Prediction of new organ onset in recurrent immunoglobulin G4-related disease during 10 years of follow-up

Zheng Liu,Yu Peng,Jieqiong Li,Hui Lu,Linyi Peng,Jiaxin Zhou,Shuang Zhou,Can Huang,Mu Wang,Liang Zhu,Hua Chen,Li Wang,Yunyun Fei,Yan Zhao,Xiaofeng Zeng,Wen Zhang
DOI: https://doi.org/10.1111/joim.13477
Abstract:Objective: Frequent relapse is a prominent challenge in managing immunoglobulin G4-related disease (IgG4-RD). According to the types of organs involved in relapse, relapse patterns were divided into recurrent organ involvement (ROI) and new organ involvement (NOI). We aimed to investigate the discrepancy in clinical relapse patterns and establish an effective prognostic nomogram for NOI. Methods: We retrospectively enrolled 125 IgG4-RD patients who experienced relapse during the follow-up period. Patients were classified into two groups: those with NOI (including NOI and NOI + ROI) and without NOI (ROI). Logistic regression analyses were used to assess the risk factors for NOI. The results were externally validated by a separate prospective cohort of 39 patients with relapse. Results: There were 81 (64.8%) and 44 (35.2%) patients without NOI and with NOI, respectively. Patients without NOI showed higher baseline disease activity. The most common ROIs were the lacrimal gland and submandibular gland, while the lung and urinary system were the most involved in NOI. Re-elevation of serum IgG4 level to 74.31% of baseline was associated with NOI. Multiple relapses, organ involvement type at baseline, glucocorticoids combined with immunosuppressive drugs (IM) or IM alone during the maintenance period, and relapse IgG4/baseline IgG4 ratio were included in the nomogram. Both internal and external validations showed good agreement and discrimination. Conclusions: About one third of IgG4-RD patients with relapse suffer from NOI. We developed a risk stratification model that can effectively predict the future risk of NOI. Glucocorticoid and IM combined therapy during maintenance is also recommended.
What problem does this paper attempt to address?