Letter: Double PBC ‐Related Autoantibody Positivity—How Essential Is It for Diagnosis?

We read with great interest the article by Liu et al. on the diagnostic value of dual positivity for AMA/AMA-M2 and anti-gp210/sp100 in primary biliary cholangitis (PBC). The large, biopsy-based cohort and systematic assessment of PBC-related antibodies are commendable 1. We would like to raise some points that may influence how clinicians interpret the reported ‘diagnostic rates’ and the perceived superiority of dual positivity. First, the cohort includes only patients positive for at least one PBC-related antibody who also underwent liver biopsy. As the authors note, biopsy is generally reserved for patients with persistent abnormalities or diagnostic uncertainty 2-4. The reported diagnostic yields of 80.05% for ‘only-AMA(s) positivity’, 53.85% for ‘only anti-gp210/sp100 positivity’ and 94.22% for ‘dual positivity’ therefore reflect a highly selected, biopsy-enriched hepatology population rather than the performance of these serological patterns in all patients undergoing autoantibody testing. Second, the handling of AMA by indirect immunofluorescence (IIF) versus AMA-M2 by immunoblot is not detailed. The composite term ‘AMA(s)’ is used, but concordance between AMA-IIF and AMA-M2 is not provided, nor is it described how discordant results (IIF-positive/immunoblot-negative or vice versa) were classified. Because AMA-M2 immunoblot using recombinant PDC-E2 antigens is highly specific for PBC and less operator-dependent than IIF on rat kidney tissue alone (instead of the recommended triple-tissue substrate), counting IIF-positive/AMA-M2-negative cases as ‘AMA-positive’ may introduce false-positive anti-mitochondrial IIF-patterns, particularly in the non-PBC group 5-7. This could lower the apparent diagnostic yield of the ‘only AMA(s)’ group and accentuate differences from dual-positive patients. Any impression of ‘limited specificity’ of AMA might therefore reflect IIF technical and interpretative issues, especially when only kidney sections are used, rather than an intrinsic limitation of AMA-M2 specificity 5-7. A comparison of AMA-IIF with AMA-M2 immunoblot across the cohort, stratified by final diagnosis (PBC/non-PBC) and a sensitivity analysis restricting ‘AMA-positive’ to AMA-M2-confirmed cases would clarify the robustness of the ‘AMA-only’ group and the incremental contribution of PBC-specific anti-nuclear autoantibodies (ANA). A notable feature of the cohort is the high rate of PBC/AIH overlap (43%) and the substantial proportion of AIH diagnoses (20%) among autoantibody-positive, non-PBC patients. Such frequencies differ markedly from most published series and may further influence the observed diagnostic yields 2, 3, 6, 8. In a previous study of 4371 consecutive sera referred to our autoimmune laboratory for heterogeneous indications, we showed that PBC-related ANA patterns (‘rim-like/membranous’ and ‘multiple nuclear dots’) corresponding to anti-gp210 and anti-Sp100 are rare but highly specific for PBC, irrespective of AMA status 8, 9. Their specificity in an unselected laboratory population aligns with the excellent ‘diagnostic rate’ of dual positivity reported by Liu et al., although the higher PBC prevalence and biopsy-based selection in their cohort necessarily amplify these yields compared with a general autoantibody-testing population. Finally, further details on the 145 autoantibody-positive patients without PBC at baseline, of whom only 2.07% progressed to PBC, would be valuable. The distribution of autoantibody profiles (AMA-positive by IIF and/or immunoblot, anti-gp210/sp100-positive, other ANA patterns or combinations) among progressors and non-progressors could help refine risk stratification and follow-up strategies in autoantibody-positive, non-PBC patients. Limiting the claim of ‘maximum diagnostic value’ to this biopsy-selected, antibody-positive population and detailing the AMA-positivity as outlined above would strengthen the clinical interpretability of this important study. Conceptualisation: Alessandro Granito. Methodology: Alessandro Granito, Georgios Pappas, Paolo Muratori. Writing original draft: Alessandro Granito. Writing – review and editing: Alessandro Granito, Georgios Pappas and Paolo Muratori. Supervision: Paolo Muratori. The authors have nothing to report. The authors declare no conflicts of interest pertaining to this manuscript. Please refer to the accompanying ICMJE Disclosure Forms for further details. This article is linked to Liu et al. paper. To view this article, visit https://doi/10.1111/apt.70614. The data that support the findings of this study are available from the corresponding author upon reasonable request.