of the proposed modules in exploiting and fusing multi-phase information.

Mitochondrial dysfunction attenuates rapid regeneration in livers with toxin-induced fibrosis.

BACKGROUND: The mechanism of associating liver partition and portal vein ligation for staged hepatectomy (ALPPS)-induced rapid liver regeneration remains poorly documented, especially in patients with fibrosis. Therefore, this study aims to investigate the underlying mechanism of ALPPS-induced accelerated regeneration in toxin-induced fibrosis models. METHODS: The ALPPS-induced regeneration model was established in livers with thioacetamide (TAA)-induced fibrosis to determine the regenerative pathways involved in rapid regeneration. Confirmatory experiments were performed in transforming growth factor beta 1 (TGFß1)-treated AML12 cells and mice with carbon tetrachloride (CCl4)-induced fibrosis. Finally, mitochondrial dysfunction was validated in fibrotic/non-fibrotic patients. RESULTS: In TAA-induced fibrotic mice, ALPPS-induced regeneration was significantly inferior to that of the control group (P=0.027 at day 2 and P<0.001 at day 7). Furthermore, mitochondria-associated genes were significantly downregulated in TAA-challenged mice. Accordingly, the reduced production of ATP and elevated levels of malondialdehyde indicated disturbances in intracellular energy metabolism during the ALPPS-induced regenerative process after TAA treatment. Further investigations were performed in TGF-ß1-treated AML12 cells and CCl4-treated mice, which indicated that mitochondrial dysfunction attenuated the capacity for rapid regeneration after ALPPS. CONCLUSIONS: In summary, this study revealed that mitochondrial dysfunction led to inferior regeneration in livers with toxin-induced fibrosis and identified new therapeutic targets to improve the feasibility and safety of the ALPPS procedure. Further studies in human patients are required in the future.

Enhanced recovery after surgery (ERAS) program in elderly patients undergoing laparoscopic hepatectomy: a retrospective cohort study.

BACKGROUND: Enhanced recovery after surgery (ERAS) has shown sufficient superiority in terms of cutting down hospital stay and costs, and reducing complications in patients undergoing laparoscopic hepatectomy (LH). However, the benefit of ERAS in elderly patients undergoing LH remains unclear, and clinical studies on this topic are still limited. METHODS: In total, 177 elderly patients (aged over 65 and underwent LH) were divided into two groups. The 107 patients in the control group received standard care, while the 70 patients in the ERAS group underwent the ERAS program after hepatectomy. The primary endpoint was the postoperative hospital stay. The secondary endpoints were resumption of oral intake, readmission rate and complications. RESULTS: ERAS had a positive effect on reducing length of hospital stay {6 [4-8] vs. 9 [7-14] days; P<0.001}. Although there was no significant reduction of overall complications in the ERAS group compared with the control group (0.500 vs. 0.626; P=0.097), the Clavien-Dindo classification of compliances in the ERAS group was lower among the patients with complications (Grade I: 0.829 vs. 0.597; P=0.018; Grade II: 0.143 vs. 0.328; P=0.044), which indicated that the patients in the control group might experience more severe complications. The readmission rates remained unaffected between the two groups. CONCLUSIONS: Consistent with younger patients, ERAS program is considered to be effective and safe, which can distinctly promote recovery after hepatectomy for elderly patients accepting LH.

eIF4EBP3 was downregulated by methylation and acted as a tumor suppressor by targeting eIF4E/ß-catenin in gastric cancer.

BACKGROUND: Epigenetic aberrations of tumor suppressor genes (TSGs), particularly DNA methylation, are frequently involved in the pathogenesis of gastric cancer (GC). Through a methylome study, we identified eIF4EBP3 as a methylated gene in GC. However, the role of eIF4EBP3 in GC progression has not been explored. METHODS: The expression and promoter region methylation of eIF4EBP3 in GC and healthy tissues were analyzed in public datasets. eIF4EBP3 expression in GC was detected by semi-quantitative RT-PCR, western blot and immunohistochemistry. We also studied epigenetic alterations and functions in GC. The effects of eIF4EBP3 on cell proliferation, migration and invasion were conducted by functional experiments in vitro and in vivo. Label-free proteomic analysis was applied to identify targets of eIF4EBP3. RESULTS: The expression level of eIF4EBP3 was downregulated in gastric cancer due to promoter region methylation, and was associated with poor survival and tumor progression. Ectopic expression of eIF4EBP3 significantly inhibited tumor cell growth, migration and invasion both in vitro and in vivo. Label-free proteomic analysis indicated eIF4EBP3 downregulated the protein level of ß-catenin, which was confirmed by western blot. Overexpression of ß-catenin reversed the inhibitory effects of eIF4EBP3 on cell growth and migration, indicating that eIF4EBP3 acts on GC cells by targeting the eIF4E/ß-catenin axis. CONCLUSION: These results suggest that eIF4EBP3 is a novel TSG methylated in gastric cancer that may play important roles in GC development and liver metastasis and indicate eIF4EBP3 as a potential metastasis and survival biomarker for GC.

TRIM59 promotes tumor growth in hepatocellular carcinoma and regulates the cell cycle by degradation of protein phosphatase 1B.

Tripartite motif 59 (TRIM59) is a member of Tripartite motif protein family, which is frequently increased in many human cancers. However, the molecular mechanism of TRIM59 in hepatocellular carcinoma (HCC) has not been fully elucidated. In this study, we report that TRIM59 plays an essential role in growth of HCC. We analyzed RNA sequencing data to explore abnormally expressed TRIM59 in HCC. The effects of TRIM59 on HCC were investigated through in vitro and in vivo assays (i.e., CCK-8 assay, colony formation assay, flow cytometry assay, xenograft model, immunohistochemistry, immunofluorescence and western blot). The mechanism of TRIM59 action was explored through co-immunoprecipitation, immunofluorescence, mass spectrometry and bioinformatics. TRIM59 expression is up-regulated in HCC tissues. A high level of TRIM59 expression is correlated with poor overall and disease-free survival of HCC patients. Knockdown of TRIM59 attenuated proliferation, induced cells arrested at G1/S phase and reduced tumor growth in the mouse xenograft model. Ectopic expression of TRIM59 had the opposite results. Mechanistically, TRIM59 promoted growth and regulated cell cycle. Further studies indicated that TRIM59 might interacted physically with PPM1B, which has been reported to negatively regulate CDKs phosphorylation. We also discovered that TRIM59 increased degradation of PPM1B. TRIM59 overexpression in HCC patients correlated with reduced expression of PPM1B and increased CDKs phosphorylation and cell cycle proteins. Our findings demonstrate that TRIM59 promotes growth by PPM1B/CDKs signaling pathway, indicating a new prognostic biomarker candidate and a potential antitumor target for HCC.

An adult gerbil model for evaluating potential coxsackievirus A16 vaccine candidates.

A suitable animal model of CVA16 infection is crucial in order to understand its pathogenesis and to help develop antiviral vaccines or screen therapeutic drugs. The neonatal mouse model has a short sensitivity period to CA16 infection, which is a major limitation. In this study, we demonstrate that adult (60-day-old) gerbils are susceptible to CVA16 infection at high doses (108.0 TCID50). A clinical isolate strain of CVA16 was inoculated intraperitoneally into adult gerbils, which subsequently developed significant clinical symptoms, including hind limb weakness, paralysis of one or both hind limbs, tremors, and eventual death from neurological disorders. Real-time RT-PCR revealed that viral loads in the spinal cord and brainstem were higher than those in other organs/tissues. Histopathological changes, such as neuronal degeneration, neuronal loss, and neuronophagia, were observed in the spinal cord, brainstem, and heart muscle, along with necrotizing myositis. Gerbils receiving both prime and boost immunizations of alum adjuvant inactivated vaccine exhibited no clinical signs of disease or mortality following challenge by CVA16, whereas 80% of control animals showed obvious clinical signs, including slowness, paralysis of one or both hind limbs, and eventual death, suggesting that the CVA16 vaccine can fully protect gerbils against CVA16 challenge. These results demonstrate that an adult gerbil model provides us with a useful tool for studying the pathogenesis and evaluating antiviral reagents of CVA16 infection. The development of this animal model would also be conducive to screening promising CVA16 vaccine candidates as well as further vaccination evaluation.

Maturity of associating liver partition and portal vein ligation for staged hepatectomy-derived liver regeneration in a rat model.

AIM: To establish a rat model for evaluating the maturity of liver regeneration derived from associating liver partition and portal vein ligation for staged hepatectomy (ALPPS). METHODS: In the present study, ALPPS, partial hepatecotmy (PHx), and sham rat models were established initially, which were validated by significant increase of proliferative markers including Ki-67, proliferating cell nuclear antigen, and cyclin D1. In the setting of accelerated proliferation in volume at the second and fifth day after ALPPS, the characteristics of newborn hepatocytes, as well as specific markers of progenitor hepatic cell, were identified. Afterwards, the detection of liver function followed by cluster analysis of functional gene expression were performed to evaluate the maturity. RESULTS: Compared with PHx and sham groups, the proliferation of FLR was significantly higher in ALPPS group (P = 0.023 and 0.001 at second day, P = 0.034 and P < 0.001 at fifth day after stage I). Meanwhile, the increased expression of proliferative markers including Ki-67, proliferating cell nuclear antigen, and cyclin D1 verified the accelerated liver regeneration derived from ALPPS procedure. However, ALPPS-induced Sox9 positive hepatocytes significantly increased beyond the portal triad, which indicated the progenitor hepatic cell was potentially involved. And the characteristics of ALPPS-induced hepatocytes indicated the lower expression of hepatocyte nuclear factor 4 and anti-tryptase in early proliferative stage. Both suggested the immaturity of ALPPS-derived liver regeneration. Additionally, the detection of liver function and functional genes expression confirmed the immaturity of renascent hepatocytes derived in early stage of ALPPS-derived liver regeneration. CONCLUSION: Our study revealed the immaturity of ALPPS-derived proliferation in early regenerative response, which indicated that the volumetric assessment overestimated the functional proliferation. This could be convincing evidence that the stage II of ALPPS should be performed prudently in patients with marginally adequate FLR, as the ALPPS-derived proliferation in volume lags behind the functional regeneration.

Overexpression of Zwint predicts poor prognosis and promotes the proliferation of hepatocellular carcinoma by regulating cell-cycle-related proteins.

INTRODUCTION: Zwint, a centromere-complex component required for the mitotic spindle checkpoint, has been reported to be overexpressed in different human cancers, but it has not been studied in human hepatocellular carcinoma (HCC). MATERIALS AND METHODS: The role of Zwint in hepatocellular carcinoma cell proliferation capacities was evaluated by using cell counting kit-8 (CCK8), flow cytometry, clone formation and tumor formation assay in nude mice. Western blot analysis and qPCR assay were performed to assess Zwint interacting with cell-cycle-related proteins. RESULTS: We report that ZWINT mRNA and protein expression were upregulated in HCC samples and cell lines. An independent set of 106 HCC-tissue pairs and corresponding noncancerous tissues was evaluated for Zwint expression using immunohistochemistry, and elevated Zwint expression in HCC tissues was significantly correlated with clinicopathological features, such as tumor size and number. Kaplan-Meier survival and Cox regression analysis revealed that high expression of Zwint was correlated with poor overall survival and a greater tendency for tumor recurrence. Ectopic expression of Zwint promoted HCC-cell proliferation, and Zwint expression affected the expression of several cell-cycle proteins, including PCNA, cyclin B1, Cdc25C and CDK1. CONCLUSION: Our findings suggest that upregulation of Zwint may contribute to the progression of HCC and may be a prognostic biomarker and potential therapeutic target for treating HCC.

Enhanced recovery care versus traditional care after laparoscopic liver resections: a randomized controlled trial.

BACKGROUND: Enhanced recovery after surgery (ERAS), with several evidence-based elements, has been shown to shorten length of hospital stay and reduce perioperative hospital costs in many operations. This randomized clinical trial was performed to compare complications and hospital stay of laparoscopic liver resection between ERAS and traditional care. METHODS: A randomized controlled trial was performed for laparoscopic liver resection from August 2015 to August 2016. Patients were randomly divided into ERAS group and traditional care group. The primary outcome was length of hospital stay (LOS) after surgery. Second outcomes included postoperative complications, hospital cost, and 30-day readmissions. Elements used in ERAS group included more perioperative education, nurse navigators, nutrition support for liver diseases, respiratory therapy, oral carbohydrate 2 h before operation, early mobilization and oral intake, goal-directed fluid therapy, less drainages, postoperative nausea and vomiting (PONV) prophylaxis and multimodal analgesia. RESULTS: The study included 58 (two conversion to laparotomy) patients in ERAS group and 61 (three conversion to laparotomy) patients in the traditional care group. Postoperative LOS was significantly shorter in the ERAS group than traditional care group (5 vs. 8 days; p < 0.001). ERAS program significantly reduced the hospital costs (CNY 45413.1 vs. 55794.1; p = 0.006) and complications (36.2 vs. 55.7%; p = 0.033). Duration till first flatus and PONV were significantly reduced in ERAS group. Pain control was better in ERAS (Visual analogue scale (VAS) POD1 (≥ 4) 19.0 vs. 39.3%, p = 0.017; VAS POD1 2.5 vs. 3.1, p = 0.010). There was no difference in the rate of 30-day readmissions (6.9 vs. 8.2%; p = 1.000). CONCLUSION: ERAS protocol is feasible and safe for laparoscopic liver resection. Patients in ERAS group have less pain and complications.

Enhanced Recovery Program Versus Traditional Care in Laparoscopic Hepatectomy.

Enhanced recovery after surgery (ERAS) has shown effectiveness in terms of reducing the hospital stay and cost associated with open liver resection. However, the benefit of ERAS in patients undergoing laparoscopic liver resection is still unclear, and clinical studies on this topic are limited.The ERAS program for laparoscopic liver resection was used in a group of 80 patients (ERAS group). The results were compared with those in a control group of 107 patients. All patients underwent laparoscopic liver resection. The primary endpoints were the postoperative hospital stay, defined as the number of days from surgery to discharge, and the hospitalization expense. The secondary endpoints were resumption of oral intake, readmissions, and complications.The median postoperative hospital stay was 6.2 ±â€Š2.6 days in the ERAS group, which was significantly shorter than that in the control group (9.9 ±â€Š5.9 d; P < 0.001). The hospitalization cost was $6871 ±â€Š2571 in the ERAS group and $7948 ±â€Š3630 in the control group (P = 0.020). The morbidity rate was 22.5% (18 of 80 patients) in the ERAS group and 43.9% (47 of 107 patients) in the control group (P = 0.002). There were no significant differences the in rate of readmission between the 2 groups.Enhanced recovery after surgery for laparoscopic liver resection is safe and effective. Patients in the ERPS group had a shorter hospital stay, fewer complications, and lower hospital costs.

Blunt Dissection: A Solution to Prevent Bile Duct Injury in Laparoscopic Cholecystectomy.

BACKGROUND: Laparoscopic cholecystectomy (LC) has been a standard operation and replaced the open cholecystectomy (OC) rapidly because the technique resulted in less pain, smaller incision, and faster recovery. This study was to evaluate the value of blunt dissection in preventing bile duct injury (BDI) in laparoscopic cholecystectomy (LC). METHODS: From 2003 to 2015, LC was performed on 21,497 patients, 7470 males and 14,027 females, age 50.3 years (14-84 years). The Calot's triangle was bluntly dissected and each duct in Calot's triangle was identified before transecting the cystic duct. RESULTS: Two hundred and thirty-nine patients (1.1%) were converted to open procedures. The postoperative hospital stay was 2.1 (0-158) days, and cases (46%) had hospitalization days of 1 day or less, and 92.8% had hospitalization days of 3 days or less; BDI was occurred in 20 cases (0.09%) including 6 cases of common BDI, 2 cases of common hepatic duct injury, 1 case of right hepatic duct injury, 1 case of accessory right hepatic duct, 1 case of aberrant BDI 1 case of biliary stricture, 1 case of biliary duct perforation, 3 cases of hemobilia, and 4 cases of bile leakage. CONCLUSION: Exposing Calot's triangle by blunt dissection in laparoscopic cholecystectomy could prevent intraoperative BDI.

Potential molecular, cellular and microenvironmental mechanism of sorafenib resistance in hepatocellular carcinoma.

Sorafenib, an orally-available kinase inhibitor, is the only standard clinical treatment against advanced hepatocellular carcinoma. However, development of resistance to sorafenib has raised concern in recent years due to the high-level heterogeneity of individual response to sorafenib treatment. The resistance mechanism underlying the impaired sensitivity to sorafenib is still elusive though some researchers have made great efforts. Here, we provide a systemic insight into the potential molecular, cellular and microenvironmental mechanism of sorafenib resistance in hepatocellular carcinoma depending on abundant previous studies and reports.

Liver cirrhosis grading Child-Pugh class B: a Goliath to challenge in laparoscopic liver resection?-prior experience and matched comparisons.

BACKGROUND: Laparoscopic hepatectomy (LH) is highly difficult in the background of liver cirrhosis. In this case series, we aimed to summarize our prior experience of LH in liver cirrhosis grading Child-Pugh class B. METHODS: In the LH database of Sir Run Run Shaw Hospital in Zhejiang, China, patients who were pathologically diagnosed with cirrhosis and graded as Child-Pugh class B or C were reviewed. RESULTS: Five patients grading Child B were included. There was no Child C case in our LH database. For included cases, median blood loss (BL) was 800 (range, 240-1,000) mL, median operative time was 135 (range, 80-170) minutes, and median length of hospital stay was 9 (range, 7-15) days. Forty percent (2/5) of patients was converted to open. The postoperative complication (PC) rate was 20.0% (1/5). When these Child B cases were compared with Child A cases undergoing LH, there was no statistical significance in BL, complication rate, operative time, open rate and hospital stay (HS) (P>0.05). This finding was confirmed by two ways of matched comparisons (a 1:2 comparison based on age and gender, and a 1:1 propensity score matching). CONCLUSIONS: Although relevant literatures had suggested feasibility of LH in cirrhotic cases grading Child A, this study was the first one to discuss the value of LH in Child B cases. Our prior experience showed that in selected patients, LH in Child B patients had the potential to be as safe as in Child A cases. The efficacy of LH in Child C patients needs further exploration.