Clinical management of relapse in type 1 autoimmune pancreatitis: a systematic review and meta-analysis

Objectives: Type 1 autoimmune pancreatitis (AIP) has a high risk of relapse, but evidence to guide management is limited. We evaluated relapse patterns, predictors, and the comparative effectiveness of maintenance and relapse treatment. Methods: We systematically reviewed studies of patients with type 1 AIP receiving maintenance treatment or relapse treatment. Single-arm, pairwise, and Bayesian network meta-analyses were performed to synthesize relapse outcomes and compare treatment strategies. Results: Thirty-seven studies involving 5,742 patients, including two randomized controlled trials and thirty-five cohort studies. Maintenance therapy reduced relapse risk (RR, 0.58; 95% CI, 0.46-0.72). Among Asian cohorts, longer-term glucocorticoid therapy was associated with a lower risk than shorter-term therapy (RR, 0.46; 95% CI, 0.23-0.89). Maintenance glucocorticoid doses of ≥ 5 mg/d showed a trend toward a lower relapse risk compared with doses of < 5 mg/d (RR, 0.61; 95% CI, 0.36-1.04). Relapse risk accumulated over time, and the pancreas was the predominant relapse site (71%). Male sex, jaundice, diabetes mellitus, elevated serum IgG4 (> 1 × ULN), bile duct involvement, and other organ involvement were associated with increased relapse risk. For relapse treatment, rituximab showed the most favorable estimates, followed by glucocorticoids plus immunomodulators, but most network comparisons were based on low- or very-low-certainty evidence. Conclusions: Maintenance therapy is associated with a lower relapse risk in type 1 AIP, with longer-term glucocorticoid therapy showing benefit in Asian cohorts. Relapse risk accumulates over time and predominantly involves the pancreatobiliary system. Rituximab may provide favorable outcomes for relapse treatment, but its comparative superiority remains uncertain because of low-certainty evidence.