Introduction
Steatotic liver disease (SLD) is silent in its early stage of disease but associated with long-term cirrhosis complications and hepatocellular carcinoma. Individuals with cardiometabolic risks are the target population for SLD screening. Fatty Liver Index (FLI) is the most widely used scoring system for identifying individuals with SLD on a population level. However, FLI incorporates waist circumference and gamma-glutamyl transferase, which are not routinely measured in clinical practice. The Dallas Steatosis Index (DSI), which utilizes more readily available clinical parameters, was recently developed to detect SLD in the US population.
Aims & Methods
This study aims to validate the DSI for diagnosing SLD in Thai individuals with cardiometabolic risk. We conducted a cross-sectional study using retrospective data from 2019 to 2023 at a tertiary care hospital in Southern Thailand. Inclusion criteria were individuals aged 18 years and older who participated in health check-up programs that included abdominal ultrasound (USG), blood tests, and had at least one cardiometabolic risk according to the latest SLD nomenclature consensus statement. SLD was diagnosed when liver steatosis was identified on the USG report. Exclusion criteria included individuals with positive HBsAg and/or HCV antibodies, as well as those with incomplete data on variables required for calculating the DSI, i.e., age, sex, ALT, triglycerides, body mass index (BMI), fasting blood sugar (FBS), diabetes, and hypertension status. The performance of DSI in diagnosing SLD in individuals with cardiometabolic risk was assessed by calculating the area under the receiver operating characteristic curve (AUROC), and sensitivity and specificity were evaluated at the given cutoffs.
Results
Of 5987 individuals without exclusion criteria, 5466 (91.3%) had at least one cardiometabolic risk and were included in this analysis. The prevalence of SLD diagnosed by USG was 48.8% (N=2667) in these at-risk individuals. DSI was calculated using published formula (https://dsi.wustl.edu/) and presented as predictive probabilities. DSI showed a very good performance in diagnosing SLD with an AUROC of 0.816 (95%CI: 0.805-0.827). At the given cutoff of 0.2 and 0.5 as the low and high cutoffs for detecting SLD, DSI <0.2 showed a sensitivity of 87.5% and NPV of 82.2% for ruling out SLD, and at the DSI >0.5, the specificity and PPV were 91.2% and 83.97% to ascertain SLD status, respectively. Among those 2667 individuals with SLD, 1358 (51%) who had available alcohol consumption data were then categorized into MASLD, MetALD, and ALD subclassifications. We found that 95.1% met the definition of MASLD, whereas 2.1% and 2.8% were categorized as MetALD, and ALD, respectively. The prevalence of advanced fibrosis (defined as FIB4 > 2.67) were 1.5%, 10.7%, and 10.5% in the MASLD, MetALD, and ALD groups, respectively (p<0.001).
Conclusion
The DSI demonstrated satisfactory diagnostic performance for diagnosing SLD in Thai individuals with cardiometabolic risk with an AUROC of 0.816 and may serve as an alternative to FLI for identifying SLD at the population level, particularly in resource-limited settings where imaging techniques such as USG and transient elastography are not readily accessible. However, additional studies are warranted to directly compare the diagnostic accuracy of DSI and FLI. Among those with SLD, MASLD were far more common than MetALD and ALD, but the presence of advanced fibrosis was significantly more prevalent in the latter two groups than those with MASLD.
References
McHenry S, Park Y, Browning JD, Sayuk G, Davidson NO. Dallas Steatosis Index Identifies Patients With Nonalcoholic Fatty Liver Disease. Clin Gastroenterol Hepatol. 2020 Aug;18(9):2073-2080.e7. doi: 10.1016/j.cgh.2020.01.020. Epub 2020 Jan 23. PMID: 31982611; PMCID: PMC7913470.