The prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD)-related advanced fibrosis and cirrhosis in the United States population utilizing AGILE 3 + and AGILE 4 scores: analysis of the NHANES 2017–2018 cycle – BMC Gastroenterology

The noninvasive blood tests such as APRI, NFS and FIB-4 are highly effective at ruling out advanced fibrosis and cirrhosis in MASLD patients but have a relatively low specificity and PPV [9]. Several studies have shown that elastography measurements (CAP, LSM) significantly improve the diagnostic accuracy for MASLD. For example, a prospective analysis of MASLD patients by Eddowes et al. demonstrated an increased prevalence for identifying liver steatosis and fibrosis using FibroScan, thereby showing increased negative predictive values (AURCO 0.87) [10]. Another study by Siddiqui et al. showed that the diagnostic accuracy of VCTE in detecting adults with MASLD accurately distinguishes advanced vs. earlier stages of fibrosis using liver histology as the standard reference [11]. Elastography is more sensitive and specific than traditional ultrasound in detecting fatty liver in the presence of moderate and high probability for fibrosis, however; combining it with blood biomarkers can further enhance the diagnostic accuracy [12, 13].

AGILE 3 + and AGILE 4 scores are an amalgam of elastography measurements (LSM), lab chemistry values (transaminases, platelets) and clinical parameters (presence of diabetes, gender and age), and are thus an attempt to enhance the overall accuracy of diagnostic testing for advance fibrosis and cirrhosis among MASLD patients, respectively. The incorporation of clinical parameters such as diabetes, gender and age into AGILE scores stems from the blatant evidence of their contribution to pathogenesis [14]. Furthermore, AGILE scores utilize a “dual cut-off algorithm” approach which has been shown to correctly classify advanced fibrosis and cirrhosis in hepatitis B patients, reduce the indeterminate grey zone and the false positive rate, and hence reducing the need for liver biopsy [15].

Similar “combining scores” have also been previously developed. For example, Newsome et al. developed FAST score by combining elastography measurements (LSM and CAP) and lab values (AST, ALT, or AST: ALT ratio) [16]. One of the shortcomings of this scoring system is that unlike AGILE scores, FAST score does not take clinical parameters into account. Furthermore, it primarily focuses on patients with fibrosis stage 2 or higher, whereas F ≥ 3 is more clinically relevant. In our study, we aimed to estimate the prevalence of MASLD-related advanced fibrosis and cirrhosis using the AGILE 3 + AGILE 4 scores; respectively, in a U.S. representative sample. In this study of 1244 patients with evidence of MASLD, the overall prevalence of AF due to MASLD based on AGILE 3 + score was 8.1%, corresponding to 4.5 million Americans, whereas based on the AGILE 4 score, the overall prevalence of MASLD cirrhosis was 1.1%, corresponding to 610,000 Americans.

The AGILE scoring system is a non-proprietary test and can be calculated using routinely collected clinical, elastography and laboratory parameters. It is crucial to balance the accuracy of the diagnostic tools and the feasibility of obtaining the diagnostic information in population-based epidemiological studies like this.

Limitations

The limitations of our study stem mainly from the nature of the AGILE scoring system itself. For instance, individuals who fall under the grey zone (11.5% using AGILE 3 + and 2.4% using AGILE 4) require further evaluation for the presence of MASLD cirrhosis, and therefore evaluating the cost-effectiveness of using AGILE3 + and AGILE 4 as a screening tool in high-risk populations is needed before implementing in clinical practice. Furthermore, despite using a nationally representative sample (NHANES database), the question whether the selected population truly represents the U.S population after applying the inclusion and exclusion criteria remains difficult to answer. The cut off for steatosis () was used based on published literature to maximize sensitivity [17], and different cut off may be used in the clinical settings.