Reply: Diagnostic accuracy of AGILE3+ score for advanced fibrosis in patients with non-alcoholic fatty liver disease: A systematic review and meta-analysis

We would like to thank Hui Liu et al1 for their interest in our systematic review and meta-analysis aiming to evaluate the diagnostic accuracy of the AGILE3+ score for detecting advanced fibrosis in patients with NAFLD.2 The issues raised by the authors merit to be addressed. First, regarding the potential use of the term "MASLD" (metabolic dysfunction-associated steatotic liver disease) for our systematic research, it is important to note that we systematically searched MEDLINE, Ovid Embase, Scopus, and Cochrane Library electronic databases from the inception to the 24th of April 2023. Consequently, it was not possible to conduct systematic research using the keyword "MASLD" as articles related to this new classification were published later. Second, it is worth acknowledging that the term "MAFLD" (metabolic dysfunction-associated fatty liver disease) has not been exclusively utilized as a keyword in any article discussing the diagnostic accuracy of the AGILE3+ score until April 2023. Indeed, Hui Liu et al should have recognized that the main cohort of study by Tang et al3 was originally included in our meta-analysis,2 while the validation cohort from that paper was effectively not included. During the process of data extraction and quality assessment, Federico Ravaioli, Andrea Dalbeni, and Alessandro Mantovani discussed the decision only to include data from the derivation cohort. The reason for this decision was twofold: first, the derivation cohort was larger than the validation cohort; second, a conservative approach was preferred due to the cutoffs used for the rule-in and rule-out. We acknowledge that this approach may be open to criticism. Therefore, to be definitive, we conducted subanalyses and sensitivity analyses that strongly suggest the accuracy of the AGILE3+ score in detecting advanced fibrosis in patients with liver disease. Moreover, we would like to clarify that the validation cohort in the study by Tang et al consisted of 142 patients, not 132, as disputed by the authors.3 Third, given that the study by Noereddin et al4 was fully published after our research, it was not reasonable to include it in our systematic review and meta-analysis. Fourth, with respect to the absence of a meta-analysis conducted on the AGILE 4 score, we would like to point out that this was not the aim of the present study. However, we encourage the authors to contemplate this aim for their next article if they believe it would be beneficial and pertinent to the scientific community. Fifth, regarding the typos observed in the graphic abstract, we would like to remind the authors that they have read a pre-edited draft and that we had already corrected the typos during the proofreading process. Before expressing any judgment, it is always best to wait for the final edited version since it is specified and known that pre-proofs could contain some typos. Unfortunately, we should all keep in mind that the Dunning-Kruger effect is always around the corner. Finally, we disagree with the authors,3 who proposed performing a network meta-analysis to improve the current meta-analysis on the AGILE3+score for detecting advanced fibrosis. The only viable option to enhance the current meta-analysis is conducting an individual patient meta-analysis, as stated in the manuscript. Nevertheless, to dispel any doubts, we have repeated the analysis, including the validation cohort of the study by Tang et al3 and the study by Noureddin et al4: by the rule-out cutoff, the overall sensitivity and specificity were 85% (95% CI: 74%–92%; I2=97%) and 58% (95% CI: 46%–70%; I2=98%), respectively; by the rule-in cutoff, the overall sensitivity and specificity were 69% (95% CI: 59%–78%; I2=90%) and 86% (95% CI: 80%–91%; I2=97%), respectively. In conclusion, our rigorous meta-analysis of over 7600 patients with biopsy-confirmed NAFLD, including 690 newly added patients, reaffirms the validity and accuracy of the AGILE3+ score as a diagnostic tool for NAFLD with advanced fibrosis. These results are significant, as they provide clear evidence that the AGILE3+ score can be used effectively in clinical practice to identify patients who may require biopsy and emerging pharmacotherapies. The conclusions of paper2 are accurate, and our findings underline the importance of utilizing the AGILE3+ score in diagnosing NAFLD with advanced fibrosis.