Objectives: Our aim was to identify the potential predictors of liver transplant in patients with autoimmune hepatitis based on data at initial presentation.
Materials and Methods: We retrospectively evaluated records of children diagnosed with autoimmune hepatitis between 2000 and 2025. We excluded 21 patients with de novo autoimmune hepatitis (n = 5) and incomplete data (n = 16). We compared demographic, clinical, biochemical, and histopathological data in 2 study groups: transplanted (n = 9) and not transplanted (n = 81).
Results: Of 90 included patients, 63 (70%) were female. Median age at diagnosis was 115 months (range, 8-215 months). Among included patients, 70 had type 1 autoimmune hepatitis, 13 had type 2 autoimmune hepatitis, and 7 had seronegative autoimmune hepatitis. Nine patients underwent liver transplant (5 with type 1 autoimmune hepatitis, 4 with type 2 autoimmune hepatitis). Rate of transplant was significantly higher in patients with type 2 autoimmune hepatitis (P = .⁰³). Compared with the nontransplant group, patients in the transplant group were more likely to have type 2 autoimmune hepatitis; higher gamma-glutamyl transferase, bilirubin, international normalized ratio, and immunoglobulin levels; and lower albumin and platelet levels. In addition, this group was more likely to have relapsing disease and moderate-to-severe fibrosis at diagnosis. Multivariate analysis showed that moderate-to-severe fibrosis and total bilirubin level were significant predictors of liver transplant (odds ratio: 21.84, 95% CI, 1.34-355.25; P = .⁰³⁰ and odds ratio: 1.22, 95% CI, 1.02-1.47; P = .⁰²⁹, respectively). Autoimmune hepatitis type 2 was not found to be associated with outcome (P = .⁸³⁹).
Conclusions: Systematic risk assessment at diagnosis is important to identify high-risk patients who may benefit from close monitoring, early escalation of therapy, and timely referral for transplant evaluation. Tailoring management strategies based on these predictive factors may improve long-term outcomes in pediatric autoimmune hepatitis.

Key words : Children, Fibrosis, Liver dysfunction, Liver transplantation, Prognosis

Introduction
Patients with pediatric autoimmune hepatitis (AIH) can present with incidental hypertransaminasemia, acute hepatitis, acute liver failure, or chronic liver disease. Diagnosis is based on clinical findings and liver enzyme abnormalities, supported by high immunoglobulin G (IgG) and autoantibody positivity, compatible liver histology, and the exclusion of other causes of hepatitis.^1-3 Pediatric AIH is commonly classified into type 1, type 2, and seronegative based on serologic markers.^3-5 Most children are responsive to immunosuppressive therapy, but a substantial proportion of patients may experience a relapsing course or treatment resistance. Liver transplantation (LT) should be considered in patients with acute liver failure, end-stage liver disease, and hepatic malignancy.^1,3,6 Previous studies have identified clinical, biochemical, and histological markers that reflect disease severity in pediatric AIH.^1,3 However, studies that have specifically focused on definitive outcomes such as LT remain limited. Moreover, findings on independent predictors of transplant are heterogeneous, and no clear consensus exists on the prognostic effects of disease subtype. Therefore, identifying independent predictors of LT in well-characterized pediatric cohorts with long-term follow-up remains important. In this study, we aimed to identify the predictors of LT in our patient group based on data available at initial presentation to assist in systematic risk assessment and individualized management strategies.

Materials and Methods
This retrospective cohort study was conducted at a tertiary-care pediatric hepatology and LT center at Gazi University (Ankara, Türkiye). We retrospectively reviewed the medical records of children diagnosed with AIH between January 2000 and October 2025. The Ethics Committee of Gazi University approved the protocol before the initiation of the study, and all procedures were conducted in accordance with the ethical standards of the institutional research committee and with the 1975 Declaration of Helsinki. Patients were eligible if they were aged <18 years at diagnosis and diagnosed with AIH based on clinical, laboratory, and histopathological evaluation, consistent with the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) guidance on pediatric AIH.³ We excluded patients with incomplete baseline data and de novo AIH after LT. Patients were categorized as type 1 AIH, type 2 AIH, or seronegative AIH based on autoantibody profiles at diagnosis. We collected data at the time of initial diagnosis from patient records using a standardized data extraction form. We recorded age at diagnosis and sex. We documented the following clinical features at diagnosis: hepatomegaly, splenomegaly, jaundice, encephalopathy, and the presence of associated autoimmune diseases. We also collected baseline laboratory evaluations, including complete blood count, liver biochemistry (aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transferase [GGT], alkaline phosphatase, bilirubin), albumin, international normalized ratio (INR), and IgG; IgG was recorded as a fold above the upper limit of normal. We obtained histopathological data from liver biopsy reports, including Ishak activity score and Ishak fibrosis stage; for comparative analyses, we categorized fibrosis as none to mild versus moderate to severe. In patients in whom liver biopsy could not be performed because of acute liver failure or clinical instability, histology variables were treated as missing for histology-based comparisons. We defined relapse and remission criteria according to ESPGHAN recommendations.³ The primary outcome of our study was LT during follow-up. For patients who underwent LT, we collected detailed transplant-related variables, including the indication for LT (acute liver failure, acute-on-chronic liver failure, or end-stage liver disease), the time interval from AIH diagnosis to LT, and donor type (living or deceased donor). Posttransplant complications were divided into early and late according to timing (events within the first year vs later).⁷ We collected posttransplant outcomes, including posttransplant survival duration; 1-, 5-, and 10-year survival rates; and mortality, specifically the timing of death (early vs late posttransplant) and the cause of death when available. We performed statistical analyses to compare transplanted versus nontransplant patients. In addition, within the transplant group, we conducted subgroup analyses according to AIH subtype (type 1 vs type 2). We presented continuous variables as median (minimum to maximum) and categorical variables as number (percent). We used the Mann-Whitney U test for between-group comparisons for continuous variables and the χ² test or Fisher exact test for categorical variables, as appropriate. We conducted analyses by using an available-case approach for variables with missing data. We performed multivariable logistic regression analysis to identify independent predictors of LT. The final model included total bilirubin level and fibrosis severity at diagnosis (moderate-to-severe vs none-to-mild), and we reported results as odds ratios (ORs) with 95% CI. P < .⁰⁵ was considered statistically significant. We used IBM SPSS Statistics for Windows version 26.0 for all statistical analyses.

Results
Among 111 children diagnosed with AIH, 21 were excluded due to de novo AIH (n = 5) and due to incomplete data (n = 16). Therefore, demographic, clinical, biochemical, and histopathological data of 90 patients were analyzed, comparing the LT group (n = 9 patients) and non-LT group (n = 81 patients). Of 90 total patients, 63 (70%) were female, and the median age at diagnosis was 115 months (range, 8-215 mo). Patients were classified as type 1 AIH (n = 70), type 2 AIH (n = 13), or seronegative AIH (n = 7) (Figure 1). In the LT group, 5 had type 1 AIH (5 of 70 [7.1%] in the type 1 group) and 4 had type 2 AIH (4 of 13 [30.8%] in the type 2 group), indicating a significantly higher transplant rate in type 2 AIH (P = .⁰³). In the LT group, 7 patients were female, and median age at diagnosis was 11.3 years (range, 0.7-17 y), which did not differ significantly between AIH types. Time from diagnosis to LT varied, with a median of 4 months overall (range, 4 days to 168 months) and no significant difference by AIH type. Among patients with available biopsy, moderate-to-severe fibrosis at diagnosis was present in 5 cases. The median posttransplant survival duration was 90 months (range, 1-156 mo). Overall, 1-, 5-, and 10-year posttransplant survival rates were 88%, 88%, and 78%. In the LT group, 2 deaths were recorded, 1 in each AIH type with causes including infection (n = 1) and vascular complication (n = 1), occurring in the early and late posttransplant periods, respectively. Liver biopsy could not be performed in 2 patients due to acute liver failure, and 1 patient was lost to follow-up (Table 1). Sex and age at diagnosis were similar between the LT and non-LT groups. Type 2 AIH was more frequent among LT patients than type 1 AIH (45% vs 11%, P = .⁰³). At diagnosis, patients in the LT group more often presented with jaundice (88% vs 37%; P = .⁰⁰⁴) and encephalopathy (44% vs 1%; P < .⁰⁰¹) than non-LT patients. In the LT group, baseline laboratory findings were consistent with a more severe phenotype, with higher GGT, total and direct bilirubin, INR, and IgG levels and lower albumin and platelet counts. No significant differences were observed in aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, hemoglobin, or white blood cell levels between the 2 groups. Histological analysis showed similar Ishak scores between the 2 groups, but moderate-to-severe fibrosis was more prevalent in LT patients compared with non-LT patients (87% vs 37%; P ≤ .⁰⁰⁸). Relapsing disease course was observed in all LT patients (100% vs 20%, P < .⁰⁰¹) (Table 2). In multivariate analysis, fibrosis severity and total bilirubin level were found to be significant predictors of LT. Moderate-to-severe fibrosis at diagnosis was associated with a markedly increased likelihood of LT (OR: 21.84; 95% CI, 1.34-355.25; P = .⁰³⁰). Similarly, elevated total bilirubin levels were independently associated with an increased risk (OR: 1.22; 95% CI, 1.02-1.47; P = .⁰²⁹). Autoimmune hepatitis type 2 was not significantly associated with the outcome (P = .⁸³⁹).

Discussion
In this retrospective study of pediatric AIH, we found that certain initial clinical features can predict which children will eventually require LT. Patients who finally required LT presented more frequently with jaundice and encephalopathy and had a biochemical profile suggesting advanced liver dysfunction, including higher GGT and bilirubin levels, impaired synthetic function, and thrombocytopenia. Histologically, patients in the LT group had more advanced fibrosis at diagnosis. Importantly, in multivariable analysis, only baseline total bilirubin and moderate-to-severe fibrosis independently predicted LT, highlighting that both acute functional impairment and established chronic structural damage contributed to risk of requiring LT in pediatric AIH. Published pediatric series have reported that the proportion of children with AIH requiring LT varies across cohorts. In selected single-center and registry-based studies, LT rates ranged from 4.5% to 19%.^8-12 Consistent with this heterogeneity, the ESPGHAN guidance has highlighted that reported LT rates in pediatric AIH can be highly variable, ranging broadly (9% to 55%).³ Discrepancies between studies may reflect differences in referral patterns (transplant center vs hepatology center), center experience, baseline disease severity at presentation, and treatment protocols.³ In our cohort, 10% of children with AIH underwent LT, falling within the range reported in the literature, with transplant performed mainly for acute liver failure, acute-on-chronic decompensation, or end-stage liver disease. Available pediatric data have also suggested generally favorable posttransplant outcomes in AIH (80% to 90%),^1,3,13 and our cohort showed similarly good mid-term survival (5-year survival rate of 88%), supporting transplant as an effective rescue strategy when progression occurs despite optimized medical therapy. In children with acute and/or chronic liver disease, biochemical and clinical indicators reflecting disease severity such as Pediatric End-Stage Liver Disease score are used to determine requirement for LT.¹⁴ In AIH, serum IgG levels are considered an important biochemical marker of disease activity. Clinical guidelines have emphasized IgG as part of treatment response assessment and follow-up, and normalization of IgG is regarded as one of the main therapeutic targets.^1,3 In our cohort, elevated bilirubin, prolonged INR, hypoalbuminemia, and encephalopathy in the LT group was consistent with the core components of established prognostic scoring systems (eg, Pediatric End-Stage Liver Disease/Model for End-Stage Liver Disease, Child-Pugh) used to determine transplant priority. Previous studies have primarily focused on predictors of native liver survival. In Maggiore and colleagues,¹⁰ lower serum GGT levels at diagnosis and a persistently normal INR during follow-up were associated with improved survival; in addition, male sex, absence of ascites, and absence of cirrhosis were also linked to better native liver survival. In another pediatric cohort, the AIH-2 subtype was associated with better native liver survival, and higher albumin levels, higher platelet counts, and normal INR at diagnosis were linked to more favorable outcomes.¹⁵ However, relapse emerged as a strong determinant of death or LT during follow-up.¹⁵ In Yang and colleagues, hypoalbuminemia and decompensated cirrhosis at presentation were associated with an increased risk of death or LT.¹⁶ In our study, the lower platelet counts observed in the transplanted group likely reflected more advanced fibrosis and portal hypertension, although platelet count is not directly included in standard prognostic scoring systems. Presence of cirrhosis at diagnosis is well-established as an indicator of poor prognosis. In line with this, patients who underwent LT in our cohort had higher Ishak fibrosis scores, and moderate-to-severe fibrosis was strongly associated with the need for LT. Although studies have suggested that type 2 AIH may have a more severe clinical course than type 1,^3,17-19 more recent reports indicated no significant prognostic difference between the 2 subtypes.^10,16 In contrast, in a large series, higher rates of LT and death occurred in patients with type 1 AIH compared with type 2.15 These findings indicate that consensus remains unclear regarding the effects of AIH subtype on prognosis. In our study, although the LT rate was higher in patients with type 2 AIH, subtype was not identified as an independent predictor in multivariate analysis. This finding suggested that the risk of transplant may be more closely related to disease severity at presentation rather than the AIH subtype itself. Previous pediatric and adult studies have shown that not achieving or maintaining complete remission is associated with fibrosis progression and an increased risk of LT.^3,15,18 Consistent with this evidence, relapse was significantly more frequent in our LT group. This finding suggested that ongoing disease activity contributes to progressive liver damage and increases the likelihood of transplant. Together, these results underscore the clinical importance of achieving and maintaining sustained remission through close monitoring and appropriate treatment optimization. Our study had several limitations. First, the retrospective and single-center design may limit the generalizability of our findings. Second, the relatively small number of transplanted patients may have reduced the statistical power of the multivariable analyses and contributed to wide confidence intervals. Third, variations in treatment protocols and follow-up duration over the extended study period may have influenced outcomes. In conclusion, the need for LT in pediatric AIH is closely linked to disease severity at diagnosis. In our study, baseline total bilirubin levels and fibrosis stage emerged as independent predictors of LT. These results highlight the importance of systematic risk assessment at presentation to identify high-risk patients who may benefit from close monitoring, timely strengthening of therapy, and early referral for transplant evaluation. Tailoring management strategies according to these baseline risk factors may ultimately improve long-term outcomes in pediatric AIH.