Medical School, National and Kapodistrian University of Athens, General Hospital of Athens “Laiko”; “Metaxa” Cancer Hospital, Piraeus, Medical School, National and Kapodistrian University of Athens, Greece
aFirst Department of Gastroenterology, Medical School, National and Kapodistrian University of Athens, General Hospital of Athens “Laiko”, Athens, Greece (Alexandra Argyrou, Stavros P. Papadakos, Andreas Koutsoumpas, Jiannis Vlachogiannakos, George V. Papatheodoridis); bDepartment of Gastroenterology, “Metaxa” Cancer Hospital, Piraeus, Greece (Stamatina Vogli); cLaboratory of Physiology, Medical School, National and Kapodistrian University of Athens, Athens, Greece (Antonios Chatzigeorgiou)
We sincerely thank Dr. Montero [1] for his thoughtful comments on the modeling strategy used in our meta-analysis of serum anti-integrin avb6 autoantibodies for primary sclerosing cholangitis (PSC) [2]. His emphasis on hierarchical diagnostic test accuracy (DTA) methodology contributes to interpreting our findings [1].
We reanalyzed the original 2×2 data using a hierarchical bivariate random-effects (Reitsma) model with restricted maximum likelihood (REML), implemented in R (v4.4.x; mada package). This approach jointly models logit sensitivity and false-positive rate, accounting for between-study heterogeneity and correlation.
In the overall PSC analysis, pooled sensitivity was 0.705 and specificity 0.894 (area under the curve receiver operating characteristic [AUROC]: 0.893); positive likelihood ratio [PLR] was 6.65 and negative likelihood ratio [NLR] 0.33. Between-study heterogeneity was substantial (standard deviation (SD) logit sensitivity: 1.462; SD logit false positive rate: 1.197), with strong negative correlation (r≈−0.97), consistent with threshold variability (Supplementary Fig. 1,2).
All studies used data-derived thresholds based on healthy controls (mean +X SD). Three studies applied mean +3 SD [3-5], whereas 1 [6] used mean +2 SD, confirming threshold heterogeneity, handled within the hierarchical model.
In the predefined PSC plus inflammatory bowel disease (IBD) subgroup, 2 studies lacked IBD-only controls [3,5]; therefore, a hypothetical comparator was constructed using externally reported specificity (88%) [7] and a 1:1 case–control ratio, acknowledging assumptions introduced. Hierarchical analysis yielded sensitivity 0.831 (95%CI 0.418-0.971) and specificity 0.740 (95%CI 0.502-0.889), AUROC 0.835, PLR 3.20, and NLR 0.23, with substantial heterogeneity (Supplementary Fig. 3,4).
In PSC without IBD, specificity remained high (0.966) with PLR 14.04 (AUROC: 0.965). For PSC vs. other cholestatic diseases, sensitivity was 0.814 and specificity 0.959 (AUROC: 0.968; PLR: 19.85) (Supplementary Fig. 5-8).
Overall, hierarchical modelling confirms moderate sensitivity and high specificity, with strong rule-in performance in selected subgroups, reinforcing the conclusions of our work.
1. Arredondo Montero J. Comments on modeling strategy and data handling in a meta-analysis of anti-integrin αvβ6 for primary sclerosing cholangitis. Ann Gastroenterol 2026;39:379-381.
2. Papadakos SP, Vogli S, Argyrou A, et al. Serum anti-integrin avb6 autoantibodies for diagnosis of primary sclerosing cholangitis:a systematic review and meta-analysis. Ann Gastroenterol 2026;39:40-47.
3. Yoshida H, Shiokawa M, Kuwada T, et al. Anti-integrin avb6 autoantibodies in patients with primary sclerosing cholangitis. J Gastroenterol 2023;58:778-789.
4. Bloemen H, Livanos AE, Martins A, et al. Anti-integrin avb6 autoantibodies are increased in primary sclerosing cholangitis patients with concomitant inflammatory bowel disease and correlate with liver disease severity. Clin Gastroenterol Hepatol 2025;23:1612-1622.
5. Yasuda M, Shiokawa M, Kuwada T, et al;Japan PSC Study Group (JPSCSG). Anti-integrin avb6 autoantibody in primary sclerosing cholangitis:a Japanese nationwide study. J Gastroenterol 2025;60:118-126.
6. Roth D, Düll MM, Horst LJ, et al. Integrin aVb6:autoantigen and driver of epithelial remodeling in colon and bile ducts in primary sclerosing cholangitis and inflammatory bowel disease. J Crohns Colitis 2025;19:jjae131.
7. Yang J, Huang MMC, Liang MMJ, Lei MMY. The diagnostic performance of serum avb6 autoantibodies for ulcerative colitis:a systematic review and meta-analysis. Clin Res Hepatol Gastroenterol 2024;48:102317.