Detection of polyreactive immunoglobulin G facilitates diagnosis in children with autoimmune hepatitis

Objective The detection of autoantibodies is essential to diagnose autoimmune hepatitis (AIH). Particularly in children, specificity of autoantibodies decreases due to lower titers being diagnostic and being present not only in AIH but also in other liver diseases. Recently, quantification of polyreactive IgG (pIgG) for detection of adult AIH showed the highest overall accuracy compared to antinuclear antibodies (ANA), anti-smooth muscle antibodies (anti-SMA), anti-liver kidney microsomal antibodies (anti-LKM) and anti-soluble liver antigen/liver pancreas antibodies (anti-SLA/LP). We aimed to evaluate the diagnostic value of pIgG for pediatric AIH. Design pIgG, quantified using HIP1R/BSA coated ELISA, and immunofluorescence on rodent tissue sections were performed centrally. The diagnostic fidelity to diagnose AIH was compared to conventional autoantibodies of AIH in training and validation cohorts from a retrospective, European multi-center cohort from nine centers from eight European countries composed of existing biorepositories from expert centers (n = 285). Results IgG from pediatric AIH patients exhibited increased polyreactivity to multiple protein and non-protein substrates compared to non-AIH liver diseases and healthy children. pIgG had an AUC of 0.900 to distinguish AIH from non-AIH liver diseases. pIgG had a 31–73% higher specificity than ANA and anti-SMA and comparable sensitivity that was 6–20 times higher than of anti-SLA/LP, anti-LC1 and anti-LKM. pIgG had a 21–34% higher accuracy than conventional autoantibodies, was positive in 43–75% of children with AIH and normal IgG and independent from treatment response. Conclusion Detecting pIgG improves the diagnostic evaluation of pediatric AIH compared to conventional autoantibodies, primarily owing to higher accuracy and specificity. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1007/s12072-024-10695-1.


Quantification of polyreactive immunoglobulin G
Patients' serum samples from Hannover Medical School were cryo-conserved at below -20°C.Serum samples from external centers were cryo-conserved according to local protocols and sent frozen to Hannover Medical School for centralized quantification of pIgG.
Samples were pseudonymized for autoantibody-testing and observers were blinded to any clinical information.
Quantification of pIgG using an ELISA to quantify reactivity to a peptide and BSA as blocking agent was performed as published 1 .In short, 0.01 µg of HIP1R fragment, 0.1µg of intersectin 1 (ITSN1) fragment or 0.01 µg of ubiquitin (UBC) per well was bound to 96 well ELISA plates over night at 4 °C.Plates were blocked with TBS and 5 % BSA for 30 minutes.Plates were washed with TBS with Tween20® 0.05 % (TBST) once.Serum samples were diluted 1:101 (v/v) in TBS and 5 % BSA and 100 µl per well were added to the ELISA plate and incubated for two hours.Plates were washed three times with TBST and incubated with a secondary rabbit anti-human anti-IgG antibody labeled with horseradish peroxidase for 30 minutes.Three washing steps with TBST were performed and 3, 3', 5, 5' tetramethyl benzidine (BioLegend, San Diego California) was added for 30 minutes for color reaction.Reaction was stopped with sulfuric acid.Optical density was read at 450 nm using an ELISA reader (Tecan Sunrise-Basic, Grödig, Austria).Sera of five patients with gradual increase in pIgG reactivity were measured in every experiment and used to compute a standard curve.Arbitrary units (AU) were calculated from the equation of the standard curve.Measurements were performed in Hannover, Germany.
As different AU dependent on center and storage duration were demonstrated, a normalization for these factors (referred to as normalized AU (nAU)) was performed as published 1 .

Liquid-Liquid Preincubation
Patient sera were diluted 1:250 (v/v) in 20 % BSA in TBS.Diluted samples were stored for preincubation at 4 °C over night.BSA preincubated samples were diluted 1:4 (v/v) due to prior dilution in TBS.These samples were added to the ELISA after blocking of the 96-well plates (100 µl/well) as described in the Methods section.

Immunofluorescence testing
IFT was performed by experienced technicians using the recommended methodology of the guidelines issued in 2004 by the Committee for Autoimmune Serology of the International Autoimmune Hepatitis Group 2 .Samples were pseudonymized for autoantibody-testing and observers were blinded to any clinical information.ANA, anti-SMA, anti-LKM and anti-LC1 were detected by IIF on sections of frozen rodent liver, stomach and kidney sections.Briefly, a commercial rodent multi-organ substrate panel (kidney, liver and stomach) was used (LKS Rat wrapped Standard Kit, Aesku.Diagnostics GmbG & Co. Wendelsheim, Germany).The sera were diluted, starting with a dilution of 1:20 up to 1:160, and applied to the slide to cover the entire tissue section and allow binding of the autoantibodies to the substrates.After washing, the sample was exposed to a second fluorochrome-labeled antibody.Finally, once washed again, the slides were examined under fluorescence microscope (Olympus BX60 Microscope, Evident Europe GmbH, Germany), and the antibody staining pattern was evaluated and interpreted accordingly to the guidelines 2 .

ELISA testing
An in-house ELISA was performed as published 3,4 .Samples were pseudonymized for autoantibody-testing and observers were blinded to any clinical information.Briefly, antibodies from defined anti-SLA or anti-LKM1 indicator sera were coated overnight in a volume per well of 50µl at room temperature in microtiter plates (Dynatech, el Paso, Texas for anti-LKM1, and Maxisorp, Nunc, Denmark for SLA).The supernatants were removed and after a washing step respective antigens were added.For the generation of the antigens rat liver was homogenized and centrifuged for 15 minutes at 3000 rpm.The pellet was discarded and the supernatant was centrifuged at 8500 rpm for 15 minutes.The pellet was discarded.The supernatant was further centrifuged for 60 min at 50000 rpm, antigens for LKM were collected from the pellet and antigens for SLA were collected from the supernatant and were added to the respective ELISA.All antigens were added at a concentration of 100 µg/ml and incubated for one hour at room temperature.Patient samples were diluted 1:10 in PBS + 10 mM EDTA and added to the microtiter plates following two washing steps.Incubation was done for one hour at room temperature.Following three additional washing steps, avidin-peroxidase and sodium perborate dissolved in citrate buffer were added to microtiter plates.The photometric reaction was stopped after five minutes, and the absorbance was measured.The percentage of inhibition of the indicator serum to its respective autoantigen was used as a surrogate for the antibody titer.

Statistical analysis was performed using SPSS (version
27, SPSS, Inc., Chicago, IL), GraphPad Prism (version 10.0.2 (232), GraphPad Software Inc., La Jolla, CA), MedCalc software (version 19.4.1, MedCalc Software Ltd, Ostend, Belgium) and easyROC package (version 1.3.1)[26] in RStudio (2021.09.0+351 "Ghost Orchid" Release, RStudio, PBC, Boston, MA).The Mann-Whitney U test was used to compare quantitative data between two groups and the Kruskal-Wallis test was used for more than two groups with Bonferroni post-hoc test.The Fisher's exact test was used to compare categorical variables.Correlation analyses were calculated with Spearman's rank correlation.The AUC and the Youden's Index were used to guide identification of respective cut-off values.AUCs were compared using DeLong's test[27].Accuracy of diagnostic test was calculated as (true positive+true negative)/total number.Sensitivities and specificities were compared with the McNemar test.Overall accuracies were compared by the comparison of the 95 % CI.P-values below 0.05 (two-tailed) were considered significant in all analyses.

Figures Supplemental figure 1 :
Figures