Machine learning-based clinical decision support system for treatment recommendation and overall survival prediction of hepatocellular carcinoma: a multi-center study.

The treatment decisions for patients with hepatocellular carcinoma are determined by a wide range of factors, and there is a significant difference between the recommendations of widely used staging systems and the actual initial treatment choices. Herein, we propose a machine learning-based clinical decision support system suitable for use in multi-center settings. We collected data from nine institutions in South Korea for training and validation datasets. The internal and external datasets included 935 and 1750 patients, respectively. We developed a model with 20 clinical variables consisting of two stages: the first stage which recommends initial treatment using an ensemble voting machine, and the second stage, which predicts post-treatment survival using a random survival forest algorithm. We derived the first and second treatment options from the results with the highest and the second-highest probabilities given by the ensemble model and predicted their post-treatment survival. When only the first treatment option was accepted, the mean accuracy of treatment recommendation in the internal and external datasets was 67.27% and 55.34%, respectively. The accuracy increased to 87.27% and 86.06%, respectively, when the second option was included as the correct answer. Harrell's C index, integrated time-dependent AUC curve, and integrated Brier score of survival prediction in the internal and external datasets were 0.8381 and 0.7767, 91.89 and 86.48, 0.12, and 0.14, respectively. The proposed system can assist physicians by providing data-driven predictions for reference from other larger institutions or other physicians within the same institution when making treatment decisions.

Effectiveness of Lenvatinib Versus Sorafenib for Unresectable Hepatocellular Carcinoma in Patients with Hepatic Decompensation.

BACKGROUND/AIM: Lenvatinib and sorafenib are currently available to treat patients with advanced hepatocellular carcinoma (HCC). However, since the clinical trials evaluating the efficacy of lenvatinib and sorafenib included only patients with Child-Pugh class A, little is known about the effectiveness of the treatments in patients with hepatic decompensation. We compared the effectiveness of lenvatinib and sorafenib in decompensated patients with unresectable HCC. METHODS: Consecutive patients who were classified as Child-Pugh class B or C and received lenvatinib or sorafenib as first-line systemic therapy for unresectable HCC between November 2018 and April 2020 at a tertiary referral center were included in this retrospective study. The primary outcome was overall survival (OS), and the secondary outcomes were progression-free survival (PFS), time-to-progression, best overall tumor response, and safety profiles. RESULTS: Among 94 patients, 34 received lenvatinib and 60 received sorafenib. The median OS was 4.1 months (95% confidence interval [CI], 2.9-5.2): 4.2 months (95% CI, 2.9-5.3) for lenvatinib and 4.1 months (95% CI, 2.7-6.4) for sorafenib. The treatment regimen was not associated with significant improvement in OS after adjusting for covariables (adjusted hazard ratio [aHR], 0.92; 95% CI, 0.54-1.54; P = 0.74). The treatment regimen was not an independent predictor of PFS (lenvatinib vs. sorafenib; aHR, 0.77; 95% CI, 0.48-1.24; P = 0.28). HRs were maintained even after balancing with the inverse probability treatment weighting method. Objective response rates were 11.8% and 6.7% in patients receiving lenvatinib and sorafenib, respectively (P = 0.45). Ten patients in both groups (five in the lenvatinib group and five in the sorafenib group) underwent dose modification due to adverse events, and significant difference was not observed between the treatment groups (P = 0.49). CONCLUSION: The effectiveness of lenvatinib and sorafenib was comparable for the treatment of unresectable HCC in decompensated patients.

Yttrium-90 Radioembolization Is Associated with Better Clinical Outcomes in Patients with Hepatocellular Carcinoma Compared with Conventional Chemoembolization: A Propensity Score-Matched Study.

BACKGROUND: Both trans-arterial radioembolization (TARE) and conventional trans-arterial chemoembolization (TACE) can effectively control hepatocellular carcinoma (HCC) in patients who are not suitable for curative resection. This study compared the effectiveness of TARE and conventional TACE as the initial trans-arterial treatment for hepatocellular carcinoma (HCC) assessed by tumor response and clinical outcomes. MATERIAL AND METHODS: Data were retrospectively analyzed the propensity score-matched cohort for overall survival (OS), progression-free survival (PFS), and intrahepatic PFS in patients who have received TARE or TACE as the first HCC treatment from March 2012 to December 2017. RESULTS: A total of 138 patients initially treated with TARE (n = 54) or TACE (n = 84) was included in this study. Of 138 patients, median age was 59 years and the mean follow-up period was 27.6 months. TARE showed better OS (hazard ratio [HR] = 0.54, 95% confidence interval [CI] = 0.31-0.92, log-rank P = 0.02), better PFS (HR = 0.51, 95% CI = 0.36-0.97, log-rank P = 0.04), and better intrahepatic PFS (HR = 0.51, 95% CI = 0.30-0.88, log-rank P = 0.01) compared with TACE. TARE was an independent prognostic factor for OS (adjusted HR [aHR] = 0.52, 95% CI = 0.30-0.90, P = 0.02), PFS (aHR = 0.57, 95% CI = 0.35-0.94, P = 0.03), and intrahepatic PFS (aHR = 0.49, 95% CI = 0.28-0.84, P = 0.01). CONCLUSION: TARE as initial trans-arterial treatment is associated with better clinical outcomes such as longer OS compared with TACE in patients with HCC.

Comparison of Overall Survival between Surgical Resection and Radiofrequency Ablation for Hepatitis B-Related Hepatocellular Carcinoma.

It remains controversial whether surgical resection, compared to radiofrequency ablation (RFA), improves overall survival (OS) in patients with early hepatocellular carcinoma (HCC). This study aimed to compare OS after RFA with that after resection for HCC. This retrospective study included patients who underwent RFA or surgical resection as initial treatment for hepatitis B virus (HBV)-related HCC at a very early or early stage. A total of 761 patients (RFA, n = 194; resection, n = 567) from Seoul National University Hospital (Seoul, South Korea) and 1277 patients (RFA, n = 352; resection, n = 925) from the Korean Primary Liver Cancer Registry were included in the hospital and nationwide cohorts, respectively. Primary and secondary endpoints were OS and recurrence-free survival (RFS), respectively. Additional analysis was performed when the history of the antiviral treatment and the type of prescribed nucleos(t)ide analogue were confirmed. The rate of complications was compared between the two treatment groups in the hospital cohort. Baseline characteristics were balanced, using inverse probability of treatment weighting (IPTW). In the hospital cohort, the RFA group had a smaller mean tumor size (1.7 vs. 3.9 cm) but a higher proportion of cirrhotic patients than the resection group (85.6% vs. 63.1%) (both p < 0.01). During 81.0 (interquartile range, 62.3-107.1) months of follow-up, there was no difference in OS (adjusted hazard ratio (aHR) = 0.870, 95% confidence interval (CI) = 0.400-1.897, p = 0.73) and RFA was associated with shorter RFS (aHR = 1.562, 95% CI = 1.099-2.219, p = 0.01) after employing IPTW. Antiviral treatment was independently associated with longer OS (aHR = 0.444, 95% CI = 0.251-0.786, p = 0.01) as well as RFS (aHR = 0.544, 95% CI = 0.391-0.757, p < 0.01) in the hospital cohort. In the nationwide cohort, there was no difference in OS (aHR = 0.981, 95% CI = 0.661-1.456, p = 0.92) between the two treatment groups when adjusted for antiviral treatment, which was a negative independent risk factor for mortality (aHR = 0.655, 95% CI = 0.451-0.952, p = 0.03) after IPTW. Among patients treated with tenofovir (n = 96) or entecavir (n = 184) in the hospital cohort, there was no difference in either OS (aHR = 0.522, 95% CI = 0.058-4.724, p = 0.56) or RFS (aHR = 1.116, 95% CI = 0.738-1.688, p = 0.60). The overall incidence of complications was higher in the resection group (26.3%) than in the RFA group (13.9%) (p < 0.01). RFA may provide comparable OS to resection in the treatment of very early or early HCC with a lower rate of complications, although RFS is marginally shorter than in the resection group after adjusting for antiviral treatment. Regardless of the type of NA, antiviral treatment in patients with HBV-related HCC is strongly associated with both OS and RFS.

Maintenance of skeletal muscle mass during FOLFIRINOX is a favorable prognostic factor in pancreatic cancer patients.

OBJECTIVE: The study aimed to investigate the effect of body composition changes during chemotherapy on clinical outcomes in patients with pancreatic cancer. RESULTS: In patients with locally advanced pancreatic cancer (LAPC), the cross-sectional area of skeletal muscle (SM) and adipose tissue (AT) at the level of third lumbar vertebra was measured. The SM and AT ratios indicated the changes during chemotherapy. The patients were classified into three groups based on these ratios: group 1, ≥ 1.00; group 2, 0.85-0.99; group 3, < 0.85. The overall survival (OS) and surgical resection rates were estimated. Fifty-eight patients with LAPC who received first-line FOLFIRINOX were analyzed. Fifteen (25.9%) patients who underwent resection showed maintained BMI, SM, and AT as compared to the patients who did not undergo resection. As the SM ratio decreased, the risk for death increased significantly. Further, the resection rate was significantly higher in patients with maintained SM compared to those with low SM ratio. On the contrary, the change in AT ratio was not associated with OS and resection rate; however, significant decrease in AT more than 15% showed poor clinical outcomes. Maintenance of SM during chemotherapy is a reliable prognostic factor indicating longer OS and higher resection rate.

Effectiveness of Transarterial Chemoembolization-First Treatment for Advanced Hepatocellular Carcinoma: A Propensity Score Matching Analysis.

BACKGROUND: Still in real-world practice, advanced hepatocellular carcinoma (HCC) patients are treated with transarterial chemoembolization (TACE). This study compared the therapeutic effectiveness of initial TACE treatment and initial sorafenib treatment in advanced HCC patients. PATIENT AND METHODS: Advanced HCC patients initially treated with sorafenib or TACE were included in this study. Treatment crossover due to an unfavorable response to initial treatment was allowed. Propensity score (PS) matching was applied for balancing baseline characteristics. The primary outcome was overall survival (OS) and the secondary outcomes included tumor response. RESULTS: A total of 554 patients were included in this study: 85 were initially treated with sorafenib (the sorafenib-first group) and 469 with TACE (the TACE-first group). In the entire cohort, the TACE-first group was associated with lower risk of death [adjusted hazard ratio (HR)=0.75, P=0.04]. In the PS-matched cohort (85 patients per group), the TACE-first group showed longer OS than the sorafenib-first group in both univariable (HR=0.68, P=0.02) and multivariable analyses (adjusted HR=0.58, P=0.002). Specifically, within both the entire and the PS-matched cohorts, the TACE-first group showed longer OS in subgroups with major portal vein tumor thrombosis (HR=0.72, P=0.048; HR=0.52, P=0.003) or infiltrative HCC (HR=0.42, P<0.001; HR=0.30, P=0.004, respectively). The objective response rate was higher in the TACE-first group (29.3% vs 14.7%, P=0.03) within the PS-matched cohort. CONCLUSION: For advanced HCC, initial TACE leads to longer OS with a more favorable tumor response than initial sorafenib treatment. Intrahepatic tumor control with initial locoregional therapy may be a potent strategy for advanced HCC.

Risk and Risk Score Performance of Hepatocellular Carcinoma Development in Patients With Hepatitis B Surface Antigen Seroclearance.

INTRODUCTION: Hepatocellular carcinoma (HCC) can develop among chronic hepatitis B patients after hepatitis B surface antigen (HBsAg) seroclearance. However, whether HCC risk after HBsAg seroclearance differs between antiviral therapy (AVT)-induced or spontaneous seroclearance cases and ways to identify at-risk populations remain unclear. METHODS: A retrospective cohort of 1,200 adult chronic hepatitis B patients who achieved HBsAg seroclearance (median age: 56 years; 824 men; 165 with cirrhosis; 216 AVT-induced cases) were analyzed. The risk of HCC after HBsAg seroclearance and the performance of 6 HCC prediction models were assessed. RESULTS: During a median of 4.8 years of follow-up (range: 0.5-17.8 years), HCC developed in 23 patients (1.9%). The HCC incidence rate was higher in the AVT-induced cases than that in the spontaneous cases (3.9% vs 0.9% at 5 years). AVT and cirrhosis were independent factors associated with HCC, with HCC incidence rates of 0.5%, 1.2%, 4.0%, and 10.5% at 5 years for spontaneous/no-cirrhosis, AVT-induced/no-cirrhosis, spontaneous/cirrhosis, and AVT-induced/cirrhosis patients, respectively. Among the 6 predictive HCC models tested, Chinese University-HCC score (0.82) showed the highest C-statistics, which was followed by guide with age, gender, HBV DNA, core promoter mutations and cirrhosis (0.81). DISCUSSION: AVT-induced HBsAg seroclearance was associated with higher HCC risk, especially for patients with cirrhosis, indicating that they need careful monitoring for HCC risk. The HCC risk models were able to stratify the HCC risk in patients with HBsAg seroclearance.

Empirical Treatment With Carbapenem vs Third-generation Cephalosporin for Treatment of Spontaneous Bacterial Peritonitis.

BACKGROUND & AIMS: Third-generation cephalosporins (TGCs) are recommended as first-line antibiotics for treatment of spontaneous bacterial peritonitis (SBP). However, antibiotics against multidrug-resistant organisms (such as carbapenems) might be necessary. We aimed to evaluate whether carbapenems are superior to TGC for treatment of SBP. METHODS: We performed a retrospective study of 865 consecutive patients with a first presentation of SBP (275 culture positive; 103 with TGC-resistant bacterial infections) treated at 7 referral centers in Korea, from September 2013 through January 2018. The primary outcome was in-hospital mortality. We made all comparisons using data from patients whose baseline characteristics were balanced by inverse probability of treatment weighting. RESULTS: Of patients who initially received empirical treatment with antibiotics, 95 (11.0%) received carbapenems and 655 (75.7%) received TGCs. Among the entire study cohort, there was no significant difference in in-hospital mortality between the carbapenem (25.8%) and TGC (25.3%) groups (adjusted odds ratio [aOR], 0.97; 95% CI, 0.85-1.11; P = .66). In the subgroup of patients with high chronic liver failure-sequential organ failure assessment (CLIF-SOFA) scores (score of 7 or greater, n = 314), carbapenem treatment was associated with lower in-hospital mortality (23.1%) than in the TGC group (38.8%) (aOR, 0.84; 95% CI, 0.75-0.94; P=.002). In contrast, among patients with lower CLIF-SOFA scores (n = 436), in-hospital mortality did not differ significantly between the carbapenem group (24.7%) and the TGC group (16.0%) (aOR, 1.06; 95% CI, 0.85-1.32; P = .58). CONCLUSIONS: For patients with a first presentation of SBP, empirical treatment with carbapenem does not reduce in-hospital mortality compared to treatment with TGCs. However, among critically ill patients (CLIF-SOFA scores ≥7), empirical carbapenem treatment was significantly associated with lower in-hospital mortality than TGCs.

Effectiveness of nivolumab versus regorafenib in hepatocellular carcinoma patients who failed sorafenib treatment.

BACKGROUND/AIMS: Several treatment options are currently available for patients with hepatocellular carcinoma (HCC) failing previous sorafenib treatment. We aimed to compare the effectiveness of regorafenib and nivolumab in these patients. METHODS: Consecutive HCC patients who received regorafenib or nivolumab after failure of sorafenib treatment were included. Primary endpoint was overall survival (OS) and secondary endpoints were time to progression, tumor response rate, and adverse events. Inverse probability of treatment weighting (IPTW) using the propensity score was conducted to reduce treatment selection bias. RESULTS: Among 150 study patients, 102 patients received regorafenib and 48 patients received nivolumab. Median OS was 6.9 (95% confidence interval [CI], 3.0-10.8) months for regorafenib and 5.9 (95% CI, 3.7-8.1) months for nivolumab (P=0.77 by log-rank test). In multivariable analysis, nivolumab was associated with prolonged OS (vs. regorafenib: adjusted hazard ratio [aHR], 0.54; 95% CI, 0.30-0.96; P=0.04). Time to progression was not significantly different between groups (nivolumab vs. regorafenib: aHR, 0.82; 95% CI, 0.51-1.30; P=0.48). HRs were maintained after IPTW. Objective response rates were 5.9% and 16.7% in patients treated with regorafenib and nivolumab, respectively (P=0.04). CONCLUSION: After sorafenib failure, the use of nivolumab may be associated with improved OS and better objective response rate as compared to using regorafenib.

Discontinuation of nucleos(t)ide analogues is not associated with a higher risk of HBsAg seroreversion after antiviral-induced HBsAg seroclearance: a nationwide multicentre study.

OBJECTIVE: Direct comparison of the clinical outcomes between nucleos(t)ide analogue (NA) discontinuation versus NA continuation has not been performed in patients with chronic hepatitis B who achieved HBsAg-seroclearance. Whether NA discontinuation was as safe as NA continuation after NA-induced surface antigen of HBV (HBsAg) seroclearance was investigated in the present study. DESIGNS: This multicentre study included 276 patients from 16 hospitals in Korea who achieved NA-induced HBsAg seroclearance: 131 (47.5%) discontinued NA treatment within 6 months after HBsAg seroclearance (NA discontinuation group) and 145 (52.5%) continued NA treatment (NA continuation group). Primary endpoint was HBsAg reversion and secondary endpoints included serum HBV DNA redetection and development of hepatocellular carcinoma (HCC). RESULTS: During follow-up (median=26.9 months, IQR=12.2-49.2 months), 10 patients (3.6%) experienced HBsAg reversion, 6 (2.2%) showed HBV DNA redetection and 8 (2.9%) developed HCC. Compared with NA continuation, NA discontinuation was not associated with HBsAg reversion in both univariable (HR=0.45, 95% CI=0.12 to 1.76, log-rank p=0.24) and multivariable analyses (adjusted HR=0.65, 95% CI=0.16 to 2.59, p=0.54). The cumulative probabilities of HBsAg reversion at 1, 3 and 5 years were 0.8%, 2.3% and 5.0% in the NA discontinuation group, and 1.5%, 6.3% and 8.4% in the NA continuation group, respectively. NA discontinuation was not associated with higher risk of either HBV redetection (HR=0.83, 95% CI=0.16 to 4.16, log-rank p=0.82) or HCC development (HR=0.53, 95% CI=0.12 to 2.23, log-rank p=0.38). CONCLUSION: The discontinuation of NA was not associated with a higher risk of either HBsAg reversion, serum HBV DNA redetection or HCC development compared with NA continuation among patients who achieved HBsAg seroclearance with NA.

Additional fibrate treatment in UDCA-refractory PBC patients.

BACKGROUND & AIMS: There is no proven treatment for ursodeoxycholic acid (UDCA)-refractory primary biliary cholangitis (PBC) other than obeticholic acid. Although fibrates have been reported to improve biochemical parameters, the long-term effects remain unclear. This study evaluated the effect of fibrate on clinical outcomes of UDCA-refractory PBC. METHODS: Patients whose alkaline phosphatase (ALP) was not normalized with at least 13 mg/kg of UDCA treatment for >1 year were included from two tertiary referral centres. The primary outcome was ALP normalization. Secondary outcomes included the development of cirrhosis and hepatic deterioration. Immortal time bias was adjusted using the Mantel-Byar method. RESULTS: A total of 100 UDCA-refractory PBC patients were included: 71 patients received UDCA alone (the UDCA group) and 29 patients received UDCA plus additional fibrate treatment of 160 mg/d fenofibrate or 400 mg/d bezafibrate (the fibrate/UDCA group). During the follow-up period, the probability of ALP normalization was significantly higher in the fibrate/UDCA group (hazard ratio [HR] = 5.00, 95% confidence interval = 2.87-8.27, P < 0.001). Among 58 non-cirrhotic patients (43 in the UDCA group and 15 in the fibrate/UDCA group), 19 patients (44.1%) in the UDCA group and none in the fibrate/UDCA group developed cirrhosis (HR = 0.12, P = 0.04). Hepatic deterioration (Child-Pugh score increase or signs of decompensated cirrhosis) occurred in 17 patients (23.9%) of the UDCA group and none in the fibrate/UDCA group in which the difference was significant (HR = 0.12, P = 0.04). CONCLUSIONS: In patients with UDCA-refractory PBC, additional fibrate treatment is associated with a higher probability of ALP normalization and a lower risk of cirrhosis development and hepatic deterioration.