Background & Aims: Although autoimmune pancreatitis (AIP) responds well to corticosteroid therapy, relapse during maintenance corticosteroid therapy or after the withdrawal of corticosteroid treatment is not uncommon. To date, the factors related to relapse of AIP have not been fully explored. Methods: To determine the clinical and genetic predictors relating to the relapse of AIP, we evaluated clinical factors, HLA polymorphisms, and the amino acid sequences in 40 patients with AIP. Results: At a median follow-up period of 40 months (range, 12–67 months), relapse developed in 13 of 40 patients with AIP (33%), in whom complete remission was achieved with oral corticosteroid therapy. Among demographics, clinical characteristics in the initial diagnosis of AIP, we could not find any clinical predictor for relapse of AIP; however, in amino acid sequence analysis for relapse of AIP, the substitution of aspartic acid to nonaspartic acid at residue 57 of DQβ1 showed a significant association with relapse of AIP (nonrelapse group, 29.6%; relapse group, 100%; P = .00003; odds ratio, 3.38; 95% confidence interval, 1.9–6.0). There was a significant difference in the timing of relapse of AIP, according to density of the nonaspartic acid residue at DQβ1 57 (nonaspartic acid homozygosity: mean ± SD, 6.7 ± 4.2 months; nonaspartic acid heterozygosity: mean ± SD, 33 ± 11 months; P < .001). Conclusions: Substitution of aspartic acid to nonaspartic acid at DQβ1 57 appears to represent a key genetic factor for relapse of AIP (ClinicalTrials.gov number, NCT00444444).
Section snippets
Study Population
Between February 2002 and May 2006, 46 consecutive patients were diagnosed with AIP. Patients treated surgically because pancreatic cancer could not be excluded in the initial diagnosis (n = 4) and patients who had an inadequate dose of corticosteroid due to poor compliance (n = 2) were excluded from the study. The remaining 40 patients (32 men and 8 women) ages 33–78 years (median age, 58.5 years) were enrolled in the study.
The diagnostic criteria23, 24, 25, 26 of AIP in the present study were
Clinical Outcome of Long-Term Follow-up for AIP With Corticosteroid Treatment
With corticosteroid therapy, complete remission was achieved in all enrolled patients. During a median follow-up period of 40 months (range, 12–67 months), 13 of 40 patients (33%) experienced relapses of AIP. These relapses occurred during maintenance corticosteroid therapy or after a complete discontinuation of corticosteroids. Seven of the 13 patients experienced a relapse on the maintenance dosage of prednisolone (2.5–7.5 mg/day; median period from initial diagnosis to relapse, 6 months;
Discussion
During the genetic analysis, we found that substitution of aspartic acid to nonaspartic acid at the 57 residue of DQβ1 may relate to relapse of AIP (Figure 1). The onset of relapse of AIP was determined, moreover, according to the density of nonaspartic acid at DQβ1 57. Patients who experienced a relapse with homozygosity of nonaspartic acid had a significant tendency toward the early onset of relapse (mean, 6.7 months), whereas those with heterozygosity of nonaspartic acid had a delayed onset
Autoimmune pancreatitis (AIP) is a rare form of pancreatitis that may lead to endocrine and exocrine insufficiency if left untreated. AIP clinically responds to glucocorticoids (GCs) therapy, but multiple GCs courses are often required to maintain remission with detrimental effects on glycaemic control.
In this systematic review and meta-analysis, we aimed to assess the rate of endocrine and of exocrine insufficiency at diagnosis and at follow up in patients with AIP as well as the impact of GC therapy on pancreatic function in the long-term.
The MEDLINE, SCOPUS, and EMBASE databases were searched from inception to August 2021 to identify studies reporting data on endocrine and exocrine insufficiency in patients with AIP. Pooled events were calculated using a random-effect model and expressed in terms of pooled prevalence rates.
A total of 6522 AIP patients and sixty-two studies were included in the analysis. The pooled estimate rate for the overall prevalence of diabetes in AIP at baseline was 37% (95% CI 32-42, I2 96%). The pooled prevalence rate of exocrine insufficiency was 45% (95%CI 32.9-57.4; I2 97%). The pooled estimate rate of diabetes at follow-up was 44% (95%CI 26.1-62.4) in studies where GCs were given to 100% of patients and 42% (95%CI 30.6-52.9) in studies where GCs were given to less than 100% of patients.
A large proportion of patients with AIP displays concomitant exocrine and endocrine insufficiency at the time of diagnosis. The incidence of diabetes at the longest available follow up tends to increase in patients treated with GCs.
Autoimmune diseases are often associated with human leukocyte antigen (HLA) haplotypes, indicating that changes in major histocompatibility complex (MHC)-dependent self-peptide or antigen presentation contribute to autoimmunity. In our study, we aimed to investigate HLA alleles in a large European cohort of autoimmune pancreatitis (AIP) patients.
Hundred patients with AIP, diagnosed and classified according to the International Consensus Diagnostic Criteria (ICDC), were prospectively enrolled in the study. Forty-four patients with chronic pancreatitis (CP) and 254 healthy subjects served as control groups. DNA was isolated from blood samples and two-digit HLA typing was performed with sequence-specific primer (SSP-) PCR. HLA allele association strength to AIP was calculated as odds ratio.
We uncovered a strong enrichment of HLA-DQB1 homozygosity in type 1 and type 2 AIP patients. Moreover, a significantly increased incidence of the HLA-DRB1∗16 and HLA-DQB1∗05 alleles and a concomitant lack of the HLA-DRB1∗13 allele was detected in AIP type 1 and type 2 patients. In contrast, the HLA-DQB1∗02 allele was underrepresented in the ‘not otherwise specified’ (NOS) AIP subtype. We detected no significant difference in the HLA-DRB3, HLA-DRB4 and HLA-DRB5 allele frequency in our cohort.
Although AIP type 1 and type 2 are characterized by distinct histopathological characteristics, both subtypes are associated with the same HLA alleles, indicating that the disease might rely on similar immunogenic mechanisms. However, AIP NOS represented another subclass of AIP.
This finding rises the possibility of familial clusters of IgG4-RD. Moreover, as FGFBP2 protein is secreted by cytotoxic T cells and binds fibroblast growth factor, an alteration in this protein might enhance the CD4+ T cell function, and consequently might play a role in the pathogenesis of IgG4-RD. In addition to HLA loci, several non-HLA gene have been identified as risk genes in IgG4-RD [31,32] (Table 2). A high-throughput sequence analysis was performed in 27 Japanese patients with AIP with the aim to identify possible gene candidates for this disease [31].
IgG4-related disease (IgG4-RD) represents an immune-mediated fibroinflammatory condition with peculiar histopathologic changes that can affect various organs. In 2012 its unified nomenclature was published, which allows to abandon other synonymous names. Up to now, only little is known about its epidemiology around the world. However, although it is generally considered a rare condition, the number of patients with IgG4-RD is increasing enormously. Likewise, the annual number of publications on this subject has increased progressively. The spectrum of clinical manifestations in IgG4-RD is highly variable, depending on the severity of the disease as well as the presence of organ(s) involvement. This review gives an overview on changing epidemiology of IgG4-RD focusing the attention on the large cohorts of patients published in the literature.
In 27 studies6,9,11,12,25–30,37,39,41,44–48,50–52,54 the length of follow-up evaluation was longer than 2 years (in 2 studies53,55 it was not clearly specified). Methodologic quality scores (scale, 0–12) ranged from 537,38,43 to 1227,45 (Supplementary Table 3). The percentage of men ranged from 45%35 to 89%.47
Risk for relapse after induction of remission with steroid therapy has been studied extensively in patients with autoimmune pancreatitis (AIP), but findings have been equivocal. We performed a systematic review and meta-analysis to estimate the relapse rate of AIP after initial remission after steroid treatment and to identify factors associated with relapse.
Three reviewers searched MEDLINE, SCOPUS, and EMBASE until July 2018 to identify studies on rate of relapse of AIP after induction of remission with steroid therapy. A pooled estimate was calculated using the DerSimonian and Laird method for a random-effects model. This study was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.
Thirty-six studies met the inclusion criteria for meta-analysis. The median follow-up time was 40.8 months. Fifty-two percent of patients were classified as having type 1 AIP. The pooled estimate of relapse rate was 33% (95% CI, 30%–37%). A higher proportion of patients with type 1 AIP had a relapse compared with patients with type 2 AIP (37.5% vs 15.9%; P < .001). We found significant heterogeneity among studies (P < .01). Long-term maintenance therapy with steroids and study quality were associated independently with AIP relapse, after we adjusted for year of publication by multivariate meta-regression.
In a systematic review and meta-analysis, we found that a large proportion of patients with AIP treated successfully with steroid induction therapy had a relapse (33%)—particularly patients with type 1 AIP (37%). Maintenance steroid therapy lasting longer than 1 year could reduce risk of relapse. However, the data characterizing relapse rates are of limited quality, indicating the need for randomized controlled trials and new immunosuppressive drugs.
