Background/Aims: The incidence of steatotic liver disease (SLD) is increasing across all age groups as the incidence of obesity increases worldwide. The existing noninvasive prediction models for SLD require laboratory tests or imaging and perform poorly in the early diagnosis of infrequently screened populations such as young adults and individuals with healthcare disparities. We developed a machine learning-based point-of-care prediction model for SLD that is readily available to the broader population with the aim of facilitating early detection and timely intervention and ultimately reducing the burden of SLD. Methods: We retrospectively analyzed the clinical data of 28,506 adults who had routine health check-ups in South Korea from January to December 2022. A total of 229,162 individuals were included in the external validation study. Data were analyzed and predictions were made using a logistic regression model with machine learning algorithms. Results: A total of 20,094 individuals were categorized into SLD and non-SLD groups on the basis of the presence of fatty liver disease. We developed three prediction models: SLD model 1, which included age and body mass index (BMI); SLD model 2, which included BMI and body fat per muscle mass; and SLD model 3, which included BMI and visceral fat per muscle mass. In the derivation cohort, the area under the receiver operating characteristic curve (AUROC) was 0.817 for model 1, 0.821 for model 2, and 0.820 for model 3. In the internal validation cohort, 86.9% of individuals were correctly classified by the SLD models. The external validation study revealed an AUROC above 0.84 for all the models. Conclusions As our three novel SLD prediction models are cost-effective, noninvasive, and accessible, they could serve as validated clinical tools for mass screening of SLD.

Background/Aims: The incidence of steatotic liver disease (SLD) is increasing across all age groups as the incidence of obesity increases worldwide. The existing noninvasive prediction models for SLD require laboratory tests or imaging and perform poorly in the early diagnosis of infrequently screened populations such as young adults and individuals with healthcare disparities. We developed a machine learning-based point-of-care prediction model for SLD that is readily available to the broader population with the aim of facilitating early detection and timely intervention and ultimately reducing the burden of SLD. Methods: We retrospectively analyzed the clinical data of 28,506 adults who had routine health check-ups in South Korea from January to December 2022. A total of 229,162 individuals were included in the external validation study. Data were analyzed and predictions were made using a logistic regression model with machine learning algorithms. Results: A total of 20,094 individuals were categorized into SLD and non-SLD groups on the basis of the presence of fatty liver disease. We developed three prediction models: SLD model 1, which included age and body mass index (BMI); SLD model 2, which included BMI and body fat per muscle mass; and SLD model 3, which included BMI and visceral fat per muscle mass. In the derivation cohort, the area under the receiver operating characteristic curve (AUROC) was 0.817 for model 1, 0.821 for model 2, and 0.820 for model 3. In the internal validation cohort, 86.9% of individuals were correctly classified by the SLD models. The external validation study revealed an AUROC above 0.84 for all the models. Conclusions As our three novel SLD prediction models are cost-effective, noninvasive, and accessible, they could serve as validated clinical tools for mass screening of SLD.

Background/Aims: The incidence of steatotic liver disease (SLD) is increasing across all age groups as the incidence of obesity increases worldwide. The existing noninvasive prediction models for SLD require laboratory tests or imaging and perform poorly in the early diagnosis of infrequently screened populations such as young adults and individuals with healthcare disparities. We developed a machine learning-based point-of-care prediction model for SLD that is readily available to the broader population with the aim of facilitating early detection and timely intervention and ultimately reducing the burden of SLD. Methods: We retrospectively analyzed the clinical data of 28,506 adults who had routine health check-ups in South Korea from January to December 2022. A total of 229,162 individuals were included in the external validation study. Data were analyzed and predictions were made using a logistic regression model with machine learning algorithms. Results: A total of 20,094 individuals were categorized into SLD and non-SLD groups on the basis of the presence of fatty liver disease. We developed three prediction models: SLD model 1, which included age and body mass index (BMI); SLD model 2, which included BMI and body fat per muscle mass; and SLD model 3, which included BMI and visceral fat per muscle mass. In the derivation cohort, the area under the receiver operating characteristic curve (AUROC) was 0.817 for model 1, 0.821 for model 2, and 0.820 for model 3. In the internal validation cohort, 86.9% of individuals were correctly classified by the SLD models. The external validation study revealed an AUROC above 0.84 for all the models. Conclusions As our three novel SLD prediction models are cost-effective, noninvasive, and accessible, they could serve as validated clinical tools for mass screening of SLD.

Background/Aims: The incidence of steatotic liver disease (SLD) is increasing across all age groups as the incidence of obesity increases worldwide. The existing noninvasive prediction models for SLD require laboratory tests or imaging and perform poorly in the early diagnosis of infrequently screened populations such as young adults and individuals with healthcare disparities. We developed a machine learning-based point-of-care prediction model for SLD that is readily available to the broader population with the aim of facilitating early detection and timely intervention and ultimately reducing the burden of SLD. Methods: We retrospectively analyzed the clinical data of 28,506 adults who had routine health check-ups in South Korea from January to December 2022. A total of 229,162 individuals were included in the external validation study. Data were analyzed and predictions were made using a logistic regression model with machine learning algorithms. Results: A total of 20,094 individuals were categorized into SLD and non-SLD groups on the basis of the presence of fatty liver disease. We developed three prediction models: SLD model 1, which included age and body mass index (BMI); SLD model 2, which included BMI and body fat per muscle mass; and SLD model 3, which included BMI and visceral fat per muscle mass. In the derivation cohort, the area under the receiver operating characteristic curve (AUROC) was 0.817 for model 1, 0.821 for model 2, and 0.820 for model 3. In the internal validation cohort, 86.9% of individuals were correctly classified by the SLD models. The external validation study revealed an AUROC above 0.84 for all the models. Conclusions As our three novel SLD prediction models are cost-effective, noninvasive, and accessible, they could serve as validated clinical tools for mass screening of SLD.

Background/Aims: Chronic hepatitis B (CHB) is related to an increased risk of extrahepatic malignancy (EHM), and antiviral treatment is associated with an incidence of EHM comparable to controls. We compared the risks of EHM and intrahepatic malignancy (IHM) between entecavir (ETV) and tenofovir disoproxil fumarate (TDF) treatment. Methods: Using data from the National Health Insurance Service of Korea, this nationwide cohort study included treatment-naïve CHB patients who initiated ETV (n=24,287) or TDF (n=29,199) therapy between 2012 and 2014. The primary outcome was the development of any primary EHM. Secondary outcomes included overall IHM development. E-value was calculated to assess the robustness of results to unmeasured confounders. Results: The median follow-up duration was 5.9 years, and all baseline characteristics were well balanced after propensity score matching. EHM incidence rate differed significantly between within versus beyond 3 years in both groups (P<0.01, Davies test). During the first 3 years, EHM risk was comparable in the propensity score-matched cohort (5.88 versus 5.84/1,000 person-years; subdistribution hazard ratio [SHR]=1.01, 95% confidence interval [CI]=0.88–1.17, P=0.84). After year 3, however, TDF was associated with a significantly lower EHM incidence compared to ETV (4.92 versus 6.91/1,000 person-years; SHR=0.70, 95% CI=0.60–0.81, P<0.01; E-value for SHR=2.21). Regarding IHM, the superiority of TDF over ETV was maintained both within (17.58 versus 20.19/1,000 person-years; SHR=0.88, 95% CI=0.81–0.95, P<0.01) and after year 3 (11.45 versus 16.20/1,000 person-years; SHR=0.68, 95% CI=0.62–0.75, P<0.01; E-value for SHR=2.30). Conclusions TDF was associated with approximately 30% lower risks of both EHM and IHM than ETV in CHB patients after 3 years of antiviral therapy.

Background: /Aim: Atezolizumab plus bevacizumab and lenvatinib are currently available as first-line therapy for the treatment of unresectable hepatocellular carcinoma (HCC). However, comparative efficacy studies are still limited. This study aimed to investigate the effectiveness of these treatments in HCC patients with portal vein tumor thrombosis (PVTT). Methods: We retrospectively included patients who received either atezolizumab plus bevacizumab or lenvatinib as first-line systemic therapy for HCC with PVTT. Primary endpoint was overall survival (OS), and secondary endpoints included progressionfree survival (PFS) and disease control rate (DCR) determined by response evaluation criteria in solid tumors, version 1.1. Results: A total of 52 patients were included: 30 received atezolizumab plus bevacizumab and 22 received lenvatinib. The median follow-up duration was 6.4 months (interquartile range, 3.9-9.8). The median OS was 10.8 months (95% confidence interval [CI], 5.7 to not estimated) with atezolizumab plus bevacizumab and 5.8 months (95% CI, 4.8 to not estimated) with lenvatinib (P=0.26 by log-rank test). There was no statistically significant difference in OS (adjusted hazard ratio [aHR], 0.71; 95% CI, 0.34-1.49; P=0.37). The median PFS was similar (P=0.63 by log-rank test), with 4.1 months (95% CI, 3.3-7.7) for atezolizumab plus bevacizumab and 4.3 months (95% CI, 2.6-5.8) for lenvatinib (aHR, 0.93; 95% CI, 0.51-1.69; P=0.80). HRs were similar after inverse probability treatment weighting. The DCRs were 23.3% and 18.2% in patients receiving atezolizumab plus bevacizumab and lenvatinib, respectively (P=0.74). Conclusion The effectiveness of atezolizumab plus bevacizumab and lenvatinib was comparable for the treatment of HCC with PVTT.