I would like to appreciate Dr. Haeryoung Kim and Dr. Sook-Hyang Jeong for their Correspondence [1] as a reply to my editorial entitled “Evaluation of the histological scoring systems of autoimmune hepatitis: A significant step towards the optimization of clinical diagnosis.” [2] As a general hepatologist, I read the Correspondence with great interest. First, although their study, titled “Comparison of four histological scoring systems for autoimmune hepatitis to improve diagnostic sensitivity,” [3] clearly validated lobular hepatitis as an important histological component in diagnosing autoimmune hepatitis (AIH), even in cases with acute presentation, they highlighted and pondered the relevance of co-occurrence of at least mild interface hepatitis in lobular hepatitis-predominant AIH cases in their Correspondence. Indeed, while lobular and portal interface hepatitis are topologically distinct pathological features, they not only may coexist but also might transit in the liver of patients with AIH [4,5]. Moreover, lymphoplasmacytic infiltrates are common components of each lesion [5]. I could paraphrase this issue into the outstanding question: how are lobular and interface hepatitis in AIH spatiotemporally associated? Spatial transcriptomic analysis of the lobular hepatitis-predominant AIH cases with mild interface hepatitis may provide us with clues to this question and deepen our understanding of the pathognomonic features of each hepatitis pattern in AIH.

Second, they highlighted another diagnostic conundrum in their Correspondence: differentiation between AIH and druginduced liver injury with AIH-like features (DI-AIH) [6], again in the context of demarcated hepatic inflammation. Although a previous study involving a relatively large cohort of patients diagnosed at the Mayo Clinic failed to distinguish between classical AIH (n=237) and DI-AIH (n=24) in histological analysis [7], the fact that suspected drugs in the study were predominantly limited to minocycline (n=11) and nitrofurantoin (n=11) raises concern about the external validity of their analysis. In that regard, I agree with their comment for the importance of reassessing histological differences between AIH and DI-AIH using existing histological scoring systems through a multicenter or multinational study. Such a study should include cases involving a wide range of suspected drugs and be coupled with sufficient clinical information.

I would like to conclude this Reply by encouraging our colleagues, hepatologists and pathologists to continue active communication in order to optimize the histological diagnosis of AIH, as underscored by Dr. Haeryoung Kim and Dr. Sook-Hyang Jeong. Bidirectional dialogue regarding laboratory findings, medication history, and histopathological interpretation of liver biopsies among us, potentially with assistance from artificial intelligence in the future, will undoubtedly serve as a driving force in addressing the unmet needs of AIH, a disease characterized by dynamic and heterogeneous manifestations.